DYNAMICS OF POVERTY, LABOR MARKET AND PUBLIC POLICIES IN
LATIN AMERICA
DYNAMICS OF POVERTY, LABOR MARKET AND PUBLIC POLICIES IN LATIN AMERICA
Luis Beccaria (Universidad Nacional de General Sarmiento)
Roxana Maurizio (Universidad Nacional de General Sarmiento - CONICET)
Ana Laura Fernández(Universidad Nacional de General Sarmiento)
Ana Paula Monsalvo (Universidad Nacional de General Sarmiento)
Mariana Álvarez (Universidad Nacional de General Sarmiento - CONICET)
Dynamics of Poverty, Labor Market and Public Policies in Latin AmericaCopyright © Partnership for Economic Policy (PEP), 2011
ALL RIGHTS RESERVED.No part of this publication may be reproduced,stored in a retrieval system, or transmittedin any form or by any means—whether virtual, electronic,mechanical, photocopying, recording, or otherwisewithout the written permission of the copyright owner.
TABLE OF CONTENTS
Acknowledgements .............................................................................. ix
Introduction ............................................................................................ 2
Economic growth, labor market and poverty in Latin America: recent trends and structural characteristics ......................................... 5
Literature Review ................................................................................. 10
Data Sources ......................................................................................... 18
Approach and Methodology ................................................................ 25
The Absolute Poverty Line Approach to Poverty .......................... 25 Methodology .................................................................................. 29
The Evolution of Poverty Incidence in Five Latin American Countries and its Characteristics ........................................................ 35
Stochastic Dominance ................................................................... 35 The Shapley Decomposition of changes in poverty ..................... 38 Poverty Status and Household Characteristics ............................ 40
Poverty Dynamics in Five Latin American Countries ....................... 46
Transition Matrix ........................................................................... 46 Factors directly associated with exit rates .................................... 51 Factors directly associated with entry rates ................................. 60 Modeling poverty dynamics ......................................................... 63 Analysis of sensitivity to the poverty line ..................................... 66 Conclusions and Policy Recommendations ........................................ 66
References ............................................................................................ 72
Annexes ................................................................................................ 80
LIST OF GRAPHS
Graph 1 Per capita GDP growth rate. Latin America 1990-2009(%) ........................................... 6
Graph 2 GDP growth and unemployment rate Latin America, 2003-2009 (%) ......................................... 6
Graph 3 Gini Coefficient Latin America Countries, in 2002 and 2008* ................. 7
Graph 4 Poverty and extreme poverty rates Latin America, 1980-2008 (% of total population) ......... 8
Graph 5 Poverty and extreme poverty rates Latin American Countries, 2009 (% of total households) .................................................... 9
Graph 6 Evolution of Poverty Incidence Five Latin American countries (selected years) .................. 36
Graph 7 Difference between FGT Curves Five Latin American countries ..................................... 37
Graph 8 Relative contribution of growth and redistribution to poverty incidence reduction Five Latin American countries ..................................... 39
Graph 9 Specific poverty rates for different household types Five Latin American countries Average ca. 2003-ca. 2008. ............................................ 42
Graph 10 Composition of total and poor households Five Latin American countries Average ca.2003-ca.2008 .............................................. 44
Graph 11 Specific poverty rate by occupational characteristics of the household head (Baseline=registered wage earner) Five Latin American countries ca. 2003- ca. 2006 .......................................................... 45
Graph 12 Poverty Entry and Exit Rates according to poverty incidence Five Latin American countries ..................................... 48
Graph 13 Relative Risk of entering and exiting poverty Households with and without children Five Latin American countries ..................................... 49
Graph 14 Relative importance of the events associated with exit rates Five Latin American countries ..................................... 53
Graph 15 Relative importance of exclusively labor events associated with exit rates Five Latin American countries ..................................... 54
Graph 16 Frequency of increases in the number of employed members by type of occupation Five Latin American countries ..................................... 56
Graph 17 Conditional probability of exiting poverty by type of occupation Five Latin American countries ..................................... 57
viii ix
ACKNOWLEDGEMENTS
This work was carried out with financial and scientific support from the Partnership for Economic Policy (PEP), which is financed by the Australian Agency for International Development (AusAID), the Canadian International Development Agency and the International Development Research Centre (IDRC).
We are very thankful to Abdelkrim Araar, Sami Bibi, John Cockburn and Jean-Yves Duclos for their technical support which was of great help in the realization of this project. We also thank the four anonymous PEP referees for their thoughtful remarks, Soledad Villafañe for her assistance in data processing and Ana Flávia Machado and Rafael Perez Ribas for their help with the income imputation methodology for Brazil.
Previous versions of this paper have been presented in a number of national and international conferences: the Annual Meeting of the Canadian Economics Association (Quebec, May 2010), the 8th PEP General Meeting (Dakar, June 2010), the Octavas Jornadas sobre Mercado de Trabajo y Equidad en Argentina (Buenos Aires, August 2010), the Conference on Dynamics of poverty, labor market and public policy in Latin America (Buenos Aires, October 2010) and the Reunión Anual de la Asociación Argentina de Economía Política (Buenos Aires, November 2010). We are very thankful to all the discussants and participants for their useful comments and remarks.
PB 1
LUIS BECCARIA, ROXANA MAURIZIO, ANA LAURA FERNÁNDEZ
ANA PAULA MONSALVO AND MARIANA ÁLVAREZ
Dynamics of Poverty, Labor Market and Public Policies in Latin America
ABSTRACT
From 2003 and up to the 2008-2009 international crisis, Latin America experienced six years of high and sustained economic growth which had a positive impact on social and labor market
indicators through the creation of jobs, the reduction of unemployment and a slight recovery of average wages. All of this resulted in both lower poverty and extreme poverty rates: in that period, the region experienced a drop of 11 percentage points (p.p.) in poverty and 6 p.p. in extreme poverty. In this context, it is highly relevant to study the flows into and out of poverty behind this significant reduction in poverty.
The main objective of this research is to study poverty dynamics in five Latin American countries, emphasizing a comparative point of view. The study specifically aims to analyze the extent to which countries with different levels of poverty incidences differ in terms of the intensity of poverty exits and entries, to identify the relative importance of events associated with poverty transitions (such as factors related to the labour market, demographic changes and public policy), and to examine the ways in which these events affect households with different characteristics.
To attain these objectives, we perform a dynamic analysis of panel data from regular household surveys. The countries under study are Argentina and Costa Rica with a relatively low poverty incidence; Brazil as an intermediate case; and Ecuador and Peru with relatively high poverty rates. This selection of countries allows having a quite complete outlook on social deprivation in the region.
This dynamic analysis is useful for policy recommendations to overcome high poverty levels in the region, both by reducing the probability of falling into poverty and increasing the chances of moving out of poverty.
2
INTRODUCTION
Although during the 2003-2008 period Latin America experienced six years of sustained growth that resulted in an improvement of labor indicators and
lower poverty and extreme poverty rates, the region still endures high levels of inequality, labor precariousness, poverty and social vulnerability.
One of the factors directly related to economic and social deprivation is the high inequality of opportunities and outcomes that still persists in these countries. Although the distribution of income has improved in the last years, inequality continues to be one of the distinctive features of Latin American countries. Yet, the dynamics of poverty resulting from the changes observed in inequality differ across countries.
The analysis of the factors associated with the level and evolution of poverty in individual countries has been the subject of an extended literature. There are also several studies that compare employment, inequality and poverty levels as well as their evolution in different Latin American countries.1 However, there is scarce research focused on poverty dynamics in individual countries of the region, especially on the factors related to poverty transitions. Furthermore, it was not possible to identify any study that compares poverty dynamics across Latin American countries.
Perhaps, the lack of adequate dynamic data in the region has been one of the reasons explaining the still small number of studies on poverty transitions. However, the supply of information that can be used to construct panel data has increased, even if it is not strictly longitudinal, and can thus be used to analyze income mobility and the factors associated with it.
This knowledge gap on poverty dynamics in the region is particularly worrisome given the high level of income mobility in Latin American countries. This is largely a result of precarious employment – that generally leads to high job instability2–
1 See, for example, Beccaria et al. (2007), ECLAC (2007), Stallings and Peres (2002), Zepeda et al. (2007 and 2009).
2 See for example Beccaria and Maurizio (2003, 2005), Kugler (2000), Paes de Barros and Leite Corseuil (1999).
2 3
representing a large share of the labour market and of low coverage for occupational risk, such as unemployment assistance or other kinds of public cash transfers. Labor instability is also high in these economies due to their relatively unstable macroeconomic situation.
The analysis of the characteristics of poverty dynamics is of great importance for policy design, for even when the level of aggregate incidence is relatively low and/or does not change, the flows of households exiting and entering poverty may still be large. Furthermore, the study of poverty transitions may offer significant evidence on the factors directly related to them. Specifically, such analyses allow identifying whether the events that trigger poverty entries and exits involve the labor market, are of a demographic character or respond to specific public policies. In order to do so, it is necessary to design an exhaustive typology of the events experienced by individuals that might take their households into or out of poverty.
It is also important to account for the fact that the level of and changes in poverty do not behave randomly and thus differ significantly between households with different characteristics. On the contrary, household composition and the employment situation of household members are among the most important observable factors associated with poverty. Characteristics of the household head and presence or absence of children are therefore dimensions which can be expected to have a strong impact on poverty dynamics.
The general objective of this paper is to study poverty dynamics in five Latin American countries, emphasizing a comparative point of view. In particular, the study aims at:
• Analyzing to what extent countries with different or similar poverty incidences may show poverty exit and entry rates of different intensities.
• Identifying the importance of different events associated with poverty transitions focusing on those related to the labor market, changes in household composition and to public policies.
• Tackling the question of whether households stay poor (or stay out of poverty) because they do not experience any positive (or negative) event, or because the impacts of the
4
event is not large enough to result in a transition into or out of poverty.
• Evaluating the relative effectiveness of different positive or negative events in taking households out of or into poverty derived from their intensity and from their conditional probability of changing the poverty status of households.
• Analyzing the effect of these events across households with different structures and characteristics. Presence or absence of children is one of the most important dimensions to be analyzed. This will help evaluate, for example, whether households with children have a greater or lesser chance of experiencing a positive event relative to other households, and/or whether those episodes are as effective at providing the household with the means to exit poverty. This is particularly relevant given the high prevalence of child poverty in Latin America.
• Suggesting general policy recommendations which either aim to reduce the incidence and impact of events which lead households into poverty or which aim to increase poverty exits, in both cases to reduce high poverty levels prevailing in the region.
As it was mentioned above, there are almost no longitudinal surveys for Latin American countries that can be used to follow a household over a long period of time.3 However, there are household surveys with rotating samples that can be used to construct panels of households who are interviewed in at least two successive periods. One panel of households enters the sample in each “wave” while another leaves. It is thus possible to compare the proportion of the sample between two successive waves, a percentage which differs among countries.
The countries that have been selected for this analysis are those for which the available information allows the analysis of poverty dynamics based on panel data from household surveys: Argentina, Brazil, Costa Rica, Ecuador and Peru. Taking into account the differences in poverty incidence exhibited by these
3 As it will be detailed later, the only exception is the Encuesta de Caracterización Socioeconómica Nacional (CASEN) of Chile.
4 5
countries, this selection provides a broad and complete overview of the poverty situation in the region. The years being analyzed are not necessarily the same for each country, but always fall between 2003 and 2008, a period of high and stable economic growth in Latin America.
The paper follows with a brief analysis of the general characteristics of the period under study in Latin America. Section 3 reviews the literature on poverty dynamics, detailing the different empirical approaches used in the international literature as well as the results obtained for Latin American countries. Section 4 presents the data sources and the panel data built for each country. Section 5 describes the theoretical approach and the methodology employed. Section 6 makes an in depth analysis of the evolution of poverty in Latin America and its characteristics. Section 7 focuses on the dynamics of poverty by first estimating the transition matrix and then by identifying the direct factors associated with entry and exit rates. An econometric modeling of poverty dynamics is also included. Finally, section 8 presents the conclusions and policy recommendations.
ECONOMIC GROWTH, LABOR MARKET AND POVERTY IN LATIN AMERICA: RECENT TRENDS
AND STRUCTURAL CHARACTERISTICS
As it was already mentioned, between 2003 and 2008, Latin America has experienced six years of sustained growth. During this period, per capita GDP rose at an annual average rate of 3.4%, a rate of unprecedented magnitude and duration in the region (Graph 1). This growth, unlike other periods of economic expansion in Latin America, was accompanied by fiscal and trade surpluses. The current account surplus in the balance of payments was linked to an improvement in the terms of trade in some countries, and to an increase in remittances in others. On the other hand, the improvement of public finances observed in these countries led to higher primary surpluses and to the elimination of operational deficits, which in turn resulted in better public debt profiles.
6
GRAPH 1Per capita GDP growth rate. Latin America 1990-2009 (%)
- 3,0
- 2,0
- 1,0
0,0
1,0
2,0
3,0
4,0
5,0
6,0
7,0
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009
3.4%
Source: ECLAC
The recent period of high economic growth experienced by Latin America –which lasted until the 2008-2009 international crisis- had a positive impact on social and labor market indicators through the creation of jobs -especially formal ones-, the reduction of unemployment and the slight recovery of average wages. Employment rates showed a positive trend, rising from 52.3% to 55% between 2003 and 2008. During this period, the regional unemployment rate fell from 11.4% to 7.5% (Graph 2).
GRAPH 2 GDP growth and unemployment rate Latin America, 2003-2009 (%)
11.0
10.3
9.1
8.6
7.4
8.3
7.9
7.0
7.5
8.0
8.5
9.0
9.5
10.0
10.5
11.0
11.5
2003 2004 2005 2006 2007 2008 2009
Une
mpl
oym
ent r
ate
-2
-1
0
1
2
3
4
5
6
7
GD
P gr
owth
GDP growth rate Unemployment ratee
Source: ECLAC and ILO
6 7
The positive macroeconomic situation seems to have facilitated the recovery of wages, a situation that in some countries was also favored by the implementation of active real minimum wage policies and the reactivation of collective bargaining, together with measures that promoted the formalization of employment. Furthermore, these measures also had a positive impact on reducing inequality by extending the coverage of labor institutions to previously excluded groups.
Indeed, these years were characterized by a reduction of income inequality in almost all the countries. As it is shown in Graph 3, with a few exceptions, the Gini coefficient is lower in 2008 than in the early years of the century.
GRAPH 3Gini Coefficient Latin America Countries, in 2002 and 2008*
Argentina
Bolivia
Brazil
El Salvador
Guatemala Honduras
Mexico
Nicaragua
Uruguay
Venezuela
Chile
Colombia
Costa Rica
Ecuador
PanamaParaguay
Perú
República Dominicana
0.40
0.45
0.50
0.55
0.60
0.65
0.70
0.40 0.45 0.50 0.55 0.60 0.65 0.70
Source: ECLAC
*Data for 2002 corresponds to the latest available information for the period between 2000 and 2002 whereas data for 2008 corresponds to the latest information for the period between 2004 and 2008.
Lower inequality and higher incomes resulted in lower poverty and extreme poverty incidence rates. As it was already mentioned, in the 2003-2008 growth period, the region experienced a drop of 11 p.p. in poverty and 6 p.p. in extreme poverty rates (Graph 4). This reduction in poverty rates was also accompanied by a fall in the number of poor people, as opposed to the situation in the eighties and nineties.
8
However, despite the progress made along the expansion phase, 32% of the population in Latin America still lived in poverty in 2010, while 13% lived in extreme poverty. The total number of poor people in that year was around 180 million, of which 72 million were indigent (ECLAC, 2010).
GRAPH 4 Poverty and extreme poverty rates Latin America, 1980-2008
(% of total population)
40.5
43.3
48.3 45.7
43.5 43.9 44.0
39.8
36.3 34.1
33.0
5
10
15
20
25
30
35
40
45
50
55
1980 1986 1990 1994 1997 1999 2002 2005 2006 2007 2008
Poverty
Extreme Poverty-11 p.p
Source: ECLAC
Behind this regional outlook, Latin America exhibits very
heterogeneous poverty incidences across countries. In this context, the countries under study cover a broad spectrum of diverse situations. As it is shown in Graph 5, according to ECLAC (2010), Argentina and Costa Rica belong to a group of relatively low poverty rates; Brazil belongs to the group of middle-low poverty and Peru and Ecuador to the group of middle-high poverty rates. This mixed panorama enriches the comparative analysis of poverty dynamics in the region.
8 9
GRAPH 5 Poverty and extreme poverty rates Latin American Countries, 2009
(% of total households)
0
10
20
30
40
50
60
70
80U
rugu
ay
Chi
le
Cos
ta R
ica
Arg
entin
a
Bra
zil
Pana
má
Vene
zuel
a
Peru
Mex
ico
Dom
.Rep
.
Ecua
dor
Col
ombi
a
El S
alva
dor
Bol
ivia
Gua
tem
ala
Para
guay
Nic
arag
ua
Hon
dura
s
Low poverty Middle - low poverty Middle - high poverty
High poverty
Source: ECLAC
In addition, the region still shows significant deficits when it comes to the labor dimension, of which The most recurrent features are high unemployment rates, underemployment, precariousness, informality, inequality and low average wages. According to ECLAC, around 7% of the active population was unemployed in 2008, while 50% of the employed were informal workers. Moreover, income distribution was still very unequal even after a period of inequality reduction: the poorest 40% of households obtained 15% of total income, while the richest 10% received 34% of income in 2008.
Within this context, certain groups of workers (among them, the less qualified, women, youths and ethnic minorities) experience the most unfavorable conditions. At the same time, given the high share of labor incomes in total household incomes, precarious labor conditions often give rise to poverty and social exclusion. The “working poor” phenomenon prevailing in these countries, even in the most developed countries of Latin America, clearly indicates that having a job is no guarantee against poverty.
High poverty and social vulnerability are generally related to three structural characteristics in the region: (1) high income
10
inequality, both of labor and family incomes; (2) very low systemic competitiveness and high productive heterogeneity; and (3) scarce protection for workers and their families due to the limited coverage of the contributory-based social security system and the insufficient development of other non-contributory schemes. In turn, labor income inequality is partly a result of productive heterogeneity, with highly efficient and competitive sectors and companies coexisting with activities and firms that operate under a survival rationale. These gaps are reflected in the distribution of labor incomes, either in an amplified or moderate manner. Therefore, the presence of a broad group of informal workers is at least partly associated with a high presence of small, unstructured productive units in which it is often more feasible to conceal labor relations. This is ultimately a result of the limited capacity of the formal sector to create jobs.
It is therefore important to point out that there is a significant percentage of workers (and households) that are not covered by social security benefits, such as unemployment insurance or those benefits related to occupational accidents or diseases. These jobs are not regulated by minimum wage legislation or working hour standards, or by rules that protect the employees against layoffs. Nor are these workers part of the mandatory social security schemes which guarantee certain income levels for the elderly, or access to low-cost health services. Informal workers are not only exposed to higher occupational and income instability and to lower remunerations, they are not protected by labor institutions either. Informality and its consequences affect the less qualified workers the most, which in turn receive lower incomes, thus leading to greater inequality.
LITERATURE REVIEW
There is a vast literature on poverty dynamics in developed countries. These studies focus on several aspects, such as those related to the identification of long duration spells of poverty, poverty traps or the differentiation between chronic and transient poverty. Other studies employ different econometric methods and models to identify the factors behind the process through which a household becomes poor, exits poverty or remains in that
10 11
state for a long period of time. Some of them model both income mobility and poverty dynamics. This is the case, for instance, of the covariance structure model developed by Lillard and Willis (1978), which based on the study of individual income mobility in the United States predicts the incidence of poverty in a given population.4
Another approach is that of Markov models. Cappellari and Jenkins (2004) propose an extension of the first order Markov matrix for the analysis of transitions between low-income situations in the United Kingdom. Of particular interest is the method proposed for obtaining unbiased estimates of transitions between poverty and non-poverty by jointly modeling entry and exit probabilities, initial poverty status and non-random attrition. In addition to modeling poverty transitions between t and t+1, the authors use a trivariate probit model to account for both sources of endogeneity (the initial status in t and panel retention between t and t+1).5
A third kind of model is applied by Bane and Ellwood (1986), who make use of duration models to estimate the conditional probability of transitions into and out of poverty. This was the first research that explicitly incorporated labor market and demographic events to the study of poverty dynamics. Then, other studies followed this line of analysis. For instance, Jenkins and Schluter (2001) compare child poverty in the United Kingdom and Germany incorporating these events. They start by defining a group of events to then decompose the probability of exiting poverty into two factors: the probability of a household to experience a given event of the above mentioned type, and the conditional probability of the household to exit poverty given the occurrence of such event.6
4 Stevens (1999) and Devicienti (2001) also employ this methodology to study poverty in the United States and in the United Kingdom, respectively.
5 Stewart and Swaffield (1999) model transitions into and out of low pay using a bivariate probit model with endogenous selection. Cantó et al. (2007) also apply a bivariate probit model to the study of poverty transitions.
6 Cantó et al. (2007) apply this methodology to the study of child poverty in Spain. Other studies which also focus on the importance of different kinds of events in poverty transitions are Ruggles and Williams (1987), Duncan (1983), Duncan and Rodgers (1988), McKernan and Ratcliffe (2002), Ballantyne et al. (2004).
12
More recent studies consider multiple poverty episodes. Stevens (1999) incorporates this possibility in the study of poverty persistence in the United States; Jenkins and Rigg (2001), Devicienti (2001), and Fertig and Tamm (2007) do the same for the United Kingdom; and Biewen (2004 and 2006) for Germany. Also, Arranz and Cantó (2007) study the influence of all past poverty episodes and their duration, on exit and re-entry rates in Spain. The results show that these models better fit the evidence than those which consider one single poverty episode.
All these studies are based on reduced models that estimate poverty transitions either directly or indirectly. However, other studies employ structural models to analyze poverty dynamics. Some examples of the latter are Burgess and Propper (1998), Burgess et al. (2002), and Aassve et al. (2004 and 2005). In these studies, poverty transitions are modeled as a result of changes in the economic and demographic variables of households. Authors try to model are the behavioral decisions that underlie poverty dynamics, focusing on those related to the labor market (the decision to work together with the estimation of an income equation), and to fertility and marriage, while emphasizing the interrelations between these dimensions. With this purpose, different correlated equations are simultaneously estimated. All these studies highlight the importance of simultaneously modeling the demographic and labor processes that underlie the results in terms of poverty status.7
Each of these approaches has advantages and drawbacks. As Cappellari and Jenkins (2004) mention, studies based on covariance structures have the advantage of allowing the decomposition of changes in household incomes into permanent and transitory components, which is highly valuable for the prediction of structural poverty. Furthermore, directly analyzing the characteristics of income without classifying households into poor and not poor makes the whole set of information useful: it allows knowing whether after a transition a household stays near
7 For example, Aassve et al. (2004) model employment, marriage and fertility decisions and derive the poverty status as a function of those variables. However, they do not model the potential effects of past decisions or past incomes on future incomes, which would allow identifying some direct or indirect status dependence. Biewen (2004) goes one step further and incorporates the effect of poverty status on future employment and household composition.
12 13
the poverty line, or if, on the contrary, the resulting distance is sufficiently large to expect more permanent changes in its status. However, these models assume the same income dynamics for all households, poor and non poor, which is a very unlikely situation. Moreover, these models do not allow explicitly considering the different demographic and income episodes that might be experienced by households.
Hazard models have the advantage of taking into account not only the effects of households’ characteristics (as well as those of their members), but also of the accumulated duration of poverty spells; they also introduce non-linearities by distinguishing between poor and non-poor households. However, these models typically ignore the types of endogeneity mentioned above.
On the other hand, trivariate and bivariate models may face identification problems because of the difficulties in finding adequate instruments.8 Finally, the most important inconvenient of the studies based on structural models might be their complexity: they require a significant quantity of equations and parameters in order to make the estimations, which implies making several assumptions that cannot always be verified.
For less developed countries, although there is a considerable body of literature about the incidence of poverty and its characteristics using cross sectional data, dynamic studies are scarce and quite recent. One of the reasons for this deficit is a lack of longitudinal data that would make it possible to observe the same household over a period of time.
In the case of Argentina, Paz (2005) applies the Cox Proportional Model to panel data based on four observations of the Encuesta Permanente de Hogares (EPH), the Argentine household survey. Three types of variables associated with poverty dynamics are identified: those which reinforce poverty either by stimulating entry into or preventing exit from that state, those affecting just one of these two types of transitions and those with the same direction of effect on both entries and exits. An example of the first of these is the household head’s level of education, higher levels
8 Given the difficulty in obtaining valid instruments for applying bivariate or trivariate models based on the variables measured in the surveys, other methods have been suggested to take into account the attrition bias. These methods are based on the re-weighting of the observations in the sample (Cantó et al. 2007; Beccaria and Maurizio, 2009).
14
of education being associated with a lower probability of entering poverty and a higher probability of exiting poverty. Belonging to a given cohort is an example of the second type of variables: a lower level of poverty exits was observed towards the end of the nineties (implying longer episodes of poverty) while the entry rates were not affected. Finally, the proportion of income-earning household members is negatively related to both the entry and exit equations, an example of the third type of variable.
Cruces and Wodon (2003) follow Jalan and Ravallion’s (2000) methodology to differentiate between chronic and transient poverty in Greater Buenos Aires (Argentina) from 1995-2002. They find that an increase in the incidence of poverty is mostly due to a rise in chronic poverty. They also find that households headed by an employer have a higher incidence of transient poverty, but not of chronic poverty. This could be due to higher income instability for the self-employed. The opposite case is found with public workers, perhaps as a consequence of lower but more stable wages.
Beccaria and Maurizio (2009) analyze the factors associated with poverty transitions in Greater Buenos Aires between 1991 and 2003. They identify and analyze the impact of different events associated with entries into and exits from poverty, paying special attention to the effect that inflation has on real income. They use data from the EPH and apply a methodology that includes a correction for sample attrition bias. They find that events related to the labor market are the most important factors behind entries and exits, both due to their relatively higher frequency and to their major impact on household incomes. Changes in household composition, however, only have a minor impact on transitions.
Applying a similar methodology, Maurizio et al. (2009) study the extent to which poverty reduction in Argentina during the 2003-2006 period has been linked to the performance of the labor market or other factors (for example, demographic). Pursuing that objective, the authors relate the transitions between poverty and non-poverty to multiple events experienced by household members. They find that both the increases in labor incomes and in the quantity of employed members are the events most frequently associated with poverty exits, while changes in household composition are less important. The results also show that households with children and those with female household
14 15
heads have more difficulty exiting poverty than all other household types.
For Brazil, Perez Ribas and Machado (2007) base their analysis of poverty dynamics among Brazilian adults in urban centres on pseudo-panels built from the Pesquisa Nacional por Amostra de Domicilios (PNAD) covering 1995–2003.9 Their main goal is to identify which individual characteristics are positively related to both movements into poverty and permanence in that condition. Then, they estimate the probability of staying in or out of poverty, or of changing from one state to the other using a bivariate probit model that distinguishes between persistent and transient components. They find that the major part (73%) of urban relative poverty10 in Brazil is chronic and is largely determined by initial poverty status, the “path dependence” effect. Non-whites, the least skilled, informal workers and residents of the northeastern region are the most likely to be chronically poor. The authors conclude that the dynamic analysis of poverty can improve anti-poverty policy design by indentifying the most accurate type of public intervention for both chronic and transient poverty. For instance, they propose an extended program of income redistribution to reduce chronic poverty.
Another document of the same authors (2010) analyzes the determinants of poverty exits in urban areas of Brazil, focusing on short run changes in the labor market and considering the duration of poverty spells. In order to do so, they apply survival models using panel data from the Pesquisa Mensal de Emprego for the 2002–2007 period. Given that this survey only collects information about labor incomes, non-labour income is imputed using the methodology proposed by Elbers et al. (2003). The main conclusions of the research are that the probability of exiting poverty falls as the duration of poverty spells increases, that the unemployment rate of household members does not directly affect the length of poverty episodes and that a rise in average wages for informal workers strongly and positively impacts the probability of exiting poverty, since poor workers are mostly employed in informal jobs.
9 They build a “pseudo-panel” of 180 cohorts according to their gender, age, race, education and region. Using six waves of the PNAD, they estimate the joint likelihood of remaining in poverty or falling into poverty.
10 Relative poor are those with incomes below 60% of the median equivalent income.
16
Additionally, the presence of pension beneficiaries in the household increases the probability of moving out of poverty. On the contrary, the size of the household, the presence of children, the low educational level of adults and the presence of a non-white household head reduce the probability of exiting poverty.
A sample of Peruvian and Madagascan urban households (1997-1999) is the basis of Herrera and Roubaud’s (2007) evaluation of poverty transitions and characteristics of transient and chronic poor households relative to non-poor households. They find that flows into and out of poverty in both countries suggest that the incidence of poverty is higher than would normally be expected when using cross-sectional data. Chronically poor households differ from transient poor both in terms of the intensity of their poverty and the incidence of associated negative factors. In particular, household size, the proportion of young children and low human and physical capital are all positively correlated with poverty, and even more so with chronic poverty. Aside from these shared features, there are also important differences between both countries. For instance, in Madagascar, working in the informal sector seems to be an important difference between chronic and transient poverty. In Peru, however, this difference does not appear to be important, probably due to higher precariousness in the formal sector resulting from labour market liberalization in this country. With regard to poverty transitions, the authors find severe limitations in the predictive capacity of the model, while they find a small number of common independent variables that might explain poverty transitions in both countries. In spite of the latter,, they are able to conclude that structural factors are not very significant for chronic poverty, while demographic and economic shocks seem to play an important role. From the point of view of policy recommendation, this analysis suggests that the significant flows between poverty and non-poverty could cause targeting errors in the anti-poverty programs.
Gómez and Román (2001) study the transitions into and out of poverty for a group of households during the 1997–2000 period in Peru, based on data from the Encuesta Nacional de Hogares (ENAHO). Their results also show that demographic and labor market variables are closely associated with poverty entries and exits. In particular, the reduction of the household’s size is positively associated with exits from poverty. On the other hand,
16 17
as it was expected, the higher the participation of household members in the labor force, the higher the probabilities of moving out of poverty.
Herrera (2001) studies changes in poverty in Peru from 1997-1999, when macroeconomic performance was negatively affected by an international financial crisis. The article analyzes poverty transition matrices using panel data observed over three years and uses the mobility index proposed by Fields and Ok (1998). Additionally, the author estimates a multinomial logit model of poverty transitions, taking as explanatory variables the characteristics of households and their members as well as other variables that reflect the provision of public goods. The results show that approximately 25% of poor and non-poor households experienced a status change every year. One interesting finding is that education is an important factor for exiting poverty, but does not prevent households from entering poverty or staying in that situation. Region of residence and availability of public services also appeared as important factors associated with poverty transitions. He finds that geographically targeted public policies such as the provision of public services (mainly health and education) have been properly oriented and benefited the most vulnerable and chronically poor households.
For Chile, Neilson et al. (2008) study poverty dynamics based on the CASEN Panel (1996-2001), distinguishing between chronic and transient poverty. They find that while 20% of the population was living below the official poverty line both in 1996 and 2001, only 9% was poor in both years. While the poverty line rose, the number of chronically poor households also increased steadily, while the transitory component of poverty remained more or less stable. They also point out that labor market events are important factors associated with poverty transitions, whereas demographic changes and other sources of income are not. Health problems suffered by the household head significantly reduce the probability of leaving poverty, although this is not a relevant event for households with higher family incomes. Presence of children in the household negatively affects the probability of leaving poverty and positively affects the risk of becoming poor. Finally, they find that although a higher educational level of the household head reduces the probability of entering into poverty, it does not significantly improve the chances of moving out of this state.
18
Slon and Zúñiga (2004 and 2006) analyze poverty entry and exit rates in Costa Rica using panel data from the Encuesta a Hogares de Propósitos Múltiples (EHPM) for the 2000–2002 period. They study the differences in households´ transition rates considering dimensions such as region, gender and educational level of the household head, the number of wage-earners per household and the households’ size. The results show that a household is more likely to exit poverty when its head has a higher level of education or works outside the agricultural sector. A positive relationship between exit rates and the number of income-generating members is also verified. They also find that transition flows were large even when macroeconomic variables such as the structure of production, inflation, unemployment and even poverty incidence remained relatively constant.
Freije (2000) studies the case of Venezuela based on panel data from the Encuesta de Hogares por Muestreo. His main findings are that educational and employment status for both the household head and other family members, cash transfers and capital income are related to poverty dynamics. He finds that education has an asymmetric effect on poverty: while it is an important factor linked to transitions out of poverty, it does not prevent households from entering into poverty. This result seems to be contrary to that found by Neilson et al. (2008) for the Chilean case, although it is similar to that found by Herrera (2001) for the Peruvian case.
Finally, a special issue of The journal of Development Studies (edited by Baulch and Hoddinot, 2000) compiles a number of studies that focus on poverty dynamics and economic mobility using longitudinal data. The results for Chile, China, Pakistan, South Africa and Zimbabwe show some regularities. A large number of households transit into and out of poverty. Changes in returns to endowments and improved capital quality both seem to be associated with income increases which may lead to an exit from poverty. Some results show that temporary shocks can have important long term consequences.
DATA SOURCES
Data used in this paper come from regular household surveys of the selected countries. In general, they have been continuously
18 19
collected for several years. While they are are mainly focused on labor market variables they also gather information on other social and demographic characteristics of the households.
In order to recognize the factors associated with poverty transitions, the databases must identify the poverty status of each household and individual along with other relevant socio-economic and demographic information measured at different points in time. These last typically include individual age, gender, education, economic activity, income, and the size and composition of households.
As mentioned above, there are almost no longitudinal surveys for Latin American countries that would allow following the households during long periods of time.11 For this research dynamic data have been constructed using the rotating sample scheme of some household surveys. This kind of scheme implies that the total sample is divided into a certain number of household groups, with each group remaining in the sample for a given number of observation periods or waves. Therefore, for each wave of the survey, one of these groups enters the sample while another one leaves. Consequently, it is possible to compare a given proportion of the sample between two or more waves.
This type of data makes it possible to know whether a given household stays in poverty or leaves it over the “n” periods the household remains in the sample. Each household may also be characterized by a number of demographic and socioeconomic attributes.
The Latin American countries under study were selected on the basis of the availability of of data from household surveys that are suitable for studying poverty dynamics with panel data. The selected countries are: Argentina, Brazil, Costa Rica, Ecuador
11 The only exception is the panel of the Encuesta de Caracterización Socioeconómica (CASEN Panel) of Chile. This survey provides information about households for three different years, each observation with a five-year interval in between (the survey was applied in 1996, 2001 and 2006). Hence, although it is a very valuable source of information for medium and long term occupational and welfare changes, it is not completely adequate for the analysis presented in this article, since most of the events considered affect the household poverty status in the short term. The results obtained from this source would therefore not be comparable to those obtained for the other countries under study. For this reason, Chile has been excluded from this research.
20
and Peru. Each survey is carried out by the respective national institutes of statistics.and collect information on households based on a probabilistic two-stage sample that is divided into different panels that enter and leave the sample in different moments of time. Their design and implementation follows the conceptual framework adopted by the International Labor Organization (ILO) and the Economic Commission for Latin America and the Caribbean (ECLAC). In all the cases it was possible to match two observations of the same household in two successive years and obtain information of a wide range of variables for each household member.
For Argentina, the data source is the Encuesta Permanente de Hogares (EPH) carried out by the Instituto Nacional de Estadística y Censos (INDEC). Micro-data are available for Greater Buenos Aires (which includes Argentina’s capital city and its surroundings) for the period that extends from 1974 to 2006, and for all major urban areas as from 1995 onwards. Also, as from 2003, after a major methodological change, the survey provides quarterly data. Households are interviewed in two successive quarters, stay out of the sample in the two following quarters and are interviewed again for two more quarters. Therefore, the transitions that can be analyzed are those that occur between two yearly observations (in the same quarter of two successive years) or between two successive quarters. The theoretic sample overlapping between one quarter and the same quarter of the following year, as well as between two successive quarters, is of 50%. For this study, we have constructed panels of households that were interviewed during the same semester of two successive years within the the 2003-2006 period.
For Brazil we employed micro-data coming from two different surveys. The Pesquisa Mensal de Emprego (PME), carried out by the Instituto Brasileiro de Geografia e Estadistica (IBGE), covers six major urban areas and provides monthly information. In this survey, 25% of the sample is replaced every month. Households are observed during four consecutive months, stay out of the sample for eight months and are interviewed again for another four months, allowing the construction of monthly and yearly panel data. Until 2003, each of these groups represented 25% of the total sampled households, and from that year on the rotation groups represent 1/8 of the sample, two of which are interviewed
20 21
each month. This methodology implies a theoretic overlapping of 75% of the sample in two successive years and of 50% every 12 months, thus avoiding “blind periods” (periods with no sample overlapping).
Given that this survey only collects information on labor incomes, we need to impute non-labor incomes in order to estimate total family incomes and the resulting poverty status of the household. For this, we apply the methodology of Machado and Perez Rivas (2010)12 using micro-data of the Pesquisa Nacional por Amostra de Domicilios (PNAD), also carried out by IBGE. This survey collects yearly information on different socio-economic aspects of the population and has national representativeness. However, since it is not possible to construct panel data from this survey, it can not be used in dynamic studies. Hence, based on variables that are present in both surveys, we assign non-labor incomes informed by PNAD to households surveyed by the PME. The estimation of non-labor incomes (pensions, capital incomes and transfers) is made through the estimation of more than 100 regressions using covariates that are common to both surveys, selected through the stepwise method. We were able to apply this methodology because both surveys use similar questionnaires and have equivalent sample representativeness of the same urban areas. After imputing non-labor incomes, we performed tests to check for the consistency of estimated incomes and poverty incidence, obtaining satisfactory results. The PME is conducted since 1982, as part of the PNAD (which is conducted since 1976), and in 2001 it went through major methodological changes that allowed the construction of comparable panels since 2003. The panels have been constructed by matching yearly micro-data corresponding to the month of September of the 2003-2006 period.
With regard to Costa Rica, the Encuesta de Hogares de Propósitos Múltiples (EHPM) that is carried out by the Instituto Nacional de Estadística y Censos (INEC) interviews households that are representative of urban and rural areas of the whole country every July. The scheme of the sample implies a rotation of 25% of the sample every year so that 75% of the sample is common to two consecutive years. The survey began in 1987, yet based on the the information available, we have constructed two panels for the 2006-2008 period.
12 Adapted from Elbers, Lanjouw and Lanjouw (2003).
22
The Encuesta Nacional de Empleo, Desempleo y Subempleo (ENEMDU) of Ecuador, carried out by the Instituto Nacional de Estadística y Censos (INEC) has a rotating sample scheme that allows building quarterly and yearly panels by interviewing each household up to four times. Its results are representative of urban and rural areas. It collects information in March, June, September and December every year in 23 provinces of the country. Households are interviewed in two consecutive quarters, stay out of the sample for two quarters and are interviewed again for two additional quarters. The sample is divided into four groups that represent 25% of the sample, ensuring an overlapping of 50% during the same month of two consecutive years. Contrary to the procedures followed by other surveys, the ENEMDU replaces households that do not respond to the survey and thus it has a very low non-response rate (of around 3% of the initial sample). For this research, panels have been constructed for the 2004-2008 period. Information from this survey is available since 2004.
Finally, for Peru there is a panel survey attached to the regular household survey, the Encuesta Nacional de Hogares sobre Condiciones de Vida y Pobreza (ENAHO) Panel, of the Instituto Nacional de Estadística e Informática (INEI). The ENAHO is conducted in urban and rural areas at the national level, and panel data are available for the 1998-2002 and 2002-2006 periods. Until 2002, the survey was conducted every year during the fourth quarter, and since 2003 it has been conducted monthly, producing monthly, quarterly and yearly estimations. Contrary to the other surveys employed in this research, this survey not only provides cross section data but also panel micro-data. Its sample has no rotating scheme, yet the same households are interviewed once a year for four consecutive years.
As it was mentioned, in order to obtain a comparable data-set for each country, transitions are defined for a one year interval between observations. The analysis is made for the years for which comparable panel data can be constructed within the period of economic growth and poverty reduction in Latin America: 2003-2006 for Argentina, 2003-2006 for Brazil, 2006-2008 for Costa Rica, 2004-2008 for Ecuador and 2002-2006 for Peru. Since not all the surveys are nationally representative, and given that poverty and labour markets may behave differently between rural areas
22 23
and urban centres, the analysis in this study will be restricted to urban areas.
With regard to the quality of the data, panel information allows checking for the consistency of many of the variables informed. In particular, in all the cases we have discarded households for which the information of at least one member presented errors regarding age and gender. Tables 1.A-1.E included in Annex A present the initial number of households and individuals sampled for each country, as well as those with valid family income information and valid data regarding the occupational category and the number of working hours for the employed household members. The same information is shown for panel data, as well as the number of households and individuals that belong to the panel.
In the case of Argentina, after controlling for valid incomes and valid information on the occupational category and the number of working hours for employed individuals, the final sample utilized in the estimations is comprised of 109,486 individuals in 32,309 households. For Brazil, after applying the consistency controls, the final sample is comprised of 53,295 individuals in 18,483 households for the 2003-2006 period. For Costa Rica, the final sample is of 17,932 individuals in 5,082 households; for Ecuador of 120,686 individuals in 32,221 households; and for Peru it is comprised of 64,516 individuals in 14,127 households. In all the cases, panel data have been pooled into one single data base in order to accumulate a larger sample of cases.
One potential problem of the analyses that compare information from different sources is that household incomes might be informed for reference periods of different lengths (i.e. monthly, quarterly or yearly incomes), which might in turn affect the measurement of poverty. As an example, households whose incomes are under the poverty line for some quarters might appear as non-poor if their yearly incomes are higher than the average poverty line. This problem also arises in static analyses, but some authors have studied the way in which results might be affected. Böheim and Jenkins (2000), for example, find no significant differences in the case of the UK, and explain this as a result of the tendency of people to answer about their usual income even if they are asked about their income in a given month. They found that the differences are more important within households that undergo demographic changes (for this group of households, they
24
find the opposite result than for the population as a whole: their annual poverty is higher than the poverty measured using their current income).
For the case of the United States, Ruggles and Williams (1987 and 1989), and Ruggles (1990), find greater differences between poverty measured in terms of annual or current incomes. They find that an important proportion of poverty transitions identified with monthly data become invisible when annual data is employed. They also find that the number of people that are poor at least once a year is four times higher than the people who are poor along the whole year.
For the case of Spain, Cantó et al. (2006) find that although poverty rates are similar using both quarterly or annual incomes, these differ significantly from other poverty measures that are more sensitive to the situation of households at the extremes of the distribution. They also find differences in the characteristics of the households that are identified as poor by each method. In adittion, the authors compare the results on income mobility derived from quarterly information with those derived from annual data (calculated as the sum of quarterly incomes). They find that income mobility is higher for current incomes (quarterly incomes) than for annual incomes, and that the bias introduced by the change in the definition of incomes differs for households with different characteristics and in different income deciles. In the particular case of poverty mobility, the differences are higher for entries than for exits.
In the present study, we use monthly incomes of a recent reference period to determine the poverty status. In this sense, the information is comparable, although some results might differ if they are drawn from incomes measured on an annual base.13
Another difficulty arises from the fact that not every transition can be captured when matching two observations. Given that a transition is identified by comparing two observations, two or more symmetrical changes in the poverty status (or in other variables) might take place between those two moments without being captured by the observed data.
13 Peru is the only case in which non-labor incomes are collected with a reference period of 6 to 12 months. These incomes have been converted into a monthly basis using information on the frequency with which households perceive incomes.
24 25
Finally, another limitation of panel information is that the proportion of households that are actually interviewed in two successive moments can be smaller than the proportion of households and individuals that should be re-interviewed according to the sample rotation scheme. This is a result of a loss of cases (“attrition”) caused by different motives (for example, because of people deciding to abandon the panel or due to difficulties in the data collection process). If this loss in not random, it could introduce biases in the sample. Different attempts have been made in order to apply a methodology that would account for the type of bias associated with this phenomenon. However, the lack of information needed to be able to differentiate the loss of data resulting from the rotating scheme, from the loss of data caused by sample attrition, prevented the application of an attrition bias correction strategy in all countries.14 Although it is not enough proof to discard the potential impact of attrition on the results, the descriptive statistics drawn from panel and cross-section data shown below are very similar.
APPROACH AND METHODOLOGY
The absolute poverty line approach to poverty identification
The absolute criterion to identifying poverty used in this study seems more appropriate than a relative criterion for Latin America as there is enough evidence that a substantial proportion of people in the region still lack resources to satisfy basic needs. The “income approach” will thus be employed, i.e. households are identified as poor if their total income is below some poverty line. This line is the value of a normative basket of goods and services that allows the satisfaction of basic needs. In order to define a normative basket, an absolute threshold for each basic need must be established first. Then, the type and quantities of the goods and services needed to meet each of those standards should
14 Beccaria and Maurizio (2009) estimated the effect of sample attrition on poverty dynamics in Argentina in the nineties. They found that it was not a very relevant effect, and thus the results obtained were not altered when this phenomenon was taken into account.
26
be specified and valuated to allow comparison with household income.
The construction of an absolute poverty line faces many difficulties and different methodologies can be applied for empirical measurement.15 The normative budgets used for each country are the same as those used by ECLAC to regularly estimate the incidence of poverty for Latin American countries.16 They were built for a given base year generally following the approach proposed by Orshansky for the official norm for the US.17 To start with, the food component of the line (or Basic Food Basket - BFB) is calculated and the value of the overall poverty line is estimated by multiplying the BFB by the inverse of the Engel’s coefficient. This simplified methodology, also employed by many countries to build their national lines, was proposed as a response to the lack of adequate information and, especially, of generally agreed norms for those basic needs not related to nutrition.
In order to calculate the normative food basket, ECLAC first estimates the caloric requirements considering the FAO/WHO/UNU recommendations for persons of different gender, age and level of physical activity. Then, the average height and weight values of the population for different age and gender strata are employed (resorting to national tables) to obtain the required kilocalories for each person and for the whole household, according to differentactivity levels. Those caloric requirements are then converted into goods and services by considering the expenditure structure of a “reference group”. The latter is defined as a group of households (usually equivalent to 20% of the population) that satisfies, on average, the caloric requirements with the lowest incomes. However, the BFB does not strictly reflect the actual composition of the reference group due to at least three reasons: (i) the consideration of other nutritional criteria (e.g. proteins or calcium) besides the caloric one. As a consequence, the initial basket which was built exclusively considering the caloric requirements is adjusted by reducing or increasing the volume
15 Much literature on poverty measurement methods also pinpoint different theoretical and the empirical difficulties. See, for example, Feres (1997), Ravallion (1994), and Rio Group (2006).
16 Estimates are usually disseminated through Social Panorama, an annual publication of the institution.
17 See Orshansky (1965).
26 27
of certain goods; (ii) the introduction of certain normative criteria that tend to reduce the weight of certain goods with a high price per calorie –or that directly excludes them; (iii) the reduction –for practical reason– in the number of products to a selected group comprised of the most representative goods of each category (i.e. cereals, vegetables, meats and so on). Finally, the selected basket is valued with the corresponding prices of the official consumer price index.
The poverty line derived from the BFB by ECLAC considers the same Engel coefficient for every countries (0.5). As it will be indicated below, this decision to use the same factor for every country was based on the need to have comparable poverty lines. These values were the average of the countries that were accounted for when the original ECLAC estimates were made during the eighties.
As it results from the previous description, the basic food basket contemplates equivalence scales between persons taking into account the differences in the caloric requirements. No consideration is made for economies of scale. Since the poverty line is computed as two times the value of the food basket, its implicit equivalence scales are those of the food component. This means that the same food scale equivalences are also used in the non-food portion of the line. Furthermore-, the overall line does not contemplate economies of scale either. These features of the ECLAC methodology are frequently criticized.
Precisely, the income approach usually faces the challenge of attaining internationally comparable estimates of poverty incidence (or other characteristics of poverty) due to difficulties in agreeing on comparable baskets or on their aggregate values. We are referring here to basic conceptual issues, for international comparability is also hindered by many empirical limitations that affect not only the construction of the poverty line, but also of the aggregate value of of households’ resources.
Even if some of the countries selected regularly produce poverty incidence estimates based on the income approach, using their own national normative baskets, we have employed ECLAC’s lines as a way to improve international comparability for they were produced resorting to the same criteria and methodology. However, even if the national lines were calculated with the same approach than that followed by ECLAC, they
28
differ in several parameters. For example, they differ in the consideration of normative criteria (e.g. some exclude meals taken outside the household, while others not). Perhaps, the most important departure is the use of each country’s actual Engel coefficient of the reference population. As it was indicated above, ECLAC uses the same methodology to build comparable normative baskets for several Latin American countries in order to compute comparable estimates of poverty incidence. An alternative approach is that employed by the World Bank, which defines a unique value in dollars of the same purchasing power to then transform it to national currencies (i.e. to national poverty lines) using purchasing power parities.
In order to compare the poverty situation in different countries we consider that building lines for each country using ECLAC’s methodology is more consistent with the theoretical underpinnings of the absolute poverty line approach. At least in theory, the values of ECLAC’s national poverty lines reflect the amount of local currency needed to buy a basket of goods and services that makes it possible to satisfy the same set of basic needs in each country, while still taking into account the specific consumption pattern of each country. In terms of Sen’s conceptualization, the different poverty lines are similar in the space of capabilities but different –i.e. country specific– in the space of goods and services.18
This is clearly the case for the Basic Food Basket component, since the same nutritional standards or thresholds are considered in each country (similarity in the space of capabilities) while the actual consumption structure of the reference population in each country is taken into account to define the list of specific goods that make up the basket. As it was mentioned above, given the existing theoretical and empirical limitations, ECLAC –as many others– resorts to the use of one single Engel coefficient for all countries. This decision, although questioned, was precisely taken as a way to reach comparable levels of the aggregate non-food components of the poverty line.
In effect, such agreement on the thresholds for the space of capabilities related to non-nutritional needs appears as very difficult to reach, like in the case of leisure. But the difficulties to reach standards –as those employed in food– also appear in
18 For a discussion on this topic, see Sen (1983 and 1985).
28 29
other goods and services that seem to be more amenable to the approach. For example, there is no clear answer for how to move from the “standard” mentioned by Sen regarding clothing: “those needed in order to not be ashamed” to a specific list of goods.
However, there was one poverty line for a Latin American country –Mexico– that was built from specific normative thresholds for all components, not only for food. And a similar attempt is being made in Canada.19 However, the normative criteria considered in these cases were ad-hoc and there is little international consensus regarding their validity.
Notwithstanding these limitations in poverty comparability derived from the specific empirical shortcuts used in the ECLAC method, the latter still appears –at least conceptually– as more adequate than the alternative of using the World Bank lines. Regarding the latter, and leaving aside any consideration on their levels (of one or two dollars a day), the reasonability of employing poverty lines that reflect baskets of goods and services of the same value that do not necessarily satisfy similar levels of needs in the various countries is not clear (e.g. in the case of the BFB, they do not necessarily imply the same level of nutrients).
Another methodological question that usually arises relates to what is the most appropriate variable to measure households’ resources, incomes or expenditures. Usually, both measurement and conceptual arguments arise in the discussion. Among the former, it is sometimes argued that expenditures appear to be better measured than incomes. From a more theoretical perspective, it is also suggested that expenditures should be used when a comparison is made with a poverty line. However, income is an adequate indicator for it reflects the capacity of a household to buy the normative basket and in that respect it should be preferred to expenditures. However, within the context of this study, the latter discussion is of an abstract nature since expenditures are not captured on a regular basis by the household surveys of the selected countries, with the exception of Peru.
MethodologyThe current study aims at estimating poverty entry and exit
rates associated with different events. Given that there are no
19 The Market Basket Measure.
30
longitudinal surveys available for the region, it is not possible to use duration models, since it is not possible to know for how long households stay in poverty or out of poverty. For these reasons, in view of the objectives of this project and the information available, the main methodological approach will be based on Jenkins and Schluter (2001), which allows a decomposition of poverty exit and entry rates associated with different kinds of events in a cross-national comparative way, as it is presented in detail in the following section.
Two points should be stressed here. As it was mentioned above, part of the literature that analyzes poverty dynamics is based on a structural model that relates different economic and demographic decisions (Burguess and Propper, 1998; Aassve et al., 2005 among others). In this research, following Bane and Ellwood’s approach, we only consider the observed episodes directly associated with poverty entries and exits,while no attempt is made to analyze the family arrangements and/or strategies that could have led to such episodes (about which no information can be drawn from the household surveys of the selected countries).20 Also, there is the possibility that some of the identified events could have ultimately been the result of some other event associated with the observed transition.21
Consequently, as events may be endogenous, they are not interpreted as the causes of transitions –exogenous events- but only as events associated with transitions. However, since a household becomes poor when its income per adult equivalent (ipae)22 falls under the poverty line per adult equivalent, either the numerator or the denominator must change for a household to enter or exit poverty. This would happen when the household experiences at least one of the episodes here identified. Precisely, analysis of the significance of those episodes is one of the objectives of this study,
20 For example, we will not take into account the effect of certain facilities that influence the decision of entering the labor market, such as child care facilities. Similarly, no considerations will be made regarding certain previous household decisions that may have made it possible for one of its members to begin to work. Only the event of a member becoming employed is identified.
21 For example, an event leading to an increase in the ipae could give rise to another episode that in turn further raises the ipae. In our analysis, both factors will be considered to be happening simultaneously.
22 Total household income/ number of equivalent adults in the household.
30 31
while no attempt is made to explore other possible factors causing such events. The short window of observation, even when a household is followed along the entire period for which it remains in the survey, is a major limitation for attempts to estimate a more structural model.
On the other hand, it should be mentioned that only the episodes associated with poverty entries and exits will be considered, but not those that prevent a transition. For example, if a member of a household that exited (entered) poverty also experienced an event that tended to reduce (increase) its ipae, such event will not be considered, as the household was able to move out of (fall into) poverty in spite of it.
However, there is some difficulty in identifying the events related to poverty transitions, because individuals can experiment multiple events simultaneously. Hence, some methodological definitions need to be made. The first refers to the mutually exclusive character of the events considered or if, on the contrary, different events are allowed to occur simultaneously. One clear difference between both approaches is that the former (proposed by Bane and Ellwood, 1986) allows decomposing poverty entry and exit rates as the sum of entries and exits related to all the identified events, while this is not possible with the second approach (used by Jenkins and Schluter, 2001, among others).23
The second methodological decision has to do with the classification of episodes. Bane and Ellwood (1986), for example, consider that a demographic event has occurred when an episode has a larger effect on the household’s needs (the number of adult equivalents) than on total household income. Any other event is classified as affecting household incomes. Jenkins and Schluter (2001) only identify a set of the most important factors, grouped according to their nature and regardless of the impact they have on incomes.
In this paper, we employ a third approach that results from the combination of the two approaches mentioned above, similar to that used in Beccaria and Maurizio (2009). Specifically, as it was mentioned, an exhaustive list of mutually exclusive events classified according to their nature is defined. However, we still need to consider categories that combine two or more events in
23 This approach is also used in Antolín et al. (1999) and Cantó et al. (2007).
32
order to cover all (i.e. 100% of) possible cases. The difference from Jenkins and Schluter (2001) is therefore that we explicity consider multiple events experienced between observations as a joint occurrence of several single events.
In order to illustrate the classification of events, we can consider the situation of a household leaving poverty. Such transition occurs if its total nominal income rises, if the household size decreases, or as a result of a combination of both episodes that leads to an increase in the ipae. These changes are the consequence of different events experienced by household members. The rise of a household’s total nominal income can be the result of one member getting a job or facing a wage increase; while, for example, the death of one member leads to a smaller household size. Therefore, we first distinguish between the latter type of events –of a demographic character– and the other types. Among non-demographic events, we consider those exclusively related to labor market events (e.g. changes in the number of employed members, changes in the number of working hours, changes in hourly earnings) or to non-labor income events (e.g. changes in incomes from pensions or in transfers, specially those related to social policies). We also take into account those episodes affecting labor and non-labor incomes simultaneously. However, some events lead to an exit from poverty by affecting both the nominal income and the size of the household –e.g. the arrival to the household of an employed person that raises the nominal ipae. Hence, this type of events are considered as demographic events resulting in labor or non-labor income changes. The procedure is similar for entries to poverty.
According to this definition, the events that could trigger exits (entries) from (into) poverty are the following:
I. Exclusively labor income events
1. Growth (reduction) in the number of employed persons not linked to an entry (exit) of labor income earners to (from) the household, maintaining the total number of household members.1.1. Growth (reduction) in the number of members who are
registered wage earners.1.2. Growth (reduction) in the number of members who are
non-registered wage earners.
32 33
1.3. Growth (reduction) in the number of members who are non-wage earners.
2. Growth (reduction) in total hourly wage of members employed in both observations, maintaining the total number of household members and worked hours.
3. Growth (reduction) in the number of working hours of members employed in both observations, maintaining the total number of household members and hourly wages.
4. Growth (reduction) in the number of working hours and in the total hourly wage of members employed in both observations, maintaining the total number of household members.
5. Growth (reduction) in the total monthly wage of members employed in both observations and in the number of employed members, not linked to an entry (exit) of labor income earners to (from) the household, maintaining the total number of household members.
II. Exclusively non-labor income events
6. Growth (reduction) in incomes from pensions not linked to the entry (exit) of pension recipients to (from) the household, maintaining the total number of household members.
7. Growth (reduction) in public monetary transfers (social policy) not linked to the entry (exit) of recipients to (from) the household, maintaining the total number of household members.
8. Growth (reduction) in other non-labor incomes not linked to the entry (exit) of non-labor income earners to (from) the household, maintaining the total number of household members.
III. Labor and non-labor income events
9. Growth (reduction) in labor and non-labor incomes not linked to an entry (exit) of labor or non-labor income earners to (from) the household, maintaining the total number of household members.
34
IV. Exclusively demographic events
10. Reduction (growth) in the total number of household members; the total nominal income remains constant.
V. Demographic events leading to income changes
11. Growth (reduction) in the number of labor or non-labor income earners due to the fact that some members enter (exit) the household.
VI. Combination of demographic and income events
12. Growth (reduction) in total nominal income (irrespective of the source of income change) and reduction (growth) in the number of household members.
VII. Events not classified
I, II and III are non-demographic events because the number of household members remains constant and there are no entries (exits) of labour or non-labour income earners to (from) the household. The remaining event types are either exclusively demographic, a combination of demographic and non-demographic factors or demographic factors leading to income change.
From the construction of mutually exclusive events it is possible to estimate the distribution of poverty transitions associated with particular events. For this, entry (S1) and exit (S2) rates are defined as the probabilities of moving from state i/j in period “t” to state j/i in “t+1”, the states being “poor” and “non–poor”. In order to quantify the impact of different events on the probability of transitions, we use the partition and additivity properties of the sample space of mutually exclusive events. Assuming that the space is partitioned in R mutually excluding events, the probability of moving from the state “i” to the state “j”, (Sij) is equal to the sum of probabilities of transition associated with each event that comprises the sample space. That is to say,
P S P S Eijr
R
j r( ) = ( )=∑
1
, [1]
34 35
where: Sij indicates the transition from state “i” in “t” to state “j”
in “t + 1”. Er indicates the occurrence of event “r”.r: 1,2,…,R i ≠ j
Following Jenkins and Schluter (2001), such distribution
can be decomposed into two factors: on the one hand, the probability of the population at risk -e.g. non-poor households when analyzing transitions to poverty– of experiencing such an event. The second factor consists of the conditional probability of the event triggering poverty entries or exits, given that the event has occurred. Therefore, this probability can be also formulated as follows:
P S P S E P Eijr
R
j r r( ) = ( ) ( )=∑
1
\ [2]
This decomposition of the probabilities of transitions allows assessing whether the importance of a given event results from its its high probability of occurrence or from its strong impact on the household’s income.
THE EVOLUTION OF POVERTY INCIDENCE IN FIVE LATIN AMERICAN COUNTRIES
AND ITS CHARACTERISTICS
Before going into the analysis of the factors associated with the dynamics of poverty, and as a background to such discussion, the recent evolution of poverty incidence in each of the five countries under study is briefly analyzed in this section using cross-sectional data.
Stochastic Dominance As it can be seen in Graph 6, during the present decade all
five countries show a reduction of poverty incidence, as it is also the case for the average of Latin American countries, as discussed above. However, this reduction was of different intensity in each
36
country. In one extreme is Argentina, where poverty fell 20 p.p., whereas in Costa Rica it decreased only 2 p.p. Due to this positive evolution of poverty along the first decade of the present century, the poverty rates for each country of the most recent years are lower than those of the beginning of the nineties. Also, contrary to the relatively homogeneous poverty performance observed during the recent years, the evolution of poverty had been very dissimilar between the countries during the nineties decade. For example, Costa Rica was already experiencing a slow reduction in poverty, during the past decade, while in Peru it was increasing. In the other countries the behavior was more erratic. Argentina stands out due to the sharp increase in poverty experienced in 2002, which then fell during the subsequent years of economic recovery.
GRAPH 6Evolution of Poverty Incidence Five Latin American countries
(selected years)
Poverty rate (households)
16.2
23.6
41.2
55.8
13.1
20.2
28.6
40.5
49.8
29.9
42.3
58
34.9
18.6
46.8
42.6
14.7 14.8
3133.1
16.3 18.2
30.7
19.9
0
10
20
30
40
50
60
70
Argentina (GBA) Costa Rica Brazil Peru Ecuador
Beginning of the nineties
Mid nineties
End of nineties
Beginning of present decade
Recent years
Low poverty poverty
Middle - low poverty Middle - high poverty
Evolution of Poverty Incidence Latin American Countries under study(selected years)
Source: ECLAC
Source: ECLAC
Although the changes in the poverty rate in most of the selected countries have been very intense, it is necessary to verify if they are statistically significant, for which we have performed a stochastic dominance exercise.24 As mentioned by Araar (2007),
24 For this section we have employed the Distributive Analysis Stata Package (Araar, 2007).
36 37
this method is useful to conduct a robust ordinal classification of the distributions according to their poverty level.
Hence, for each country we compare the distributions of total household incomes in the initial and final years of the period under study. Incomes of the final year are adjusted by the initial/final poverty line ratio. Graph 7 shows the difference between the FGT25 curves (and the confidence interval) for the initial and final years where the parameter α is equal to zero.
GRAPH 7Difference between FGT Curves Five Latin American countries
25 Foster, Greer and Thorbecke (1984)
-.2-.1
5-.1
-.05
0
100 160 220 280 340 400Poverty line (z)
Confidence interval (95 %) Estimated difference
(alpha = 0)Difference between FGT curves: (2006)-(2003) and the statistical robustness. Argentina
Note: The initial value of the poverty line per adult equivalent is expressed in local currency: AR$265.5 for Argentina, Rs 224.2 for Brazil, C 60583 for Costa Rica, $ 89 for Ecuador and PEN 283 for Peru.
Source: Authors’ elaboration based on national household surveys.
38
As it can be seen in these graphs, the results obtained show that it is possible to achieve robust conclusions about the temporal variations of poverty incidence. Specifically, in all the countries a reduction of poverty rates is verified during the period under analysis. Such behavior is –as it was mentioned above- associated with the relatively positive macroeconomic and labor market performance exhibited by these countries during the 2000s. Based on these results, in the following section we investigate whether these changes were exclusively related to the growth of average incomes derived from the positive macroeconomic performance, or if they were also a consequence of distributive improvements.
The Shapley Decomposition of changes in poverty26
Decomposition exercises aim at identifying the relative significance of the variations in the mean and in the distribution of household incomes, on the evolution of poverty. In particular, changes in poverty levels are decomposed into two effects: on the one hand, the changes occurred as a consequence of variations in the average of households’ total real income maintaining the distribution of income constant –“real growth effect”-; and, on the other hand, the changes occurred as a consequence of variations in the distribution of income, maintaining the average real income constant –“distribution effect”. The real growth effect can in turn be decomposed into an “inflation effect” and a “nominal income effect”. The inflation effect shows what would be the change in poverty incidence if nominal incomes and their distribution remained constant, while the nominal income effect quantifies the impact of income changes under the assumption of constant distribution and prices. Alternative methodologies such as those proposed by Mahmoudi (1998), Datt and Ravallion (1992) and Shapley (see Araar, 2003; Kakwani, 1997 or Shorrocks, 1999) were also explored.
However, we only present the Shapley Decomposition because it is a more satisfactory method to identify the different contributions of the explanatory factors. This procedure consists of first estimating the marginal effect of each factor as they are eliminated in succession, and then averaging these marginal effects over all the possible elimination sequences (Shorrocks, 1999).
26 The poverty change decomposition presented in this section was made using a Stata do file developed by Abdelkrim Araar.
38 39
The Shapley Decomposition has three advantages in comparison to the alternative decomposition methods: (1) symmetry in the treatment of the different factors into which the poverty variation is decomposed; (2) exact decomposition (no unexplained residue is left); and (3) the contribution of each factor can be easily interpreted as its expected marginal effect.
Graph 8 and Tables 2.A-2.E included in Annex B show the results of applying the Shapley decomposition to the changes of poverty incidence between the initial and final years. The reduction of poverty occurs, to a large extent, as a result of the economic growth experienced by these countries, whereas the effect of the improvement in income distribution is clearly less significant. This is not an unexpected result, given that usually the distribution changes take place at a slower pace, while the period under analysis is relatively short. Moreover, as it can be observed in the graph, in all cases inflation has had a significant negative impact, considerably reducing the possibilities of a further decline in poverty incidence.
GRAPH 8Relative contribution of growth and redistribution to poverty incidence reduction
Five Latin American countries
-10%
10%
30%
50%
70%
90%
110%
Argentina Brazil Costa Rica Ecuador Peru
Growth effect Distribution effect
Source: Authors’ elaboration based on national household surveys.
40
Beyond this general outlook, it is possible to distinguish three different situations: (1) Costa Rica, where the fall in poverty was exclusively associated with economic growth, given that income distribution remained almost stable; (2) Argentina, Ecuador and Peru, where the real growth component explains a larger proportion (about 60%-70%) of total poverty reduction yet redistribution is also relatively important; (3) Brazil, where distributional changes had a larger impact on poverty reduction when compared to the other countries.
In any case, the region continues to show a very unequal income distribution which seems difficult to improve. In view of this, a conclusion drawn from these results is that the persistent high income inequality determines the need to grow at very high rates so as to reduce the incidence of poverty, at least to some extent.
Poverty Status and Household Characteristics The incidence of poverty analyzed above is not homogeneous
across households. Rather,it has been more intense among families with certain characteristics. Household composition and the labor situation of the household head are two very relevant variables for the differing incidence of poverty among types of households (Tables 3.A – 3.E in the Annex B).
A general feature that is observed in the five countries is that households with children are overrepresented among the poor, and this occurs notwithstanding the differences in the composition of households observed for the population between countries. Households headed by a woman and belonging to this group of households appear more likely to be poor than the population as a whole. For instance, in Argentina (Table 3.A), it is observed that while in the second half of 2006 households with children under 18 years old represent approximately 47% of total households (column 7), such percentage is 75% within poor households (column 9). For Brazil (Table 3.B) these figures are 49% and 72%, respectively; and in Costa Rica (Table 3.C), 57% of the households have children younger than 18 years old among their members (column 7), while this figure increases to 66% within poor households in 2008 (column 9). In the case of Ecuador, in 2008 these proportions are 63% and 82%, while in Peru, they are 67% and 84%, respectively.
40 41
In Argentina, the specific poverty rate for households with children in 2006 is 32%, while it is 10% for the rest of the households (column 8). In Costa Rica, the specific poverty rate for this kind of households is 25%, 7 p.p. higher than that of households without children (column 8). In Brazil (Table 3.B), these percentages are 42% and 16%, respectively in 2006. These gaps also appear in the cases of Ecuador (Table 3.D) and Peru (Table 3.E). However, this average is a result of even more heterogeneous situations: within the households with children, the incidence is higher for those with only one of the parents present (incomplete nuclear households) and female household heads. For instance, in Ecuador while poverty incidence in complete nuclear households was 39% in 2008, it was 43% in the case of incomplete nuclear households with female household head. This result is verified in all the five countries (Graph 9 below).
Within the households without children, poverty incidence differs between countries according to the size of the households, the age of their heads and whether they are complete nuclear households or not (Graph 9). For example, in Argentina, one-person households experienced the lowest poverty rates in the period considered, especially if their member is a woman and also if they are older than 65 years old.27 In Costa Rica, poverty incidence in 2008 was the lowest among one-person and complete nuclear households with male household heads younger than 65 years old. Even in the absence of children, households with female heads in one-person households present the highest incidence (around 31%), followed by one-person households with male household heads over 65 years old (26%). A similar panorama to that of Costa Rica is also verified in Ecuador. In Brazil, the differences within households without children are smaller than in the other countries.
27 A more detailed analysis of this group shows that in the case of the elders, the relatively small incidence is basically explained by the low value of the poverty line for their age given that, on average, their incomes are significantly lower than those of the other groups of households.
42
GRAPH 9Specific poverty rates for different household types
Five Latin American countriesAverage ca. 2003-ca. 2008.
Specific poverty rates by type of household. Argentina. 2003-2006
15%
6%
17%
9%
17% 18%21% 21%
48%
55%
43%
0%
10%
20%
30%
40%
50%
60%
One person.Male older
than 65 years.
One person.Female older
than 65 years.
One person.Male youngerthan 65 years.
One person.Female
younger than65 years.
Completenuclear
household.Head older
than 65 years.
Completenuclear
household.Head youngerthan 65 years.
Incompletenuclear
household.Head older
than 65 years.
Incompletenuclear
household.Head youngerthan 65 years.
Incompletenuclear
household.Male
householdhead.
Incompletenuclear
household.Female
householdhead.
Completenuclear
household
Households with children
Specific poverty rates by type of household. Brazil. 2003-2006.
13%17%
19% 18%14%
17% 18% 20%
47%51%
44%
0%
10%
20%
30%
40%
50%
60%
One person. Maleolder than 65
years.
One person.Female older than
65 years.
One person. Maleyounger than 65
years.
One person.Female youngerthan 65 years.
Complete nuclearhousehold. Head
older than 65years.
Complete nuclearhousehold. Headyounger than 65
years.
Incompletenuclear
household. Headolder than 65
years.
Incompletenuclear
household. Headyounger than 65
years.
Incompletenuclear
household. Malehousehold head.
Incompletenuclear
household.Female
household head.
Complete nuclearhousehold
Households with children
Specific poverty rates by type of household. Costa Rica. 2006-2008.
33% 34%
13%
17%
28%
12%
27%
18%
31%
37%
27%
0%
5%
10%
15%
20%
25%
30%
35%
40%
One person. Maleolder than 65
years.
One person.Female older than
65 years.
One person. Maleyounger than 65
years.
One person.Female youngerthan 65 years.
Complete nuclearhousehold. Head
older than 65years.
Complete nuclearhousehold. Headyounger than 65
years.
Incompletenuclear
household. Headolder than 65
years.
Incompletenuclear
household. Headyounger than 65
years.
Incompletenuclear
household. Malehousehold head.
Incompletenuclear
household.Female
household head.
Complete nuclearhousehold
Households with children
Specific poverty rates by type of household. Ecuador. 2004-2008.
15%
6%
17%
9%
17% 18%21% 21%
48%
55%
43%
0%
10%
20%
30%
40%
50%
60%
One person. Maleolder than 65
years.
One person.Female older than
65 years.
One person. Maleyounger than 65
years.
One person.Female youngerthan 65 years.
Complete nuclearhousehold. Head
older than 65years.
Complete nuclearhousehold. Headyounger than 65
years.
Incompletenuclear
household. Headolder than 65
years.
Incompletenuclear
household. Headyounger than 65
years.
Incompletenuclear
household. Malehousehold head.
Incompletenuclear
household.Female
household head.
Complete nuclearhousehold
Households with children
Specific poverty rates by type of household. Peru. 2002-2006.
17% 15%11% 11%
16% 16%13% 13%
48%
62%
52%
0%
10%
20%
30%
40%
50%
60%
70%
One person.Male older than
65 years.
One person.Female olderthan 65 years.
One person.Male youngerthan 65 years.
One person.Female youngerthan 65 years.
Completenuclear
household. Headolder than 65
years.
Completenuclear
household. Headyounger than 65
years.
Incompletenuclear
household. Headolder than 65
years.
Incompletenuclear
household. Headyounger than 65
years.
Incompletenuclear
household. Malehousehold head.
Incompletenuclear
household.Female
household head.
Completenuclear
household
Households with children
Source: Authors’ elaboration based on national household surveys.
Some characteristics of the heads of household also help explain differing levels of poverty. An inverse relation between the household head’s level of education and the probability of being poor is found in all five countries.
Poverty does not behave monotonically with respect to the age of the household head. In some cases, such as Argentina, Ecuador and Peru, poverty initially increases with age (or remains relatively stable) to then decrease, showing the lowest rates for individuals older than 64 years old (column 8). In other cases, such as in Brazil, a negative correlation between poverty and age is verified. The relationship between age and poverty is not so clear in Costa Rica, however, and that country is the only exception to the rule that poverty is lowest among people over the age of 65. This general rule for the age group may indicate the existence of some “selection bias”, since only elder adults earning a minimum income that guarantees their subsistence can live on their own; otherwise, they would be forced to live in extended households.
42 43
This might also be a result of the methodology used for poverty identification. In particular, equivalences among age groups are derived from the differences observed in their nutritional requirements without taking into account the differences existing in other types of needs. As the nutritional requirements are lower for elder people, the poverty lines for households with members of such age groups might be underestimated as they do not consider the relatively larger requirements they have in other areas such as health.
Gender is another important dimension to characterize poverty. Results on this variable are not homogeneous across countries: in Peru, the specific poverty rate for households with a female head is lower than for households with a male head. In Costa Rica, the incidence of poverty in households with female heads is 28% in 2008, whereas in the case of households with male heads it is 19%. In Argentina, Brazil and Ecuador households with female heads show a slightly higher poverty rate than households with a male head.
As expected, poverty incidence is higher within the group of households headed by a person who is unemployed or out of the labor force without a pension (column 8). However, perhaps the most important result is that having a job is not a guarantee against poverty: a high proportion (between 60% and 80%) of poor household heads are employed in the countries under study. Job characteristics appear to be important variables associated with the existence of the working poor. In particular, as it is shown in Graph 10 below, households whose heads are non-registered wage earners (and also own-account workers) are overrepresented among the poor.28
28 Non- registered employees are those wage earners not covered by labor legislation, that is to say, wage earners whose employer does not abide by labor laws and regulations. In this paper, the lack of registration of the employee in the social security system is the proxy variable used to identified non- registered workers.
44
Graph 10Composition of total and poor households
Five Latin American countriesAverage ca.2003-ca.2008.
Occupational category of household head. Argentina. 2003-2006.
5%
20%
8%
31%
19%67%
3%
15%
Unemployed
Inactive with pension
Inactive without pension
Registered wage earner
Non-registered wage earner
Own account
Employer 11%
10%
14%15%
30%
19%
1%
65%
Unemployed
Inactive with pension
Inactive without pension
Registered wage earner
Non-registered wage earner
Own account
Employer
Occupational category of poor household head. Argentina. 2003-2006
Occupational category of household head. Brazil. 2003-2006.
4%
22%
8%
34%
11%
15%
4%
65%
Unemployed
Inactive with pension
Inactive without pension
Registered wage earner
Non-registered wage earner
Own account
Employer
Occupational category of poor household head. Brazil. 2003-2006.
10%
12%
20%
26%
14%
17%
1%
58%
Unemployed
Inactive with pension
Inactive without pension
Registered wage earner
Non-registered wageearnerOwn account
Employer
Occupational category of household head. Costa Rica. 2006-2008.
2%
12%
10%40%
11%
16%
8%
76%
Unemployed
Inactive with pension
Inactive without pension
Registered wage earner
Non-registered wageearnerOwn account
Employer
4%
9%
21% 26%
14%
21%
5%
66%
Unemployed
Inactive with pension
Inactive without pension
Registered wage earner
Non-registered wageearnerOwn account
Employer
Occupational category of poor household head. Costa Rica. 2006-2008.
Occupational category of household head. Ecuador. 2004-2008.
5%
20%
8%
31%
19%
15%
3%
67%
Unemployed
Inactive with pension
Inactive without pension
Registered wage earner
Non-registered wageearnerOwn account
Employer
Occupational category of poor household head. Ecuador. 2004-2008.
11%
10%
14%15%
30%
19%
1%
65%
Unemployed
Inactive with pension
Inactive without pension
Registered wage earner
Non-registered wageearnerOwn account
Employer
Occupational category of household head. Peru. 2002-2006.
5%
11%
9% 20%
18%
29%
7%
74%
Unemployed
Inactive with pension
Inactive without pension
Registered wage earner
Non-registered wage earner
Own account
Employer
Occupational category of poor household head. Peru. 2002-2006
8%
3%
14%
9%
25%
37%
4%
75%
Unemployed
Inactive with pension
Inactive without pension
Registered wage earner
Non-registered wageearnerOwn account
Employer
Source: Authors’ elaboration based on national household surveys.
Another way to visualize the relationship between job characteristics and poverty is analizing the specific poverty rates of households whose heads are non-registered wage earners or own account workers. As it can be seen in Graph 11, these specific rates are about three times higher than the specific poverty rate of households with registered-wage earner heads for Argentina,
44 45
Ecuador and Peru. For instance, in Ecuador, 45% of the households headed by a no-registered wage earner were poor and 41% of households whose heads were own account workers were poor in 2008, while this figure was only 11% for households whose heads were employed in jobs registered in the social security system (column 8). This means that around 80% of poor households have an employed head (column 9). This scenario is very similar in the rest of the countries.
GRAPH 11Specific poverty rate by occupational characteristics of the household head
(Baseline=registered wage earner)Five Latin American countries
Ca. 2003- ca. 2006.
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
Non-registered wage earner Own account Employer
Argentina
Brazil
Costa Rica
Ecuador
Peru
Source: Authors’ elaboration based on national household surveys.
Therefore, as it was mentioned above, the lack of jobs is not the only factor associated with the ongoing prevalence of poverty in these countries. The poor quality of jobs is perhaps on of the most relevant factors associated with poverty in Latin America. Instead of “unemployed poor”, the more frequent feature is that of “working poor”. This phenomenon is is related to both supply factors (such as low educational levels) and demand factors (such as the insufficient formal jobs generated in the economy, or widespread low productivity). Moreover, it is observed that in all cases poverty incidence is lower for households with full-time employed heads than for those with over- or under-employed
46
heads. may indicate that underemployment and a precarious job both negatively affect the ability to attain a high enough income to fulfill the basic bundle of consumption needs. It is interesting to note, however, that the relationship between the incidence of poverty and working hours is not monotonic, since there is an important proportion of household heads who work more than 45 hours a week and are still poor. This again may reflect the precarious conditions that poor households work in, with little or no enforcement of labour regulations and long working days needed to reach a minimum income.
The information analyzed above originates in cross section data. However, as it is shown in Tables 4.A-4.E included in Annex B, these results are corroborated by those from the the pool of panels used for each country. This is particularly important as it suggests that the reduction in the number of observations derived from the construction of panels does not alter the results regarding the characteristics of households and individuals under analysis.
POVERTY DYNAMICS IN FIVE LATIN AMERICAN COUNTRIES
Transition Matrix Starting with the dynamic approach, entry and exit poverty rates
will be measured and analyzed in order to assess the importance of these movements on poverty incidence performance.
The entry rates are calculated as the proportion of non-poor households in “t” that become poor in “t+1”. Exit rates are computed as the proportion of poor-households in “t” that become non-poor in “t+1”. These transition rates between poverty and non-poverty may be interpreted as the conditional probability that a household in a stationary population experiences a transition, given its initial situation. In this sense, the poverty status of households describes a first-order Markov process.29
29 The characteristics of the samples allow associating a multinomial distribution with the quantities of entries (to) and exits (from) poverty. Therefore, consistent estimation of the probabilities of entering or exiting poverty (pij) are reached by optimizing the likelihood function, which results in the relative frequencies of these transitions.
46 47
Tables 5.A-5.E included Annex C present the annual transition rates. Approximately, between 33% and 45% of households that were poor in the initial year were no longer in that situation one year later. However, the transition matrix also shows that around 10%-25% of households that were initially not poor fell into poverty during the following year. In particular, the entry and exit rates between two observations with one year in between are, on average for the considered period, 7.9% and 33.4%, for Argentina; 13.2% and 42% for Brazil, 10.9% and 45.2% for Costa Rica; 15.3% and 36.8% for Ecuador, and 20.8% and 38.5% for Peru. These findings show the importance of entry and exit flows in these countries, even in Costa Rica, where the incidence of poverty is relatively low.
As expected, the probability of being poor in a given moment is strongly conditioned by the poverty situation in the previous observation, with poverty in the current period being more likely among households that were poor in the previous period. More information would be necessary, however, to make any conclusive statements about true dependence on the initial state.
From the comparison of households ordered according to the average poverty incidence during the period under analysis, it is possible to conclude that although there is some positive relationship between poverty incidence and entry rates, and some negative correlation between poverty and exit rates, these regularities are not verified in all the cases. For instance, Argentina has the lowest entry and exit rates and Ecuador has the highest poverty rates, although entry rates are lower than those observed, for example, in Peru (Graph 12).
48
GRAPH 12Poverty Entry and Exit Rates according to poverty incidence
Five Latin American countries
11.210.0
13.2
21.0
15.0
46.9
32.0
42.0
38.0 37.0
0
5
10
15
20
25
30
35
40
45
50
Costa Rica (24%) Argentina (27%) Brazil (29%) Peru (35%) Ecuador (37%)
Entry Rate
Exit Rate
Source: Authors’ elaboration based on national household surveys.
Tables 5.A-5.E included in Annex C also present the same transition matrices separately for households with and without children. As mentioned before, higher poverty rates among households with children are due to both higher entry rates and lower exit rates, amounting to longer episodes of poverty. This holds true for every country in this study except for Costa Rica, where households with children have both higher entry and exit rates. However, The size of the gap between households with children and all other households is different for entries and exits, except in Brazil where both relative risks are similar. In other countries, the biggest differences between households with and without children are found in terms of poverty entries: households with children are about twice as likely to enter poverty as those without (in Argentina and Peru). These results give greater relevance to the dynamic analysis, for they show the heterogeneity of situations that may be taking place in terms of entry and exit flows behind a given poverty rate.
48 49
Graph 13Relative Risk of entering and exiting poverty
Households with and without childrenFive Latin American countries
0
0.5
1
1.5
2
2.5
Argentina Peru Ecuador Costa Rica Brazil
Entries Exits
Source: Authors’ elaboration based on national household surveys.
These results could be directly associated with a specific value
of the poverty line. Hence, to rule out this possibility and assess the robustness of the results, we performed a sensitivity analysis of poverty dynamics to changes in the poverty line. As shown in Tables 5.A-5.E included in Annex C, we have constructed the same transition matrices but this time reducing and increasing the poverty line by 10% (poverty line at 90% and 110%, respectively). In almost all the cases, entry rates show a monotonically increasing behavior when the poverty line increases, while the opposite occurs with exit rates.30 It is also important to mention that the behavior observed for households with and without children remains unchanged for different values of the poverty line.
Other interesting indicators can be derived from the transition matrices, such as the proportion of new poor, defined as those households that are not poor in the initial period and are poor in the final period. As it is shown in the tables below, new poor households represent on average 30% or more of those poor
30 The only exception to this regularity is Argentina, where entry rates first fall and then grow as the poverty line increases.
50
households in the second observation. It is also possible to assess the relative importance of those households that are poor in both observations, and those that are poor in only one of the periods observed (households entering or exiting poverty). As it can be seen, the proportion of households that are poor in only one observation is between 16% and 28%, while 13% to 25% of households are poor in both years. Furthermore, between 32% and 53% of total households are poor in at least one observation, thus indicating that the incidence of poverty drawn from these data is significantly higher than that coming from static data.
Poverty situation of households in both observationsFive Latin American countries
100% Poverty lineArgentina Brazil Costa Rica Ecuador Peru
NEW POOR 21% 36% 40% 30% 33%POOR IN ONLY ONE OBSERVATION 16% 21% 20% 23% 28%POOR IN BOTH OBSERVATIONS 21% 17% 13% 23% 25%NEVER POOR 63% 62% 68% 54% 47%POOR IN AT LEAST ONE OBSERVATION 37% 38% 32% 46% 53%
90% Poverty lineArgentina Brazil Costa Rica Ecuador Peru
NEW POOR 23% 39% 44% 34% 37%POOR IN ONLY ONE OBSERVATION 15% 20% 18% 22% 27%POOR IN BOTH OBSERVATIONS 17% 14% 10% 19% 21%NEVER POOR 67% 66% 10% 59% 52%POOR IN AT LEAST ONE OBSERVATION 33% 34% 28% 41% 48%
110% Poverty lineArgentina Brazil Costa Rica Ecuador Peru
NEW POOR 19% 33% 37% 27% 29%POOR IN ONLY ONE OBSERVATION 16% 22% 21% 24% 28%POOR IN BOTH OBSERVATIONS 24% 19% 15% 27% 30%NEVER POOR 60% 58% 64% 50% 58%POOR IN AT LEAST ONE OBSERVATION 40% 42% 36% 50% 59%
Source: Authors’ elaboration based on national household surveys.
Finally, as it is shown in the table above, while the percentage of households that remain poor in both observations increases monotonically with the poverty line, the percentage of households that are poor in only one observation has a different behavior. This is due to the opposite effect the increase of the poverty line has on poverty entries and exits, as it was just explained above.
50 51
Factors directly associated with exit rates General overview
This section focuses on the analysis of the events associated with poverty exits using the decomposition stated in equation [2]. Factors directly associated with exits are examined first because, as seen above, the level of poverty in these countries declined during the timeframe studied. Hence, it is interesting to analyze the factors that have been directly associated with these positive dynamics.
Tables 6.A–6.E present poverty exit rates disaggregated by the type of event experienced by the household. The result, presented in column (3), is equal to the product of the frequency of each event (column 1) and the conditional probability of exiting poverty given that the event has taken place (column 2). The decomposition of exit rates is also reproduced in the table below:
Decomposition of exit rates: event frequency and conditional probability of poverty transition
Five Latin American countries
Decomposition of the entry rates: event frequency and conditional probability ofFive Latin American countries
P (event) x P (S/E) = Entry rate
Argentina 38% x 21% = 8%
Brazil 50% x 25% = 13%
Costa Rica 49% x 23% = 11%
Ecuador 57% x 27% = 15%
Peru 67% x 32% = 21%
Source: Authors’ elaboration based on national household surveys.
An important finding is that in all countries, a very high proportion of the initially poor households experienced some positive event that would potentially allow them to exit poverty (column 1). In particular, around 75% of these households experienced an increase in their income per adult equivalent (ipae), as a result of increases in total household incomes or due to
52
a reduction of the household’s size. This positive finding is at least partly linked to specific characteristics of the analyzed period, when economic growth and poverty reduction prevailed.
Of households that experienced a positive event, however, only around 50% (or less) actually moved out from poverty, while the other half experienced increases in incomes that were not sufficient to change their poverty status (column 2). Therefore, the high levels of poverty incidence do not seem to be exclusively associated with the absence of “positive” events but rather they are also related, at least to some extent, to the weakness of these events, which does not allow these families to escape poverty.
When comparing these results across countries, it becomes evident that the largest divergences in terms of the conditional probability, that is to say the “effectiveness” of the events to take households out of poverty. This variable practically explains the whole of the differences observed in the exit rates, given that the probability that a poor household experiences some positive event is very similar in all the countries analyzed. One of the reasons for this finding is that the poverty gap tends to be larger in countries with a higher poverty rate, making it more difficult to exit poverty even after experiencing an increase in household income.
The events exclusively related to the labor market are the most relevant among those associated with poverty exits (Tables 6.A-6.E included in Annex C, column 3). As summarized in Graph 14, between 30% and 70% of the poverty episodes that come to an end are related to a labor event. The second most significant group of events is the combination of non-labor and labor incomes’ growth. This clearly means that the labor market has played a very important role in the improvement of the living conditions of households in recent years, both through events exclusively related to the labor market, or through a combination of the latter with increases in other non-labor incomes. Together, these events account for 60%-80% the exit rate in these countries during the period of study.
52 53
GRAPH 14Relative importance of the events associated with exit rates
Five Latin American countries
30.8
67.0
49.6
40.1
22.62
15.17
20.89
28.62
11.31
26.41
4.38
6.904.01
10.46 9.91 12.6615.96
19.97
51.5
28.88
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Argentina Brazil Costa Rica Ecuador Peru
Exclusively labor events Labor and non labor events Exclusively non-labor events
Exclusively demographic Demographic leading to income events Demographic and income events
Source: Authors’ elaboration based on national household surveys.
Depending on the country, these events are followed in
importance by exclusively non-labor income events (for instance, Brazil and Argentina) or those wich combine demographic and income events (Costa Rica, Ecuador and Peru).
As for exclusively demographic events (such as a reduction in the number of household members), they barely matter for changes in poverty in all of the countries. This evidence seems to be consistent with that found in other studies that use a similar methodology. These studies generally find that changes in income or the number of employed household members are the events most frequently associated with exits from poverty, while the changes in the composition of households are less important.31 This is not a surprising result because the yearly observation window is probably too short to observe household demographic change because such events are typically less frequent.
31 Bane and Ellwood (1986), Ruggles and Williams (1987) and Duncan and Rodgers (1988), among others.
54
Given the relevance of the labor market in the transits out of poverty, taking a more detailed look at the events associated with these changes. The most frequent among the latter events are either wage growth (Argentina and Ecuador) or an increase in the number of employed household members (Brazil, Costa Rica and Peru) (Tables 6.A-6.E included in Annex C, and Graph 15).
GRAPH 15Relative importance of exclusively labor events associated with exit rates
Five Latin American countries
49%
12% 15%
15%
9%
39%
Argentina
Growth in labor incomes
Growth in the number of employed members
Growth in the number of working hours
Registered wage earners
Non registered wage earners
Non-wage earners
37%
6% 27%
16%
15%
57%
Brazil
Growth in labor incomes
Growth in the number of employed members
Growth in the number of working hours
Registered wage earners
Non registered wage earners
Non-wage earners
34%
5% 28%
16%
61%
17%
Costa Rica
Growth in labor incomes
Growth in the number of employed members
Growth in the number of working hours
Registered wage earners
Non registered wage earners
Non-wage earners
61%
12%
4%
11%
12%
27%
Ecuador
Growth in labor incomes
Growth in the number of employed members
Growth in the number of working hours
Registered wage earners
Non registered wage earners
Non-wage earners
37%
21%
8%
18%
16%
42%
Peru
Growth in labor incomes
Growth in the number of employed members
Growth in the number of working hours
Registered wage earners
Non registered wage earners
Non-wage earners
Source: Authors’ elaboration based on national household surveys.
Regarding the latter event, it is worth noting that in most cases, the newly employed members get a job as wage-earners not registered in the social security system. In fact, in Ecuador and Peru, exits from poverty associated with getting a job of this type are more frequent than exits linked to household members obtaining a registered job. These exits, together with those derived from an increase in the number of employed members who get a self-employed job, explain most of the increases of employed individuals in initially poor households that exit poverty. In both
54 55
countries, however, getting a registered job has relatively low relevance. In the other three countries, even if the importance of registered occupations in taking households out of poverty is greater than in the other countries, the probability of obtaining non-registered jobs is still high. This result brings to light the still high levels of labor precariousness prevailing in the region and its relation to poverty.
The increase in incomes from pensions is the most important non-labor event in Argentina, Brazil and Costa Rica, while the growth of other non-labor incomes is the most frequent event in Ecuador and Peru. The second type mostly includes donations from one household to another. In particular, in Ecuador these are generally remittances from migrants working in foreign countries, an expected result considering the high relevance of this type of flow of incomes.
Another important finding is the low relevance of public transfers in explaining exits from poverty. This is particularly worrisome given the presence and extension of conditional cash transfers (CCTs) such as Bolsa Familia in Brazil, Programa Jefes in Argentina, Plan Juntos in Peru, Bono de Desarrollo Humano in Ecuador and Avancemos in Costa Rica, in recent years. We investigate possible explanations for this result below..
When decomposing the exit probabilities according to equation [2], it becomes evident that most of the exits from porverty that take place jointly with labor events (simple or combined) are mainly derived from the relatively higher frequency of these events compared to others (column 1). To a less extent, they also respond to the relatively high conditional probability of these events, i.e. the probability of exiting poverty given that the event has actually taken place (column 2).
In the particular case of wage increases, their relevance in poverty exits is mainly explained by their relatively higher frequency compared to other events, whereas the probability of exiting poverty given that this event has taken palce is lower than that associated with employment growth. This is another expected result given that the total increase in household income associated with a family member obtaining a job is typically larger than an increase in income for an already employed household member.
56
Then, it is important to highlight the differences in this decomposition depending on the type of the job found. It was already mentioned that the generation of wage-earning employment not covered by social security has been highly predominant during the period under analysis. In particular, as it is shown in Graph 16, with the only exception of Brazil, the probability of a poor household member of obtaining a non-registered job is significantly higher than the probability of obtaining a registered occupation.
GRAPH 16Frequency of increases in the number of employed members
by type of occupationFive Latin American countries
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
Argentina Brazil Costa Rica Ecuador Peru
Registered wage earners
Non.registered wage earners
Non-wage earners
Source: Authors’ elaboration based on national household surveys.
This is an extremely important result because starting a registered job is associated with a higher probability of exiting poverty than other types of employment in every country studied, given higher average wages for registered jobs (Graph 17). This means that the possibilities of getting a job covered by social security are scarce but the probability of exiting poverty when this type of episode takes place is higher. For example, a member of a poor Argentinian household who finds a registered job is nearly three times as likely to bring their household out of poverty than
56 57
one who finds a non-registered job. This clearly shows that labour precariousness and informality reduce opportunities to escape poverty in Latin America.
GRAPH 17
Conditional probability of exiting poverty by type of occupationFive Latin American countries
1
1.2
1.4
1.6
1.8
2
2.2
2.4
2.6
2.8
3
Argentina Brazil Costa Rica Ecuador Peru
Registered/ Non registered wage earners
Registered wage earners/ Non wage earners
Source: Authors’ elaboration based on national household surveys.
Coming back to public cash transfers, their limited importance as a source of exits from poverty might derive from both the low frequency of this kind of events and their low conditional probability which, in turn, seem to be indicating the small coverage and magnitude of these transfers, respectively. Different factors could explain the small role of public transfers in exits from poverty. First of all, this type of income flow could be underreported in the surveys. Furthermore, as it was mentioned above, in the case of Brazil non-labor incomes have been computed based on information from the PNAD, and hence it has not been possible to identify the households that actually receive monetary transfers, which could result in an underestimation of the impact of these incomes on transitions between poverty and non-poverty. Secondly, studies have generally found that these transfers are more effective at reducing extreme poverty than poverty,32 the
32 See, for example, ILO/IPEC (2007), Perez Ribas et al. (2008), Villatoro (2007, 2008), ILO (2009), Román (2010), Perova and Vakis (2009), Vera Soares et al. (2006).
58
second of which is the variable analyzed here. Third, our analysis only deals with urban areas, while some of these programs are focused in rural areas. Fourth, the time period is also important because some of the CCTs in these countries began quite recently, so the panels built for this study were unable to capture them. Other programs also started well before the timeframe of the current study so it is unlikely for the data to capture new entries into the program. Furthermore, households who are beneficiaries of these programs saw their incomes increase when they entered the program, but not necessarily during the period under study. For example in Argentina, poor households had already started receiving the Plan Jefes, and there have not been many new beneficiaries in the period under observation, while the amount of the cash transfers did not change either. As a result, there have hardly been any observable events (changes) of this kind during the period under analysis Fifth, up to this point we have analyzed the association between these transfers and aggregate exit rates while, CCTs in the region generally focus on households with children.
Finally, there is one aspect related to the methodological approach employed that could also explain these findings. In particular, it is worth remembering that the analysis is based on an exhaustive list of mutually exclusive events, meaning that the evaluation of the role of CCTs results from the frequency and the conditional probability of experiencing an increase in the amount of this kind of income only. This means that the role we find for CCTs results from the frequency and the conditional probability of experiencing only an increase the amount of this kind of income. If another source of income also changes between observations, these changes are classified together as a combined event, reducing the visibility of these public transfers.
Households with and without children
As it was shown in section 6.3 and section 7.1 above, there appears to be a strong relationship between the composition of households -especially the presence of children- and poverty status. In particular, it has been shown that the incidence of poverty in households with children is higher than in households without children and that this is associated with lower exit rates and higher entry rates. Consequently, we will now analyze the
58 59
events associated with poverty transitions, distinguishing between households with and without children. A priori, it would be expected that households of differing size and composition may be more or less likely to be exposed to particular kinds of events, both of demographic and labor market types.33
The results, shown in Table 6.A-6.E included Annex C, suggest that those episodes exclusively related to the labor market are the most important ones for both types of households, with and without children. However, in all cases their relative importance is higher within the first group. The opposite occurs with non-labor events, which turn out to be very significant in the case of households without children, a situation associated with the significant increase in incomes from pensions occurred in Argentina, Brazil and Costa Rica. This seems to be a reasonable result, for it is more probable to find elder members in households without children and, thus, more probable for these households to experience increases in incomes from pensions. In Peru and Ecuador, the increase in other non-labor incomes is the most significant non-labor event for households without children.
These findings are partially consistent with those obtained by Cantó et al. (2007) for Spain. The authors show that poverty exit rates for households with children are more related to labor market events, whereas non-labor events are the most important among households without children. They conclude that the dynamics of the first type of households is more related to the economic cycle, while the behavior of the second group is more closely linked to the social security system.34 In any case, it is worth stressing that non-labour events in the countries included in this study tend to involve pensions or other non-labour income rather than cash transfers in relation to a social policy program.
33 For example, labor market episodes would be expected to be less frequent among the households with the majority of its members at retirement age than among households with members at working age.
34 This is related to the fact that the social security system in this country is more directed towards the fight against poverty of the elder (who are mainly in households without children) than of the young and children. In this sense, this contrasts with the situation of other European countries in which the universal child care system is more developed and thus appears as a particularly important factor explaining transitions from poverty.
60
The relative importance of the increase in the number of employed members vis à vis the growth of wages is higher in the case of households without children than in those with children. This might be at least in part associated with the differences observed in the demographic composition of both types of households, since it is more probable to find young individuals who are entering the labor market within the first group. Again, members of both household types tend to find jobs that are not registered in the social security system.
As already mentioned, changes in household composition are relatively unimportant for poverty exits. In this context, and similarly to the results obtained for other countries35, household composition changes more frequently for households without children in Argentina and Peru, suggesting that households with children are more stable in their structure. The opposite situation is found in the other countries in the present study. Finally, public transfers other than pensions have no significant impact on either type of households.
One important aspect to be highlighted from the decomposition of exit rates from poverty is that the conditional probabilities associated with each type of event are systematically higher for households without children, whereas the proportion of poor households experiencing some event is very similar for both types of households. The first of these two factors therefore explains much of the difference in exit rates between these groups.36
Factors directly associated with entry rates
General overview Tables 7.A-7.E included in Annex C show the factors associated
with poverty entries. A first result to be highlighted is the very high proportion of non-poor households that go through a “negative” event that reduces their ipae (between 38% and 67 %) (see table below). Then, in approximately one third of such households these events effectively lead them to poverty, which explains the high frequency of movements towards poverty analyzed above.
35 For the Spanish case see, for example, Cantó et al. (2007).36 In the case of Costa Rica, households with children experience events
that might take them out of poverty more frequently than households without children. However, like in other countries, the former households register a lower conditional probability of exiting poverty.
60 61
Decomposition of the entry rates: event frequency and conditional probability of poverty transition
Five Latin American countries
P (event) x P (S/E) = Entry rate
Argentina 38% x 21% = 8%
Brazil 50% x 25% = 13%
Costa Rica 49% x 23% = 11%
Ecuador 57% x 27% = 15%
Peru 67% x 32% = 21%
Source: Authors’ elaboration based on national household surveys.
This finding shows that there is a non negligible group of households moving into poverty even when poverty is in overall decline, as happened in recent years in the analyzed countries. It also stresses the importance, once again, of analyzing poverty flows that underlie static indicators of poverty incidence.
Unlike the case of exit rates, the most important divergences across countries are explained by the frequency of the events, while the conditional probabilities of the latter are more similar for all of them. For example, in Peru the probability of entering poverty is more than twice as high as in Argentina, and this is mostly a result of the relatively higher frequency of negative events observed in the former country (67%) compared to the latter (38%). In fact, initially non-poor households in Argentina are the least likely to face events which reduce their ipae. This positive situation can probably be explained by significant improvements in social and labour conditions over recent years. In any case, as it was mentioned above, at least 40% of non-poor families in all five countries suffered an event with the potential to push them into poverty.
As in the case of exits from poverty, exclusively labor events are the most important source of poverty entries: between 30%-50% of the transitions into poverty take place together with a negative
62
labor event (column 3). The only exception is Brazil, where that proportion is of only 25%, while most entries are related to the reduction of non-labor incomes. The joint reduction of labor and non-labor incomes is also a significant event in some of the countries.
Exclusively demographic events are also relatively unimportant for poverty entries, but appear to play somewhat more of a role here than in the case of exits from poverty: an increase in the number of household members is more relevant for poverty mobility than its reduction. In this context, demographic events seem to be somewhat more significant in Peru and Ecuador. This result could be at least in part related to the higher fertility rates observed in comparison to other countries, as well as to the different composition of households. At the same time, like in the case of exits, the reduction of incomes from cash transfer policies has no relevance whatsoever to explain entries to poverty.
Loss of a job by a household member is the most important event among exclusively labour events, except in Ecuador, where a wage reduction is the single labour market event most associated with entry into poverty. The analysis of the two components affecting the probability of entering poverty makes it clear that in all cases, as expected, the conditional probability associated with the loss of an occupation is significantly higher than the probability linked to the reduction of labor incomes. As a matter of fact, in Argentina, Brazil and Costa Rica, the relevance of such event is exclusively related to its stronger impact in terms of reducing the household income, given that this event occurs with less frequency than the fall of nominal wages.
Again, occupation type must also be considered when looking at entry rates. Specifically, the high frequency of losses of non-registered jobs among non-poor households suggests both the high prevalence of this kind of occupations and also their higher instability. Additionally, in Argentina and Costa Rica, the conditional probability of entering poverty associated with the loss of a job is, as expected, higher in the case of registered jobs than in the case of non-registered jobs. This is due to the higher wages generally associated with registered occupations.
62 63
Households with and without children
An interesting result appears when comparing poverty entries for households with and without children: there is no substantial difference in the total frequency of events for each household type. These events’ impact (conditional probability) in poverty entries, however, differ substantially. In particular, the probability of entering into poverty given the occurrence of a negative event in households with children is approximately 10 p.p. higher than for households without children. This is observed for every type of event.
Finally, like in the case of exits, labor events are the most significant type of event in both types of households, while they are more relevant for households with children. In the case of households without children, the reduction of non-labor incomes acquires more importance. At the same time, households with children seem to be more stable in their structure, which is reflected in a higher incidence of exclusively demographic events in households without children.
Modeling poverty dynamics37 The relative importance of the effects analyzed might in
turn be affected by the characteristics of households and their members. In order to consider all these factors simultaneously, we have modeled poverty dynamics with a Multinomial Logit model, taking into account that from two observations it is possible to identify four mutually exclusive possibilities: (i) being poor in both periods, (ii) being non-poor in the first period and poor in the second period, (iii) being poor in the first period and non-poor in the second period and (iv) being non-poor in both periods.38
Thus, it is possible to compute Relative Risk Ratios –RRR- which indicate the risk of a household experiencing one of the four mutually exclusive alternatives,and these probabilities can then be normalized dividing them by the probability experiencing any other state. Hence, on the one hand, we will present the risk of entering into poverty faced by a household in relation to never being poor and, on the other hand, the risk of exiting poverty in relation to being poor in both observations.
37 The collaboration of Abdelkrim Araar was of great importance for the elaboration of this section.
38 A similar approach is followed by Justino and Litchfield (2003).
64
The covariates are comprised of the different possible events as well as of an exhaustive list of the attributes of households and of their members in the initial observation. An important difference in relation to the previously presented approach is that the events used in the regressions are now defined in a non mutually-exclusive way. In particular, for one household (poor or not poor), a dummy variable will take the value 1 if such household experienced a given event (that might potentially trigger an exit from or an entry into poverty), regardless of whether it has additionally experienced another event or not. This allows indentifying not only those events that lead to changes in the household status, but also those that might prevent transitions.
The results are shown in Tables 8.A-8E included in Annex C. Coefficients with values lower than 1 indicate that the corresponding variables raise the probability of the household being in the base category (never being poor in the case of entry rates, and being poor in both observations in the case of exit rates), whereas coefficients with values higher than 1 mean that the corresponding variables raise the probability of the household being in the alternative category.
In most cases, we confirm that the coefficients corresponding to the events that might potentially trigger changes in households’ status are statistically significant for the explanation of the movements into and out of poverty, and have the expected signs even when controlling for the socio-demographic characteristics of households, year and region. In particular, “positive” events actually raise the probability of exiting poverty while “negative” events have the opposite effect.
It is also observed from the analysis that increases (reductions) in the number of household members employed in wage earning occupations registered in the social security system are associated with a higher conditional probability than increases (reductions) in the number of members employed in not registered wage earning or independent occupations. This means that when a household member obtains (looses) an occupation registered in the social security system, the odds of exiting (entering) poverty increase (fall) more than in the case of obtaining (loosing) an occupation of another type. This was an expected result given the higher income levels usually obtained in registered jobs.
64 65
Regarding the characteristics of household members, we also confirm that the educational level of the household head significantly raises the probability of exiting poverty and reduces the probability of entering into that situation. Similar results are attained in relation to the educational level of the spouse.
At the same time, the results show that in relation to being employed in a registered salaried position, households whose head or spouse are non-registered wage earners have relatively higher probabilities of entering poverty and lower probabilities of exiting poverty. However, it is important to point out that both coefficients are larger for the head than for the spouse, which is consistent with the fact that usually the head’s income is more important in the households’ total income. In general terms, a similar panorama is also found when household heads or spouses are either unemployed, independent workers or economically inactive individuals who do not receive a pension: all of these situations are associated with larger probabilities of entering into poverty and lower probabilities of moving out of poverty.
On the other hand, in all the cases, the presence of children raises the probability of entering poverty and reduces the probability of exiting poverty. The inverse of the dependency rate (the proportion of employed members relative to the size of the household) has an impact of the opposite sign given that in all cases, as it is expected, the larger the number of household members that receive an income, the greater the resources available for the household to avoid entering into poverty or to rapidly move out of that state.
Finally, one important issue to be considered is that including the events as covariates may cause endogeneity problems given that they can simultaneously be a cause and a consequence of poverty transitions. However, the information available does not provide us with adequate instruments to address this problem. Bearing in mind these limitations, we employ these regressions in order to corroborate if the results regarding the previously analyzed conditional probabilities hold valid even when considering the characteristics of the households and their members.
66
Analysis of sensitivity to the poverty lineLastly, an analysis of the sensitivity of poverty dynamics to
changes in the poverty line was performed to assess the robustness of the results. In particular, the same decomposition shown in equation [2] but only considering those transitions that take the households to positions 10% below and 10% above the value of the poverty line was carried out.
Tables 9.A-9.E included in Annex C present the results for exit rates considering a 10% reduction of the poverty line, whereas Tables 10.A-10.E included in Annex E present these results considering a 10% increase of the poverty line. Tables 11.A-11.E and Tables 12.A-12.E included in Annex C, respectively, present the same results for entry rates. As it is shown, the results regarding the relative importance of the events do not change substantially, which indicates that the conclusions are robust to changes in the value of the poverty line.
CONCLUSIONS AND POLICY RECOMMENDATIONS
The research summarized in this document had three objectives: (1) to estimate the role of the labor market, social policies and demographic changes in transitions into and out of poverty; (2) to evaluate whether the observed differences in households’ poverty flows can be mostly explained by differences in the probability of certain types of events occurring or by these events’ differing impacts, i.e., the conditional probability that their poverty status will change after a given event has taken place; (3) to determine to what extent the composition of households and the characteristics of their members affect both probabilities. The results of this study allow us to draw the following conclusions.
In methodological terms, the relevance of the dynamic analysis is demonstrated not only for evaluation of the intensity of poverty entry and exit flows, but also for identification of factors directly associated with these transitions. In particular, this approach allows us to: analyze to what extent countries with different or similar poverty incidences may show exit and entry rates to poverty of different intensity; identify the importance of different
66 67
events associated with poverty movements; assess the relative significance of movements happening as a result of labor market events, demographic changes and public policies; address the question of whether households stay poor (or stay out of poverty) because they do not experience any positive (or negative) event or because the events do not have enough impact to generate a transition out or into poverty; and analyze the differential effects of these events on households of diverse structures and characteristics.
Regarding the results, a very important finding is that a high proportion of initially poor households in every country experienced a positive event that may help them exit poverty. However, only a small share of these households actually exited poverty, while the others experienced increases in income that were not sufficient to change their poverty status. This suggests that the difficulty of exiting poverty does not have so much to do with household members’ inability to obtain new income (such as by getting a new job) as it has to do with the fact that the additional income is not sufficient to escape their situation.
Another relevant outcome is that events exclusively related to the labour market are the most important in every country in the study. This tends to occur via changes in the number of employed household members in some countries, while income changes are more important in other countries. Labour precariousness also appears as an important factor in each of the countries in this study. In particular, a significant proportion of the new jobs obtained by poor households members are non-registered in the social security system. This implies a lack of social benefits and considerably reduces the positive impact that getting a job has for families’ incomes, contributing to the ongoing phenomenon of the working poor. In fact, around 20%-30% of poor household heads in the countries analyzed are employed as informal workers. This means that unemployment is not the only problem of the Latin American labor market; low wages and precarious labor conditions also appear as important issues.
The analysis also shows that public transfers have scarce importance in explaining exits from poverty. This is particularly worrisome given the significant expansion of cash transfers in Latin American countries during the recent years. However, this result might not be conclusive given that there are certain
68
additional factors that need to be considered. In the first place, there could be some underreporting in the surveys, and in the particular case of Brazil, the available dynamic information does not allow identifying the households that are actually receiving public transfers. Moreover, some programs started in recent years, and are can therefore only be partially captured in the last observations of the panel built for each country. On the contrary, some CCTs had already been working for several years and no new beneficiaries were added during the period under analysis. Changes in incomes from social policy transfers can also have taken place in combination with other events, in a period of labor market improvement. Furthermore, impact evaluations of CCTs have generally proven that their effect has been more important on extreme poverty than on poverty changes.
Finally, the events associated with demographic changes –increase or reduction in the number of household members- did not prove to be important in any case.
When the events associated with poverty entries and exits are analyzed separately for households with and without children, we find that the most important differences between both types of households are observed in the conditional probability of the events rather than in the frequency of their occurrence. In particular, a similar share of households with and without children experience a positive event that could lead to an exit from poverty, but the effect on household income is much greater for households without children, increasing the odds of households in this group exiting poverty.
For the case of entries, however, it is households with children who register a higher probability of entering poverty after experiencing an event that has reduced their income per adult equivalent.
These results reinforce the idea that households with children in Latin America are among the most vulnerable population, not only because they have higher probabilities of entering into poverty when exposed to negative shocks, but also because they do not have the tools needed to rapidly exit that situation. The high incidence of child poverty that characterizes the region is a result of this situation.
68 69
Finally, it can be said that the dynamic of poverty in the region is strongly linked to income mobility, which is in turn linked with occupational instability and hourly wage changes. These results indicate that households are exposed to macroeconomic and labour market cycles, while public policies that limit their negative effects or strengthen their positive effects appear to be limited.
Regarding policy recommendations, the approach used in this study can be used to identify possible interventions which reduce the odds of entering poverty and increase the chances of moving out of poverty. Two pillars arise as clear guides for policy design: (1) labour conditions should be systematically improved along with a significant reduction of informality and unemployment, and (2) a social protection system should be strengthened, at a minimum including an extended unemployment insurance scheme, cash transfers for households with children and pensions for the elderly.
These results support an expansion of antipoverty strategies, through labour market policies and other approaches of more universal character. On the one hand, priority should be given to efforts which aim to prevent low and medium-low income workers from facing income-reducing events, especially those with a greater chance of dragging the household into poverty. A central preoccupation of those strategies should be to reduce the share of –the highly unstable- informal and precarious employment. This implies adopting both demand and supply side approaches to the issue by stimulating the creation of jobs that are suitable for these workers and increasing their chances of finding available jobs (through training and/or better employment services, for example).
Reducing the negative effects of labor episodes appears to be another possible component of an antipoverty policy. Precisely, when low-earning workers are covered by the social security system, the impact of job separations can be mitigated, at least in part, by employment-protection legislation and unemployment insurance. Nevertheless, most unstable jobs in these economies are precarious and not registered in the social security system. Hence, an important proportion of workers, especially those receiving low earnings, face frequent changes of jobs, mediated or not by unemployment episodes.
70
However, unlike developed countries, the scope and coverage of unemployment insurance in Latin America has been historically limited. Even in those few countries that do have policies of this kind (Argentina, Brazil, Chile, Ecuador, Uruguay and Venezuela), coverage rates among the unemployed population are low. This is mainly a result of high labor precariousness, reduced registration rates, high occupational instability and –probably to a lesser extent than in Europe- long term unemployment. Therefore, it would be positive to extend some kind of unemployment benefits to those leaving non-covered jobs.
Increasing the probabilities of leaving poverty should also be an important part of such policies. This implies the same actions noted above -at both the supply and demand sides of employment- in order to improve the quality of jobs. Wage levels should also be considered as an objective, since getting a job is no guarantee of leaving poverty, particularly when a large portion of jobs are informal. In this regard, reinforcing the minimum wage, as it has been done in Brazil and Argentina, constitutes a valuable instrument, especially if it also has an impact on the wages of informal workers.
If enough jobs are created, especially decent work that generate a sufficient income, and if there is unemployment insurance to support the unemployed who are actively looking for a job, there is less need for members of poor households to accept precarious and low-paid work, resulting in reduced flows into informality.
Even if cash transfer programs for the unemployed are put into place or extended, households with low and unstable labor incomes will still be facing major difficulties. As a result of this, various countries of the region have been implementing conditional cash transfers focused on low-income households with children. The “Bolsa Familia” in Brazil, the “Progama Familias” in Argentina, “Juntos” in Peru, “Avancemos” in Costa Rica and “Bono de Desarrollo Humano” in Ecuador, are examples of conditional cash transfers focused on poor households. Although the results of these programs in these countries are generally satisfactory in terms of their focus and success in reducing extreme poverty, the magnitude of their results has been low. For this reason, and in parallel with other policies, it is necessary that these countries reinforce this kind of transfers (both in coverage and in the amount of the transfer, which is generally very low) at
70 71
least until the labour market generates enough jobs with enough income to escape poverty. An important initiative to extend family allowances recently began in Argentina. These were traditionally only received by children of registered wage earners and were extended to cover children of the unemployed, independent workers and non-registered wage earners with low income, bringing these groups into the social security system.
An important aspect that should be taken into account is that a comprehensive public policy for children cannot be exclusively grounded on monetary transfers. Countries must move towards a greater integration with other programs that aim at tackling the social risks faced by households in which children and adolescents live. This requires ensuring access to good quality health and education services to effectively support the human and economic capabilities of future generations.
Finally, all the above-mentioned policies should be complemented by schemes of non-contributive pensions and other actions aimed at improving the quality and accessibility to the public health system.
72
REFERENCES
Aassve, A., S. Burgess, C. Propper and M. Dickson (2004) “Employment, Family Union and Childbearing Decisions in Great Britain”, CASE Discussion Paper No. 84.
Aassve, A., S. Burgess, C. Propper, and M. Dickson (2005) “Modelling Poverty by Not Modelling Poverty: A Simultaneous Hazard Approach to the UK”, ISER Working paper 2005 – 26.
Antolín, P., T. Dang and H. Oxley (1999) “Poverty dynamics in four OCED countries”, Economics Department Working Papers N° 212, OECD.
Araar, A. (2003) “The Shapley value”, Sisera training workshop on poverty dynamics, Kampala, Uganda.
Araar, A. (2007) “Poverty, Inequality and Stochastic Dominance, Theory and Practice: Illustration with Burkina Faso Surveys”, PMMA Working Paper 2007-08.
Araar A. and J. Duclos (2007) “Poverty dynamics PEP-PMMA training workshop”, Lima, Peru
Arranz, J. and O. Cantó (2007) “The Dynamics of Poverty in Spain: The relevance of considering multiple-spells and their accumulation”, The Society for the Study of Economic Inequality (ECINEQ).
Ballantyne, S., S. Chapple, D. Maré and J. Timmins (2004) “Triggering movements into and out of child poverty: a comparative study of New Zealand, Britain and West Germany”, Social Policy Journal of New Zealand, N° 22.
Bane, M. and D. Ellwood (1986) “Slipping into and out of poverty: the dynamics of spells”, Journal of Human Resources, 21 (1).
Baulch, B. and J. Hoddinott (2000) “Economic Mobility and Poverty Dynamics in Developing Countries”, Journal of Development Studies, Vol.36, No. 6, August, pp.1–24.
Beccaria, L. and R. Maurizio (2003) “Movilidad ocupacional en Argentina”, Informe de investigación, No. 18, Universidad Nacional de General Sarmiento, Argentina.
72 73
Beccaria, L. and R. Maurizio (2005) “Changes in Occupational Mobility, Labor Regulations and rising precariousness in Argentina”, Universidad Nacional de General Sarmiento, Argentina.
Beccaria, L., M. González and R. Maurizio (2007) “Desigualdad y polarización en los ingresos de los trabajadores. Diferentes tendencias en América Latina”, presented at the VI Jornadas sobre Mercado de trabajo y equidad, Universidad Nacional de General Sarmiento, Argentina.
Beccaria, L. and R. Maurizio (2009) “Factors associated to poverty mobility in Greater Buenos Aires”, Journal of Income Distribution. An International Quarterly, Vol. 18, No. 2, June 2009, York University, Canada. ISSN: 0926-6437.
Biewen, M. (2004) “Measuring State Dependence in Individual Poverty Status: Are there Feedback Effects to Employment Decisions and Household Composition?”, Discussion Papers of DIW Berlin 429, DIW Berlin, German Institute for Economic Research.
Biewen, M. (2006) “Who are the chronic poor? An econometric analysis of chronic poverty in Germany”, Research on Economic Inequality, Vol. 13, pp. 31-62
Böheim, R. and S. Jenkins (2000) “Do current income and annual income measures provide different pictures of Britain’s income distribution?”, Working Paper 2000-16, Institute for Social and Economic Research, University of Essex.
Breen, R. and P. Moisio (2004) “Overestimated Poverty Mobility: Poverty Dynamics Corrected for Measurement Error”, Journal of Economic Inequality, 2 (3).
Burgess, S. and C. Propper (1998) “An Economic Model of Household Income Dynamics, with an Application to Poverty Dynamics among American Women”, CASE Discussion Paper No. 9.
Burgess, S., C. Propper and M. Dickson (2002) “The Analysis of Poverty Data with Endogenous Transitions”, University of Bristol.
Cantó, O., C. del Río and C. Gradín (2006) “Poverty statics and dynamics: does the accounting period matter?”, International Journal of Social Welfare, Vol. 15 (3).
74
Cantó, O, C. Del Río and C. Gradín (2007) “What helps households with children in leaving poverty?: Evidence from Spain”, Research on Economic Inequality, Vol. 14, pp. 1-29.
Cappellari, L. and S. Jenkins (2004) “Modelling low income transitions”, Journal of Applied Econometrics, vol. 19, Issue 5, 593–610.
Creedy, J. and V. Borooah (1998) “Income mobility, temporary and permanent poverty”, Australian Economic papers, vol. 37, n°1.
Cruces, G. and Q. Wodon (2003) “Transient and chronic poverty in turbulent times: Argentina 1995-2002”, Economic Bulletin, Vol. 9, N° 3
Datt, G. and M. Ravallion (1992) “Growth and redistribution components of change in poverty measures. A decomposition with applications to Brazil and India in the 1980s”, LSMS Working Paper N° 83.
Devicienti, F. (2001) “Poverty persistence in Britain: a multivariate analysis using the BHPS, 1991-1997”, ISER Working paper 2001-02, University of Essex, Colchester.
Duncan, G. (1983) “The implications of changing family composition for the dynamic analysis of family economic well-being, in Atkinson and Cowel (eds.) Panel Data on Incomes, Occasional Paper N° 2, London School of Economics, London.
Duncan, G. and W. Rodgers (1988) ‘‘Longitudinal Aspects of Childhood Poverty’’, Journal of Marriage and the Family 50(4):1007–22.
ECLAC (2007) Social Panorama of Latin America 2007, Santiago de Chile.
ECLAC (2010) Social Panorama of Latin America 2010, Santiago de Chile.
Elbers, C., J. Lanjouw and P. Lanjouw (2003) “Micro-level estimation of Poverty and Inequality”, Econometrica N° 71, pp. 355-364.
Feres, J. (1997) “Notes on the measurement of poverty by the income method”, CEPAL Review, No. 61, pp. 119-133.
74 75
Fertig, M. and M. Tamm (2007) “Always poor or never poor and nothing in between? Duration of child poverty in Germany”, Working Papers 59, ECINEQ.
Fields, G. and E. Ok (1998) “Measuring movement of incomes”, Economica, 66, pp.455-471.
Foster, J., J. Greer and E. Thorbecke (1984) “A Class of Decomposable Poverty Measures”, Econometrica 52 (3), 761-766.
Freije, S. (2000) “Income, positional and poverty dynamics in Venezuela”, in 5th Meetings of the Latin American and Caribbean (LACEA), Rio de Janeiro.
Gómez, J. and C. Román (2001) “Movilidad de Ingresos y Transiciones fuera de la Pobreza un análisis dinámico para el Perú”, Informe Final Concurso de Proyectos Breves de Investigación del CIES, Grupo de Análisis para el Desarrollo.
Herrera, J. (1999) “Ajuste económico, desigualdad y movilidad”, in R. Webb; M. Ventocilla (1999) (eds) Pobreza y economía social. Análisis de una encuesta (ENNIV-1997), USAID, Instituto Cuanto, UNICEF, pp 101-142.
Herrera, J. (2001) “Poverty Dynamics in Peru, 1997-1999”, Document De Travail, DT/2001/09, Paris, Développements, Institutions and Analyses de Long Terme (DIAL).
Herrera J. and F. Roubaud (2007) “Urban poverty dynamics in Peru and Madagascar”, International Planning Studies, 75(1), 2007, pp.70-95.
IBGE (Instituto Brasileiro de Geografía e Estadística) (2007) “Pesquisa Mensal de Emprego”, Serie Relatorios Metodologicos N°23.
IBGE (Instituto Brasileiro de Geografía e Estadística) (2007) “Pesquisa Nacional por Amostra de Domicilios”.
ILO / IPEC (2007) Trabajo Infantil y los programas de transferencias en efectivo condicionadas en América Latina.
ILO (2009) “Bolsa Familia en Brasil: Contexto, Concepto e Impacto”, Geveva.
INDEC (Instituto Nacional de Estadística y Censos) “Mercado de Trabajo: principales indicadores”, Encuesta Permanente de Hogares, INDEC, Argentina.
76
INEC (Instituto Nacional de Estadística y Censos) (2004) “Documento metodológico. Encuesta de Hogares de Propósitos Múltiples”, Área de Censos y Encuestas, INEC, Costa Rica.
Jalan, J. and M. Ravallion (2000) “Is Transient Poverty Different? Evidence for Rural China”, Journal of Development Studies Nº 36.
Jalan, J. and M. Ravallion (2001) “Household Income Dynamics in Rural China”, World Bank Policy Research Working Paper N° 2706.
Jenkins, S. (1999) “Modelling household income dynamics”, ESRC Research Centre on Micro-Social Change, Working Paper 99-1, ISER, University of Essex.
Jenkins, S. and C. Schluter (2001) “Why are child poverty rates higher in Britain than in Germany? A longitudinal perspective”, Journal of Human Resources XXXIV.
Jenkins, S. and J. Rigg. (2001) “The Dynamics of Poverty in Britain, Department for Work and Pensions”, Research Report N° 157.
Justino, P. and J. Litchfield (2003) “Poverty dynamics in rural Vietnam: winners and losers during reform”, Poverty Research Unit at Sussex, Department of Economics, University of Sussex.
Kakwani, N. (1997), “On measuring Growth and Inequality Components of Changes in Poverty with Application to Thailand” Discussion paper 97/16, The University of New South Wales.
Kugler, A. (2000) “The Incidence of Job Security Regulations on Labor Market Flexibility and Compliance in Colombia: Evidence from the 1990 Reform”, IADB, N° R.393.
Lillard, L. and R. Willis (1978) “Dynamic aspects of earnings mobility”, Econometrica 46, 985-1012.
Machado, A. and R. Perez Ribas (2010) “Do changes in the labor market take families out of poverty? Determinants of exiting poverty in Brazilian metropolitan regions”, Journal of Development Studies, Vol. 46, No. 9, 1503-1522.
Mahmoudi, V. (1998) “Growth-equity decomposition of change in poverty: An application to Iran”, University of Essex.
76 77
Maurizio, R. (2009) “Macroeconomic regime, trade openness, unemployment and inequality: the Argentine Experience”, The IDEAs Working Paper Series, Nº 03/2009, International Development Economics Associates, India.
Maurizio, R., B. Perrot and S. Villafañe (2009) “How important were social and labor market policies in reducing poverty across different families in Argentina? A dynamic approach”, Journal of Income Distribution, An International Quarterly, vol. 18, Nº 3-4, septiembre-diciembre 2009, York University, Canadá. ISSN: 0926-6437.
McKernan, S. and C. Ratcliffe (2002) “Events that Trigger Poverty Entries and Exits”, Social Science Quarterly 86 (s1), 1146–1169.
Neilson, C., D. Contreras, R. Cooper and J. Hermann (2008) “The Dynamics of Poverty in Chile”, Journal of Latin American Studies 40, 251–273, Cambridge University Press.
Orshansky, M. (1965) “Counting the Poor: Another Look at the Poverty Profile”, Social Security Bulletin, Vol. 28, 1, pp. 3-29
Paes de Barros, R. and C. Leite Corseuil (1999) “Labor Market Regulations and the Duration of Employment in Brazil”, IPEA Working Paper N° 676.
Paz, J. (2005) “Pobres pobres, cada vez más pobres. Una visión global de la pobreza”, in Mercado de trabajo y Equidad en Argentina, Beccaria, L. and R. Maurizio (eds), Argentina, Prometeo.
Perez Ribas, R. and F. Machado (2007) “Distinguishing chronic poverty from transient poverty in Brazil: developing a model for pseudo-panel data”, International Poverty Centre Working Paper Nº 46, April.
Perez Ribas, R., F. V. Soares and G. Hirata (2008) “The Impact of CCTs. What we know and what we are not sure about”, in International Poverty Centre, Poverty in Focus “Cash Transfers. Lesson from Africa and Latin America”, Brazil.
Perez Ribas, R. and S. Suarez Dillon Soares (2008) “Sobre o painel da Pesquisa Mensal de Emprego (PME) do IBGE”, Texto para Discussao No. 1348, IPEA.
78
Perova, E. and R. Vakis (2009) “Welfare impacts of the “Juntos” Program in Peru: Evidence from a non-experimental evaluation”, The World Bank.
Ravallion, M. (1994) “Poverty Comparisons”, The World Bank, Harwood Academic Publishers, Switzerland.
Rio Group (2006) “Compendium of best practices in poverty measurement”, Expert Group in poverty statistics, Rio de Janeiro.
Román, I. (2010) “Sustentabilidad de los programas de transferencias condicionadas: la experiencia del Instituto Mixto de Ayuda Social y “Avancemos” en Costa Rica”, Serie Políticas Sociales, Social Development Division, ECLAC, Santiago de Chile.
Ruggles, P. and R. Williams (1987) “Transitions In and Out of Poverty: New Data from the Survey of Income and Program Participation”, U.S. Department of Commerce, Bureau of the Census, SIPP Working Paper N° 8716, Washington, DC.
Ruggles, P. and R. Williams (1989) “Longitudinal measures of poverty: accounting for income and assets over time”, Review of Income and Wealth, 353.
Ruggles, P. (1990) Drawing the line: alternative poverty measures and their implications for public policy, The Urban Institute Press, Washington.
Sen, A. (1983) “Poor, relatively speaking”, Oxford Economic Papers, Vol. 35, pp.153-169
Sen, A. (1985) “A sociological approach to the measurement of poverty: a reply to professor Peter Townsend”, Oxford Economic Papers, Vol 37, pp 669-676
Shorrocks, A. (1999) “Decomposition Procedures for Distributional Analysis: A Unified Framework Based on the Shapley Value”, Department of Economics, University of Essex
Slon, P. and E. Zúñiga (2004) “Aspectos dinámicos de la pobreza en Costa Rica”, Thesis, San José, Escuela de Economía, Universidad de Costa Rica.
Slon, P. and E. Zúñiga (2006) “Poverty dynamics in Costa Rica with panel data from cross-sections”, ECLAC Review No. 89, August.
78 79
Stallings, B. and W. Peres (2002) “Crecimiento, empleo y equidad. El impacto de las reformas económicas en América Latina y el Caribe”, ECLAC, Santiago de Chile.
Stevens, A. (1999) “Climbing out of poverty, falling back in. Measuring the persistence of poverty over multiple spells”, Journal of Human Resources, XXXIV: 557-588.
Stewart, M. and J. Swaffield (1999) “Low pay dynamics and transition probabilities”, Economica, 66, 23–42.
UNICEF (2005) “Pobreza infantil en países ricos”, Report Card No. 6, Instituto de Investigaciones Innocenti, Italia.
Veras Soares, F., S. Soares, M. Medeiros and R. Guerreiro Osório (2006) “Cash transfer programmes in Brazil: Impacts on inequality and poverty”, Working Paper number 21, International Poverty Centre, UNDP/IPEA.
Villatoro, P. (2007) “Las transferencias condicionadas en América Latina: luces y sombras”, CEPAL, Chile.
Villatoro, P. (2008) “CCTs in Latin America: Human Capital Accumulation and Poverty Reduction”, International Poverty Centre, Poverty in Focus “Cash Transfers. Lesson from Africa an Latin America”, Brazil.
Whelan Ch. and B. Maître (2006) “Comparing Poverty and Deprivation Dynamics: Issues of Reliability and Validity”, The Journal of Economic Inequality, Volume 4, No. 3, 303-323.
Yacub, S. (2000) Poverty dynamics in developing countries, Brighton, Sussex, England: Institute of Development Studies.
Zepeda, E., D. Alarcón, F. Veras Soares and R. Guerreiro Osório (2007) “Growth, poverty and employment in Brazil, Chile and Mexico”, Working Paper No. 42, International Poverty Center, December.
Zepeda, E., D. Alarcón, F. Veras Soares and R. Guerreiro Osório (2009) “Changes in earnings in Brazil, Chile, and Mexico: Disentangling the forces behind pro-poor change in labor markets”, Working Paper Nº 51, International Poverty Center, December.
80
ANNEX A
Characteristics of the data
Households 2003-2004 2004-2005 2005-2006households in panels 6,536 6,942 6,802households with valid income 6,047 6,617 6,573employed with valid occupational category 6,047 6,617 6,573employed with valid hours 5,808 6,345 6,330
Individuals 2003-2004 2004-2005 2005-2006individuals in panels 19,415 20,258 19,456individuals with valid income 17,796 19,227 18,750employed with valid occupational category 17,796 19,227 18,750employed with valid hours 16,968 18,329 17,998Source: Authors’ elaboration based on data PME & PNAD (IBGE)
Table 1.B. (cont.) Pool of panels.Brazil. Urban areas. Period 2003-2006.
Households 2003 2004 2005 2006Total 30,257 30,709 31,107 32,038households with valid income 28,523 29,910 30,315 31,522employed with valid occupational category 28,522 29,910 30,315 31,522employed with valid hours 27,981 29,157 29,657 30,797
Individuals 2003 2004 2005 2006Total 100,459 100,683 100,817 102,916individuals with valid income 93,859 97,922 97,858 101,212employed with valid occupational category 93,857 97,922 97,858 101,212employed with valid hours 91,781 95,103 95,449 98,546Source: Authors’ elaboration based on data PME & PNAD (IBGE)
Table 1.B. Cross section.Brazil. Urban areas. Period 2003-2006.
Households 22003-22004 12004-12005 22004-2005 12005-12006 22005-22006households in panels 9,983 9,945 10,293 9,983 10,068households with valid income 6,933 7,001 7,623 7,686 7,760employed with valid occupational category 6,890 6,945 7,577 7,611 7,702employed with valid hours 6,286 6,064 6,969 6,638 7,068
Individuals 22003-22004 12004-12005 22004-2005 12005-12006 22005-22006individuals in panels 35,769 34,648 35,416 34,842 35,090households with valid income 24,525 24,251 26,060 26,836 26,917employed with valid occupational category 24,332 24,008 25,883 26,501 26,678employed with valid hours 21,753 20,516 23,521 22,564 24,078Source: Authors’ elaboration based on data from Encuesta Permanente de Hogares (EPH-INDEC).
Table 1.A. (cont.) Pool of panels. Argentina. Urban areas. Period 2003-2006.
Households 2nd. 2003 1st. 2004 2nd. 2004 1st. 2005 2nd. 2005 1st. 2006 2nd. 2006Total 26,502 26,625 27,303 27,108 27,511 27,277 37,521households with valid income 20,931 21,364 22,305 22,623 23,205 23,444 32,371employed with valid occupational category 20,923 21,364 22,165 22,437 23,044 23,230 32,133employed with valid hours 20,151 20,071 20,940 20,515 21,802 21,252 30,456
Individuals 2nd. 2003 1st. 2004 2nd. 2004 1st. 2005 2nd. 2005 1st. 2006 2nd. 2006Total 93,244 93,158 94,772 93,942 94,813 93,558 129,410individuals with valid income 73,091 74,485 76,812 78,401 79,715 80,176 111,217employed with valid occupational category 73,058 74,485 76,205 77,647 79,068 79,232 110,261employed with valid hours 69,894 69,338 71,023 69,795 73,952 71,263 103,327Source: Authors’ elaboration based on data from Encuesta Permanente de Hogares (EPH-INDEC).
Table 1.A. Cross section.Argentina. Urban areas. Period 2003-2006.
80 81
Households 7 2006 - 7 2007 72007 - 72008households in panels 2,616 2,660households with valid income 2,564 2,583employed with valid occupational category 2,564 2,583employed with valid hours 2,531 2,551
Individuals 7 2006 - 72007 7 2007 - 7 2008individuals in panels 9,278 9,360individuals with valid income 9,084 9,090employed with valid occupational category 9,084 9,090employed with valid hours 8,976 8,956
Source: Authors’ elaboration based on data from Encuesta de Hogares de Propósitos Múltiples (INEC).
Table 1.C. (cont.) Pool of panelsCosta Rica. Urban areas. Period 2006-2008.
Households 7 2006 7 2007 7 2008Total 11,991 12,361 12,531households with valid income 11,873 12,196 12,331employed with valid occupational category 11,873 12,196 12,331employed with valid hours 11,799 12,107 12,239
Individuals 7 2006 7 2007 7 2008Total 45,139 46,278 46,101individuals with valid income 44,699 45,645 45,362employed with valid occupational category 44,699 45,645 45,362employed with valid hours 44,390 45,250 44,986
Table 1.C. Costa Rica. Cross sectionCosta Rica. Urban areas. Period 2006-2008.
Source: Authors’ elaboration based on data from Encuesta de Hogares de Propósitos Múltiples (INEC).
82
Hou
seho
lds
3200
4-32
005
6200
4-62
005
9200
5-92
006
1220
05-1
2200
632
006-
3200
762
006-
6200
792
007-
9200
812
2007
-122
008
hous
ehol
ds in
pan
els
2,51
92,
683
2,74
56,
592
2,27
92,
323
2,80
47,
731
hous
ehol
ds w
ith v
alid
inco
me
2,33
22,
457
2,68
46,
312
2,00
32,
116
2,61
87,
072
empl
oyed
with
val
id o
ccup
atio
nal c
ateg
ory
2,32
72,
452
2,68
46,
285
1,99
72,
010
2,60
97,
033
empl
oyed
with
val
id h
ours
2,
260
2,40
02,
609
6,17
91,
941
1,97
52,
550
6,94
8
Indi
vidu
als
3200
4-32
005
6200
4-62
005
9200
5-92
006
1220
05-1
2200
632
006-
3200
762
006-
6200
792
007-
9200
812
2007
-122
008
indi
vidu
als
in p
anel
s9,
761
10,3
9310
,471
25,6
198,
800
8,77
810
,370
29,2
33in
divi
dual
s w
ith v
alid
inco
me
8,91
29,
378
10,2
0224
,374
7,53
97,
921
9,56
526
,494
empl
oyed
with
val
id o
ccup
atio
nal c
ateg
ory
8,89
19,
360
10,2
0224
,240
7,51
17,
476
9,52
026
,317
empl
oyed
with
val
id h
ours
8,
598
9,14
59,
856
23,7
387,
248
7,32
69,
264
25,9
36So
urce
: Aut
hors
’ ela
bora
tion
base
d on
dat
a fro
m E
ncue
sta
de E
mpl
eo, D
esem
pleo
y S
ubem
pleo
(IN
EC).
Tabl
e 1.
D. (
cont
.) Po
ol o
f pan
els.
Ecua
dor.
Urb
an a
reas
. Per
iod
2004
-200
8.
hous
ehol
ds in
pa
nels
hous
ehol
ds
with
val
id
inco
me
empl
oyed
with
va
lid
occu
patio
nal
empl
oyed
with
va
lid h
ours
in
divi
dual
s in
pa
nels
indi
vidu
als
with
va
lid in
com
eem
ploy
ed w
ith
valid
oc
cupa
tiona
l
empl
oyed
with
va
lid h
ours
32
004
1905
918
083
1802
917
855
8193
076
993
7670
675
797
6200
458
5554
0653
9153
5324
304
2199
721
931
2172
832
005
5876
5651
5645
5549
2428
523
089
2306
322
606
6200
559
3757
5957
5956
6524
424
2354
023
540
2307
992
005
5851
5742
5742
5646
2410
523
541
2354
123
058
1220
0518
357
1765
417
599
1744
277
050
7369
073
403
7258
832
006
5748
5200
5190
5087
2367
620
897
2084
520
357
9200
658
4057
8757
8757
0723
867
2361
623
616
2318
562
006
5721
5521
5518
5453
2344
022
450
2243
622
081
1220
0618
484
1822
318
193
1806
477
964
7664
476
483
7580
832
007
5808
5381
5378
5302
2361
421
414
2139
620
967
6200
766
5361
4857
6656
9426
774
2435
522
564
2220
392
007
6620
6232
6217
6116
2618
024
388
2430
223
802
1220
0718
933
1741
117
354
1724
976
922
7003
769
756
6925
732
008
6624
6025
6012
5945
2616
123
447
2337
923
056
6200
891
7488
1687
8987
1537
869
3617
636
039
3566
592
008
6623
6495
6478
6419
2633
925
783
3771
925
398
1220
0819
394
1910
019
061
1891
078
742
7734
577
143
7640
3S
ourc
e: A
utho
rs’ e
labo
ratio
n ba
sed
on d
ata
from
Enc
uest
a de
Em
pleo
, Des
empl
eo y
Sub
empl
eo (I
NE
C).
Hou
seho
lds
Indi
vidu
als
Tabl
e 1.
D. C
ross
sec
tion.
Ecu
ador
. Urb
an a
reas
. Per
iod
2004
-200
8.
82 83
Households 2002 2003 2004 2005 2006Total 18,598 18,912 19,502 19,895 20,577households with valid income 18,597 18,911 19,502 19,895 20,575employed with valid occupational category 18,597 18,911 19,502 19,895 20,575employed with valid hours 18,241 18,304 18,320 18,626 19,471
Individuals 2002 2003 2004 2005 2006Total 83,529 85,364 87,697 87,876 90,440individuals with valid income 83,528 85,363 87,697 87,876 90,438employed with valid occupational category 83,528 85,363 87,697 87,876 90,438employed with valid hours 81,813 82,320 81,242 81,087 84,591Source: Authors’ elaboration based on data from Encuesta Nacional de Hogares (ENAHO-INEI).
Table 1.E. Cross section.Peru. Urban areas. Period 2002- 2006.
Households 2002-2003 2003-2004 2004-2005 2005-2006households in panels 3,206 2,883 4,301 4,657households with valid income 3,158 2,844 4,244 4,579employed with valid occupational category 3,158 2,844 4,244 4,579employed with valid hours 3,026 2,726 4,038 4,337
Individuals 2002-2003 2003-2004 2004-2005 2005-2006individuals in panels 14,716 13,441 19,634 20,971households with valid income 14,567 13,311 19,391 20,711employed with valid occupational category 14,567 13,311 19,391 20,711employed with valid hours 13,949 12,729 18,383 19,455Source: Authors’ elaboration based on data from Encuesta Nacional de Hogares (ENAHO-INEI).
Table 1.E. (cont.) Pool of panels Peru. Urban areas. Period 2002- 2006.
84
ANNEX B
Shapley decomposition and household characteristics
Estimate STD % Estimate STD % Growth (real) -0.127413 0.006653 70% Growth (real) -0.07196 0.008542 99% Growth (Nominal) -0.22587 125% Growth (Nominal)-0.100311 151% Inflation 0.098457 -54% Inflation 0.028351 -43%Redistribution -0.053631 0.007343 30% Redistribution 0.001 0.0114 1%Difference: (d2-d1) -0.181002 0.00673 100% Difference: (d2-d1)-0.066308 0.009355 100%
Estimate STD % Estimate STD % Growth (real) -0.027233 0.002889 58% Growth (real) -0.060834 0.022672 63% Growth (Nominal) -0.099111 212% Growth (Nominal)-0.144577 150% Inflation 0.071878 -154% Inflation 0.083743 -87%Redistribution -0.019481 0.003916 42% Redistribution -0.036457 0.024023 38%Difference: (d2-d1) -0.046714 0.002845 100% Difference: (d2-d1)-0.096475 0.006747 100%
Estimate STD % Growth (real) -0.016118 0.021052 62% Growth (Nominal)-0.0582720 92% Inflation 0.0421540 -66%Redistribution -0.047688 0.02524 38%Difference: (d2-d1)-0.063643 0.010123 100%
Ecuador. Urban areas. Period 2004-2008.
Source: Authors’ elaboration based on data from Encuesta de Empleo, Desempleo y Subempleo (INEC).
Table 2.E. Shapley value decompositionPeru. Urban areas. Period 2002-2006.
Table 2.C. Shapley value decompositionCosta Rica. Urban areas. Period 2006-2008.
Source: Authors’ elaboration based on data from Encuesta de Hogares de Propósitos Múltiples (INEC).
Table 2.D. Shapley value decomposition
Table 2.A. Shapley value decompositionArgentina. Urban areas. Period 2003-2006.
Source: Authors’ elaboration based on data from Encuesta Permanente de Hogares (EPH-INDEC).
Table 2.B. Shapley value decompositionBrazil. Urban areas. Period 2003-2006.
Source: Authors’ elaboration based on data PME & PNAD (IBGE)
Source: Authors’ elaboration based on data from Encuesta Nacional de Hogares (ENEI)
84 85
CH
ARAC
TER
ISTI
CS
OF
THE
HO
USE
HO
LD H
EAD
SC
ompo
sitio
n of
tota
l ho
useh
olds
Spec
ific
pove
rty
rate
Com
posi
tion
of p
oor
hous
ehol
dsC
ompo
sitio
n of
tota
l ho
useh
olds
Spec
ific
pove
rty
rate
Com
posi
tion
of p
oor
hous
ehol
dsC
ompo
sitio
n of
tota
l ho
useh
olds
Spec
ific
pove
rty
rate
Com
posi
tion
of p
oor
hous
ehol
ds
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
Educ
atio
nal l
evel
Low
42%
42%
62%
45%
53%
59%
40%
32%
64%
Med
ium
33%
28%
31%
33%
39%
32%
34%
17%
29%
Hig
h24
%10
%8%
23%
15%
8%26
%5%
7%A
ge24
yea
rs o
ld o
r you
nger
5%32
%5%
5%42
%5%
4%23
%5%
25 -
40 y
ears
old
29%
34%
33%
28%
45%
32%
30%
23%
35%
41 -
65 y
ears
old
45%
33%
49%
44%
43%
49%
46%
22%
50%
Old
er th
an 6
5 ye
ars
old
21%
18%
12%
23%
25%
14%
20%
11%
11%
Gen
der
Mal
e68
%30
%69
%68
%41
%71
%69
%20
%68
%Fe
mal
e32
%28
%31
%32
%36
%29
%31
%21
%32
%
OC
CU
PATI
ON
AL V
ARIA
BLE
SC
ompo
sitio
n of
tota
l ho
useh
olds
Spec
ific
pove
rty
rate
Com
posi
tion
of p
oor
hous
ehol
dsC
ompo
sitio
n of
tota
l ho
useh
olds
Spec
ific
pove
rty
rate
Com
posi
tion
of p
oor
hous
ehol
dsC
ompo
sitio
n of
tota
l ho
useh
olds
Spec
ific
pove
rty
rate
Com
posi
tion
of p
oor
hous
ehol
ds (1
)(2
)(3
)(4
)(5
)(6
)(7
)(8
)(9
)O
ccup
atio
nal c
ateg
ory
of th
e ho
useh
old
head
Reg
iste
red
wag
e ea
rner
31%
16%
15%
29%
23%
17%
33%
8%13
%N
on-r
egis
tere
d w
age
earn
er19
%46
%30
%19
%60
%30
%18
%32
%29
%O
wn
acco
unt
15%
36%
19%
14%
46%
17%
15%
27%
21%
Empl
oyer
3%11
%1%
2%15
%1%
4%6%
1%U
nem
ploy
ed5%
64%
11%
7%73
%12
%3%
54%
10%
Inac
tive
with
pen
sion
20%
15%
10%
22%
21%
11%
19%
9%8%
Inac
tive
with
out p
ensi
on8%
52%
14%
7%60
%11
%8%
44%
17%
Hou
rs w
orke
d by
the
hous
ehol
d he
adU
nder
empl
oyed
28%
43%
42%
31%
57%
45%
25%
29%
39%
Full
empl
oyed
27%
24%
22%
26%
31%
20%
29%
17%
23%
Ove
r em
ploy
ed44
%30
%36
%43
%33
%35
%46
%28
%37
%
HO
USE
HO
LD C
OM
POSI
TIO
NC
ompo
sitio
n of
tota
l ho
useh
olds
Spec
ific
pove
rty
rate
Com
posi
tion
of p
oor
hous
ehol
dsC
ompo
sitio
n of
tota
l ho
useh
olds
Spec
ific
pove
rty
rate
Com
posi
tion
of p
oor
hous
ehol
dsC
ompo
sitio
n of
tota
l ho
useh
olds
Spec
ific
pove
rty
rate
Com
posi
tion
of p
oor
hous
ehol
ds (1
)(2
)(3
)(4
)(5
)(6
)(7
)(8
)(9
)H
ouse
hold
s w
ithou
t chi
ldre
n53
%16
%27
%53
%22
%29
%53
%10
%25
%O
ne p
erso
n. M
ale
olde
r tha
n 65
yea
rs.
2%15
%1%
2%18
%1%
2%12
%1%
One
per
son.
Fem
ale
olde
r tha
n 65
yea
rs.
6%6%
1%6%
8%1%
5%3%
1%O
ne p
erso
n. M
ale
youn
ger t
han
65 y
ears
.4%
17%
2%4%
22%
2%4%
12%
3%O
ne p
erso
n. F
emal
e yo
unge
r tha
n 65
yea
rs.
5%9%
2%5%
12%
2%5%
7%2%
Com
plet
e nu
clea
r hou
seho
ld. H
ead
olde
r tha
n 65
yea
rs.
9%17
%5%
9%26
%6%
9%8%
4%C
ompl
ete
nucl
ear h
ouse
hold
. Hea
d yo
unge
r tha
n 65
yea
rs.
16%
18%
9%15
%26
%10
%17
%9%
8%In
com
plet
e nu
clea
r hou
seho
ld. H
ead
olde
r tha
n 65
yea
rs.
4%21
%2%
4%30
%3%
4%12
%2%
Inco
mpl
ete
nucl
ear h
ouse
hold
. Hea
d yo
unge
r tha
n 65
yea
rs.
8%21
%5%
7%27
%5%
8%15
%6%
Hou
seho
lds
with
chi
ldre
n47
%45
%73
%47
%59
%71
%47
%32
%75
%In
com
plet
e nu
clea
r hou
seho
ld. M
ale
hous
ehol
d he
ad.
2%48
%3%
2%60
%2%
2%37
%3%
Inco
mpl
ete
nucl
ear h
ouse
hold
. Fem
ale
hous
ehol
d he
ad.
9%55
%18
%10
%68
%17
%9%
41%
19%
Com
plet
e nu
clea
r hou
seho
ld36
%43
%52
%36
%56
%51
%36
%29
%52
%
POVE
RTY
RA
TE (H
ouse
hold
s)
Sour
ce: A
utho
rs’ e
labo
ratio
n ba
sed
on d
ata
from
Enc
uest
a Pe
rman
ente
de
Hog
ares
(EPH
-IND
EC).
Tabl
e 3.
A. H
ouse
hold
com
posi
tion
and
spec
ific
pove
rty
rate
s. C
ross
sec
tion.
Ave
rage
200
3-20
0620
0320
06
Arg
entin
a. U
rban
are
as. 2
003-
2006
Ave
rage
200
3-20
0620
0320
06
Ave
rage
200
3-20
0620
0320
06
30%
39%
20%
86
CH
AR
AC
TE
RIS
TIC
S O
F T
HE
HO
US
EH
OL
D
HE
AD
S
Co
mp
os
itio
n
of
tota
l h
ou
se
ho
lds
Sp
ec
ific
p
ove
rty
rate
Co
mp
os
itio
n
of
po
or
ho
us
eh
old
s
Co
mp
os
itio
n
of
tota
l h
ou
se
ho
lds
Sp
ec
ific
p
ove
rty
rate
Co
mp
os
itio
n
of
po
or
ho
us
eh
old
s
Co
mp
os
itio
n
of
tota
l h
ou
se
ho
lds
Sp
ec
ific
p
ove
rty
rate
Co
mp
os
itio
n
of
po
or
ho
us
eh
old
s (
1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
Ed
uc
ati
on
al
leve
lL
ow
5
2%
43
%7
0%
54
%4
5%
70
%5
0%
40
%6
9%
Me
diu
m3
2%
26
%2
6%
31
%2
9%
25
%3
3%
24
%2
7%
Hig
h1
6%
8%
4%
16
%1
0%
4%
17
%6
%3
%A
ge
24
ye
ars
old
or
you
nge
r4
%4
6%
5%
4%
50
%6
%3
%4
1%
5%
25
- 4
0 y
ea
rs o
ld3
2%
40
%4
0%
33
%4
2%
40
%3
1%
38
%4
0%
41
- 6
5 y
ea
rs o
ld4
9%
29
%4
6%
49
%3
2%
45
%5
0%
27
%4
6%
Old
er
tha
n 6
5 y
ea
rs o
ld1
5%
19
%9
%1
4%
22
%9
%1
6%
16
%9
%G
en
de
rM
ale
65
%3
1%
64
%6
7%
34
%6
6%
63
%2
9%
61
%F
em
ale
35
%3
3%
36
%3
3%
36
%3
4%
37
%3
0%
39
%
OC
CU
PA
TIO
NA
L V
AR
IAB
LE
SC
om
po
sit
ion
o
f to
tal
ho
us
eh
old
s
Sp
ec
ific
p
ove
rty
rate
Co
mp
os
itio
n
of
po
or
ho
us
eh
old
s
Co
mp
os
itio
n
of
tota
l h
ou
se
ho
lds
Sp
ec
ific
p
ove
rty
rate
Co
mp
os
itio
n
of
po
or
ho
us
eh
old
s
Co
mp
os
itio
n
of
tota
l h
ou
se
ho
lds
Sp
ec
ific
p
ove
rty
rate
Co
mp
os
itio
n
of
po
or
ho
us
eh
old
s (
1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
Oc
cu
pa
tio
na
l c
ate
go
ry o
f th
e h
ou
se
ho
ld h
ea
dR
egis
tere
d w
age
ea
rne
r3
4%
24
%2
6%
34
%2
6%
26
%3
4%
22
%2
6%
No
n-r
egis
tere
d w
age
ea
rne
r1
1%
38
%1
4%
11
%4
1%
14
%1
1%
36
%1
4%
Ow
n a
cco
un
t1
5%
35
%1
7%
16
%3
7%
17
%1
5%
32
%1
7%
Em
plo
yer
4%
9%
1%
4%
11
%1
%5
%8
%1
%U
ne
mp
loye
d4
%7
2%
10
%5
%7
5%
11
%4
%6
8%
9%
Ina
ctiv
e w
ith p
en
sio
n2
2%
17
%1
2%
22
%1
9%
12
%2
3%
15
%1
2%
Ina
ctiv
e w
itho
ut
pe
nsi
on
8%
76
%2
0%
8%
78
%1
9%
9%
73
%2
2%
Ho
urs
wo
rke
d b
y th
e h
ou
se
ho
ld h
ea
dU
nd
ere
mp
loye
d1
5%
33
%1
7%
13
%3
7%
15
%1
7%
28
%1
8%
Fu
ll e
mp
loye
d4
9%
26
%4
5%
50
%2
9%
46
%4
9%
23
%4
4%
Ove
r e
mp
loye
d3
6%
31
%3
9%
38
%3
3%
39
%3
4%
29
%3
8%
HO
US
EH
OL
D C
OM
PO
SIT
ION
Co
mp
os
itio
n
of
tota
l h
ou
se
ho
lds
Sp
ec
ific
p
ove
rty
rate
Co
mp
os
itio
n
of
po
or
ho
us
eh
old
s
Co
mp
os
itio
n
of
tota
l h
ou
se
ho
lds
Sp
ec
ific
p
ove
rty
rate
Co
mp
os
itio
n
of
po
or
ho
us
eh
old
s
Co
mp
os
itio
n
of
tota
l h
ou
se
ho
lds
Sp
ec
ific
p
ove
rty
rate
Co
mp
os
itio
n
of
po
or
ho
us
eh
old
s (
1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
Fa
mil
ies
wit
ho
ut
ch
ild
ren
49
%1
8%
27
%4
7%
19
%2
6%
51
%1
6%
28
%O
ne
-pe
rso
n h
ou
seh
old
s. M
en
old
er
tha
n 6
5 y
ea
r o
ld1
%1
3%
0%
1%
14
%0
%1
%1
3%
1%
On
e-p
ers
on
ho
use
ho
lds.
Fe
ma
le o
lde
r th
an
65
ye
ar
3%
17
%2
%3
%1
9%
2%
4%
15
%2
%O
ne
-pe
rso
n h
ou
seh
old
s. M
en
yo
un
ge
r th
an
65
ye
ar
5%
19
%3
%5
%2
1%
3%
5%
18
%3
%O
ne
-pe
rso
n h
ou
seh
old
s. F
em
ale
yo
un
ge
r th
an
65
5
%1
8%
3%
5%
19
%3
%5
%1
7%
3%
Co
mp
lete
nu
cle
ar
ho
use
ho
ld.
He
ad
old
er
tha
n 6
5
5%
14
%2
%5
%1
7%
3%
6%
11
%2
%C
om
ple
te n
ucl
ea
r h
ou
seh
old
. H
ea
d y
ou
nge
r th
an
65
1
8%
17
%1
0%
18
%1
8%
9%
18
%1
6%
10
%In
com
ple
te n
ucl
ea
r h
ou
seh
old
. H
ea
d o
lde
r th
an
65
3
%1
8%
2%
3%
21
%2
%4
%1
6%
2%
Inco
mp
lete
nu
cle
ar
ho
use
ho
ld.
He
ad
yo
un
ge
r th
an
8
%2
0%
5%
8%
22
%5
%9
%1
9%
6%
Fa
mil
ies
wit
h c
hil
dre
n5
1%
46
%7
3%
53
%4
9%
74
%4
9%
42
%7
2%
Inco
mp
lete
nu
cle
ar
ho
use
ho
ld .
Ma
le h
ou
seh
old
h
ea
d2
%4
7%
2%
2%
49
%2
%2
%4
4%
2%
Inco
mp
lete
nu
cle
ar
ho
use
ho
ld .
Fe
ma
le h
ou
seh
old
h
ea
d1
1%
51
%1
8%
11
%5
4%
17
%1
1%
49
%1
9%
Co
mp
lete
nu
cle
ar
ho
use
ho
ld
38
%4
4%
52
%4
0%
47
%5
4%
37
%4
0%
51
%
PO
VE
RT
Y R
AT
E (
Ho
us
eh
old
s)
31
.9%
34
.6%
29
.2%
So
urc
e:
Au
tho
rs’ e
lab
ora
tion
ba
sed
on
da
ta P
ME
& P
NA
D (
IBG
E)
Ta
ble
3.B
. H
ou
se
ho
ld c
om
po
sit
ion
an
d s
pe
cif
ic p
ove
rty
rate
s.
Cro
ss
se
cti
on
.B
razi
l. U
rba
n a
rea
s.
Pe
rio
d 2
00
3-2
00
6.
Ave
rag
e 2
00
3-2
00
6S
ep
. 2
00
3S
ep
. 2
00
6
Ave
rag
e 2
00
3-2
00
6S
ep
. 2
00
3S
ep
. 2
00
6
Ave
rag
e 2
00
3-2
00
6S
ep
. 2
00
3S
ep
. 2
00
6
86 87
CH
ARAC
TER
ISTI
CS
OF
THE
HO
USE
HO
LD H
EAD
SC
ompo
sitio
n of
tota
l ho
useh
olds
Spec
ific
pove
rty
rate
Com
posi
tion
of p
oor
hous
ehol
dsC
ompo
sitio
n of
tota
l ho
useh
olds
Spec
ific
pove
rty
rate
Com
posi
tion
of p
oor
hous
ehol
dsC
ompo
sitio
n of
tota
l ho
useh
olds
Spec
ific
pove
rty
rate
Com
posi
tion
of p
oor
hous
ehol
ds
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
Educ
atio
nal l
evel
Com
plet
e pr
imar
y or
less
42%
38%
63%
43%
43%
65%
41%
32%
61%
Seco
ndar
y32
%21
%27
%33
%23
%27
%32
%19
%28
%U
nive
rsity
or m
ore
26%
9%10
%25
%10
%8%
27%
9%11
%A
ge24
yea
rs o
ld o
r you
nger
3%23
%3%
3%28
%3%
3%19
%3%
25 -
40 y
ears
old
29%
25%
29%
30%
29%
30%
27%
22%
27%
41 -
65 y
ears
old
54%
23%
50%
53%
27%
49%
54%
20%
50%
Old
er th
an 6
5 ye
ars
old
14%
32%
18%
14%
35%
17%
15%
29%
20%
Gen
der
Mal
e68
%23
%61
%68
%27
%64
%67
%19
%59
%Fe
mal
e32
%30
%39
%32
%32
%36
%33
%28
%41
%
OC
CU
PATI
ON
AL V
ARIA
BLE
SC
ompo
sitio
n of
tota
l ho
useh
olds
Spec
ific
pove
rty
rate
Com
posi
tion
of p
oor
hous
ehol
dsC
ompo
sitio
n of
tota
l ho
useh
olds
Spec
ific
pove
rty
rate
Com
posi
tion
of p
oor
hous
ehol
dsC
ompo
sitio
n of
tota
l ho
useh
olds
Spec
ific
pove
rty
rate
Com
posi
tion
of p
oor
hous
ehol
ds (1
)(2
)(3
)(4
)(5
)(6
)(7
)(8
)(9
)O
ccup
atio
nal c
ateg
ory
of th
e ho
useh
old
head
Reg
iste
red
wag
e ea
rner
40%
17%
26%
41%
20%
28%
40%
13%
24%
Non
-reg
iste
red
wag
e ea
rner
11%
32%
14%
11%
34%
13%
11%
30%
15%
Ow
n ac
coun
t16
%31
%21
%17
%34
%20
%16
%29
%21
%Em
ploy
er8%
16%
5%8%
19%
5%8%
14%
5%U
nem
ploy
ed2%
61%
4%2%
73%
4%1%
49%
3%In
activ
e w
ith p
ensi
on12
%20
%9%
12%
24%
10%
13%
15%
9%In
activ
e w
ithou
t pen
sion
10%
51%
21%
10%
55%
19%
10%
48%
22%
Hou
rs w
orke
d by
the
hous
ehol
d he
adU
nder
empl
oyed
12%
32%
18%
13%
34%
18%
12%
30%
19%
Full
empl
oyed
16%
23%
14%
15%
19%
12%
17%
27%
17%
Ove
r em
ploy
ed72
%64
%68
%72
%25
%71
%71
%10
4%64
%
HO
USE
HO
LD C
OM
POSI
TIO
NC
ompo
sitio
n of
tota
l ho
useh
olds
Spec
ific
pove
rty
rate
Com
posi
tion
of p
oor
hous
ehol
dsC
ompo
sitio
n of
tota
l ho
useh
olds
Spec
ific
pove
rty
rate
Com
posi
tion
of p
oor
hous
ehol
dsC
ompo
sitio
n of
tota
l ho
useh
olds
Spec
ific
pove
rty
rate
Com
posi
tion
of p
oor
hous
ehol
ds (1
)(2
)(3
)(4
)(5
)(6
)(7
)(8
)(9
)Fa
mili
es w
ithou
t chi
ldre
n41
%19
%31
%40
%20
%28
%43
%18
%34
%O
ne-p
erso
n ho
useh
olds
. Men
old
er th
an 6
5 ye
ars
1%33
%1%
1%40
%1%
1%26
%1%
One
-per
son
hous
ehol
ds. F
emal
e ho
useh
old
head
old
er th
an 6
5 ye
ars
2%34
%3%
2%36
%2%
2%31
%3%
One
-per
son
hous
ehol
ds. M
en y
oung
er th
an 6
5 ye
ars
3%13
%2%
3%12
%1%
3%14
%2%
One
-per
son
hous
ehol
ds. F
emal
e ho
useh
old
head
you
nger
than
65
2%17
%2%
2%18
%1%
3%16
%2%
Com
plet
e nu
clea
r hou
seho
ld. H
ead
olde
r tha
n 65
yea
rs6%
28%
7%6%
31%
6%7%
24%
7%C
ompl
ete
nucl
ear h
ouse
hold
. Hea
d yo
unge
r tha
n 65
yea
rs15
%12
%7%
14%
12%
6%16
%12
%9%
Inco
mpl
ete
nucl
ear h
ouse
hold
. Hea
d ol
der t
han
65 y
ears
4%27
%4%
4%29
%4%
4%26
%4%
Inco
mpl
ete
nucl
ear h
ouse
hold
. Hea
d yo
unge
r tha
n 65
yea
rs8%
18%
6%8%
19%
5%8%
18%
7%Fa
mili
es w
ith c
hild
ren
59%
30%
69%
60%
34%
72%
57%
25%
66%
Inco
mpl
ete
nucl
ear h
ouse
hold
. M
ale
hous
ehol
d he
ad1%
31%
2%1%
34%
2%1%
27%
2%In
com
plet
e nu
clea
r hou
seho
ld .
Fem
ale
hous
ehol
d he
ad15
%37
%22
%15
%40
%21
%15
%35
%23
%C
ompl
ete
nucl
ear h
ouse
hold
42
%27
%45
%43
%32
%49
%41
%22
%41
%
POVE
RTY
RA
TE (H
ouse
hold
s)
Sour
ce: A
utho
rs’ e
labo
ratio
n ba
sed
on d
ata
from
Enc
uest
a de
Hog
ares
de
Prop
ósito
s Múl
tiple
s (IN
EC).
2006
2008
Tab
le 3
.C. H
ouse
hold
com
posi
tion
and
spec
ific
pove
rty
rate
s. C
ross
sec
tion.
Ave
rage
200
6-20
0820
0620
08
25%
29%
22%
Cos
ta R
ica.
Urb
an a
reas
. 200
6-20
08.
Ave
rage
200
6-20
0820
0620
08
Ave
rage
200
6-20
08
88
CH
ARAC
TER
ISTI
CS
OF
THE
HO
USE
HO
LD H
EAD
SC
ompo
sitio
n of
tota
l ho
useh
olds
Spec
ific
pove
rty
rate
Com
posi
tion
of p
oor
hous
ehol
dsC
ompo
sitio
n of
tota
l ho
useh
olds
Spec
ific
pove
rty
rate
Com
posi
tion
of p
oor
hous
ehol
dsC
ompo
sitio
n of
tota
l ho
useh
olds
Spec
ific
pove
rty
rate
Com
posi
tion
of p
oor
hous
ehol
ds
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
Educ
atio
nal l
evel
Com
plet
e pr
imar
y or
less
41%
51%
57%
42%
59%
57%
41%
43%
57%
Seco
ndar
y35
%37
%35
%35
%44
%35
%35
%31
%35
%U
nive
rsity
or m
ore
24%
13%
8%23
%16
%8%
25%
10%
8%A
ge24
yea
rs o
ld o
r you
nger
4%39
%4%
5%49
%5%
3%30
%3%
25 -
40 y
ears
old
30%
43%
34%
34%
48%
36%
27%
37%
32%
41 -
65 y
ears
old
51%
36%
48%
48%
43%
46%
53%
30%
50%
Old
er th
an 6
5 ye
ars
old
15%
36%
14%
14%
43%
13%
17%
28%
15%
Gen
der
Mal
e75
%38
%74
%77
%44
%76
%73
%32
%72
%Fe
mal
e25
%39
%26
%23
%47
%24
%27
%32
%28
%
OC
CU
PATI
ON
AL V
ARIA
BLE
SC
ompo
sitio
n of
tota
l ho
useh
olds
Spec
ific
pove
rty
rate
Com
posi
tion
of p
oor
hous
ehol
dsC
ompo
sitio
n of
tota
l ho
useh
olds
Spec
ific
pove
rty
rate
Com
posi
tion
of p
oor
hous
ehol
dsC
ompo
sitio
n of
tota
l ho
useh
olds
Spec
ific
pove
rty
rate
Com
posi
tion
of p
oor
hous
ehol
ds (1
)(2
)(3
)(4
)(5
)(6
)(7
)(8
)(9
)O
ccup
atio
nal c
ateg
ory
of th
e ho
useh
old
head
Reg
iste
red
wag
e ea
rner
23%
19%
11%
23%
26%
14%
23%
11%
8%N
on-r
egis
tere
d w
age
earn
er23
%50
%31
%23
%56
%29
%23
%45
%32
%O
wn
acco
unt
28%
47%
35%
28%
54%
34%
28%
41%
37%
Empl
oyer
8%21
%5%
9%21
%4%
8%21
%5%
Une
mpl
oyed
2%60
%3%
2%78
%4%
2%41
%3%
Inac
tive
with
pen
sion
5%16
%2%
5%24
%3%
5%7%
1%In
activ
e w
ithou
t pen
sion
10%
50%
13%
10%
58%
13%
10%
41%
14%
Hou
rs w
orke
d by
the
hous
ehol
d he
adU
nder
empl
oyed
15%
53%
21%
15%
59%
21%
14%
47%
21%
Full
empl
oyed
35%
32%
30%
35%
40%
32%
35%
25%
27%
Ove
r em
ploy
ed51
%37
%50
%50
%41
%48
%51
%32
%52
%
HO
USE
HO
LD C
OM
POSI
TIO
NC
ompo
sitio
n of
tota
l ho
useh
olds
Spec
ific
pove
rty
rate
Com
posi
tion
of p
oor
hous
ehol
dsC
ompo
sitio
n of
tota
l ho
useh
olds
Spec
ific
pove
rty
rate
Com
posi
tion
of p
oor
hous
ehol
dsC
ompo
sitio
n of
tota
l ho
useh
olds
Spec
ific
pove
rty
rate
Com
posi
tion
of p
oor
hous
ehol
ds (1
)(2
)(3
)(4
)(5
)(6
)(7
)(8
)(9
)Fa
mili
es w
ithou
t chi
ldre
n33
%22
%18
%31
%27
%19
%35
%16
%18
%O
ne-p
erso
n ho
useh
olds
. Men
old
er th
an 6
5 ye
ars
1%33
%1%
1%37
%1%
1%28
%1%
One
-per
son
hous
ehol
ds. F
emal
e ho
useh
old
head
old
er th
an 6
5 ye
ars
2%26
%1%
2%33
%1%
2%20
%1%
One
-per
son
hous
ehol
ds. M
en y
oung
er th
an 6
5 ye
ars
4%13
%1%
3%18
%1%
4%8%
1%
One
-per
son
hous
ehol
ds. F
emal
e ho
useh
old
head
you
nger
than
65
2%19
%1%
1%18
%1%
2%20
%1%
Com
plet
e nu
clea
r hou
seho
ld. H
ead
olde
r tha
n 65
yea
rs5%
29%
4%5%
37%
4%5%
22%
4%C
ompl
ete
nucl
ear h
ouse
hold
. Hea
d yo
unge
r tha
n 65
yea
rs11
%20
%6%
11%
24%
6%11
%16
%6%
Inco
mpl
ete
nucl
ear h
ouse
hold
. Hea
d ol
der t
han
65 y
ears
3%23
%2%
3%31
%2%
3%16
%1%
Inco
mpl
ete
nucl
ear h
ouse
hold
. Hea
d yo
unge
r tha
n 65
yea
rs6%
20%
3%5%
27%
3%6%
12%
2%Fa
mili
es w
ith c
hild
ren
64%
46%
82%
65%
53%
81%
63%
40%
82%
Inco
mpl
ete
nucl
ear h
ouse
hold
. M
ale
hous
ehol
d he
ad3%
36%
3%4%
41%
3%2%
30%
2%In
com
plet
e nu
clea
r hou
seho
ld .
Fem
ale
hous
ehol
d he
ad14
%51
%18
%13
%59
%17
%14
%43
%20
%C
ompl
ete
nucl
ear h
ouse
hold
50
%46
%61
%52
%52
%61
%49
%39
%60
%
POVE
RTY
RA
TE (H
ouse
hold
s)
Sour
ce: A
utho
rs’ e
labo
ratio
n ba
sed
on d
ata
from
Enc
uest
a de
Em
pleo
, Des
empl
eo y
Sub
empl
eo (I
NEC
).
2008
Tabl
e 3.
D. H
ouse
hold
com
posi
tion
and
spec
ific
pove
rty
rate
s. C
ross
sec
tion.
Ave
rage
200
4-20
0820
0420
08
38%
45%
32%
Ecua
dor.
Urb
an a
reas
. 20
04-2
008
Ave
rage
200
4-20
0820
0420
08
Ave
rage
200
4-20
0820
04
88 89
CH
AR
AC
TE
RIS
TIC
S O
F T
HE
HO
US
EH
OL
D H
EA
DS
Co
mp
osi
tio
n o
f to
tal
ho
use
ho
lds
Sp
ecif
ic
po
vert
y ra
te
Co
mp
osi
tio
n
of
po
or
ho
use
ho
lds
Co
mp
osi
tio
n o
f to
tal
ho
use
ho
lds
Sp
ecif
ic
po
vert
y ra
te
Co
mp
osi
tio
n
of
po
or
ho
use
ho
lds
Co
mp
osi
tio
n o
f to
tal
ho
use
ho
lds
Sp
ecif
ic
po
vert
y ra
te
Co
mp
osi
tio
n
of
po
or
ho
use
ho
lds
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
Ed
uca
tio
nal
leve
lLo
w
28%
41%
39%
29%
49%
40%
27%
33%
38%
Med
ium
44%
33%
49%
44%
39%
48%
44%
28%
50%
Hig
h28
%13
%12
%27
%16
%12
%29
%11
%13
%A
ge
24 y
ears
old
or
youn
ger
3%30
%3%
3%36
%3%
3%24
%3%
25 -
40
year
s ol
d29
%36
%36
%31
%41
%36
%27
%31
%35
%41
- 6
5 ye
ars
old
52%
29%
50%
51%
35%
51%
53%
22%
50%
Old
er th
an 6
5 ye
ars
old
17%
20%
11%
15%
25%
11%
18%
16%
12%
Gen
der
Mal
e75
%36
%76
%76
%36
%77
%74
%36
%75
%F
emal
e25
%28
%24
%24
%34
%23
%26
%23
%25
%
OC
CU
PA
TIO
NA
L V
AR
IAB
LE
SC
om
po
siti
on
of
tota
l h
ou
seh
old
s
Sp
ecif
ic
po
vert
y ra
te
Co
mp
osi
tio
n
of
po
or
ho
use
ho
lds
Co
mp
osi
tio
n o
f to
tal
ho
use
ho
lds
Sp
ecif
ic
po
vert
y ra
te
Co
mp
osi
tio
n
of
po
or
ho
use
ho
lds
Co
mp
osi
tio
n o
f to
tal
ho
use
ho
lds
Sp
ecif
ic
po
vert
y ra
te
Co
mp
osi
tio
n
of
po
or
ho
use
ho
lds
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
Occ
up
atio
nal
cat
ego
ry o
f th
e h
ou
seh
old
hea
dR
egis
tere
d w
age
earn
er20
%14
%9%
18%
17%
9%23
%10
%10
%N
on-r
egis
tere
d w
age
earn
er18
%39
%25
%22
%45
%28
%15
%33
%21
%O
wn
acco
unt
29%
37%
37%
29%
42%
35%
29%
32%
40%
Em
ploy
er7%
16%
4%6%
20%
4%7%
12%
4%U
nem
ploy
ed5%
43%
8%6%
50%
8%5%
37%
8%In
activ
e w
ith p
ensi
on11
%10
%3%
11%
14%
4%11
%5%
2%In
activ
e w
ithou
t pen
sion
9%42
%14
%9%
48%
13%
10%
35%
15%
Ho
urs
wo
rked
by
the
ho
use
ho
ld h
ead
Und
erem
ploy
ed34
%34
%39
%34
%41
%39
%34
%27
%39
%F
ull e
mpl
oyed
29%
24%
23%
30%
29%
24%
29%
19%
23%
Ove
r em
ploy
ed10
0%30
%10
0%10
0%35
%10
0%10
0%24
%10
0%
HO
US
EH
OL
D C
OM
PO
SIT
ION
Co
mp
osi
tio
n o
f to
tal
ho
use
ho
lds
Sp
ecif
ic
po
vert
y ra
te
Co
mp
osi
tio
n
of
po
or
ho
use
ho
lds
Co
mp
osi
tio
n o
f to
tal
ho
use
ho
lds
Sp
ecif
ic
po
vert
y ra
te
Co
mp
osi
tio
n
of
po
or
ho
use
ho
lds
Co
mp
osi
tio
n o
f to
tal
ho
use
ho
lds
Sp
ecif
ic
po
vert
y ra
te
Co
mp
osi
tio
n
of
po
or
ho
use
ho
lds
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
Fam
ilies
wit
ho
ut
child
ren
31%
14%
15%
29%
17%
14%
33%
11%
16%
One
-per
son
hous
ehol
ds. M
en o
lder
than
65
year
old
1%17
%0%
1%23
%1%
1%11
%0%
One
-per
son
hous
ehol
ds. F
emal
e ol
der
than
65
year
old
1%15
%1%
1%14
%0%
2%15
%1%
One
-per
son
hous
ehol
ds. M
en y
oung
er th
an 6
5 ye
ar o
ld4%
11%
1%3%
15%
1%4%
6%1%
One
-per
son
hous
ehol
ds. F
emal
e yo
unge
r th
an 6
5 ye
ar o
ld1%
11%
1%1%
11%
0%1%
12%
1%C
ompl
ete
nucl
ear
hous
ehol
d. H
ead
olde
r th
an 6
5 ye
ar o
ld6%
16%
3%5%
20%
3%6%
12%
3%C
ompl
ete
nucl
ear
hous
ehol
d. H
ead
youn
ger
than
65
year
old
8%16
%4%
8%21
%5%
8%12
%4%
Inco
mpl
ete
nucl
ear
hous
ehol
d. H
ead
olde
r th
an 6
5 ye
ars
3%13
%1%
3%14
%1%
3%12
%1%
Inco
mpl
ete
nucl
ear
hous
ehol
d. H
ead
youn
ger
than
65
year
s7%
13%
3%6%
14%
2%7%
13%
4%F
amili
es w
ith
ch
ildre
n69
%36
%85
%71
%43
%86
%67
%29
%84
%In
com
plet
e nu
clea
r ho
useh
old
. Mal
e ho
useh
old
head
4%48
%5%
4%48
%5%
4%48
%5%
Inco
mpl
ete
nucl
ear
hous
ehol
d . F
emal
e ho
useh
old
head
14%
62%
17%
13%
62%
17%
14%
62%
18%
Com
plet
e nu
clea
r ho
useh
old
52%
52%
63%
54%
52%
65%
50%
52%
62%
PO
VE
RT
Y R
AT
E (
Ho
use
ho
lds)
30.4
%35
.6%
25.1
%
So
urc
e: A
utho
rs’ e
labo
ratio
n ba
sed
on d
ata
from
Enc
uest
a N
acio
nal d
e H
ogar
es (
EN
AH
O-I
NE
I).
Ave
rag
e 20
02 -
200
620
0220
06
Ave
rag
e 20
02 -
200
620
0220
06
Tab
le 3
.E. H
ou
seh
old
co
mp
osi
tio
n a
nd
sp
ecif
ic p
ove
rty
rate
s. C
ross
sec
tio
n.
Per
u. U
rban
are
as. P
erio
d 2
002-
2006
.
Ave
rag
e 20
02 -
200
620
0220
06
90
CH
AR
AC
TER
ISTI
CS
OF
THE
HO
USE
HO
LD H
EAD
SC
ompo
sitio
n of
tota
l ho
useh
olds
Spec
ific
pove
rty
rate
Com
posi
tion
of p
oor
hous
ehol
ds
Com
posi
tion
of to
tal
hous
ehol
ds
Spec
ific
pove
rty
rate
Com
posi
tion
of p
oor
hous
ehol
ds
Com
posi
tion
of to
tal
hous
ehol
ds
Spec
ific
pove
rty
rate
Com
posi
tion
of p
oor
hous
ehol
dsEx
it ra
teEn
try
rate
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
(10)
(11)
Educ
atio
nal l
evel
Low
44
%43
%61
%47
%51
%60
%43
%31
%67
%29
%11
%M
ediu
m33
%30
%31
%32
%38
%31
%32
%17
%27
%36
%9%
Hig
h23
%11
%8%
22%
17%
9%25
%4%
5%51
%3%
Age
24 y
ears
old
or y
oung
er4%
37%
4%5%
44%
5%3%
25%
3%32
%12
%25
- 40
yea
rs o
ld29
%39
%35
%29
%45
%33
%29
%27
%39
%28
%9%
41 -
65 y
ears
old
44%
34%
47%
42%
44%
47%
45%
21%
48%
35%
9%O
lder
than
65
year
s ol
d23
%17
%13
%24
%23
%14
%23
%9%
10%
45%
5%G
ende
rM
ale
69%
33%
72%
69%
42%
73%
67%
20%
68%
34%
9%Fe
mal
e31
%29
%28
%31
%34
%27
%33
%19
%32
%32
%6%
OC
CU
PATI
ON
AL
VAR
IAB
LES
Com
posi
tion
of to
tal
hous
ehol
ds
Spec
ific
pove
rty
rate
Com
posi
tion
of p
oor
hous
ehol
ds
Com
posi
tion
of to
tal
hous
ehol
ds
Spec
ific
pove
rty
rate
Com
posi
tion
of p
oor
hous
ehol
ds
Com
posi
tion
of to
tal
hous
ehol
ds
Spec
ific
pove
rty
rate
Com
posi
tion
of p
oor
hous
ehol
dsEx
it ra
teEn
try
rate
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
(10)
(11)
Occ
upat
iona
l cat
egor
y of
the
hous
ehol
d he
adR
egis
tere
d w
age
earn
er30
%17
%17
%30
%26
%20
%32
%9%
14%
46%
6%N
on-r
egis
tere
d w
age
earn
er19
%51
%31
%19
%59
%29
%18
%35
%33
%23
%15
%O
wn
acco
unt
15%
40%
20%
14%
48%
18%
16%
25%
20%
33%
13%
Em
ploy
er3%
13%
1%2%
18%
1%3%
6%1%
35%
6%U
nem
ploy
ed4%
64%
9%5%
66%
9%3%
49%
7%32
%9%
Inac
tive
with
pen
sion
22%
15%
11%
23%
21%
12%
22%
8%8%
45%
4%In
activ
e w
ithou
t pen
sion
7%53
%12
%7%
62%
12%
7%46
%17
%34
%14
%H
ours
wor
ked
by th
e ho
useh
old
head
Und
erem
ploy
ed27
%49
%40
%29
%58
%42
%25
%29
%36
%24
%9%
Full
empl
oyed
27%
25%
21%
28%
33%
22%
29%
20%
25%
35%
7%O
ver e
mpl
oyed
46%
27%
39%
43%
35%
36%
46%
30%
38%
37%
9%
HO
USE
HO
LD C
OM
POSI
TIO
NC
ompo
sitio
n of
tota
l ho
useh
olds
Spec
ific
pove
rty
rate
Com
posi
tion
of p
oor
hous
ehol
ds
Com
posi
tion
of to
tal
hous
ehol
ds
Spec
ific
pove
rty
rate
Com
posi
tion
of p
oor
hous
ehol
ds
Com
posi
tion
of to
tal
hous
ehol
ds
Spec
ific
pove
rty
rate
Com
posi
tion
of p
oor
hous
ehol
dsEx
it ra
teEn
try
rate
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
(10)
(11)
Fam
ilies
with
out c
hild
ren
52%
16%
26%
50%
20%
26%
53%
8%21
%48
%5%
One
-per
son
hous
ehol
ds. M
en o
lder
than
65
year
old
2%13
%1%
2%28
%2%
2%11
%1%
29%
3%O
ne-p
erso
n ho
useh
olds
. Fem
ale
olde
r tha
n 65
yea
r old
7%5%
1%7%
6%1%
7%3%
1%56
%3%
One
-per
son
hous
ehol
ds. M
en y
oung
er th
an 6
5 ye
ar o
ld4%
16%
2%4%
21%
2%5%
10%
2%45
%6%
One
-per
son
hous
ehol
ds. F
emal
e yo
unge
r tha
n 65
yea
r old
4%9%
1%4%
7%1%
5%4%
1%52
%3%
Com
plet
e nu
clea
r hou
seho
ld. H
ead
olde
r tha
n 65
yea
r old
10%
19%
6%9%
22%
5%9%
8%4%
51%
6%C
ompl
ete
nucl
ear h
ouse
hold
. Hea
d yo
unge
r tha
n 65
yea
r old
15%
19%
9%13
%25
%8%
14%
10%
7%46
%7%
Inco
mpl
ete
nucl
ear h
ouse
hold
. Hea
d ol
der t
han
65 y
ears
4%18
%2%
3%28
%2%
4%10
%2%
55%
7%In
com
plet
e nu
clea
r hou
seho
ld. H
ead
youn
ger t
han
65 y
ears
7%19
%4%
6%23
%4%
8%9%
3%51
%6%
Fam
ilies
with
chi
ldre
n48
%49
%74
%50
%59
%74
%47
%33
%79
%28
%12
%In
com
plet
e nu
clea
r hou
seho
ld .
Mal
e ho
useh
old
head
2%49
%3%
1%62
%2%
2%38
%3%
32%
18%
Inco
mpl
ete
nucl
ear h
ouse
hold
. Fe
mal
e ho
useh
old
head
9%57
%16
%10
%64
%16
%9%
42%
19%
24%
14%
Com
plet
e nu
clea
r hou
seho
ld
37%
47%
55%
38%
58%
56%
37%
31%
57%
29%
12%
POVE
RTY
RA
TE (H
ouse
hold
s)33
%8%
Sour
ce: A
utho
rs’ e
labo
ratio
n ba
sed
on d
ata
from
Enc
uest
a P
erm
anen
te d
e H
ogar
es (E
PH
-IND
EC
).
Tabl
e 4.
A. H
ouse
hold
com
posi
tion
and
spec
ific
pove
rty
rate
s. P
ool o
f pan
els.
Arg
entin
a. U
rban
are
as. P
erio
d 20
03-2
006.
Ave
rage
200
3-20
062n
d. 2
003
2nd.
200
6 A
vera
ge 2
003-
2006
Ave
rage
200
3-20
062n
d. 2
003
2nd.
200
6 A
vera
ge 2
003-
2006
31%
39%
20%
Ave
rage
200
3-20
062n
d. 2
003
2nd.
200
6 A
vera
ge 2
003-
2006
90 91
CH
AR
AC
TER
ISTI
CS
OF
THE
HO
USE
HO
LD H
EAD
SC
ompo
sitio
n of
tota
l ho
useh
olds
Spec
ific
pove
rty
rate
Com
posi
tion
of p
oor
hous
ehol
ds
Com
posi
tion
of to
tal
hous
ehol
ds
Spec
ific
pove
rty
rate
Com
posi
tion
of p
oor
hous
ehol
ds
Com
posi
tion
of to
tal
hous
ehol
ds
Spec
ific
pove
rty
rate
Com
posi
tion
of p
oor
hous
ehol
dsEx
it ra
teEn
try
rate
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
(10)
(11)
Educ
atio
nal l
evel
Low
51
%39
%69
%52
%42
%68
%49
%36
%69
%38
%20
%M
ediu
m32
%25
%28
%31
%28
%28
%33
%22
%28
%47
%11
%H
igh
18%
6%4%
17%
8%4%
18%
5%4%
77%
4%A
ge24
yea
rs o
ld o
r you
nger
3%42
%4%
3%42
%4%
2%42
%4%
31%
18%
25 -
40 y
ears
old
30%
36%
38%
33%
40%
41%
27%
33%
35%
29%
13%
41 -
65 y
ears
old
50%
27%
48%
48%
30%
45%
52%
25%
51%
46%
13%
Old
er th
an 6
5 ye
ars
old
18%
17%
10%
16%
20%
10%
19%
14%
11%
78%
13%
Gen
der
Mal
e67
%29
%67
%70
%31
%69
%65
%26
%65
%39
%12
%Fe
mal
e33
%29
%33
%30
%33
%31
%35
%25
%35
%47
%15
%
OC
CU
PATI
ON
AL
VAR
IAB
LES
Com
posi
tion
of to
tal
hous
ehol
ds
Spec
ific
pove
rty
rate
Com
posi
tion
of p
oor
hous
ehol
ds
Com
posi
tion
of to
tal
hous
ehol
ds
Spec
ific
pove
rty
rate
Com
posi
tion
of p
oor
hous
ehol
ds
Com
posi
tion
of to
tal
hous
ehol
ds
Spec
ific
pove
rty
rate
Com
posi
tion
of p
oor
hous
ehol
dsEx
it ra
teEn
try
rate
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
(10)
(11)
Occ
upat
iona
l cat
egor
y of
the
hous
ehol
d he
adR
egis
tere
d w
age
earn
er33
%22
%26
%33
%24
%25
%33
%20
%27
%32
%9%
Non
-reg
iste
red
wag
e ea
rner
11%
35%
13%
11%
39%
14%
10%
30%
12%
33%
17%
Ow
n ac
coun
t16
%34
%19
%17
%38
%20
%15
%31
%19
%31
%13
%E
mpl
oyer
4%6%
1%4%
6%1%
5%7%
1%38
%6%
Une
mpl
oyed
3%65
%8%
4%73
%10
%3%
57%
6%46
%23
%In
activ
e w
ith p
ensi
on25
%13
%12
%24
%15
%11
%26
%12
%13
%58
%18
%In
activ
e w
ithou
t pen
sion
7%77
%21
%7%
79%
18%
8%74
%23
%56
%12
%H
ours
wor
ked
by th
e ho
useh
old
head
Und
erem
ploy
ed15
%30
%16
%13
%36
%15
%17
%24
%18
%37
%11
%Fu
ll em
ploy
ed48
%24
%44
%48
%27
%44
%48
%22
%44
%33
%10
%O
ver e
mpl
oyed
37%
29%
40%
39%
32%
41%
34%
27%
38%
29%
14%
HO
USE
HO
LD C
OM
POSI
TIO
NC
ompo
sitio
n of
tota
l ho
useh
olds
Spec
ific
pove
rty
rate
Com
posi
tion
of p
oor
hous
ehol
ds
Com
posi
tion
of to
tal
hous
ehol
ds
Spec
ific
pove
rty
rate
Com
posi
tion
of p
oor
hous
ehol
ds
Com
posi
tion
of to
tal
hous
ehol
ds
Spec
ific
pove
rty
rate
Com
posi
tion
of p
oor
hous
ehol
dsEx
it ra
teEn
try
rate
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
(10)
(11)
Fam
ilies
with
out c
hild
ren
54%
18%
33%
52%
19%
32%
56%
16%
34%
64%
12%
One
-per
son
hous
ehol
ds. M
en o
lder
than
65
year
old
2%16
%1%
2%16
%1%
2%17
%1%
75%
15%
One
-per
son
hous
ehol
ds. F
emal
e ol
der t
han
65 y
ear o
ld5%
16%
3%5%
20%
3%5%
13%
3%89
%14
%O
ne-p
erso
n ho
useh
olds
. Men
you
nger
than
65
year
old
6%21
%4%
6%25
%4%
6%17
%4%
53%
12%
One
-per
son
hous
ehol
ds. F
emal
e yo
unge
r tha
n 65
yea
r old
5%17
%3%
5%18
%3%
6%16
%4%
63%
14%
Com
plet
e nu
clea
r hou
seho
ld. H
ead
olde
r tha
n 65
yea
r old
7%12
%3%
7%15
%3%
8%9%
3%85
%11
%C
ompl
ete
nucl
ear h
ouse
hold
. Hea
d yo
unge
r tha
n 65
yea
r old
18%
17%
11%
18%
18%
10%
18%
16%
11%
57%
10%
Inco
mpl
ete
nucl
ear h
ouse
hold
. Hea
d ol
der t
han
65 y
ears
4%18
%2%
3%19
%2%
4%18
%3%
61%
12%
Inco
mpl
ete
nucl
ear h
ouse
hold
. Hea
d yo
unge
r tha
n 65
yea
rs8%
22%
6%7%
24%
5%8%
20%
6%57
%11
%Fa
mili
es w
ith c
hild
ren
46%
42%
67%
48%
45%
68%
44%
38%
66%
31%
16%
Inco
mpl
ete
nucl
ear h
ouse
hold
. M
ale
hous
ehol
d he
ad1%
38%
1%1%
47%
2%1%
29%
1%37
%15
%In
com
plet
e nu
clea
r hou
seho
ld .
Fem
ale
hous
ehol
d he
ad9%
50%
15%
9%55
%15
%9%
44%
15%
33%
23%
Com
plet
e nu
clea
r hou
seho
ld
36%
40%
50%
38%
43%
51%
34%
37%
49%
30%
14%
POVE
RTY
RA
TE (H
ouse
hold
s)28
.7%
31.8
%25
.5%
42%
13%
Sour
ce: A
utho
rs’ e
labo
ratio
n ba
sed
on d
ata
PM
E &
PN
AD
(IB
GE
)
Tabl
e 4.
B. H
ouse
hold
com
posi
tion
and
spec
ific
pove
rty
rate
s. P
ool o
f pan
els.
Bra
zil.
Urb
an a
reas
. Per
iod
2003
-200
6.
Ave
rage
200
3 - 2
006
Sep.
200
3Se
p. 2
006
Ave
rage
200
3-20
06
Ave
rage
200
3-20
06Se
p. 2
003
Sep.
200
6 A
vera
ge 2
003-
2006
Ave
rage
200
3-20
06Se
p. 2
003
Sep.
200
6 A
vera
ge 2
003-
2006
92
CHAR
ACTE
RIST
ICS
OF
THE
HOUS
EHO
LD H
EADS
Com
posi
tion
of to
tal
hous
ehol
dsSp
ecifi
c po
verty
ra
teCo
mpo
sitio
n of
poo
r ho
useh
olds
Com
posi
tion
of to
tal
hous
ehol
dsSp
ecifi
c po
verty
ra
teCo
mpo
sitio
n of
poo
r ho
useh
olds
Com
posi
tion
of to
tal
hous
ehol
dsSp
ecifi
c po
verty
ra
teCo
mpo
sitio
n of
poo
r ho
useh
olds
Exit
rate
Entry
rate
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
(10)
(11)
Educ
atio
nal l
evel
Com
plet
e pr
imar
y or
less
40%
38%
65%
40%
40%
64%
40%
30%
63%
40%
16%
Seco
ndar
y32
%18
%24
%33
%17
%23
%30
%15
%24
%60
%11
%Un
ivers
ity o
r mor
e28
%10
%11
%28
%12
%13
%30
%8%
13%
59%
6%Ag
e24
yea
rs o
ld o
r you
nger
2%23
%2%
2%25
%2%
1%45
%3%
49%
23%
25 -
40 y
ears
old
27%
28%
31%
27%
28%
31%
25%
19%
25%
49%
13%
41 -
65 y
ears
old
56%
20%
46%
55%
22%
49%
58%
15%
46%
52%
10%
Old
er th
an 6
5 ye
ars
old
16%
31%
20%
16%
28%
18%
16%
30%
26%
33%
11%
Gen
der
Mal
e68
%22
%64
%68
%24
%67
%70
%17
%61
%51
%11
%Fe
mal
e32
%27
%36
%32
%26
%33
%30
%25
%39
%39
%11
%
OCC
UPAT
IONA
L VA
RIAB
LES
Com
posi
tion
of to
tal
hous
ehol
dsSp
ecifi
c po
verty
ra
teCo
mpo
sitio
n of
poo
r ho
useh
olds
Com
posi
tion
of to
tal
hous
ehol
dsSp
ecifi
c po
verty
ra
teCo
mpo
sitio
n of
poo
r ho
useh
olds
Com
posi
tion
of to
tal
hous
ehol
dsSp
ecifi
c po
verty
ra
teCo
mpo
sitio
n of
poo
r ho
useh
olds
Exit
rate
Entry
rate
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
(10)
(11)
Occ
upat
iona
l cat
egor
y of
the
hous
ehol
d he
adRe
gist
ered
wag
e ea
rner
40%
14%
24%
40%
18%
29%
41%
11%
24%
61%
10%
Non-
regi
ster
ed w
age
earn
er10
%30
%13
%10
%30
%12
%10
%23
%11
%34
%19
%O
wn a
ccou
nt15
%34
%21
%15
%33
%20
%17
%25
%22
%46
%12
%Em
ploy
er9%
16%
6%9%
15%
6%8%
10%
4%66
%11
%Un
empl
oyed
1%69
%4%
2%66
%4%
1%74
%4%
40%
24%
Inac
tive
with
pen
sion
14%
17%
10%
15%
14%
9%15
%16
%12
%43
%8%
Inac
tive
with
out p
ensio
n9%
53%
21%
9%57
%20
%9%
46%
22%
37%
14%
Hour
s w
orke
d by
the
hous
ehol
d he
adUn
dere
mpl
oyed
12%
32%
19%
12%
28%
15%
10%
16%
11%
44%
12%
Full e
mpl
oyed
18%
13%
11%
18%
14%
12%
20%
32%
21%
52%
9%O
ver e
mpl
oyed
70%
20%
70%
71%
23%
74%
70%
103%
68%
52%
12%
HOUS
EHO
LD C
OM
POSI
TIO
NCo
mpo
sitio
n of
tota
l ho
useh
olds
Spec
ific
pove
rty
rate
Com
posi
tion
of p
oor
hous
ehol
dsCo
mpo
sitio
n of
tota
l ho
useh
olds
Spec
ific
pove
rty
rate
Com
posi
tion
of p
oor
hous
ehol
dsCo
mpo
sitio
n of
tota
l ho
useh
olds
Spec
ific
pove
rty
rate
Com
posi
tion
of p
oor
hous
ehol
dsEx
it ra
teEn
try ra
te
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
(10)
(11)
Fam
ilies
with
out c
hild
ren
43%
19%
35%
44%
18%
33%
45%
17%
40%
44%
9%O
ne-p
erso
n ho
useh
olds
. Men
old
er th
an 6
5 ye
ars
1%34
%1%
1%29
%1%
2%21
%2%
8%7%
One
-per
son
hous
ehol
ds. F
emal
e ho
useh
old
head
old
er th
an 6
5 ye
ars
2%31
%3%
3%32
%4%
2%22
%3%
13%
9%O
ne-p
erso
n ho
useh
olds
. Men
you
nger
than
65
year
s3%
9%1%
4%10
%2%
3%9%
2%68
%5%
One
-per
son
hous
ehol
ds. F
emal
e ho
useh
old
head
you
nger
than
65
3%21
%3%
3%17
%2%
3%15
%3%
61%
7%Co
mpl
ete
nucle
ar h
ouse
hold
. Hea
d ol
der t
han
65 y
ears
7%29
%9%
8%26
%8%
8%28
%12
%33
%12
%Co
mpl
ete
nucle
ar h
ouse
hold
. Hea
d yo
unge
r tha
n 65
yea
rs14
%11
%7%
13%
8%4%
16%
8%7%
75%
7%In
com
plet
e nu
clear
hou
seho
ld. H
ead
olde
r tha
n 65
yea
rs4%
26%
4%4%
24%
4%4%
23%
5%40
%13
%In
com
plet
e nu
clear
hou
seho
ld. H
ead
youn
ger t
han
65 y
ears
7%21
%6%
8%22
%8%
7%22
%8%
43%
11%
Fam
ilies
with
chi
ldre
n57
%27
%65
%56
%30
%67
%55
%21
%60
%48
%13
%In
com
plet
e nu
clear
hou
seho
ld .
Mal
e ho
useh
old
head
1%17
%1%
1%24
%1%
1%8%
0%69
%6%
Inco
mpl
ete
nucle
ar h
ouse
hold
. Fe
mal
e ho
useh
old
head
13%
30%
17%
13%
31%
16%
13%
32%
21%
36%
12%
Com
plet
e nu
clear
hou
seho
ld
42%
27%
47%
42%
30%
50%
41%
18%
39%
52%
14%
POVE
RTY
RATE
(Hou
seho
lds)
47%
11%
Sour
ce: A
utho
rs’ e
labo
ratio
n ba
sed
on d
ata f
rom
Enc
uesta
de H
ogar
es d
e Pro
pósit
os M
últip
les (
INEC
).
Ave
rage
200
6-20
0820
0620
08 A
vera
ge 2
006-
2008
Ave
rage
200
6-20
0820
0620
08 A
vera
ge 2
006-
2008
Tabl
e 4.
C. H
ouse
hold
com
posi
tion
and
spec
ific
pove
rty ra
tes.
Poo
l of p
anel
s.
24%
25%
19%
Cost
a Ri
ca. U
rban
are
as. 2
006-
2008
.
Ave
rage
200
6-20
0820
0620
08 A
vera
ge 2
006-
2008
92 93
CH
ARAC
TER
ISTI
CS
OF
THE
HO
USE
HO
LD H
EAD
SC
ompo
sitio
n of
tota
l ho
useh
olds
Spec
ific
pove
rty
rate
Com
posi
tion
of p
oor
hous
ehol
dsC
ompo
sitio
n of
tota
l ho
useh
olds
Spec
ific
pove
rty
rate
Com
posi
tion
of p
oor
hous
ehol
dsC
ompo
sitio
n of
tota
l ho
useh
olds
Spec
ific
pove
rty
rate
Com
posi
tion
of p
oor
hous
ehol
dsEx
it ra
teEn
try
rate
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
(10)
(11)
Educ
atio
nal l
evel
Com
plet
e pr
imar
y or
less
37%
50%
53%
37%
60%
53%
39%
42%
55%
33%
23%
Seco
ndar
y37
%36
%38
%38
%41
%38
%36
%30
%36
%40
%16
%U
nive
rsity
or m
ore
25%
12%
9%25
%16
%9%
25%
10%
9%57
%7%
Age
24 y
ears
old
or y
oung
er3%
37%
3%3%
43%
3%2%
23%
2%48
%15
%25
- 40
yea
rs o
ld33
%44
%40
%37
%49
%42
%30
%39
%38
%31
%18
%41
- 65
yea
rs o
ld50
%33
%46
%48
%39
%44
%52
%28
%47
%41
%14
%O
lder
than
65
year
s ol
d14
%31
%12
%12
%38
%10
%16
%24
%13
%41
%14
%G
ende
rM
ale
77%
36%
77%
79%
42%
78%
75%
31%
75%
37%
15%
Fem
ale
23%
37%
23%
21%
45%
22%
25%
31%
25%
38%
16%
OC
CU
PATI
ON
AL V
ARIA
BLE
SC
ompo
sitio
n of
tota
l ho
useh
olds
Spec
ific
pove
rty
rate
Com
posi
tion
of p
oor
hous
ehol
dsC
ompo
sitio
n of
tota
l ho
useh
olds
Spec
ific
pove
rty
rate
Com
posi
tion
of p
oor
hous
ehol
dsC
ompo
sitio
n of
tota
l ho
useh
olds
Spec
ific
pove
rty
rate
Com
posi
tion
of p
oor
hous
ehol
dsEx
it ra
teEn
try
rate
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
(10)
(11)
Occ
upat
iona
l cat
egor
y of
the
hous
ehol
d he
adR
egis
tere
d w
age
earn
er25
%18
%12
%26
%25
%15
%24
%10
%8%
51%
8%N
on-r
egis
tere
d w
age
earn
er22
%50
%30
%22
%54
%29
%22
%47
%34
%28
%21
%O
wn
acco
unt
29%
45%
36%
28%
51%
34%
29%
38%
37%
34%
22%
Empl
oyer
9%21
%5%
9%22
%5%
7%18
%4%
50%
14%
Une
mpl
oyed
2%63
%3%
2%76
%4%
2%43
%3%
42%
5%In
activ
e w
ith p
ensi
on5%
13%
2%5%
22%
3%6%
8%1%
65%
4%In
activ
e w
ithou
t pen
sion
8%49
%11
%8%
58%
11%
9%40
%12
%37
%22
%H
ours
wor
ked
by th
e ho
useh
old
head
Und
erem
ploy
ed12
%49
%20
%14
%55
%18
%13
%48
%21
%32
%20
%Fu
ll em
ploy
ed38
%32
%31
%34
%39
%32
%36
%25
%29
%38
%13
%O
ver e
mpl
oyed
51%
35%
49%
52%
40%
50%
51%
31%
51%
36%
17%
HO
USE
HO
LD C
OM
POSI
TIO
NC
ompo
sitio
n of
tota
l ho
useh
olds
Spec
ific
pove
rty
rate
Com
posi
tion
of p
oor
hous
ehol
dsC
ompo
sitio
n of
tota
l ho
useh
olds
Spec
ific
pove
rty
rate
Com
posi
tion
of p
oor
hous
ehol
dsC
ompo
sitio
n of
tota
l ho
useh
olds
Spec
ific
pove
rty
rate
Com
posi
tion
of p
oor
hous
ehol
dsEx
it ra
teEn
try
rate
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
(10)
(11)
Fam
ilies
with
out c
hild
ren
33%
20%
18%
30%
26%
18%
36%
16%
19%
50%
11%
One
-per
son
hous
ehol
ds. M
en o
lder
than
65
year
s1%
31%
1%2%
34%
1%2%
27%
1%40
%22
%O
ne-p
erso
n ho
useh
olds
. Fem
ale
hous
ehol
d he
ad o
lder
than
65
year
s2%
30%
2%2%
34%
1%3%
18%
2%46
%9%
One
-per
son
hous
ehol
ds. M
en y
oung
er th
an 6
5 ye
ars
4%15
%2%
4%20
%2%
4%9%
1%55
%13
%
One
-per
son
hous
ehol
ds. F
emal
e ho
useh
old
head
you
nger
than
65
2%17
%1%
2%29
%1%
2%22
%2%
41%
10%
Com
plet
e nu
clea
r hou
seho
ld. H
ead
olde
r tha
n 65
yea
rs5%
24%
4%4%
33%
3%6%
18%
4%44
%11
%C
ompl
ete
nucl
ear h
ouse
hold
. Hea
d yo
unge
r tha
n 65
yea
rs11
%17
%5%
11%
22%
6%11
%16
%6%
57%
10%
Inco
mpl
ete
nucl
ear h
ouse
hold
. Hea
d ol
der t
han
65 y
ears
2%21
%1%
2%19
%1%
3%19
%2%
50%
13%
Inco
mpl
ete
nucl
ear h
ouse
hold
. Hea
d yo
unge
r tha
n 65
yea
rs5%
20%
3%4%
26%
3%6%
15%
3%55
%10
%Fa
mili
es w
ith c
hild
ren
65%
44%
82%
67%
50%
82%
62%
39%
81%
34%
18%
Inco
mpl
ete
nucl
ear h
ouse
hold
. M
ale
hous
ehol
d he
ad2%
40%
2%2%
38%
2%2%
26%
2%57
%18
%In
com
plet
e nu
clea
r hou
seho
ld .
Fem
ale
hous
ehol
d he
ad12
%49
%16
%11
%55
%15
%12
%42
%17
%34
%23
%C
ompl
ete
nucl
ear h
ouse
hold
53
%43
%63
%56
%50
%65
%50
%38
%62
%33
%18
%
POVE
RTY
RA
TE (H
ouse
hold
s)37
%15
%
Sour
ce: A
utho
rs’ e
labo
ratio
n ba
sed
on d
ata
from
Enc
uest
a de
Em
pleo
, Des
empl
eo y
Sub
empl
eo (I
NEC
).
Ave
rage
200
4-20
0820
0420
08 A
vera
ge 2
004-
2008
Ave
rage
200
4-20
0820
0420
08 A
vera
ge 2
004-
2008
Tabl
e 4.
D. H
ouse
hold
com
posi
tion
and
spec
ific
pove
rty
rate
s. P
ool o
f pan
els.
36%
43%
31%
Ecua
dor.
Urb
an a
reas
. 200
4-20
08
Ave
rage
200
4-20
0820
0420
08 A
vera
ge 2
004-
2008
94
CH
AR
AC
TER
ISTI
CS
OF
THE
HO
US
EH
OLD
HE
AD
SC
ompo
sitio
n of
to
tal
hous
ehol
ds
Spe
cific
po
vert
y ra
te
Com
posi
tion
of
poor
ho
useh
olds
Com
posi
tion
of
tota
l ho
useh
olds
Spe
cific
po
vert
y ra
te
Com
posi
tion
of
poor
ho
useh
olds
Com
posi
tion
of
tota
l ho
useh
olds
Spe
cific
po
vert
y ra
te
Com
posi
tion
of
poor
ho
useh
olds
Exi
t rat
eE
ntry
rat
e
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
(10)
(11)
Edu
catio
nal l
evel
Low
30
%49
%41
%29
%54
%41
%30
%43
%42
%33
%29
%M
ediu
m43
%36
%45
%43
%41
%46
%43
%32
%44
%39
%23
%H
igh
27%
18%
14%
28%
18%
13%
26%
18%
15%
51%
13%
Age
24 y
ears
old
or
youn
ger
2%28
%2%
2%34
%2%
2%22
%2%
56%
14%
25 -
40
year
s ol
d28
%39
%31
%32
%42
%34
%24
%36
%28
%38
%22
%41
- 6
5 ye
ars
old
54%
34%
54%
53%
38%
53%
55%
31%
55%
37%
22%
Old
er th
an 6
5 ye
ars
old
15%
29%
13%
13%
32%
11%
18%
26%
15%
45%
15%
Gen
der
Mal
e76
%35
%75
%79
%38
%79
%73
%31
%72
%37
%20
%F
emal
e27
%28
%27
%21
%38
%21
%33
%19
%32
%43
%22
%
OC
CU
PA
TIO
NA
L V
AR
IAB
LES
Com
posi
tion
of
tota
l ho
useh
olds
Spe
cific
po
vert
y ra
te
Com
posi
tion
of
poor
ho
useh
olds
Com
posi
tion
of
tota
l ho
useh
olds
Spe
cific
po
vert
y ra
te
Com
posi
tion
of
poor
ho
useh
olds
Com
posi
tion
of
tota
l ho
useh
olds
Spe
cific
po
vert
y ra
te
Com
posi
tion
of
poor
ho
useh
olds
Exi
t rat
eE
ntry
rat
e
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
(10)
(11)
Occ
upat
iona
l cat
egor
y of
the
hous
ehol
d he
adR
egis
tere
d w
age
earn
er19
%19
%11
%18
%22
%10
%21
%16
%11
%42
%15
%N
on-r
egis
tere
d w
age
earn
er20
%41
%25
%24
%46
%30
%16
%36
%19
%31
%26
%O
wn
acco
unt
28%
44%
37%
28%
48%
36%
28%
40%
38%
34%
28%
Em
ploy
er7%
19%
4%7%
18%
4%7%
20%
5%47
%23
%U
nem
ploy
ed5%
47%
6%4%
50%
6%5%
43%
7%47
%17
%In
activ
e w
ith p
ensi
on11
%14
%4%
10%
17%
5%12
%10
%4%
50%
11%
Inac
tive
with
out p
ensi
on10
%43
%13
%9%
43%
10%
10%
43%
15%
43%
22%
Hou
rs w
orke
d by
the
hous
ehol
d he
adU
nder
empl
oyed
22%
37%
23%
21%
37%
19%
23%
37%
27%
38%
17%
Ful
l em
ploy
ed18
%27
%14
%20
%31
%16
%16
%24
%12
%30
%17
%O
ver
empl
oyed
60%
38%
63%
60%
43%
65%
61%
32%
61%
38%
21%
HO
US
EH
OLD
CO
MP
OS
ITIO
NC
ompo
sitio
n of
to
tal
hous
ehol
ds
Spe
cific
po
vert
y ra
te
Com
posi
tion
of
poor
ho
useh
olds
Com
posi
tion
of
tota
l ho
useh
olds
Spe
cific
po
vert
y ra
te
Com
posi
tion
of
poor
ho
useh
olds
Com
posi
tion
of
tota
l ho
useh
olds
Spe
cific
po
vert
y ra
te
Com
posi
tion
of
poor
ho
useh
olds
Exi
t rat
eE
ntry
rat
e
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
(10)
(11)
Fam
ilies
with
out c
hild
ren
29%
19%
16%
26%
19%
13%
31%
20%
20%
49%
13%
One
-per
son
hous
ehol
ds. M
en o
lder
than
65
year
old
1%22
%0%
1%35
%1%
1%9%
0%44
%10
%O
ne-p
erso
n ho
useh
olds
. Fem
ale
olde
r th
an 6
5 ye
ar o
ld1%
48%
1%1%
56%
1%1%
40%
2%25
%15
%O
ne-p
erso
n ho
useh
olds
. Men
you
nger
than
65
year
old
3%6%
0%3%
8%1%
3%3%
0%70
%8%
One
-per
son
hous
ehol
ds. F
emal
e yo
unge
r th
an 6
5 ye
ar o
ld1%
23%
1%1%
21%
1%1%
25%
1%36
%11
%C
ompl
ete
nucl
ear
hous
ehol
d. H
ead
olde
r th
an 6
5 ye
ar o
ld5%
20%
3%4%
21%
2%6%
20%
4%39
%10
%C
ompl
ete
nucl
ear
hous
ehol
d. H
ead
youn
ger
than
65
year
old
8%23
%5%
8%24
%5%
8%23
%6%
52%
14%
Inco
mpl
ete
nucl
ear
hous
ehol
d. H
ead
olde
r th
an 6
5 ye
ars
3%14
%1%
3%11
%1%
3%16
%2%
50%
10%
Inco
mpl
ete
nucl
ear
hous
ehol
d. H
ead
youn
ger
than
65
year
s6%
18%
4%5%
14%
2%7%
23%
5%57
%19
%Fa
mili
es w
ith c
hild
ren
71%
41%
84%
74%
45%
87%
69%
37%
80%
37%
25%
Inco
mpl
ete
nucl
ear
hous
ehol
d . M
ale
hous
ehol
d he
ad4%
41%
4%3%
41%
4%4%
42%
5%40
%33
%In
com
plet
e nu
clea
r ho
useh
old
. Fem
ale
hous
ehol
d he
ad14
%43
%17
%12
%50
%16
%15
%37
%17
%42
%26
%C
ompl
ete
nucl
ear
hous
ehol
d 54
%40
%63
%59
%44
%67
%50
%36
%58
%45
%25
%
PO
VE
RTY
RA
TE (H
ouse
hold
s)34
.8%
38.3
%31
.2%
38%
21%
Sou
rce:
Aut
hors
’ ela
bora
tion
base
d on
dat
a fr
om E
ncue
sta
Nac
iona
l de
Hog
ares
(E
NA
HO
-IN
EI)
.
Ave
rage
200
2-20
0620
0220
06 A
vera
ge 2
002-
2006
Ave
rage
200
2-20
0620
0220
06 A
vera
ge 2
002-
2006
Tabl
e 4.
E. H
ouse
hold
com
posi
tion
and
spec
ific
pove
rty
rate
s. P
ool o
f pan
els.
Per
u. U
rban
are
as. 2
002-
2006
.
Ave
rage
200
2-20
0620
0220
06 A
vera
ge 2
002-
2006
94 95
ANNEX C
Poverty transitions and trigger events
Rates
Total households
Households with children
Households without children
Total households
Households with children
Households without children
Total households
Households with children
Households without children
Entry 0.0795*** 0.124*** 0.0548*** 0.0710*** 0.112*** 0.0467*** 0.0878*** 0.135*** 0.0633***[0.00265] [0.00536] [0.00281] [0.00248] [0.00498] [0.00258] [0.00299] [0.00634] [0.00306]
Exit 0.334*** 0.281*** 0.485*** 0.369*** 0.314*** 0.536*** 0.298*** 0.242*** 0.450***[0.00684] [0.00769] [0.0138] [0.00756] [0.00856] [0.0151] [0.00625] [0.00691] [0.0128]
Standard errors in brackets*** p<0.01, ** p<0.05, * p<0.1Source: Authors’ elaboration based on data from Encuesta Permanente de Hogares (INDEC).
Rates
Total households
Households with children
Households without children
Total households
Households with children
Households without children
Total households
Households with children
Households without children
Entry 0.132*** 0.159*** 0.115*** 0.119*** 0.145*** 0.102*** 0.142*** 0.169*** 0.127***[0.00328] [0.00579] [0.00392] [0.00304] [0.00528] [0.00365] [0.00348] [0.00617] [0.00416]
Exit 0.420*** 0.310*** 0.636*** 0.446*** 0.331*** 0.669*** 0.393*** 0.283*** 0.606***[0.00759] [0.00870] [0.0126] [0.00812] [0.00937] [0.0131] [0.00714] [0.00811] [0.0121]
Standard errors in brackets*** p<0.01, ** p<0.05, * p<0.11/ Metropolitan areas of Recife, Salvador, Belo Horizonte, Rio de Janeiro, São Paulo and Porto Alegre.Source: Authors’ elaboration based on data PME & PNAD (IBGE)
Rates
Total households
Households with children
Households without children
Total households
Households with children
Households without children
Total households
Households with children
Households without children
Entry 0.109*** 0.128*** 0.0879*** 0.0957*** 0.111*** 0.0776*** 0.121*** 0.142*** 0.0994***[0.00791] [0.0117] [0.0104] [0.00728] [0.0107] [0.00967] [0.00850] [0.0126] [0.0111]
Exit 0.452*** 0.456*** 0.444*** 0.500*** 0.530*** 0.450*** 0.433*** 0.439*** 0.424***[0.0223] [0.0282] [0.0369] [0.0244] [0.0306] [0.0406] [0.0207] [0.0253] [0.0357]
Standard errors in brackets*** p<0.01, ** p<0.05, * p<0.1Source: Authors’ elaboration based on data from Encuesta de Hogares de Propósitos Múltiples (INEC).
Rates
Total households
Households with children
Households without children
Total households
Households with children
Households without children
Total households
Households with children
Households without children
Entry 0.153*** 0.184*** 0.110*** 0.138*** 0.166*** 0.0962*** 0.168*** 0.199*** 0.126***[0.00376] [0.00527] [0.00506] [0.00347] [0.00482] [0.00469] [0.00405] [0.00566] [0.00554]
Exit 0.368*** 0.337*** 0.503*** 0.403*** 0.378*** 0.514*** 0.338*** 0.307*** 0.468***[0.00660] [0.00720] [0.0154] [0.00720] [0.00789] [0.0168] [0.00616] [0.00671] [0.0144]
Standard errors in brackets*** p<0.01, ** p<0.05, * p<0.1Source: Authors’ elaboration based on data from Encuesta de Empleo, Desempleo y Subempleo (INEC).
Rates
Total households
Households with children
Households without children
Total households
Households with children
Households without children
Total households
Households with children
Households without children
Entry 0.208*** 0.254*** 0.126*** 0.189*** 0.230*** 0.113*** 0.229*** 0.273*** 0.156***[0.00880] [0.0116] [0.0119] [0.00829] [0.0108] [0.0114] [0.00966] [0.0127] [0.0136]
Exit 0.385*** 0.367*** 0.494*** 0.427*** 0.411*** 0.534*** 0.347*** 0.327*** 0.463***[0.0132] [0.0141] [0.0328] [0.0143] [0.0153] [0.0359] [0.0121] [0.0128] [0.0329]
Standard errors in brackets*** p<0.01, ** p<0.05, * p<0.1Source: Authors’ elaboration based on data from Encuesta Nacional de Hogares (INEI)
Table 5.E. Intensity of poverty transitions: entry and exit rates.Peru. Urban areas. Period 2002-2006.
Poverty line at 100% Poverty line at 90% Poverty line at 110%
Table 5.D. Intensity of poverty transitions: entry and exit rates.Ecuador. Urban areas. Period 2004-2008.
Poverty line at 100% Poverty line at 90% Poverty line at 110%
Table 5.C. Intensity of poverty transitions: entry and exit rates.Costa Rica. Urban areas. Period 2006-2008.
Poverty line at 100% Poverty line at 90% Poverty line at 110%
Table 5.B. Intensity of poverty transitions: entry and exit rates.Brazil. Urban areas. Period 2003-2006. 1/
Poverty line at 100% Poverty line at 90% Poverty line at 110%
Poverty line at 90% Poverty line at 110%Poverty line at 100%
Table 5.A. Intensity of poverty transitions: entry and exit rates.Argentina. Urban areas. Period 2003-2006.
96
N°(1
)SE
(2)
SE(1
)*(2)
=(3)
SE(1
)SE
(2)
SE(1
)*(2)
=(3)
SE(1
)SE
(2)
SE(1
)*(2)
=(3)
SETO
TAL
LABO
R EV
ENTS
37.2
7(0
.71)
40.8
0(1
.19)
15.1
7(0
.51)
40.9
9(0
.85)
35.9
8(1
.31)
14.7
3(0
.60)
26.7
6(1
.19)
61.7
3(2
.52)
16.4
1(0
.99)
1G
rowt
h in
the
num
ber o
f em
ploy
ed m
embe
rs7.
0(0
.38)
37.7
(2.7
7)2.
6(0
.25)
6.4
(0.4
4)27
.1(3
.28)
1.7
(0.2
5)8.
5(0
.75)
60.3
(4.4
4)5.
1(0
.61)
1.1
Gro
wth
in th
e nu
mbe
r of r
egis
tere
d w
age
earn
ers
1.3
(0.1
8)75
.8(4
.83)
1.0
(0.1
7)1.
2(0
.21)
69.2
(6.6
7)0.
8(0
.20)
1.7
(0.3
7)89
.7(6
.46)
1.5
(0.3
6)1.
2G
rowt
h in
the
num
ber o
f non
-regi
ster
ed w
age
earn
ers
3.8
(0.2
9)26
.2(3
.31)
1.0
(0.1
4)3.
8(0
.34)
15.6
(3.1
2)0.
6(0
.13)
3.7
(0.5
1)56
.8(6
.84)
2.1
(0.4
1)1.
3G
rowt
h in
the
num
ber o
f non
-wag
e ea
rner
s1.
9(0
.19)
33.5
(4.6
7)0.
6(0
.11)
1.4
(0.2
0)21
.8(5
.79)
0.3
(0.0
9)3.
1(0
.44)
48.8
(7.1
6)1.
5(0
.30)
2G
rowt
h in
tota
l hou
rly w
age
of m
embe
rs e
mpl
oyed
in b
oth
obse
rvat
ions
10.8
(0.4
4)31
.0(1
.94)
3.3
(0.2
4)12
.7(0
.55)
28.7
(2.0
7)3.
6(0
.31)
5.3
(0.5
9)46
.7(5
.64)
2.5
(0.3
5)
3G
rowt
h in
the
num
ber o
f wor
king
hou
rs o
f mem
bers
em
ploy
ed in
bo
th o
bser
vatio
ns3.
6(0
.26)
22.8
(3.0
4)0.
8(0
.12)
3.9
(0.3
3)19
.7(3
.32)
0.8
(0.1
5)2.
5(0
.41)
36.6
(7.4
4)0.
9(0
.22)
4G
rowt
h in
the
num
ber o
f wor
king
hou
rs a
nd in
the
tota
l hou
rly
wag
e of
mem
bers
em
ploy
ed in
bot
h ob
serv
atio
ns8.
7(0
.48)
43.4
(2.8
0)3.
8(0
.28)
10.0
(0.6
1)38
.1(3
.01)
3.8
(0.3
4)5.
1(0
.55)
73.6
(4.4
5)3.
7(0
.50)
5G
rowt
h in
the
tota
l mon
thly
wag
e of
mem
bers
em
ploy
ed in
bot
h ob
serv
atio
ns a
nd in
the
num
ber o
f em
ploy
ed m
embe
rs7.
2(0
.36)
64.2
(2.4
9)4.
6(0
.29)
7.9
(0.4
4)60
.5(2
.82)
4.8
(0.3
5)5.
4(0
.63)
79.8
(4.8
4)4.
2(0
.56)
TOTA
L NO
N-LA
BOR
EVEN
TS9.
3(0
.39)
35.9
(2.0
7)3.
3(0
.24)
4.6
(0.2
8)10
.2(1
.80)
0.5
(0.0
9)22
.7(1
.14)
50.4
(2.8
7)11
.4(0
.84)
6G
rowt
h in
the
inco
me
from
pen
sion
s 5.
8(0
.32)
42.2
(2.7
5)2.
4(0
.21)
1.6
(0.1
6)11
.9(2
.77)
0.2
(0.0
5)17
.6(1
.05)
50.2
(3.2
9)8.
8(0
.75)
7G
rowt
h in
pub
lic m
onet
ary
trans
fers
(soc
ial p
olicy
) 1.
5(0
.15)
3.8
(1.9
8)0.
1(0
.03)
1.8
(0.1
8)1.
1(0
.58)
0.0
(0.0
1)0.
8(0
.22)
19.7
(11.
86)
0.2
(0.1
1)8
Gro
wth
in o
ther
non
-labo
r inc
omes
2.
0(0
.18)
42.1
(4.5
9)0.
8(0
.12)
1.1
(0.1
5)22
.0(5
.50)
0.2
(0.0
7)4.
3(0
.55)
57.1
(6.5
5)2.
5(0
.40)
III -
Labo
r and
non
-la
bor i
ncom
e ev
ents
9G
rowt
h in
labo
r and
non
-labo
r inc
omes
11.6
(0.4
4)57
.6(1
.97)
6.7
(0.3
6)12
.4(0
.54)
51.9
(2.2
7)6.
4(0
.42)
9.3
(0.7
8)79
.1(3
.44)
7.3
(0.7
1)
IV E
xclu
sivel
y de
mog
raph
ic ev
ents
10Re
duct
ion
in th
e to
tal n
umbe
r of h
ouse
hold
mem
bers
; the
tota
l no
min
al in
com
e re
mai
ns c
onst
ant
5.9
(0.3
6)18
.6(2
.72)
0.8
(0.1
3)6.
3(0
.45)
14.8
(2.7
5)0.
7(0
.14)
4.7
(0.5
6)34
.0(7
.33)
1.1
(0.3
1)
V -
Dem
ogra
phic
even
ts le
adin
g to
in
com
e ch
ange
s11
Gro
wth
in th
e nu
mbe
r of l
abor
or n
on-la
bor i
ncom
e ea
rner
s du
e to
the
entra
nce
of m
embe
rs to
the
hous
ehol
d1.
6(0
.18)
27.7
(5.3
0)0.
4(0
.09)
1.3
(0.2
0)21
.7(6
.46)
0.3
(0.0
9)2.
4(0
.44)
37.5
(9.1
0)0.
8(0
.22)
VI -
Com
bina
tion
of
dem
ogra
phic
and
inco
me
even
ts12
Gro
wth
in th
e to
tal n
omin
al in
com
e an
d gr
owth
in th
e nu
mbe
r of
hous
ehol
d m
embe
rs.
6.4
(0.3
3)58
.4(2
.95)
3.1
(0.2
3)6.
2(0
.38)
50.8
(3.4
6)2.
7(0
.24)
7.0
(0.6
9)78
.9(5
.00)
4.3
(0.5
3)
6.0
(0.3
7)48
.3(3
.22)
2.8
(0.2
3)5.
5(0
.44)
37.0
(3.7
9)1.
9(0
.22)
7.5
(0.6
9)70
.8(4
.26)
5.3
(0.6
0)TO
TAL
HOUS
EHO
LDS
WIT
H E
VENT
S78
.1(0
.59)
43.1
(0.8
3)32
.3(0
.67)
77.3
(0.7
0)36
.6(0
.95)
27.2
(0.7
5)80
.4(1
.11)
60.7
(1.5
1)46
.6(1
.35)
TOTA
L HO
USEH
OLD
S W
ITHO
UT E
VENT
S21
.9(0
.59)
0.0
(0.0
3)0.
0(0
.01)
22.7
(0.7
0)0.
00.
000.
00.
0019
.6(1
.11)
0.1
(0.1
4)0.
0(0
.03)
TOTA
L HO
USEH
OLD
S10
0.0
0.00
0.0
0.00
32.3
0.00
100.
00.
000.
00.
0027
.20.
0010
0.0
0.00
0.0
0.00
46.7
0.00
TOTA
L NU
MBE
R of
HO
USEH
OLD
S W
ITH
EVE
NTS
2,37
6,05
40.
000
0.00
00.
001,
736,
783
0.00
00.
000
0.00
639,
271
0.00
00.
000
0.00
TOTA
L NU
MBE
R of
HO
USEH
OLD
S W
ITHO
UT E
VENT
S66
4,89
20.
000
0.00
00.
0050
8,86
60.
000
0.00
00.
0015
6,02
60.
000
0.00
00.
00TO
TAL
NUM
BER
of H
OUS
EHO
LDS
3,04
0,94
60.
000
0.00
00.
002,
245,
649
0.00
00.
000
0.00
795,
297
0.00
00.
000
0.00
1/ D
ecom
posit
ion
bas
ed o
n eq
uatio
n [2
]So
urce
: Aut
hors
’ ela
bora
tion
base
d on
dat
a fro
m E
ncue
sta
Perm
anen
te d
e Ho
gare
s (IN
DEC)
.
D E M O G R A P H I C &
C O M B I N E D
E V E N T S
VII
- Eve
nts
not c
lass
ified
Exit
Even
ts
N O N D E M O G R A P H I C
I - E
xclu
sivel
y la
bor
inco
me
even
ts
II - E
xclu
sivel
y no
n-la
bor i
ncom
e ev
ents
P(S/
E)Ex
itP(
even
t)P(
S/E)
Tabl
e 6.
A. D
ecom
posi
tion
of th
e ex
it ra
te fr
om p
over
ty. 1
/Ar
gent
ina.
Urb
an a
reas
. Per
iod
2003
200
6.
Tota
l hou
seho
lds
Hous
ehol
ds w
ith c
hild
ren
Hous
ehol
ds w
ithou
t chi
ldre
nP(
even
t)P(
S/E)
Exit
P(ev
ent)
96 97
N°(1
)SE
(2)
SE(1
)*(2)
=(3)
SE(1
)SE
(2)
SE(1
)*(2)
=(3)
SE(1
)SE
(2)
SE(1
)*(2)
=(3)
SETO
TAL
LABO
R EV
ENTS
30.56
(0.71
)37
.93(1
.37)
11.59
(0.50
)34
.76(0
.90)
31.40
(1.52
)10
.91(0
.60)
22.34
(1.10
)57
.85(2
.72)
12.92
(0.89
)1
Grow
th in
the n
umbe
r of e
mpl
oyed
mem
bers
7.7(0
.42)
46.5
(2.84
)3.6
(0.30
)6.7
(0.49
)31
.8(3
.54)
2.1(0
.29)
9.5(0
.78)
66.8
(3.99
)6.4
(0.67
)1.1
Grow
th in
the n
umbe
r of r
egist
ered
wag
e ear
ners
3.0(0
.27)
56.3
(4.50
)1.7
(0.20
)2.6
(0.30
)37
.6(5
.78)
1.0(0
.19)
3.7(0
.52)
81.7
(5.28
)3.0
(0.47
)1.2
Grow
th in
the n
umbe
r of n
on-re
gist
ered
wag
e ear
ners
2.7(0
.26)
36.1
(4.78
)1.0
(0.17
)2.8
(0.34
)29
.8(5
.72)
0.8(0
.20)
2.5(0
.40)
50.3
(8.24
)1.2
(0.30
)1.3
Grow
th in
the n
umbe
r of n
on-w
age e
arne
rs2.0
(0.22
)46
.0(5
.40)
0.9(0
.15)
1.3(0
.21)
24.6
(6.54
)0.3
(0.10
)3.3
(0.48
)62
.6(6
.98)
2.1(0
.39)
2Gr
owth
in to
tal h
ourly
wag
e of m
embe
rs em
ploye
d in b
oth
obse
rvatio
ns10
.8(0
.48)
21.4
(1.96
)2.3
(0.24
)13
.5(0
.65)
18.9
(2.07
)2.6
(0.31
)5.3
(0.58
)34
.1(5
.33)
1.8(0
.35)
3Gr
owth
in th
e num
ber o
f wor
king
hour
s of m
embe
rs em
ploye
d in
both
obse
rvatio
ns2.1
(0.23
)18
.7(4
.27)
0.4(0
.10)
2.6(0
.32)
14.4
(4.20
)0.4
(0.11
)1.3
(0.30
)34
.5(1
1.32)
0.5(0
.20)
4Gr
owth
in th
e num
ber o
f wor
king
hour
s and
in th
e tota
l hou
rly w
age
of m
embe
rs em
ploye
d in b
oth ob
serva
tions
4.1(0
.30)
42.4
(3.74
)1.7
(0.21
)4.9
(0.40
)39
.1(4
.20)
1.9(0
.27)
2.6(0
.42)
54.5
(8.13
)1.4
(0.31
)
5Gr
owth
in th
e tota
l mon
thly
wage
of m
embe
rs em
ploye
d in b
oth
obse
rvatio
ns an
d in t
he nu
mbe
r of e
mplo
yed m
embe
rs5.9
(0.36
)60
.8(3
.09)
3.6(0
.28)
7.1(0
.49)
55.8
(3.54
)4.0
(0.36
)3.6
(0.47
)80
.4(5
.19)
2.9(0
.43)
TOTA
L NO
N-LA
BOUR
EVE
NTS
14.4
(0.54
)69
.1(1
.84)
9.9(0
.47)
8.5(0
.52)
44.4
(3.18
)3.8
(0.36
)25
.9(1
.17)
84.9
(1.78
)22
.0(1
.12)
6Gr
owth
in the
inco
me f
rom
pen
sions
11
.8(0
.50)
77.8
(1.82
)9.2
(0.45
)5.6
(0.43
)55
.8(3
.90)
3.1(0
.33)
24.0
(1.15
)87
.8(1
.72)
21.0
(1.10
)8
Grow
th in
othe
r inc
ome
2.6(0
.24)
29.4
(4.22
)0.8
(0.13
)2.9
(0.31
)22
.7(4
.54)
0.7(0
.15)
2.0(0
.34)
48.9
(8.72
)1.0
(0.25
)III
- Lab
or an
d non
-lab
or in
com
e eve
nts9
Grow
th in
labo
r and
non-
labor
inco
mes
15.1
(0.54
)72
.0(1
.69)
10.9
(0.48
)15
.1(0
.66)
61.6
(2.25
)9.3
(0.54
)15
.0(0
.96)
92.4
(1.71
)13
.9(0
.93)
IV E
xclus
ively
dem
ogra
phic
even
ts10
Redu
ction
in th
e tota
l num
ber o
f hou
seho
ld m
embe
rs; th
e tota
l no
mina
l inco
me r
emain
s con
stant
3.1(0
.26)
22.2
(3.59
)0.7
(0.13
)3.6
(0.34
)22
.6(4
.10)
0.8(0
.17)
2.1(0
.36)
20.8
(7.40
)0.4
(0.18
)
V -
Dem
ogra
phic
even
ts lea
ding t
o inc
ome c
hang
es11
Grow
th in
the n
umbe
r of l
abor
or n
on-la
bor i
ncom
e ear
ners
due t
o the
entra
nce o
f mem
bers
to th
e hou
seho
ld1.7
(0.19
)47
.0(5
.48)
0.8(0
.13)
1.0(0
.17)
20.6
(6.80
)0.2
(0.07
)3.2
(0.44
)62
.8(6
.58)
2.0(0
.36)
VI -
Com
binati
on of
de
mog
raph
ic an
d inc
ome e
vents
12Gr
owth
in th
e tota
l nom
inal
inco
me a
nd gr
owth
in the
num
ber o
f ho
useh
old
mem
bers
.4.7
(0.32
)79
.9(2
.79)
3.7(0
.29)
4.7(0
.39)
73.7
(3.70
)3.5
(0.34
)4.6
(0.56
)92
.2(3
.51)
4.2(0
.54)
6.7(0
.39)
65.5
(2.88
)4.4
(0.31
)4.7
(0.39
)53
.1(4
.15)
2.5(0
.28)
10.6
(0.85
)76
.4(3
.87)
8.1(0
.73)
TOTA
L HO
USEH
OLDS
WIT
H E
VENT
S76
.20.0
055
.1(0
.87)
42.0
(0.76
)72
.4(0
.85)
42.8
(1.09
)31
.0(0
.87)
83.7
(0.95
)75
.9(1
.22)
63.6
(1.26
)TO
TAL
HOUS
EHOL
DS W
ITHO
UT E
VENT
S23
.8(0
.65)
0.00.0
0.
0.00
27.6
(0.85
)0.0
0.00
.0.0
016
.3(0
.95)
0.00.0
0.
0.00
TOTA
L HO
USEH
OLDS
100.0
(0.65
)0.0
0.00
.0.0
010
0.00.0
00.0
0.00
.0.0
010
0.00.0
00.0
0.00
.0.0
0TO
TAL
NUMB
ER o
f HOU
SEHO
LDS
WIT
H E
VENT
S1,9
37,67
80.0
00
0.00
00.0
01,2
18,50
30.0
00
0.00
00.0
071
9,175
0.00
00.0
00
0.00
TOTA
L NU
MBER
of H
OUSE
HOLD
S W
ITHO
UT E
VENT
S60
4,587
0.00
00.0
00
0.00
465,0
040.0
00
0.00
00.0
013
9,583
0.00
00.0
00
0.00
TOTA
L NU
MBER
of H
OUSE
HOLD
S2,5
42,26
50.0
00
0.00
00.0
01,6
83,50
70.0
00
0.00
00.0
085
8,758
0.00
00.0
00
0.00
1/ De
com
posit
ion b
ased
on e
quati
on [2
]2/
Metro
polita
n are
as of
Rec
ife, S
alvad
or, B
elo H
orizo
nte, R
io de
Jane
iro, S
ão P
aulo
and P
orto
Aleg
re.
Tabl
e 6.B
. Dec
ompo
sitio
n of
the e
xit ra
te fr
om p
over
ty. 1
/Br
azil.
Urba
n ar
eas.
Perio
d 20
03-2
006.
2/
Tota
l hou
seho
lds
Hous
ehol
ds w
ith ch
ildre
nHo
useh
olds
with
out c
hild
ren
Exit
Even
ts
N O N D E M O G R A P H I C
I - E
xclus
ively
labor
inc
ome e
vents
P(S/
E)Ex
itP(
even
t)P(
S/E)
P(ev
ent)
P(S/
E)
II - E
xclus
ively
non-
labor
inco
me e
vents
VII -
Eve
nts no
t clas
sified
P(ev
ent)
D E M O G R A P H
I C &
C O M B I N E D
E V E N T S
Exit
98
N°(1
)SE
(2)
SE(1
)*(2)
=(3)
SE(1
)SE
(2)
SE(1
)*(2)
=(3)
SE(1
)SE
(2)
SE(1
)*(2)
=(3)
SETO
TAL
LABO
R EV
ENTS
40.55
(2.15
)63
.64(3
.30)
25.81
(1.92
)45
.54(2
.65)
58.52
(4.01
)26
.65(2
.40)
31.32
(3.54
)77
.41(5
.37)
24.25
(3.23
)1
Grow
th in
the n
umbe
r of e
mpl
oyed
mem
bers
10.0
(1.31
)74
.7(5
.94)
7.5(1
.17)
8.9(1
.57)
73.8
(8.05
)6.6
(1.39
)12
.1(2
.36)
75.9
(8.74
)9.1
(2.13
)1.1
Grow
th in
the n
umbe
r of r
egist
ered
wag
e ear
ners
3.7(0
.85)
91.9
(5.98
)3.4
(0.82
)2.7
(0.89
)95
.0(4
.99)
2.5(0
.88)
5.6(1
.76)
89.2
(10.2
2)5.0
(1.66
)1.2
Grow
th in
the n
umbe
r of n
on-re
gist
ered
wag
e ear
ners
2.9(0
.72)
70.9
(11.1
0)2.1
(0.62
)3.0
(0.93
)69
.9(1
4.09)
2.1(0
.80)
2.9(1
.12)
73.0
(17.9
5)2.1
(0.95
)1.3
Grow
th in
the n
umbe
r of n
on-w
age e
arne
rs3.4
(0.80
)59
.0(1
1.88)
2.0(0
.63)
3.2(1
.00)
59.8
(15.5
2)1.9
(0.78
)3.6
(1.32
)57
.5(1
8.35)
2.1(1
.06)
2Gr
owth
in to
tal h
ourly
wag
e of m
embe
rs em
ploye
d in b
oth
obse
rvatio
ns9.2
(1.26
)45
.3(7
.17)
4.2(0
.87)
12.3
(1.76
)36
.7(7
.35)
4.5(1
.10)
3.6(1
.44)
100.0
0.00
3.6(1
.44)
3Gr
owth
in th
e num
ber o
f wor
king
hour
s of m
embe
rs em
ploye
d in
both
obse
rvatio
ns3.1
(0.74
)21
.4(9
.94)
0.7(0
.34)
2.8(0
.86)
8.6(8
.22)
0.2(0
.24)
3.5(1
.37)
40.6
(19.3
4)1.4
(0.87
)
4Gr
owth
in th
e num
ber o
f wor
king
hour
s and
in th
e tota
l hou
rly
wage
of m
embe
rs em
ploye
d in b
oth ob
serva
tions
8.2(1
.19)
51.9
(7.63
)4.3
(0.86
)10
.4(1
.63)
51.1
(8.38
)5.3
(1.19
)4.1
(1.47
)56
.0(1
8.45)
2.3(1
.07)
5Gr
owth
in th
e tota
l mon
thly
wage
of m
embe
rs em
ploye
d in b
oth
obse
rvatio
ns an
d in t
he nu
mbe
r of e
mplo
yed m
embe
rs10
.1(1
.33)
91.9
(3.70
)9.3
(1.29
)11
.2(1
.71)
90.0
(4.91
)10
.0(1
.64)
8.1(2
.09)
96.6
(3.39
)7.8
(2.07
)
TOTA
L NO
N-LA
BOR
EVEN
TS8.1
(1.24
)20
.7(6
.16)
1.7(0
.57)
3.5(0
.96)
10.0
(7.37
)0.3
(0.26
)16
.8(2
.93)
24.8
(7.83
)4.2
(1.53
)6
Grow
th in
the i
ncom
e fro
m p
ensio
ns
4.9(1
.02)
26.1
(8.47
)1.3
(0.49
)0.6
(0.36
)59
.0(3
0.39)
0.3(0
.26)
12.8
(2.70
)23
.3(8
.70)
3.0(1
.30)
7Gr
owth
in pu
blic m
onet
ary t
rans
fers
(soc
ial po
licy)
2.2(0
.62)
0.00.0
0.
0.00
2.9(0
.89)
0.00.0
0.
0.00
0.9(0
.66)
0.00.0
0.
0.00
8Gr
owth
in ot
her n
on-la
bor in
com
es
1.1(0
.46)
38.1
(20.9
1)0.4
(0.29
)0.0
0.00
0.00.0
00.0
0.00
3.1(1
.30)
38.1
(20.9
1)1.2
(0.83
)
III - L
abor
and n
on-
labor
inco
me e
vents
9Gr
owth
in la
bor a
nd no
n-lab
or in
com
es9.2
(1.25
)63
.2(6
.55)
5.8(0
.99)
9.6(1
.52)
50.1
(8.39
)4.8
(1.10
)8.5
(2.19
)90
.6(6
.49)
7.7(1
.94)
IV E
xclus
ively
dem
ogra
phic
even
ts10
Redu
ction
in th
e tota
l num
ber o
f hou
seho
ld m
embe
rs; th
e tota
l no
mina
l inco
me r
emain
s con
stant
2.8(0
.71)
24.7
(10.5
0)0.7
(0.33
)2.7
(0.89
)25
.6(1
3.78)
0.7(0
.43)
2.9(1
.17)
23.2
(15.8
2)0.7
(0.50
)
V -
Dem
ogra
phic
even
ts lea
ding t
o inc
ome c
hang
es11
Grow
th in
the n
umbe
r of l
abor
or n
on-la
bor i
ncom
e ear
ners
du
e to t
he en
tranc
e of m
embe
rs to
the ho
useh
old0.7
(0.37
)40
.5(2
5.09)
0.3(0
.22)
0.7(0
.46)
0.00.0
00.0
0.00
0.8(0
.63)
100.0
0.00
0.8(0
.63)
VI -
Com
binati
on of
de
mog
raph
ic an
d inc
ome e
vents
12Gr
owth
in th
e tota
l nom
inal
inco
me a
nd gr
owth
in the
num
ber o
f ho
useh
old
mem
bers
.6.8
(1.09
)71
.2(7
.68)
4.9(0
.92)
8.5(1
.50)
72.1
(8.43
)6.2
(1.27
)3.7
(1.37
)67
.3(1
8.37)
2.5(1
.09)
10.5
(1.32
)73
.7(5
.79)
7.7(1
.16)
12.3
(1.76
)76
.9(6
.24)
9.5(1
.58)
7.2(1
.87)
63.6
(13.2
4)4.6
(1.50
)TO
TAL
HOUS
EHOL
DS W
ITH
EVE
NTS
78.8
(1.75
)59
.6(2
.36)
46.9
(2.15
)82
.9(1
.98)
58.1
(2.89
)48
.2(2
.67)
71.2
(3.35
)62
.7(4
.12)
44.7
(3.65
)TO
TAL
HOUS
EHOL
DS W
ITHO
UT E
VENT
S21
.2(1
.75)
0.00.0
00.0
0.00
17.1
(1.98
)0.0
0.00
.0.0
028
.8(3
.35)
0.00.0
0.
0.00
TOTA
L HO
USEH
OLDS
100.0
0.00
0.00.0
00.0
0.00
100.0
0.00
0.00.0
0.
0.00
100.0
0.00
0.00.0
0.
0.00
TOTA
L NU
MBER
of H
OUSE
HOLD
S W
ITH
EVE
NTS
67,63
40.0
00
0.00
00.0
046
,201
0.00
00.0
00
0.00
21,43
30.0
00
0.00
00.0
0TO
TAL
NUMB
ER o
f HOU
SEHO
LDS
WIT
HOUT
EVE
NTS
18,19
80.0
00
0.00
00.0
09,5
240.0
00
0.00
00.0
08,6
740.0
00
0.00
00.0
0TO
TAL
NUMB
ER o
f HOU
SEHO
LDS
85,83
20.0
00
0.00
00.0
055
,725
0.00
00.0
00
0.00
30,10
70.0
00
0.00
00.0
0
1/ De
com
posit
ion b
ased
on e
quati
on [2
]So
urce
: Auth
ors’
elabo
ratio
n bas
ed on
data
from
Enc
uesta
de H
ogar
es de
Pro
pósit
os M
últipl
es (I
NEC)
.
D E M O G R A P H I C &
C O M B I N E D
E V E N T S
VII -
Eve
nts no
t clas
sified
Exit
Even
ts
N O N D E M O G R A P H I C
I - E
xclus
ively
labor
inc
ome e
vents
II - E
xclus
ively
non-
labor
inco
me e
vents
P(S/
E)Ex
itP(
even
t)P(
S/E)
Tabl
e 6.C
. Dec
ompo
sitio
n of
the e
xit ra
te fr
om p
over
ty. 1/
Cost
a Rica
. Urb
an ar
eas.
Perio
d 20
06-2
008.
Tota
l hou
seho
lds
Hous
ehol
ds w
ith ch
ildre
nHo
useh
olds
with
out c
hild
ren
P(ev
ent)
P(S/
E)Ex
itP(
even
t)
98 99
N°(1
)SE
(2)
SE(1
)*(2)
=(3)
SE(1
)SE
(2)
SE(1
)*(2)
=(3)
SE(1
)SE
(2)
SE(1
)*(2)
=(3)
SETO
TAL
LABO
R EV
ENTS
36.84
(0.65
)44
.56(1
.08)
16.41
(0.49
)38
.20(0
.72)
41.97
(1.17
)16
.04(0
.55)
30.73
(1.43
)58
.92(2
.71)
18.10
(1.17
)1
Grow
th in
the
num
ber o
f em
ploy
ed m
embe
rs4.
9(0
.28)
40.8
(2.9
2)2.
0(0
.18)
4.5
(0.3
0)36
.9(3
.31)
1.7
(0.1
9)6.
5(0
.75)
52.8
(5.9
2)3.
4(0
.56)
1.1
Grow
th in
the
num
ber o
f reg
ister
ed w
age e
arne
rs0.
5(0
.09)
67.3
(8.9
3)0.
3(0
.07)
0.4
(0.0
8)53
.5(1
1.34
)0.
2(0
.06)
0.9
(0.2
9)92
.5(7
.34)
0.8
(0.2
9)1.
2Gr
owth
in th
e nu
mbe
r of n
on-re
gist
ered
wag
e ear
ners
2.2
(0.1
9)37
.6(4
.29)
0.8
(0.1
2)2.
2(0
.21)
32.5
(4.5
8)0.
7(0
.12)
2.2
(0.4
6)60
.2(1
0.20
)1.
4(0
.36)
1.3
Grow
th in
the
num
ber o
f non
-wag
e ear
ners
2.2
(0.1
9)38
.5(4
.34)
0.9
(0.1
2)1.
9(0
.21)
38.9
(5.2
2)0.
8(0
.13)
3.4
(0.5
3)37
.6(7
.78)
1.3
(0.3
3)
2Gr
owth
in to
tal h
ourly
wag
e of m
embe
rs e
mplo
yed
in bo
th
obse
rvat
ions
11.9
(0.4
3)37
.9(1
.84)
4.5
(0.2
7)12
.6(0
.48)
35.9
(1.9
6)4.
5(0
.30)
8.8
(0.8
5)50
.4(5
.09)
4.4
(0.6
2)
3Gr
owth
in th
e nu
mbe
r of w
orkin
g ho
urs o
f mem
bers
em
ploye
d in
both
obs
erva
tions
4.2
(0.2
6)21
.0(2
.60)
0.9
(0.1
2)4.
3(0
.29)
15.3
(2.5
2)0.
7(0
.12)
3.8
(0.5
8)49
.5(7
.76)
1.9
(0.4
2)
4Gr
owth
in th
e nu
mbe
r of w
orkin
g ho
urs a
nd in
the
tota
l hou
rly
wage
of m
embe
rs e
mplo
yed
in bo
th o
bser
vatio
ns9.
0(0
.37)
51.6
(2.1
8)4.
6(0
.28)
9.4
(0.4
2)48
.9(2
.36)
4.6
(0.3
1)7.
0(0
.76)
68.2
(5.3
0)4.
8(0
.64)
5Gr
owth
in th
e to
tal m
onth
ly wa
ge o
f mem
bers
em
ploye
d in
both
ob
serv
ation
s and
in th
e nu
mbe
r of e
mplo
yed
mem
bers
6.9
(0.3
3)64
.0(2
.38)
4.4
(0.2
7)7.
4(0
.38)
62.1
(2.5
7)4.
6(0
.31)
4.6
(0.6
4)77
.4(5
.90)
3.6
(0.5
7)
TOTA
L NO
N-LA
BOR
EVEN
TS5.8
(0.32
)39
.6(2
.71)
2.3(0
.20)
3.8(0
.29)
26.4
(3.47
)1.0
(0.15
)14
.7(1
.13)
54.8
(3.88
)8.0
(0.85
)6
Grow
th in
the
inco
me f
rom
pen
sions
0.
7(0
.11)
53.7
(7.4
9)0.
4(0
.08)
0.4
(0.0
9)39
.9(1
0.50
)0.
2(0
.06)
2.1
(0.4
5)66
.7(1
0.42
)1.
4(0
.36)
7Gr
owth
in p
ublic
mon
etar
y tra
nsfe
rs (s
ocial
poli
cy)
0.9
(0.1
2)0.
00.
00.
0.00
0.8
(0.1
3)0.
00.
00.
0.00
1.2
(0.3
3)0.
00.
00.
0.00
8Gr
owth
in o
ther
non
-labo
r inc
omes
4.
1(0
.28)
45.8
(3.2
7)1.
9(0
.19)
2.5
(0.2
4)32
.8(4
.60)
0.8
(0.1
4)11
.3(1
.03)
58.4
(4.3
1)6.
6(0
.78)
III -
Labo
r and
non
-lab
or in
com
e ev
ents
9Gr
owth
in la
bor a
nd n
on-la
bor i
ncom
es12
.5(0
.44)
55.2
(1.8
7)6.
9(0
.34)
12.5
(0.4
9)50
.0(2
.10)
6.3
(0.3
6)12
.5(1
.01)
78.6
(3.4
1)9.
8(0
.91)
IV E
xclus
ively
dem
ogra
phic
even
ts10
Redu
ction
in th
e to
tal n
umbe
r of h
ouse
hold
mem
bers
; the
tota
l no
mina
l inco
me
rem
ains c
onsta
nt5.
6(0
.30)
18.0
(2.0
4)1.
0(0
.13)
6.1
(0.3
5)18
.3(2
.20)
1.1
(0.1
5)3.
5(0
.58)
15.7
(5.2
1)0.
6(0
.20)
V -
Dem
ogra
phic
even
ts lea
ding
to
incom
e ch
ange
s11
Grow
th in
the
num
ber o
f lab
or o
r non
-labo
r inc
ome e
arne
rs
due
to th
e en
tranc
e of
mem
bers
to th
e ho
useh
old2.
4(0
.20)
49.6
(4.3
0)1.
2(0
.15)
2.2
(0.2
2)46
.9(4
.90)
1.0
(0.1
5)3.
1(0
.55)
58.0
(8.8
7)1.
8(0
.42)
VI -
Com
binat
ion o
f de
mog
raph
ic an
d inc
ome
even
ts12
Grow
th in
the
tota
l nom
inal
inco
me a
nd g
rowt
h in
the
num
ber o
f ho
useh
old
mem
bers
.7.
6(0
.35)
69.4
(2.2
0)5.
3(0
.30)
7.9
(0.3
9)67
.4(2
.44)
5.3
(0.3
3)6.
4(0
.78)
80.4
(4.5
3)5.
1(0
.73)
7.1
(0.3
4)53
.4(2
.48)
3.8
(0.2
5)6.
2(0
.36)
50.2
(2.9
7)3.
1(0
.25)
11.1
(0.9
4)61
.3(4
.39)
6.8
(0.7
6)TO
TAL
HOUS
EHOL
DS W
ITH
EVE
NTS
77.8
(0.5
5)47
.4(0
.76)
36.9
(0.6
5)76
.9(0
.62)
44.1
(0.8
4)33
.9(0
.71)
82.0
(1.1
7)61
.3(1
.63)
50.2
(1.5
3)TO
TAL
HOUS
EHOL
DS W
ITHO
UT E
VENT
S22
.2(0
.55)
0.1
(0.0
8)0.
0(0
.02)
23.1
(0.6
2)0.
00.
000.
00.
0018
.0(1
.17)
0.5
(0.5
3)0.
1(0
.10)
TOTA
L HO
USEH
OLDS
100.
036
.910
0.0
33.9
100.
050
.3TO
TAL
NUM
BER
of H
OUSE
HOLD
S W
ITH
EVE
NTS
1,82
3,09
91,
471,
673
351,
426
TOTA
L NU
MBE
R of
HOU
SEHO
LDS
WIT
HOUT
EVE
NTS
519,
960
442,
694
77,2
66TO
TAL
NUM
BER
of H
OUSE
HOLD
S2,
343,
059
1,91
4,36
742
8,69
2
1/ D
ecom
posit
ion b
ased
on
equa
tion
[2]
Sour
ce: A
utho
rs’ e
labor
ation
bas
ed o
n da
ta fr
om E
ncue
sta d
e Em
pleo,
Des
emple
o y S
ubem
pleo
(INEC
).
D E M O G R A P H I C &
C O M B I N E D
E V E N T S
VII
- Eve
nts n
ot cl
assif
ied
Exit
Even
ts
N O N D E M O G R A P H I C
I - E
xclus
ively
labor
inc
ome
even
ts
II - E
xclus
ively
non-
labor
inco
me
even
ts
P(S/
E)Ex
itP(
even
t)P(
S/E)
Tabl
e 6.D
. Dec
ompo
sitio
n of
the e
xit ra
te fr
om p
over
ty. 1
/Ec
uado
r. Ur
ban
area
s. Pe
riod
2004
-200
8.
Tota
l hou
seho
lds
Hous
ehol
ds w
ith ch
ildre
nHo
useh
olds
with
out c
hild
ren
P(ev
ent)
P(S/
E)Ex
itP(
even
t)
100
N°(1
)SE
(2)
SE(1
)*(2)
=(3)
SE(1
)SE
(2)
SE(1
)*(2)
=(3)
SE(1
)SE
(2)
SE(1
)*(2)
=(3)
SETO
TAL
LABO
R EV
ENTS
27.64
(1.11
)42
.58(2
.49)
11.77
(0.84
)28
.11(1
.20)
40.30
(2.66
)11
.33(0
.91)
24.70
(2.90
)58
.61(6
.30)
14.48
(2.13
)1
Grow
th in
the
num
ber o
f em
ploy
ed m
embe
rs6.
0(0
.58)
38.2
(4.8
2)2.
3(0
.37)
5.9
(0.6
2)32
.8(5
.12)
1.9
(0.3
6)6.
8(1
.72)
66.5
(11.
26)
4.6
(1.3
3)1.
1Gr
owth
in th
e nu
mbe
r of r
egist
ered
wag
e ear
ners
0.8
(0.2
2)55
.4(1
3.62
)0.
4(0
.16)
0.8
(0.2
3)44
.9(1
5.29
)0.
3(0
.16)
1.1
(0.6
5)10
0.0
0.00
1.1
(0.6
5)1.
2Gr
owth
in th
e nu
mbe
r of n
on-re
gist
ered
wag
e ear
ners
2.8
(0.4
0)34
.5(6
.83)
1.0
(0.2
4)2.
6(0
.40)
29.6
(7.4
4)0.
8(0
.23)
4.1
(1.4
7)53
.7(1
5.35
)2.
2(0
.97)
1.3
Grow
th in
the
num
ber o
f non
-wag
e ear
ners
2.4
(0.3
7)36
.6(7
.65)
0.9
(0.2
3)2.
5(0
.42)
32.5
(7.8
6)0.
8(0
.24)
1.7
(0.7
7)75
.6(1
7.42
)1.
3(0
.72)
2Gr
owth
in to
tal h
ourly
wag
e of m
embe
rs e
mplo
yed
in bo
th
obse
rvat
ions
5.8
(0.5
6)35
.6(5
.03)
2.1
(0.3
6)5.
9(0
.61)
30.4
(5.2
3)1.
8(0
.36)
5.3
(1.4
0)71
.0(1
2.69
)3.
8(1
.22)
3Gr
owth
in th
e nu
mbe
r of w
orkin
g ho
urs
of m
embe
rs e
mplo
yed
in bo
th o
bser
vatio
ns3.
8(0
.48)
30.1
(5.9
8)1.
1(0
.27)
4.0
(0.5
3)28
.1(6
.17)
1.1
(0.2
8)2.
8(1
.12)
48.0
(20.
59)
1.3
(0.7
8)
4Gr
owth
in th
e nu
mbe
r of w
orkin
g ho
urs
and
in th
e to
tal h
ourly
wa
ge o
f mem
bers
em
ploye
d in
both
obs
erva
tions
4.9
(0.5
1)51
.3(5
.44)
2.5
(0.3
8)5.
1(0
.55)
54.8
(5.6
7)2.
8(0
.43)
4.0
(1.3
1)23
.8(1
3.26
)0.
9(0
.57)
5Gr
owth
in th
e to
tal m
onth
ly wa
ge o
f mem
bers
em
ploye
d in
both
ob
serv
ation
s and
in th
e nu
mbe
r of e
mpl
oyed
mem
bers
7.1
(0.6
2)52
.7(4
.58)
3.7
(0.4
7)7.
3(0
.68)
50.9
(4.9
0)3.
7(0
.51)
5.8
(1.3
9)66
.8(1
1.88
)3.
8(1
.20)
TOTA
L NO
N-LA
BOR
EVEN
TS3.6
(0.47
)32
.7(6
.07)
1.2(0
.27)
2.7(0
.41)
31.3
(7.33
)0.8
(0.24
)9.2
(2.10
)35
.1(1
0.45)
3.2(1
.24)
6Gr
owth
in th
e in
com
e fro
m p
ensio
ns
0.2
(0.1
2)57
.0(3
4.68
)0.
1(0
.09)
0.1
(0.0
8)0.
00.
000.
00.
000.
7(0
.66)
100.
00.
000.
7(0
.66)
7Gr
owth
in p
ublic
mon
etar
y tra
nsfe
rs (s
ocial
poli
cy)
0.1
(0.0
8)0.
00.
00.
0.00
0.1
(0.0
9)0.
00.
00.
0.00
0.0
0.00
0.0
0.00
0.0
0.00
8Gr
owth
in o
ther
non
-labo
r inc
omes
3.
3(0
.45)
32.5
(6.1
9)1.
1(0
.26)
2.5
(0.3
9)33
.8(7
.72)
0.8
(0.2
4)8.
6(2
.01)
30.1
(10.
00)
2.6
(1.0
6)
III -
Labo
r and
non
-lab
or in
com
e ev
ents
9Gr
owth
in la
bor a
nd n
on-la
bor i
ncom
es13
.0(0
.83)
64.9
(3.1
5)8.
4(0
.70)
13.3
(0.9
1)63
.5(3
.39)
8.5
(0.7
7)10
.7(1
.87)
75.5
(7.4
0)8.
1(1
.71)
IV E
xclus
ively
dem
ogra
phic
even
ts10
Redu
ction
in th
e to
tal n
umbe
r of h
ouse
hold
mem
bers
; the
tota
l no
mina
l inco
me
rem
ains c
onsta
nt6.
2(0
.57)
21.2
(4.0
0)1.
3(0
.28)
6.4
(0.6
2)18
.1(3
.98)
1.1
(0.2
8)5.
0(1
.36)
45.6
(14.
40)
2.3
(1.0
4)
V -
Dem
ogra
phic
even
ts lea
ding
to
incom
e ch
ange
s11
Grow
th in
the
num
ber o
f labo
r or n
on-la
bor i
ncom
e ea
rner
s due
to
the
entra
nce
of m
embe
rs to
the
hous
ehold
2.1
(0.3
3)39
.5(7
.83)
0.8
(0.2
2)2.
0(0
.33)
32.1
(8.3
2)0.
6(0
.20)
3.2
(1.1
5)67
.4(1
5.93
)2.
2(1
.00)
VI -
Com
binat
ion o
f de
mog
raph
ic an
d inc
ome
even
ts12
Grow
th in
the
tota
l nom
inal
inco
me
and
grow
th in
the
num
ber o
f ho
useh
old
mem
bers
.8.
5(0
.67)
68.9
(3.7
2)5.
9(0
.58)
8.8
(0.7
3)66
.8(4
.00)
5.8
(0.6
2)7.
0(1
.68)
85.4
(9.4
6)6.
0(1
.57)
10.7
(0.7
7)62
.0(3
.58)
6.6
(0.6
5)9.
8(0
.78)
60.8
(4.0
0)6.
0(0
.65)
15.9
(2.5
8)66
.5(8
.01)
10.6
(2.2
2)TO
TAL
HOUS
EHOL
DS W
ITH
EVE
NTS
71.6
(1.0
3)50
.2(1
.52)
36.0
(1.2
7)71
.0(1
.13)
48.2
(1.6
4)34
.2(1
.36)
75.8
(2.5
7)61
.8(3
.58)
46.8
(3.2
0)TO
TAL
HOUS
EHOL
DS W
ITHO
UT E
VENT
S28
.4(1
.03)
3.6
(0.8
4)1.
0(0
.24)
29.0
(1.1
3)3.
8(0
.92)
1.1
(0.2
7)24
.2(2
.57)
2.0
(2.0
1)0.
5(0
.49)
TOTA
L HO
USEH
OLDS
100.
00.
000.
00.
0037
.00.
0010
0.0
0.00
0.0
0.00
35.3
0.00
100.
00.
000.
00.
0047
.30.
00TO
TAL
NUM
BER
of H
OUSE
HOLD
S W
ITH
EVE
NTS
3,23
7,27
90.
000
0.00
00.
002,
759,
411
0.00
00.
000
0.00
477,
868
0.00
00.
000
0.00
TOTA
L NU
MBE
R of
HOU
SEHO
LDS
WIT
HOUT
EVE
NTS
1,28
0,93
20.
000
0.00
00.
001,
127,
986
0.00
00.
000
0.00
152,
946
0.00
00.
000
0.00
TOTA
L NU
MBE
R of
HOU
SEHO
LDS
4,51
8,21
10.
000
0.00
00.
003,
887,
397
0.00
00.
000
0.00
630,
814
0.00
00.
000
0.00
1/ D
ecom
posit
ion b
ased
on
equa
tion
[2]
Sour
ce: A
utho
rs’ e
labor
ation
bas
ed o
n da
ta fr
om E
ncue
sta N
acion
al de
Hog
ares
(INE
I)
P(S/
E)
Tabl
e 6.E
. Dec
ompo
sitio
n of
the e
xit ra
te fr
om p
over
ty. 1
/Pe
ru. U
rban
area
s. Pe
riod
2002
-200
6.
Tota
l hou
seho
lds
Hous
ehol
ds w
ith ch
ildre
nHo
useh
olds
with
out c
hild
ren
P(ev
ent)
P(S/
E)Ex
itP(
even
t)
D E M O G R A P H I C &
C O M B I N E D
E V E N T S
VII
- Eve
nts n
ot cl
assif
ied
Exit
Even
ts
N O N D E M O G R A P H I C
I - E
xclus
ively
labor
inc
ome
even
ts
II - E
xclus
ively
non-
labor
inco
me
even
ts
P(S/
E)Ex
itP(
even
t)
100 101
N°(1
)SE
(2)
SE(1
)*(2)
=(3)
SE(1
)SE
(2)
SE(1
)*(2)
=(3)
SE(1
)SE
(2)
SE(1
)*(2)
=(3)
SETO
TAL
LABO
R EV
ENTS
14.2
(0.35
)23
.81(1
.2)3.3
4(0
.18)
18.8
(0.66
)29
.89(1
.78)
5.58
(0.38
)11
.62(0
.41)
18.35
(1.50
)2.1
0(0
.19)
1Re
ducti
on in
the
num
ber o
f em
ploy
ed m
embe
rs3.
2(0
.18)
37.1
(2.8
8)1.
2(0
.12)
3.3
(0.3
1)49
.1(4
.88)
1.6
(0.2
3)3.
1(0
.22)
30.1
(3.4
5)0.
9(0
.13)
1.1
Redu
ction
in th
e nu
mbe
r of r
egist
ered
wag
e ear
ners
0.9
(0.1
1)37
.8(6
.04)
0.3
(0.0
7)1.
1(0
.18)
52.6
(8.6
1)0.
5(0
.14)
0.8
(0.1
3)27
.6(7
.94)
0.2
(0.0
8)
1.2
Redu
ction
in th
e nu
mbe
r of n
on-re
gist
ered
wag
e ear
ners
1.2
(0.1
2)34
.2(4
.49)
0.4
(0.0
7)1.
5(0
.22)
38.3
(7.2
7)0.
6(0
.14)
1.1
(0.1
3)31
.1(5
.65)
0.3
(0.0
8)
1.3
Redu
ction
in th
e nu
mbe
r of n
on-w
age e
arne
rs1.
0(0
.10)
40.0
(4.6
1)0.
4(0
.06)
0.7
(0.1
3)66
.8(8
.07)
0.5
(0.1
2)1.
2(0
.13)
30.7
(4.9
5)0.
4(0
.07)
2Re
ducti
on in
tota
l hou
rly w
age o
f mem
bers
em
ploye
d in
both
ob
serv
ation
s4.
6(0
.21)
15.7
(1.7
1)0.
7(0
.09)
6.6
(0.4
2)21
.6(2
.70)
1.4
(0.2
0)3.
5(0
.24)
9.7
(2.0
1)0.
3(0
.07)
3Re
ducti
on in
the
num
ber o
f wor
king
hour
s of m
embe
rs e
mplo
yed
in bo
th o
bser
vatio
ns2.
4(0
.15)
13.6
(2.0
0)0.
3(0
.05)
3.2
(0.3
0)17
.1(3
.19)
0.5
(0.1
1)1.
9(0
.17)
10.3
(2.4
6)0.
2(0
.05)
4Re
ducti
on in
the
num
ber o
f wor
king
hour
s and
in th
e to
tal h
ourly
wa
ge o
f mem
bers
em
ploye
d in
both
obs
erva
tions
2.6
(0.1
6)23
.0(2
.49)
0.6
(0.0
7)4.
0(0
.32)
30.0
(3.6
3)1.
2(0
.17)
1.9
(0.1
7)14
.7(3
.25)
0.3
(0.0
7)
5Re
ducti
on in
the
tota
l mon
thly
wage
of m
embe
rs e
mplo
yed
in bo
th
obse
rvat
ions a
nd in
the
num
ber o
f em
ploy
ed m
embe
rs1.
4(0
.13)
40.0
(4.5
5)0.
5(0
.08)
1.8
(0.2
2)49
.0(6
.41)
0.9
(0.1
6)1.
2(0
.15)
32.7
(6.2
8)0.
4(0
.09)
TOTA
L NO
N-LA
BOR
EVEN
TS7.6
(0.25
)14
.5(1
.17)
1.1(0
.09)
2.9(0
.25)
37.4
(4.20
)1.1
(0.15
)10
.1(0
.36)
10.8
(1.12
)1.1
(0.12
)6
Redu
ction
in th
e in
com
e fro
m p
ensio
ns
4.1
(0.1
8)10
.1(1
.36)
0.4
(0.0
6)0.
7(0
.11)
31.5
(7.7
0)0.
2(0
.06)
6.0
(0.2
8)8.
7(1
.34)
0.5
(0.0
8)
7Re
ducti
on in
pub
lic m
onet
ary t
rans
fers
(soc
ial p
olicy
) 0.
4(0
.06)
39.8
(7.1
4)0.
2(0
.04)
0.8
(0.1
4)49
.1(9
.09)
0.4
(0.1
0)0.
2(0
.04)
17.6
(6.6
4)0.
0(0
.01)
8Re
ducti
on in
oth
er n
on-la
bor i
ncom
es
3.1
(0.1
7)17
.2(1
.95)
0.5
(0.0
6)1.
5(0
.18)
34.1
(5.6
4)0.
5(0
.10)
4.0
(0.2
4)13
.7(1
.99)
0.5
(0.0
8)
III -
Labo
r and
non
-lab
or in
com
e ev
ents
9Re
ducti
on in
labo
r and
non
-labo
r inc
omes
2.3
(0.1
4)31
.3(2
.86)
0.7
(0.0
8)2.
7(0
.25)
42.1
(4.4
3)1.
2(0
.15)
2.0
(0.1
7)23
.0(3
.71)
0.5
(0.0
8)
IV E
xclus
ively
dem
ogra
phic
even
ts10
Grow
th in
the
tota
l num
ber o
f hou
seho
ld m
embe
rs; t
he to
tal
nom
inal in
com
e re
main
s con
stant
5.7
(0.2
2)8.
1(0
.97)
0.5
(0.0
6)6.
2(0
.39)
7.5
(1.5
6)0.
5(0
.10)
5.4
(0.2
6)8.
4(1
.24)
0.4
(0.0
7)
V -
Dem
ogra
phic
even
ts lea
ding
to
incom
e ch
ange
s11
Redu
ction
in th
e nu
mbe
r of l
abor
or n
on-la
bor i
ncom
e ear
ners
du
e to
the
entra
nce
of m
embe
rs to
the
hous
ehold
3.2
(0.1
7)9.
7(1
.87)
0.2
(0.0
4)2.
8(0
.27)
14.4
(3.7
5)0.
3(0
.08)
3.4
(0.2
2)7.
3(2
.07)
0.2
(0.0
5)
VI -
Com
binat
ion o
f de
mog
raph
ic an
d inc
ome
even
ts12
Redu
ction
in th
e to
tal n
omin
al in
com
e and
gro
wth
in th
e nu
mbe
r of
hou
seho
ld m
embe
rs.
1.8
(0.1
3)47
.3(3
.64)
0.8
(0.0
9)2.
4(0
.25)
62.5
(5.2
2)1.
5(0
.20)
1.4
(0.1
4)32
.5(4
.50)
0.4
(0.0
7)
3.0
(0.1
8)28
.5(2
.84)
0.7
(0.0
8)3.
3(0
.30)
50.2
(5.1
7)1.
3(0
.19)
2.9
(0.2
2)16
.3(2
.84)
0.4
(0.0
8)TO
TAL
HOUS
EHOL
DS W
ITH
EVE
NTS
37.7
(0.4
8)20
.5(0
.65)
7.3
(0.2
5)39
.2(0
.81)
30.3
(1.2
2)11
.3(0
.51)
36.9
(0.5
9)14
.6(0
.73)
5.1
(0.2
7)TO
TAL
HOUS
EHOL
DS W
ITHO
UT E
VENT
S62
.3(0
.48)
0.7
(0.1
0)0.
4(0
.06)
60.8
(0.8
1)1.
2(0
.22)
0.7
(0.1
3)63
.1(0
.59)
0.3
(0.0
9)0.
2(0
.06)
TOTA
L HO
USEH
OLDS
100.
00.
000.
00.
007.
70.
0010
0.0
0.00
0.0
0.00
12.1
0.00
100.
00.
000.
00.
005.
30.
00TO
TAL
NUM
BER
of H
OUSE
HOLD
S W
ITH
EVE
NTS
2,49
5,09
60.
000
0.00
00.
0092
4,60
90.
000
0.00
00.
001,
570,
487
0.00
00.
000
0.00
TOTA
L NU
MBE
R of
HOU
SEHO
LDS
WIT
HOUT
EVE
NTS
4,12
2,42
30.
000
0.00
00.
001,
431,
730
0.00
00.
000
0.00
2,69
0,69
30.
000
0.00
00.
00TO
TAL
NUM
BER
of H
OUSE
HOLD
S6,
617,
519
0.00
00.
000
0.00
2,35
6,33
90.
000
0.00
00.
004,
261,
180
0.00
00.
000
0.00
1/ D
ecom
posit
ion b
ased
on
equa
tion
[2]
Sour
ce: A
utho
rs’ e
labor
ation
bas
ed o
n da
ta fr
om E
ncue
sta P
erm
anen
te d
e Ho
gare
s (IN
DEC)
.
P(S/
E)
Tabl
e 7.A
. Dec
ompo
sitio
n of
the e
ntry
rate
from
pov
erty
. 1/
Arge
ntin
a. Ur
ban
area
s. Pe
riod
2003
-200
6.
Tota
l hou
seho
lds
Hous
ehol
ds w
ith ch
ildre
nHo
useh
olds
with
out c
hild
ren
P(ev
ent)
P(S/
E)En
tryP(
even
t)
D E M O G R A P H I C &
C O M B I N E D
E V E N T S
VII
- Eve
nts n
ot cl
assif
ied
Entry
Even
ts
N O N D E M O G R A P H I C
I - E
xclus
ively
labor
inc
ome
even
ts
II - E
xclus
ively
non-
labor
inco
me
even
ts
P(S/
E)En
tryP(
even
t)
102
N°(1
)SE
(2)
SE(1
)*(2)
=(3)
SE(1
)SE
(2)
SE(1
)*(2)
=(3)
SE(1
)SE
(2)
SE(1
)*(2)
=(3)
SETO
TAL
labor
EVE
NTS
12.16
(0.33
)23
.87(1
.22)
2.90
(0.17
)17
.99(0
.63)
28.41
(1.74
)5.1
1(0
.36)
8.60
(0.36
)18
.07(1
.61)
1.56
(0.15
)1
Redu
ction
in th
e num
ber o
f em
ploy
ed m
embe
rs2.4
(0.15
)45
.9(3
.17)
1.1(0
.10)
3.1(0
.28)
49.7
(4.59
)1.5
(0.20
)2.0
(0.18
)42
.2(4
.35)
0.8(0
.11)
1.1Re
ducti
on in
the n
umbe
r of r
egist
ered
wag
e ear
ners
1.2(0
.10)
46.2
(4.50
)0.5
(0.07
)1.8
(0.21
)47
.7(6
.09)
0.8(0
.15)
0.8(0
.11)
44.1
(6.62
)0.3
(0.07
)1.2
Redu
ction
in th
e num
ber o
f non
-regi
ster
ed w
age e
arne
rs0.7
(0.09
)46
.9(6
.21)
0.3(0
.06)
0.9(0
.16)
46.2
(8.78
)0.4
(0.11
)0.6
(0.10
)47
.6(8
.78)
0.3(0
.07)
1.3Re
ducti
on in
the n
umbe
r of n
on-w
age e
arne
rs0.6
(0.07
)44
.0(6
.38)
0.3(0
.05)
0.4(0
.10)
65.3
(10.3
4)0.3
(0.08
)0.7
(0.10
)35
.6(7
.39)
0.2(0
.06)
2Re
ducti
on in
total
hou
rly w
age o
f mem
bers
emplo
yed i
n both
ob
serva
tions
4.5(0
.21)
13.7
(1.55
)0.6
(0.07
)6.5
(0.40
)17
.9(2
.33)
1.2(0
.16)
3.2(0
.23)
8.5(1
.88)
0.3(0
.06)
3Re
ducti
on in
the n
umbe
r of w
orkin
g ho
urs o
f mem
bers
emplo
yed
in bo
th ob
serva
tions
1.6(0
.12)
13.3
(2.76
)0.2
(0.05
)2.3
(0.24
)20
.4(4
.38)
0.5(0
.12)
1.1(0
.13)
4.1(2
.25)
0.0(0
.03)
4Re
ducti
on in
the n
umbe
r of w
orkin
g ho
urs a
nd in
the t
otal h
ourly
wa
ge of
mem
bers
emplo
yed i
n both
obse
rvatio
ns2.4
(0.16
)17
.8(2
.68)
0.4(0
.07)
3.8(0
.33)
20.5
(3.74
)0.8
(0.16
)1.5
(0.15
)13
.6(3
.49)
0.2(0
.06)
5Re
ducti
on in
the t
otal m
onth
ly wa
ge of
mem
bers
emplo
yed i
n both
ob
serva
tions
and i
n the
num
ber o
f em
ploy
ed m
embe
rs1.4
(0.12
)40
.6(4
.15)
0.5(0
.07)
2.3(0
.24)
50.8
(5.45
)1.1
(0.17
)0.8
(0.11
)23
.1(5
.60)
0.2(0
.05)
TOTA
L NO
N-LA
BOR
EVEN
TS17
.4(0
.38)
25.7
(1.03
)4.5
(0.20
)11
.1(0
.50)
35.1
(2.26
)3.9
(0.30
)21
.2(0
.52)
22.7
(1.14
)4.8
(0.27
)6
Redu
ction
in th
e inc
ome f
rom
pen
sions
14
.6(0
.35)
27.3
(1.14
)4.0
(0.19
)8.0
(0.43
)37
.7(2
.69)
3.0(0
.27)
18.6
(0.49
)24
.6(1
.25)
4.6(0
.26)
8Re
ducti
onin
othe
r non
-labo
ur in
com
es
2.8(0
.17)
17.4
(2.23
)0.5
(0.07
)3.1
(0.28
)28
.2(4
.09)
0.9(0
.15)
2.7(0
.21)
9.9(2
.34)
0.3(0
.07)
III - L
abor
and n
on-
labor
inco
me e
vents
9Re
ducti
on in
labo
r and
non-
labor
inco
mes
5.9(0
.23)
32.3
(1.84
)1.9
(0.13
)6.5
(0.39
)41
.0(2
.98)
2.7(0
.24)
5.5(0
.28)
26.0
(2.28
)1.4
(0.15
)
IV E
xclus
ively
dem
ogra
phic
even
ts10
Grow
th in
the to
tal n
umbe
r of h
ouse
hold
mem
bers
; the t
otal
nom
inal in
com
e rem
ains c
onsta
nt4.6
(0.21
)9.9
(1.39
)0.5
(0.07
)3.0
(0.28
)9.5
(3.02
)0.3
(0.10
)5.6
(0.29
)10
.0(1
.55)
0.6(0
.09)
V -
Dem
ogra
phic
even
ts lea
ding t
o inc
ome c
hang
es11
Redu
ction
in th
e num
ber o
f lab
or o
r non
-labo
r inc
ome e
arne
rs
due t
o the
entra
nce o
f mem
bers
to the
hous
ehold
2.1(0
.14)
14.6
(2.36
)0.3
(0.05
)2.2
(0.24
)15
.0(3
.93)
0.3(0
.10)
2.1(0
.17)
14.3
(2.94
)0.3
(0.07
)
VI -
Com
binati
on of
de
mog
raph
ic an
d inc
ome e
vents
12Re
ducti
on in
the t
otal
nom
inal
inco
me a
nd gr
owth
in the
num
ber
of h
ouse
hold
mem
bers
.2.6
(0.15
)42
.2(2
.92)
1.1(0
.10)
1.9(0
.22)
52.2
(5.83
)1.0
(0.16
)3.0
(0.21
)38
.4(3
.31)
1.2(0
.12)
5.2(0
.22)
28.2
(1.87
)1.5
(0.11
)4.2
(0.31
)34
.2(3
.39)
1.5(0
.17)
5.8(0
.30)
25.6
(2.25
)1.5
(0.15
)TO
TAL
HOUS
EHOL
DS W
ITH
EVE
NTS
50.0
(0.50
)25
.2(0
.60)
12.6
(0.32
)47
.0(0
.81)
31.4
(1.07
)14
.7(0
.56)
51.8
(0.63
)21
.8(0
.71)
11.3
(0.39
)TO
TAL
HOUS
EHOL
DS W
ITHO
UT E
VENT
S50
.0(0
.50)
1.1(0
.15)
0.6(0
.07)
53.0
(0.81
)2.3
(0.33
)1.2
(0.17
)48
.2(0
.63)
0.4(0
.11)
0.2(0
.05)
TOTA
L HO
USEH
OLDS
100.0
0.00
0.00.0
013
.20.0
064
.10.0
00.0
0.00
5.10.0
010
0.00.0
00.0
0.00
11.5
0.00
TOTA
L NU
MBER
of H
OUSE
HOLD
S W
ITH
EVE
NTS
3,176
,849
0.00
00.0
00
0.00
1,130
,552
0.00
00.0
00
0.00
2,046
,297
0.00
00.0
00
0.00
TOTA
L NU
MBER
of H
OUSE
HOLD
S W
ITHO
UT E
VENT
S3,1
77,61
80.0
00
0.00
00.0
01,2
76,39
00.0
00
0.00
00.0
01,9
01,22
80.0
00
0.00
00.0
0TO
TAL
NUMB
ER o
f HOU
SEHO
LDS
6,354
,467
0.00
00.0
00
0.00
2,406
,942
0.00
00.0
00
0.00
3,947
,525
0.00
00.0
00
0.00
1/ De
com
posit
ion b
ased
on e
quati
on [2
]So
urce
: Auth
ors’
elabo
ratio
n bas
ed on
data
PME
& PN
AD (I
BGE)
P(S/
E)
Tabl
e 7.B
. Dec
ompo
sitio
n of
the e
ntry
rate
to p
over
ty. 1/
Braz
il. Ur
ban
area
s. Pe
riod
2003
-200
6. 2/
Tota
l hou
seho
lds
Hous
ehol
ds w
ith ch
ildre
nHo
useh
olds
with
out c
hild
ren
P(ev
ent)
P(S/
E)En
tryP(
even
t)
D E M O G R A P H I
C &
C O M B I N E D
E V E N T S
VII -
Eve
nts no
t clas
sified
Entry
Even
ts
N O N D E M O G R A P H I C
I - E
xclus
ively
labor
inc
ome e
vents
II - E
xclus
ively
non-
labor
inco
me e
vents
P(S/
E)En
tryP(
even
t)
102 103
N°(1
)SE
(2)
SE(1
)*(2)
=(3)
SE(1
)SE
(2)
SE(1
)*(2)
=(3)
SE(1
)SE
(2)
SE(1
)*(2)
=(3)
SETO
TAL
LABO
R EV
ENTS
24.8
7(1
.06)
23.1
3(2
.08)
5.75
(0.5
7)28
.40
(1.4
9)27
.91
(2.8
2)7.
93(0
.90)
20.6
5(1
.50)
15.2
5(2
.83)
3.15
(0.6
2)1
Redu
ctio
n in
the
num
ber o
f em
ploy
ed m
embe
rs6.
4(0
.61)
37.0
(4.7
1)2.
4(0
.37)
5.3
(0.7
5)51
.9(7
.21)
2.8
(0.5
5)7.
7(0
.98)
24.6
(5.6
6)1.
9(0
.49)
1.1
Redu
ctio
n in
the
num
ber o
f reg
iste
red
wag
e ea
rner
s3.
3(0
.44)
43.5
(6.8
1)1.
4(0
.30)
2.5
(0.5
2)68
.2(9
.66)
1.7
(0.4
4)4.
2(0
.74)
26.0
(7.7
3)1.
1(0
.37)
1.2
Redu
ctio
n in
the
num
ber o
f non
-regi
ster
ed w
age
earn
ers
1.7
(0.3
2)24
.6(7
.84)
0.4
(0.1
5)1.
6(0
.43)
26.7
(11.
18)
0.4
(0.2
1)1.
8(0
.49)
22.4
(10.
92)
0.4
(0.2
2)1.
3Re
duct
ion
in th
e nu
mbe
r of n
on-w
age
earn
ers
1.4
(0.3
0)36
.8(1
0.01
)0.
5(0
.18)
1.2
(0.3
6)52
.4(1
4.77
)0.
6(0
.26)
1.7
(0.4
8)23
.2(1
2.63
)0.
4(0
.24)
2Re
duct
ion
in to
tal h
ourly
wag
e of
mem
bers
em
ploy
ed in
bot
h ob
serv
atio
ns8.
7(0
.71)
12.7
(2.7
5)1.
1(0
.25)
11.3
(1.0
7)17
.4(3
.75)
2.0
(0.4
6)5.
6(0
.86)
1.2
(1.2
0)0.
1(0
.07)
3Re
duct
ion
in th
e nu
mbe
r of w
orki
ng h
ours
of m
embe
rs e
mpl
oyed
in
bot
h ob
serv
atio
ns2.
5(0
.39)
21.1
(6.3
0)0.
5(0
.18)
3.2
(0.5
9)25
.9(8
.25)
0.8
(0.3
0)1.
8(0
.49)
10.7
(8.0
5)0.
2(0
.15)
4Re
duct
ion
in th
e nu
mbe
r of w
orki
ng h
ours
and
in th
e to
tal h
ourly
w
age
of m
embe
rs e
mpl
oyed
in b
oth
obse
rvat
ions
3.0
(0.4
1)12
.2(4
.39)
0.4
(0.1
4)3.
4(0
.61)
15.1
(6.4
4)0.
5(0
.24)
2.4
(0.5
4)7.
3(4
.27)
0.2
(0.1
0)
5Re
duct
ion
in th
e to
tal m
onth
ly w
age
of m
embe
rs e
mpl
oyed
in b
oth
obse
rvat
ions
and
in th
e nu
mbe
r of e
mpl
oyed
mem
bers
4.2
(0.5
0)32
.5(5
.61)
1.4
(0.2
9)5.
1(0
.74)
35.9
(7.0
6)1.
8(0
.44)
3.2
(0.6
6)25
.9(9
.01)
0.8
(0.3
3)
TOTA
L NO
N-LA
BOR
EVEN
TS3.
7(0
.49)
13.2
(4.3
5)0.
5(0
.18)
2.3
(0.5
2)5.
3(5
.18)
0.1
(0.1
2)5.
4(0
.87)
17.2
(5.8
8)0.
9(0
.36)
6Re
duct
ion
in th
e in
com
e fro
m p
ensi
ons
1.6
(0.3
1)18
.4(7
.65)
0.3
(0.1
3)0.
4(0
.21)
0.0
0.00
0.0
0.00
3.0
(0.6
3)20
.9(8
.57)
0.6
(0.2
9)7
Redu
ctio
n in
pub
lic m
onet
ary
trans
fers
(soc
ial p
olicy
) 0.
1(0
.09)
0.0
0.00
.0.
000.
3(0
.16)
0.0
0.00
.0.
000.
00.
000.
00.
000.
00.
008
Redu
ctio
n in
oth
er n
on-la
bor i
ncom
es
2.0
(0.3
8)10
.0(5
.32)
0.2
(0.1
2)1.
7(0
.45)
7.3
(7.0
4)0.
1(0
.12)
2.4
(0.6
3)12
.3(7
.68)
0.3
(0.2
1)
III -
Labo
r and
non
-la
bor i
ncom
e ev
ents
9Re
duct
ion
in la
bor a
nd n
on-la
bor i
ncom
es3.
6(0
.47)
24.0
(5.5
5)0.
9(0
.23)
2.6
(0.5
1)17
.6(7
.95)
0.5
(0.2
3)4.
8(0
.84)
28.1
(7.4
9)1.
4(0
.43)
IV E
xclu
sivel
y de
mog
raph
ic ev
ents
10G
rowt
h in
the
tota
l num
ber o
f hou
seho
ld m
embe
rs; t
he to
tal
nom
inal
inco
me
rem
ains
con
stan
t6.
9(0
.62)
3.5
(1.5
3)0.
2(0
.11)
6.1
(0.8
0)0.
00.
000.
00.
007.
9(0
.97)
6.6
(2.8
9)0.
5(0
.24)
V -
Dem
ogra
phic
even
ts le
adin
g to
in
com
e ch
ange
s11
Redu
ctio
n in
the
num
ber o
f lab
or o
r non
-labo
r inc
ome
earn
ers
due
to th
e en
tranc
e of
mem
bers
to th
e ho
useh
old
0.3
(0.1
2)9.
4(9
.35)
0.0
(0.0
3)0.
1(0
.12)
0.0
0.00
0.0
0.00
0.5
(0.2
3)12
.1(1
1.98
)0.
1(0
.06)
VI -
Com
bina
tion
of
dem
ogra
phic
and
inco
me
even
ts12
Redu
ctio
n in
the
tota
l nom
inal
inco
me
and
grow
th in
the
num
ber
of h
ouse
hold
mem
bers
.2.
9(0
.41)
36.1
(6.8
8)1.
0(0
.25)
2.3
(0.5
0)43
.2(1
0.83
)1.
0(0
.33)
3.6
(0.6
8)30
.7(8
.71)
1.1
(0.3
7)
6.2
(0.6
0)43
.2(4
.98)
2.7
(0.4
0)7.
4(0
.89)
46.3
(6.2
4)3.
4(0
.61)
4.8
(0.7
8)37
.5(8
.18)
1.8
(0.4
9)TO
TAL
HOUS
EHO
LDS
WIT
H E
VENT
S48
.5(1
.23)
22.9
(1.4
8)11
.1(0
.77)
49.2
(1.6
7)26
.3(2
.10)
12.9
(1.1
3)47
.7(1
.83)
18.7
(2.0
2)8.
9(1
.03)
TOTA
L HO
USEH
OLD
S W
ITHO
UT E
VENT
S51
.5(1
.23)
0.1
(0.1
2)0.
1(0
.06)
50.8
(1.6
7)0.
2(0
.23)
0.1
(0.1
2)52
.3(1
.83)
0.0
0.00
0.0
0.00
TOTA
L HO
USEH
OLD
S10
0.0
0.00
0.0
0.00
11.2
0.00
100.
00.
000.
00.
0013
.10.
0010
0.0
0.00
0.0
0.00
8.9
0.00
TOTA
L NU
MBE
R of
HO
USEH
OLD
S W
ITH
EVE
NTS
133,
090
0.00
00.
000
0.00
73,5
320.
000
0.00
00.
0059
,558
0.00
00.
000
0.00
TOTA
L NU
MBE
R of
HO
USEH
OLD
S W
ITHO
UT E
VENT
S14
1,32
60.
000
0.00
00.
0075
,924
0.00
00.
000
0.00
65,4
020.
000
0.00
00.
00TO
TAL
NUM
BER
of H
OUS
EHO
LDS
274,
416
0.00
00.
000
0.00
149,
456
0.00
00.
000
0.00
124,
960
0.00
00.
000
0.00
1/ D
ecom
posit
ion
bas
ed o
n eq
uatio
n [2
]So
urce
: Aut
hors
’ ela
bora
tion
base
d on
dat
a fro
m E
ncue
sta
de H
ogar
es d
e Pr
opós
itos
Múl
tiple
s (IN
EC).
P(S/
E)
Tabl
e 7.
C. D
ecom
posi
tion
of th
e en
try ra
te to
pov
erty
. 1/
Cost
a Ri
ca. U
rban
are
as. P
erio
d 20
06-2
008.
Tota
l hou
seho
lds
Hous
ehol
ds w
ith c
hild
ren
Hous
ehol
ds w
ithou
t chi
ldre
nP(
even
t)P(
S/E)
Entry
P(ev
ent)
D E M O G R A P H I C &
C O M B I N E D
E V E N T S
VII
- Eve
nts
not c
lass
ified
Entry
Even
ts
N O N D E M O G R A P H I C
I - E
xclu
sivel
y lab
or
inco
me
even
ts
II - E
xclu
sivel
y non
-la
bor i
ncom
e ev
ents
P(S/
E)En
tryP(
even
t)
104
N°(1
)SE
(2)
SE(1
)*(2)
=(3)
SE(1
)SE
(2)
SE(1
)*(2)
=(3)
SE(1
)SE
(2)
SE(1
)*(2)
=(3)
SETO
TAL
LABO
R EV
ENTS
21.9
0(0
.43)
29.3
4(0
.98)
6.42
(0.2
5)25
.25
(0.5
8)33
.89
(1.2
5)8.
56(0
.37)
17.1
1(0
.61)
19.7
6(1
.48)
3.38
(0.2
8)1
Red
uctio
n in
the
num
ber o
f em
ploy
ed m
embe
rs3.
0(0
.18)
35.6
(2.7
6)1.
1(0
.10)
3.2
(0.2
4)37
.6(3
.56)
1.2
(0.1
4)2.
8(0
.27)
32.4
(4.3
6)0.
9(0
.15)
1.1
Red
uctio
n in
the
num
ber o
f reg
iste
red
wag
e ea
rner
s0.
6(0
.08)
29.1
(6.1
4)0.
2(0
.05)
0.7
(0.1
2)33
.3(8
.08)
0.2
(0.0
7)0.
6(0
.12)
21.7
(8.9
7)0.
1(0
.06)
1.2
Red
uctio
n in
the
num
ber o
f non
-regi
ster
ed w
age
earn
ers
1.0
(0.1
0)37
.0(4
.75)
0.4
(0.0
6)1.
1(0
.14)
36.6
(5.8
5)0.
4(0
.08)
0.9
(0.1
5)37
.6(8
.10)
0.3
(0.0
9)
1.3
Red
uctio
n in
the
num
ber o
f non
-wag
e ea
rner
s1.
4(0
.12)
37.5
(4.0
8)0.
5(0
.07)
1.4
(0.1
6)40
.4(5
.39)
0.6
(0.1
0)1.
4(0
.19)
33.2
(6.1
5)0.
5(0
.10)
2R
educ
tion
in to
tal h
ourly
wag
e of
mem
bers
em
ploy
ed in
bot
h ob
serv
atio
ns8.
8(0
.29)
25.7
(1.5
0)2.
3(0
.15)
10.2
(0.4
0)31
.5(1
.93)
3.2
(0.2
3)6.
8(0
.40)
13.4
(2.0
3)0.
9(0
.15)
3R
educ
tion
in th
e nu
mbe
r of w
orki
ng h
ours
of m
embe
rs e
mpl
oyed
in
bot
h ob
serv
atio
ns2.
7(0
.16)
17.6
(2.2
7)0.
5(0
.07)
3.1
(0.2
3)16
.8(2
.73)
0.5
(0.0
9)2.
3(0
.23)
19.1
(4.0
6)0.
4(0
.10)
4R
educ
tion
in th
e nu
mbe
r of w
orki
ng h
ours
and
in th
e to
tal h
ourly
w
age
of m
embe
rs e
mpl
oyed
in b
oth
obse
rvat
ions
4.7
(0.2
1)28
.1(2
.02)
1.3
(0.1
1)5.
3(0
.29)
32.5
(2.5
7)1.
7(0
.16)
3.8
(0.3
1)19
.1(3
.08)
0.7
(0.1
3)
5R
educ
tion
in th
e to
tal m
onth
ly w
age
of m
embe
rs e
mpl
oyed
in b
oth
obse
rvat
ions
and
in th
e nu
mbe
r of e
mpl
oyed
mem
bers
2.7
(0.1
6)48
.4(3
.11)
1.3
(0.1
1)3.
5(0
.24)
54.4
(3.5
3)1.
9(0
.18)
1.4
(0.1
9)27
.4(5
.49)
0.4
(0.0
9)
TOTA
L NO
N-LA
BOR
EVEN
TS5.
1(0
.23)
18.9
(1.7
0)1.
0(0
.10)
3.6
(0.2
4)19
.2(2
.66)
0.7
(0.1
0)7.
4(0
.42)
18.7
(2.2
2)1.
4(0
.18)
6R
educ
tion
in th
e in
com
e fro
m p
ensi
ons
1.0
(0.1
1)15
.0(3
.59)
0.2
(0.0
4)0.
5(0
.09)
29.8
(8.5
8)0.
2(0
.05)
1.7
(0.2
2)8.
9(3
.24)
0.2
(0.0
6)
7R
educ
tion
in p
ublic
mon
etar
y tra
nsfe
rs (s
ocia
l pol
icy)
0.
1(0
.03)
4.7
(4.7
2)0.
0(0
.00)
0.1
(0.0
5)6.
3(6
.37)
0.0
(0.0
1)0.
1(0
.04)
0.0
0.00
0.0
0.00
8R
educ
tion
in o
ther
non
-labo
r inc
omes
4.
0(0
.20)
20.2
(1.9
6)0.
8(0
.09)
2.9
(0.2
2)17
.9(2
.79)
0.5
(0.0
9)5.
6(0
.37)
22.0
(2.7
1)1.
2(0
.17)
III -
Labo
r and
non
-la
bor i
ncom
e ev
ents
9R
educ
tion
in la
bor a
nd n
on-la
bor i
ncom
es6.
5(0
.25)
37.3
(1.9
3)2.
4(0
.16)
7.0
(0.3
4)43
.5(2
.46)
3.1
(0.2
2)5.
8(0
.38)
26.4
(2.9
4)1.
5(0
.20)
IV E
xclu
sive
ly de
mog
raph
ic e
vent
s10
Gro
wth
in th
e to
tal n
umbe
r of h
ouse
hold
mem
bers
; the
tota
l no
min
al in
com
e re
mai
ns c
onst
ant
8.4
(0.2
9)8.
4(0
.96)
0.7
(0.0
8)6.
8(0
.34)
7.5
(1.3
3)0.
5(0
.09)
10.8
(0.5
1)9.
1(1
.37)
1.0
(0.1
5)
V -
Dem
ogra
phic
ev
ents
lead
ing
to
inco
me
chan
ges
11R
educ
tion
in th
e nu
mbe
r of l
abor
or n
on-la
bor i
ncom
e ea
rner
s du
e to
the
entra
nce
of m
embe
rs to
the
hous
ehol
d2.
8(0
.17)
24.7
(2.5
5)0.
7(0
.08)
2.6
(0.2
1)32
.7(3
.74)
0.9
(0.1
2)3.
2(0
.28)
15.3
(3.2
4)0.
5(0
.11)
VI -
Com
bina
tion
of
dem
ogra
phic
and
in
com
e ev
ents
12R
educ
tion
in th
e to
tal n
omin
al in
com
e an
d gr
owth
in th
e nu
mbe
r of
hou
seho
ld m
embe
rs.
5.4
(0.2
4)43
.5(2
.16)
2.4
(0.1
6)5.
5(0
.31)
45.5
(2.8
2)2.
5(0
.21)
5.4
(0.3
7)40
.6(3
.36)
2.2
(0.2
3)
6.8
(0.2
6)25
.3(1
.67)
1.7
(0.1
3)7.
1(0
.34)
30.0
(2.2
4)2.
1(0
.19)
6.3
(0.4
0)17
.7(2
.38)
1.1
(0.1
6)TO
TAL
HOUS
EHO
LDS
WIT
H E
VENT
S57
.1(0
.52)
26.8
(0.6
0)15
.3(0
.37)
57.9
(0.6
7)31
.6(0
.81)
18.3
(0.5
2)55
.9(0
.82)
19.8
(0.8
5)11
.1(0
.50)
TOTA
L HO
USEH
OLD
S W
ITHO
UT E
VENT
S42
.9(0
.52)
0.6
(0.1
2)0.
3(0
.05)
42.1
(0.6
7)0.
9(0
.19)
0.4
(0.0
8)44
.1(0
.82)
0.2
(0.0
9)0.
1(0
.04)
TOTA
L HO
USEH
OLD
S10
0.0
15.6
100.
018
.710
0.0
11.1
TOTA
L NU
MBE
R of
HO
USEH
OLD
S W
ITH
EVE
NTS
2,35
8,52
11,
405,
627
952,
894
TOTA
L NU
MBE
R of
HO
USEH
OLD
S W
ITHO
UT E
VENT
S1,
774,
813
1,02
3,10
875
1,70
5TO
TAL
NUM
BER
of H
OUS
EHO
LDS
4,13
3,33
42,
428,
735
1,70
4,59
9
1/ D
ecom
posi
tion
bas
ed o
n eq
uatio
n [2
]So
urce
: Aut
hors
’ ela
bora
tion
base
d on
dat
a fro
m E
ncue
sta
de E
mpl
eo, D
esem
pleo
y S
ubem
pleo
(IN
EC).
P(S/
E)
Tabl
e 7.
D. D
ecom
posi
tion
of th
e en
try ra
te to
pov
erty
. 1/
Ecua
dor.
Urba
n ar
eas.
Per
iod
2004
-200
8.
Tota
l hou
seho
lds
Hous
ehol
ds w
ith c
hild
ren
Hous
ehol
ds w
ithou
t chi
ldre
nP(
even
t)P(
S/E)
Entry
P(ev
ent)
D E M O G R A P H I C &
C O M B I N E D
E V E N T S
VII
- Eve
nts
not c
lass
ified
Entry
Even
ts
N O N D E M O G R A P H I C
I - E
xclu
sive
ly la
bor
inco
me
even
ts
II - E
xclu
sive
ly no
n-la
bor i
ncom
e ev
ents
P(S/
E)En
tryP(
even
t)
104 105
N°
(1)
SE(2
)SE
(1)*
(2)=
(3)
SE(1
)SE
(2)
SE(1
)*(2
)=(3
)SE
(1)
SE(2
)SE
(1)*
(2)=
(3)
SETO
TAL
LAB
OR
EVE
NTS
18.9
1(0
.85)
32.8
7(2
.26)
6.22
(0.5
0)20
.15
(1.0
9)41
.28
(2.8
7)8.
32(0
.72)
16.7
3(1
.35)
15.1
1(2
.85)
2.53
(0.5
1)1
Red
uctio
n in
the
num
ber o
f em
ploy
ed m
embe
rs4.
0(0
.41)
47.4
(5.3
9)1.
9(0
.27)
3.8
(0.4
9)57
.9(6
.61)
2.2
(0.3
8)4.
2(0
.75)
30.7
(7.7
8)1.
3(0
.36)
1.1
Red
uctio
n in
the
num
ber o
f reg
iste
red
wag
e ea
rner
s1.
2(0
.23)
40.1
(9.2
9)0.
5(0
.14)
1.1
(0.2
7)60
.4(1
2.49
)0.
6(0
.20)
1.4
(0.4
1)12
.3(8
.54)
0.2
(0.1
2)
1.2
Red
uctio
n in
the
num
ber o
f non
-reg
iste
red
wag
e ea
rner
s1.
2(0
.23)
59.4
(9.4
7)0.
7(0
.17)
1.2
(0.2
7)61
.5(1
1.37
)0.
7(0
.21)
1.4
(0.4
2)56
.1(1
6.12
)0.
8(0
.27)
1.3
Red
uctio
n in
the
num
ber o
f non
-wag
e ea
rner
s1.
5(0
.26)
43.4
(8.5
1)0.
7(0
.17)
1.6
(0.3
3)53
.4(1
0.59
)0.
8(0
.24)
1.5
(0.4
2)24
.3(1
1.96
)0.
4(0
.20)
2R
educ
tion
in to
tal h
ourly
wag
e of
mem
bers
em
ploy
ed in
bot
h ob
serv
atio
ns5.
7(0
.52)
24.1
(3.6
8)1.
4(0
.24)
6.1
(0.6
5)29
.5(4
.74)
1.8
(0.3
3)5.
1(0
.84)
12.8
(5.2
7)0.
7(0
.29)
3R
educ
tion
in th
e nu
mbe
r of w
orki
ng h
ours
of m
embe
rs e
mpl
oyed
in
bot
h ob
serv
atio
ns1.
9(0
.28)
25.5
(6.4
6)0.
5(0
.14)
2.0
(0.3
5)38
.4(8
.82)
0.8
(0.2
2)1.
8(0
.45)
0.0
0.00
0.0
0.00
4R
educ
tion
in th
e nu
mbe
r of w
orki
ng h
ours
and
in th
e to
tal h
ourly
w
age
of m
embe
rs e
mpl
oyed
in b
oth
obse
rvat
ions
3.7
(0.4
0)29
.3(5
.23)
1.1
(0.2
4)4.
3(0
.54)
37.4
(6.5
5)1.
6(0
.36)
2.8
(0.5
5)7.
7(4
.75)
0.2
(0.1
4)
5R
educ
tion
in th
e to
tal m
onth
ly w
age
of m
embe
rs e
mpl
oyed
in b
oth
obse
rvat
ions
and
in th
e nu
mbe
r of e
mpl
oyed
mem
bers
3.6
(0.3
9)38
.5(5
.34)
1.4
(0.2
4)4.
0(0
.52)
49.1
(6.7
1)2.
0(0
.35)
2.9
(0.5
6)13
.0(6
.45)
0.4
(0.2
0)
TOTA
L N
ON
-LAB
OR
EVE
NTS
4.8
(0.4
9)14
.8(3
.39)
0.7
(0.1
7)4.
1(0
.56)
19.0
(4.9
4)0.
8(0
.22)
5.9
(0.8
6)9.
7(4
.43)
0.6
(0.2
7)6
Red
uctio
n in
the
inco
me
from
pen
sion
s 0.
8(0
.18)
8.8
(5.1
9)0.
1(0
.04)
0.4
(0.1
6)20
.8(1
4.07
)0.
1(0
.06)
1.4
(0.4
0)2.
9(2
.96)
0.0
(0.0
4)
7R
educ
tion
in p
ublic
mon
etar
y tr
ansf
ers
(soc
ial p
olic
y)
0.0
0.00
0.0
0.00
0.0
0.00
0.0
0.00
0.0
0.00
0.0
0.00
0.0
0.00
0.0
0.00
0.0
0.00
8R
educ
tion
in o
ther
non
-labo
r inc
omes
4.
0(0
.44)
15.9
(3.8
5)0.
6(0
.16)
3.7
(0.5
1)18
.8(5
.16)
0.7
(0.2
1)4.
5(0
.76)
11.9
(5.6
5)0.
5(0
.27)
III -
Labo
r and
non
-la
bor i
ncom
e ev
ents
9R
educ
tion
in la
bor a
nd n
on-la
bor i
ncom
es11
.0(0
.68)
40.1
(3.1
3)4.
4(0
.44)
12.1
(0.8
6)44
.9(3
.75)
5.5
(0.6
0)9.
1(1
.07)
28.8
(5.4
3)2.
6(0
.56)
IV E
xclu
sive
ly de
mog
raph
ic e
vent
s10
Gro
wth
in th
e to
tal n
umbe
r of h
ouse
hold
mem
bers
; the
tota
l no
min
al in
com
e re
mai
ns c
onst
ant
9.4
(0.6
1)9.
1(1
.84)
0.9
(0.1
8)8.
1(0
.72)
8.2
(2.3
7)0.
7(0
.20)
11.6
(1.1
5)10
.3(2
.86)
1.2
(0.3
5)
V -
Dem
ogra
phic
ev
ents
lead
ing
to
inco
me
chan
ges
11R
educ
tion
in th
e nu
mbe
r of l
abor
or n
on-la
bor i
ncom
e ea
rner
s du
e to
the
entra
nce
of m
embe
rs to
the
hous
ehol
d2.
4(0
.31)
23.1
(5.5
3)0.
5(0
.15)
2.2
(0.3
8)34
.4(8
.05)
0.8
(0.2
3)2.
7(0
.55)
6.9
(4.4
8)0.
2(0
.12)
VI -
Com
bina
tion
of
dem
ogra
phic
and
in
com
e ev
ents
12R
educ
tion
in th
e to
tal n
omin
al in
com
e an
d gr
owth
in th
e nu
mbe
r of
hou
seho
ld m
embe
rs.
11.9
(0.7
0)46
.1(3
.00)
5.5
(0.5
0)12
.7(0
.89)
49.5
(3.5
9)6.
3(0
.66)
10.5
(1.1
4)38
.7(5
.19)
4.1
(0.7
1)
8.2
(0.5
8)39
.8(3
.59)
3.2
(0.3
8)8.
6(0
.73)
45.6
(4.2
6)3.
9(0
.51)
7.3
(0.9
9)27
.8(6
.16)
2.0
(0.5
2)TO
TAL
HO
USE
HO
LDS
WIT
H E
VEN
TS66
.5(0
.91)
32.3
(1.2
1)21
.5(0
.90)
68.0
(1.1
8)38
.5(1
.52)
26.2
(1.1
8)63
.8(1
.43)
20.7
(1.7
8)13
.2(1
.22)
TOTA
L H
OU
SEH
OLD
S W
ITH
OU
T EV
ENTS
33.5
(0.9
1)1.
2(0
.38)
0.4
(0.1
3)32
.0(1
.18)
1.2
(0.4
8)0.
4(0
.16)
36.2
(1.4
3)1.
2(0
.63)
0.5
(0.2
3)TO
TAL
HO
USE
HO
LDS
100.
00.
000.
00.
0021
.90.
0010
0.0
0.00
0.0
0.00
26.6
0.00
100.
00.
000.
00.
0013
.70.
00TO
TAL
NU
MB
ER o
f HO
USE
HO
LDS
WIT
H E
VEN
TS4,
291,
020
0.00
00.
000
0.00
2,79
5,66
60.
000
0.00
00.
001,
495,
354
0.00
00.
000
0.00
TOTA
L N
UM
BER
of H
OU
SEH
OLD
S W
ITH
OU
T EV
ENTS
2,16
1,06
10.
000
0.00
00.
001,
313,
840
0.00
00.
000
0.00
847,
221
0.00
00.
000
0.00
TOTA
L N
UM
BER
of H
OU
SEH
OLD
S6,
452,
081
0.00
00.
000
0.00
4,10
9,50
60.
000
0.00
00.
002,
342,
575
0.00
00.
000
0.00
1/ D
ecom
posi
tion
bas
ed o
n eq
uatio
n [2
]So
urce
: Aut
hors
’ ela
bora
tion
base
d on
dat
a fro
m E
ncue
sta
Nac
iona
l de
Hog
ares
(IN
EI)
D E M O G R A P H I C &
C O M B I N E D
E V E N T S
VII
- Eve
nts
not c
lass
ified
Entr
yEv
ents
N O N D E M O G R A P H I C
I - E
xclu
sive
ly la
bor
inco
me
even
ts
II - E
xclu
sive
ly no
n-la
bor i
ncom
e ev
ents
P(S/
E)En
try
P(ev
ent)
P(S/
E)
Tabl
e 7.
E. D
ecom
posi
tion
of th
e en
try
rate
from
pov
erty
. 1/
Peru
. Urb
an a
reas
. Per
iod
2002
-200
6.
Tota
l hou
seho
lds
Hou
seho
lds
with
chi
ldre
nH
ouse
hold
s w
ithou
t chi
ldre
nP(
even
t)P(
S/E)
Entr
yP(
even
t)
106
Entry Rates Exit RatesNonPoor-
Poor / Never Poor
Poor-NonPoor /
Always PoorCharacteristics of head in t
Men 1.394*** 1.049Education
Medium 0.639*** 1.424***High 0.336*** 2.617***
AgeF24to40 1.157 1.244*F41to65 1.196 1.306**More65years 0.981 1.442**
Occupational CategoryNon registered wage earner 2.313*** 0.396***OwnAccount 3.019*** 0.456***Employer 1.043 0.798Unemployed 2.181*** 0.453***Inactive with pension 0.773* 0.767**Inactive without pension 2.840*** 0.583***
Characteristics of spouse in tWithout Spouse 0.691* 0.841Men 1.708*** 0.927
EducationMedium 0.676*** 1.249***High 0.337*** 1.589***
AgeF24to40 0.728* 1.109F41to65 0.520*** 1.142More65years 0.616* 1.076
Occupational CategoryNon registered wage earner 2.268*** 0.607***OwnAccount 1.728*** 0.870Employer 1.173 1.104Unemployed 2.434*** 0.673***Inactive with pension 0.769 0.695*Inactive without pension 2.207*** 0.721***
Other characteristicsRegion NOA 1.620*** 0.627*** NEA 1.979*** 0.550*** Cuyo 1.123 1.037 Pampeana 1.016 0.962 Patagonia 0.593*** 1.935***Household characteristics
Dependency Rate 0.373*** 3.696***Incomplete Hous. with children 1.610*** 0.837*Number of children 1.469*** 0.644***Number of elderly 0.946 0.939Number of adults 1.095** 0.793***
Year2004 0.875* 1.145**2005 0.646*** 1.607***
Changes between t and t+1Increase Register Wage Earners 0.0928*** 13.71***Reduction Register Wage Earners 6.341*** 0.286***Increase Non-Register Wage Earners 0.480*** 2.786***Reduction Non-Register Wage Earners 3.031*** 0.361***Increase Non-Wage Earners 0.430*** 2.948***Reduction Non-Wage Earners 5.067*** 0.321***Increase Hourly Wages 0.142*** 3.524***Reduction Hourly Wages 0.949 0.731**Increase Working Hours 0.498*** 1.901***Reduction Working Hours 1.851*** 0.697***Increase Pensions 0.225*** 3.533***Reduction Pensions 1.558*** 0.545***Increase Monetary Transfers 0.985 1.552***Reduction Monetary Transfers 3.091*** 0.533***Increase Other Non-labour income 0.317*** 3.638***Reduction Other Non-labour income 2.131*** 0.749***Reduction Members 0.402*** 4.629***Increase Members 3.361*** 0.628***Constant 0.127*** 0.355***
Observations 31,053 31,053
*** p<0.01, ** p<0.05, * p<0.1
Table 8.A. Relative Risk RatiosArgentina. Urban areas. Period 2003-2006.
Source: Authors’ elaboration based on data from Encuesta Permanente de Hogares (EPH-INDEC).
106 107
Entry Rates Exit Rates
NonPoor-Poor / Never
Poor
Poor-NonPoor /
Always Poor
Characteristics of head in tMen 0.995 1.142
EducationMedium 0.502*** 1.803***High 0.168*** 5.222***
AgeF24to40 0.630** 1.335*F41to65 0.450*** 1.784***More65years 0.264*** 3.562***
Occupational CategoryNon registered wage earner 1.827*** 0.708**OwnAccount 1.733*** 0.715***Employer 0.743* 1.087Unemployed 3.048*** 0.719**Inactive 1.729*** 0.878
Characteristics of spouse in tWithout Spouse 1.166 0.791Men 0.973 0.718*
EducationMedium 0.593*** 1.615***High 0.303*** 3.930***
AgeF24to40 0.768* 0.946F41to65 0.855 1.479**More65years 1.438*** 0.811*
Occupational CategoryNon registered wage earner 1.332** 1.012OwnAccount 1.669*** 0.797Employer 0.696 3.598Unemployed 1.396* 0.720**Inactive 1.438*** 0.811*
Other characteristicsRegion
Recife 4.515*** 0.261***Salvador 2.856*** 0.323***Belo Horizonte 0.941 1.270**Rio de Janeiro 2.127*** 0.498***Porto Alegre 1.547*** 0.777**
Household characteristicsDependency Rate 0.157*** 1.743**Incomplete Hous. with children 1.360** 0.774*Number of children 1.356*** 0.540***Number of elderly 0.559*** 1.373*Number of adults 0.00000 0.00000
Year2004 0.711*** 1.288***2005 1.009 0.967
Changes between t and t+1Increase Register Wage Earners 0.196*** 5.593***Reduction Register Wage Earners 5.431*** 0.316***Increase Non-Register Wage Earners 0.452*** 2.782***Reduction Non-Register Wage Earners 3.472*** 0.420***Increase Non-Wage Earners 0.458*** 3.192***Reduction Non-Wage Earners 3.948*** 0.426***Increase Hourly Wages 0.345*** 1.757***Reduction Hourly Wages 1.180 0.709**Increase Working Hours 0.684*** 1.630***Reduction Working Hours 1.249** 0.906Increase Pensions 0.0843*** 8.376***Reduction Pensions 1.190* 0.896Increase Other Non-labour income 0.405*** 2.127***Reduction Other Non-labour income 1.372*** 0.777**Reduction Members 0.330*** 4.506***Increase Members 2.757*** 0.362***Constant 0.901 0.650
Observations 18,476 18,476
*** p<0.01, ** p<0.05, * p<0.1
Table 8.B. Relative Risk RatiosBrazil. Urban areas. Period 2003 - 2006.
Source: Authors’ elaboration based on data PME & PNAD (IBGE)
108
Entry Rates Exit Rates
NonPoor-Poor / Never Poor
Poor-NonPoor / Always Poor
Characteristics of head in tMen 0.728 1.573
EducationMedium 0.566** 2.040***High 0.359*** 1.521
AgeF24to40 0.597 1.378F41to65 0.407 0.814More65years 0.339 2.650
Occupational CategoryNon registered wage earner 1.383 0.268***OwnAccount 1.515 0.365***Employer 1.471 1.142Unemployed 0.284 1.196Inactive with pension 0.380** 0.424Inactive without pension 0.562 0.561
Characteristics of spouse in tWithout Spouse 0.435 2.033Men 1.069 0.993
EducationMedium 0.782 1.212High 0.399** 3.751**
AgeF24to40 0.620 2.513F41to65 0.416 3.933More65years 0.480 9.977*
Occupational CategoryNon registered wage earner 1.228 1.101OwnAccount 1.659 1.034Employer 0.810 0.467Unemployed 0.755 0.570Inactive with pension 7.45e-07*** 1.553Inactive without pension 1.960** 0.883
Other characteristicsRegion
Chorotega 0.971 0.980Pacífico 1.236 0.859Brunca 1.150 1.194Huetar Atlántica 0.601 1.206Huetar Norte 0.526 0.446
Household characteristicsDependency Rate 3.301** 0.227*Incomplete Hous. with children 0.676 1.575Number of children 1.656*** 0.471***Number of elderly 1.851 0.121***Number of adults 1.036 1.056
Year2007 1.036 0.227*
Changes between t and t+1Increase Register Wage Earners 0.542 2.627***Reduction Register Wage Earners 2.836*** 0.950Increase Non-Register Wage Earners 0.901 2.073***Reduction Non-Register Wage Earners 1.031 1.700Increase Non-Wage Earners 1.296 2.692***Reduction Non-Wage Earners 1.421 1.286Increase Hourly Wages 0.0182*** 4.507***Reduction Hourly Wages 0.103*** 0.862Increase Working Hours 0.424*** 1.662Reduction Working Hours 1.432 0.919Increase Pensions 0.163*** 8.338***Reduction Pensions 1.037 4.734**Increase Monetary Transfers 0.865 0.716Reduction Monetary Transfers 0.938 0.655Increase Other Non-labour income 0.339** 2.454*Reduction Other Non-labour income 0.630 0.305Reduction Members 0.524** 6.002***Increase Members 1.602 1.032Constant 3.208 0.253
Observations 2,229 2,229
*** p<0.01, ** p<0.05, * p<0.1
Table 8.C. Relative Risk RatiosCosta Rica. Urban areas. Period 2006 - 2008.
Source: Authors’ elaboration based on data from Encuesta de Hogares de Propósitos Múltiples (INEC).
108 109
Entry Rates Exit Rates
NonPoor-Poor / Never
Poor
Poor-NonPoor /
Always Poor
Characteristics of head in tMen 1.083 1.528***
EducationMedium 0.602*** 1.448***High 0.383*** 2.027***
AgeF24to40 1.476** 0.598***F41to65 1.487** 0.620***More65years 1.107 0.775
Occupational CategoryNon registered wage earner 2.185*** 0.470***OwnAccount 2.930*** 0.565***Employer 1.719*** 1.104Unemployed 2.358*** 0.571***Inactive with pension 0.411*** 1.369Inactive without pension 1.415** 0.661***
Characteristics of spouse in tWithout Spouse 1.302 0.828Men 1.238 0.939
EducationMedium 0.831** 1.124High 0.465*** 1.515***
AgeF24to40 0.858 0.972F41to65 0.943 0.823More65years 0.872 0.804
Occupational CategoryNon registered wage earner 1.227** 0.961OwnAccount 1.352*** 0.941Employer 0.676** 2.042***Unemployed 1.423* 0.603***Inactive with pension 0.821 4.634*Inactive without pension 1.666*** 0.592***
Other characteristicsHousehold characteristics
Dependency Rate 0.210*** 1.497*Incomplete Hous. with children 0.990 1.275**Number of children 1.335*** 0.661***Number of elderly 0.951 0.878Number of adult 0.896*** 1.029
Year2005 1.028 0.9412006 0.765*** 1.177*2007 0.965 0.943
Changes between t and t+1Increase Register Wage Earners 0.169*** 4.987***Reduction Register Wage Earners 5.012*** 0.438***Increase Non-Register Wage Earners 0.382*** 2.426***Reduction Non-Register Wage Earners 3.544*** 0.447***Increase Non-Wage Earners 0.429*** 2.029***Reduction Non-Wage Earners 3.268*** 0.399***Increase Hourly Wages 0.143*** 3.644***Reduction Hourly Wages 0.863 0.638***Increase Working Hours 0.644*** 1.580***Reduction Working Hours 1.471*** 0.669***Increase Pensions 0.202*** 3.201***Reduction Pensions 1.174 0.661**Increase Monetary Transfers 1.918*** 0.698***Reduction Monetary Transfers 1.984*** 0.768**Increase Other Non-labour income 0.253*** 3.059***Reduction Other Non-labour income 1.143* 0.686***Reduction Members 0.445*** 3.043***Increase Members 2.381*** 0.539***Constant 0.450*** 0.609*
Observations 20,771 20,771
*** p<0.01, ** p<0.05, * p<0.1
Table 8.D. Relative Risk RatiosEcuador. Urban areas. Period 2004- 2008.
Source: Authors’ elaboration based on data from Encuesta de Empleo, Desempleo y Subempleo (INEC).
110
Entry Rates Exit Rates
NonPoor-Poor / Never Poor
Poor-NonPoor / Always Poor
Characteristics of head in tMen 1.180 1.089
EducationMedium 0.651*** 1.553***High 0.418*** 1.792***
AgeF24to40 1.143 0.786F41to65 1.054 0.543*More65years 1.095 0.523
Occupational CategoryNon-registered wage earner 2.335*** 0.742OwnAccount 2.656*** 0.829Employer 1.196 1.303Unemployed 1.222 1.025Inactive with pension 0.816 1.013Inactive without pension 1.256 1.005
Characteristics of spouse in tWithout Spouse 1.272 0.733Men 1.425 1.153
EducationMedium 0.769* 1.497***High 0.354*** 2.165***
AgeF24to40 1.422 0.575*F41to65 1.911* 0.672More65years 1.044 0.364**
Occupational CategoryNon-registered wage earner 1.017 0.865OwnAccount 1.124 1.149Employer 0.797 1.668Unemployed 1.381 0.934Inactive with pension 0.565 1.450Inactive without pension 1.244 0.957
Other characteristicsRegion
Costa Norte 0.713** 1.240Costa Centro 0.864 0.904Costa Sur 0.902 0.643Sierra Norte 0.488 2.650Sierra Centro 1.212 0.459***Sierra Sur 1.484** 0.664**Selva 0.884 0.642***
Household characteristicsDependency Rate 0.330*** 2.027*Incomplete Hous. with children 1.406 1.344Number of children 1.328*** 0.742***Number of elderly 0.869 1.172Number of adults 0.992 0.993
Year2003 1.378** 0.560***2004 1.397** 0.617***2005 0.864 1.001
Changes between t and t+1Increase Register Wage Earners 0.262*** 3.129***Reduction Register Wage Earners 4.846*** 0.420***Increase Non-Register Wage Earners 0.499*** 1.823***Reduction Non-Register Wage Earners 2.922*** 0.520***Increase Non-Wage Earners 0.736* 1.756***Reduction Non-Wage Earners 2.631*** 0.501***Increase Hourly Wages 0.318*** 1.548***Reduction Hourly Wages 1.003 0.534***Increase Working Hours 0.564*** 2.278***Reduction Working Hours 1.029 0.911Increase Pensions 0.291*** 3.491***Reduction Pensions 0.888 1.808*Increase Monetary Transfers 0.894 0.678Reduction Monetary Transfers 0.612 0.609Increase Other Non-labour income 0.573*** 1.619***Reduction Other Non-labour income 0.969 0.937Reduction Members 1.226 1.469***Increase Members 0.408*** 3.108***Constant 0.276** 0.598
Observations 5,566 5,566
*** p<0.01, ** p<0.05, * p<0.1
Table 8.E. Relative Risk RatiosPeru. Urban areas. Period 2002-2006.
Source: Authors’ elaboration based on data from Encuesta Nacional de Hogares (ENAHO-INEI).
110 111
N°(1
)SE
(2)
SE(1
)*(2)
=(3)
SE(1
)SE
(2)
SE(1
)*(2)
=(3)
SE(1
)SE
(2)
SE(1
)*(2)
=(3)
SETO
TAL
LABO
R EV
ENTS
37.52
(0.76
)44
.97(1
.29)
16.84
(0.57
)40
.52(0
.90)
39.94
(1.41
)16
.16(0
.65)
28.40
(1.32
)66
.93(2
.57)
18.89
(1.16
)1
Grow
th in
the n
umbe
r of e
mpl
oyed
mem
bers
7.5(0
.41)
41.6
(2.80
)3.1
(0.27
)6.7
(0.46
)30
.5(3
.24)
2.1(0
.26)
9.9(0
.89)
64.8
(4.33
)6.4
(0.76
)1.1
Grow
th in
the n
umbe
r of r
egist
ered
wag
e ear
ners
1.3(0
.17)
79.6
(4.45
)1.1
(0.16
)1.1
(0.18
)70
.1(6
.44)
0.8(0
.16)
1.9(0
.44)
97.3
(2.00
)1.8
(0.44
)1.2
Grow
th in
the n
umbe
r of n
on-re
gist
ered
wag
e ear
ners
4.1(0
.32)
30.6
(3.63
)1.3
(0.18
)4.0
(0.37
)21
.4(3
.80)
0.9(0
.17)
4.3(0
.61)
56.9
(7.03
)2.4
(0.49
)1.3
Grow
th in
the n
umbe
r of n
on-w
age e
arne
rs2.1
(0.21
)39
.1(5
.06)
0.8(0
.14)
1.6(0
.22)
25.1
(6.49
)0.4
(0.12
)3.7
(0.53
)57
.3(7
.10)
2.1(0
.42)
2Gr
owth
in to
tal h
ourly
wag
e of m
embe
rs em
ploye
d in b
oth
obse
rvatio
ns10
.4(0
.45)
34.9
(2.18
)3.6
(0.28
)12
.1(0
.56)
32.7
(2.33
)4.0
(0.34
)5.2
(0.66
)50
.2(6
.50)
2.6(0
.42)
3Gr
owth
in th
e num
ber o
f wor
king
hour
s of m
embe
rs em
ploye
d in
both
obse
rvatio
ns3.5
(0.28
)25
.0(3
.39)
0.9(0
.14)
3.9(0
.35)
21.4
(3.57
)0.8
(0.16
)2.3
(0.41
)44
.1(9
.08)
1.0(0
.27)
4Gr
owth
in th
e num
ber o
f wor
king
hour
s and
in th
e tota
l hou
rly
wage
of m
embe
rs em
ploye
d in b
oth ob
serva
tions
8.7(0
.52)
48.4
(3.15
)4.2
(0.31
)9.8
(0.65
)43
.1(3
.40)
4.2(0
.37)
5.2(0
.61)
79.1
(4.39
)4.1
(0.56
)
5Gr
owth
in th
e tota
l mon
thly
wage
of m
embe
rs em
ploye
d in b
oth
obse
rvatio
ns an
d in t
he nu
mbe
r of e
mplo
yed m
embe
rs7.4
(0.39
)68
.0(2
.58)
5.0(0
.33)
7.9(0
.46)
64.3
(2.94
)5.1
(0.37
)5.9
(0.72
)83
.5(4
.77)
4.9(0
.66)
TOTA
L NO
N-LA
BOR
EVEN
TS9.0
(0.40
)40
.8(2
.30)
3.7(0
.27)
4.9(0
.31)
12.9
(2.26
)0.6
(0.12
)21
.7(1
.23)
59.7
(3.20
)13
.0(0
.98)
6Gr
owth
in th
e inc
ome f
rom
pen
sions
5.3
(0.32
)51
.3(3
.11)
2.7(0
.23)
1.6(0
.17)
13.8
(3.09
)0.2
(0.05
)16
.3(1
.11)
62.8
(3.69
)10
.2(0
.88)
7Gr
owth
in pu
blic m
onet
ary t
rans
fers
(soc
ial po
licy)
1.7(0
.16)
1.7(0
.77)
0.0(0
.01)
2.0(0
.20)
0.5(0
.38)
0.0(0
.01)
0.8(0
.23)
10.4
(5.81
)0.1
(0.05
)8
Grow
th in
othe
r non
-labo
r inco
mes
2.1
(0.20
)45
.6(4
.86)
1.0(0
.14)
1.3(0
.17)
31.0
(6.63
)0.4
(0.11
)4.6
(0.63
)57
.5(7
.03)
2.7(0
.46)
III - L
abor
and n
on-
labor
inco
me e
vents
9Gr
owth
in la
bor a
nd no
n-lab
or in
com
es13
.0(0
.50)
61.9
(1.93
)8.0
(0.41
)13
.7(0
.59)
56.0
(2.25
)7.6
(0.47
)11
.0(0
.91)
84.0
(2.99
)9.2
(0.85
)
IV E
xclus
ively
dem
ogra
phic
even
ts10
Redu
ction
in th
e tota
l num
ber o
f hou
seho
ld m
embe
rs; th
e tota
l no
mina
l inco
me r
emain
s con
stant
5.7(0
.38)
18.0
(2.77
)0.8
(0.12
)6.3
(0.47
)15
.8(3
.04)
0.7(0
.15)
4.0(0
.51)
30.1
(6.46
)0.8
(0.19
)
V -
Dem
ogra
phic
even
ts lea
ding t
o inc
ome c
hang
es11
Grow
th in
the n
umbe
r of l
abor
or n
on-la
bor i
ncom
e ear
ners
du
e to t
he en
tranc
e of m
embe
rs to
the h
ouse
hold
1.6(0
.20)
33.6
(6.05
)0.5
(0.10
)1.3
(0.21
)27
.4(7
.36)
0.3(0
.11)
2.5(0
.50)
44.3
(10.4
2)1.0
(0.27
)
VI -
Com
binati
on of
de
mog
raph
ic an
d inc
ome e
vents
12Gr
owth
in th
e tota
l nom
inal
inco
me a
nd gr
owth
in the
num
ber o
f ho
useh
old
mem
bers
.6.5
(0.36
)60
.9(3
.13)
3.3(0
.24)
6.3(0
.40)
55.0
(3.63
)2.9
(0.26
)7.2
(0.77
)78
.0(5
.69)
4.3(0
.58)
5.1(0
.38)
53.5
(3.83
)2.7
(0.32
)4.8
(0.46
)45
.2(5
.27)
2.1(0
.38)
6.2(0
.70)
72.4
(5.19
)4.5
(0.60
)TO
TAL
HOUS
EHOL
DS W
ITH
EVE
NTS
78.5
(0.63
)47
.3(0
.89)
35.6
(0.74
)77
.7(0
.74)
40.8
(1.04
)30
.4(0
.84)
81.0
(1.19
)66
.7(1
.59)
51.6
(1.48
)TO
TAL
HOUS
EHOL
DS W
ITHO
UT E
VENT
S21
.5(0
.63)
0.0(0
.02)
0.0(0
.01)
22.3
(0.74
)0.0
0.00
0.00.0
019
.0(1
.19)
0.2(0
.11)
0.0(0
.02)
TOTA
L HO
USEH
OLDS
100.0
0.00
0.00.0
035
.60.0
010
0.00.0
00.0
0.00
30.4
0.00
100.0
0.00
0.00.0
051
.60.0
0TO
TAL
NUMB
ER o
f HOU
SEHO
LDS
WIT
H E
VENT
S2,1
05,69
90.0
00
0.00
00.0
01,5
68,50
30.0
00
0.00
00.0
053
7,196
0.00
00.0
00
0.00
TOTA
L NU
MBER
of H
OUSE
HOLD
S W
ITHO
UT E
VENT
S57
6,364
0.00
00.0
00
0.00
450,5
830.0
00
0.00
00.0
012
5,781
0.00
00.0
00
0.00
TOTA
L NU
MBER
of H
OUSE
HOLD
S2,6
82,06
30.0
00
0.00
00.0
02,0
19,08
60.0
00
0.00
00.0
066
2,977
0.00
00.0
00
0.00
1/ De
com
posit
ion b
ased
on e
quati
on [2
]So
urce
: Auth
ors’
elabo
ratio
n bas
ed on
data
from
Enc
uesta
Per
man
ente
de H
ogar
es (I
NDEC
).
P(S/
E)
Tabl
e 9.A
. Dec
ompo
sitio
n of
the e
xit ra
te fr
om p
over
ty. S
ensit
ivity
analy
sis to
pov
erty
line (
Redu
ctio
n o
f 10%
in p
over
y lin
e) 1/
Arge
ntin
a. Ur
ban
area
s. Pe
riod
2003
2006
.
Tota
l hou
seho
lds
Hous
ehol
ds w
ith ch
ildre
nHo
useh
olds
with
out c
hild
ren
P(ev
ent)
P(S/
E)Ex
itP(
even
t)
D E M O G R A P H I C &
C O M B I N E D
E V E N T S
VII -
Eve
nts no
t clas
sified
Exit
Even
ts
N O N D E M O G R A P H I C
I - E
xclus
ively
labor
inc
ome e
vents
II - E
xclus
ively
non-
labor
inco
me e
vents
P(S/
E)Ex
itP(
even
t)
112
N°(1)
SE(2)
SE(1)
*(2)=
(3)SE
(1)SE
(2)SE
(1)*(2
)=(3)
SE(1)
SE(2)
SE(1)
*(2)=
(3)SE
TOTA
L LAB
OR E
VENT
S17
.88(0.
51)
36.18
(1.54
)6.4
7(0.
33)
18.71
(0.61
)28
.83(1.
65)
5.39
(0.36
)15
.49(0.
95)
61.67
(3.24
)9.5
6(0.
78)
1Gr
owth
in th
e num
ber o
f emp
loyed
mem
bers
4.8(0
.29)
47.1
(3.12
)2.2
(0.20
)3.9
(0.31
)33
.0(3
.75)
1.3(0
.18)
7.1(0
.68)
69.7
(4.47
)4.9
(0.59
)1.1
Grow
th in
the n
umbe
r of r
egist
ered
wag
e ear
ners
1.8(0
.18)
58.7
(4.86
)1.1
(0.14
)1.5
(0.19
)45
.3(6
.29)
0.7(0
.13)
2.7(0
.43)
80.3
(5.97
)2.2
(0.40
)1.2
Grow
th in
the n
umbe
r of n
on-re
gister
ed w
age e
arne
rs1.6
(0.17
)33
.5(5
.17)
0.5(0
.10)
1.5(0
.20)
26.7
(5.95
)0.4
(0.11
)1.9
(0.36
)48
.8(9
.42)
0.9(0
.26)
1.3Gr
owth
in th
e num
ber o
f non
-wag
e ear
ners
1.3(0
.15)
47.4
(5.98
)0.6
(0.11
)0.9
(0.15
)22
.6(6
.58)
0.2(0
.07)
2.4(0
.42)
74.3
(7.42
)1.8
(0.37
)
2Gr
owth
in to
tal ho
urly
wage
of m
embe
rs em
ploye
d in b
oth
obse
rvatio
ns6.7
(0.34
)19
.6(2
.10)
1.3(0
.16)
7.7(0
.42)
15.3
(2.09
)1.2
(0.17
)3.9
(0.51
)43
.6(6
.53)
1.7(0
.34)
3Gr
owth
in th
e num
ber o
f wor
king h
ours
of m
embe
rs em
ploye
d in
both
obse
rvatio
ns1.2
(0.15
)16
.8(4
.73)
0.2(0
.06)
1.2(0
.17)
9.0(4
.16)
0.1(0
.05)
1.1(0
.28)
42.7
(12.9
0)0.5
(0.19
)
4Gr
owth
in th
e num
ber o
f wor
king h
ours
and i
n the
total
hour
ly wa
ge of
mem
bers
emplo
yed i
n both
obse
rvatio
ns2.4
(0.20
)38
.8(4
.20)
0.9(0
.13)
2.7(0
.25)
34.5
(4.46
)0.9
(0.15
)1.5
(0.33
)61
.7(1
1.01)
0.9(0
.25)
5Gr
owth
in th
e tota
l mon
thly
wage
of m
embe
rs em
ploye
d in b
oth
obse
rvatio
ns an
d in t
he nu
mber
of em
ploye
d mem
bers
2.8(0
.22)
63.3
(3.86
)1.8
(0.18
)3.1
(0.27
)59
.8(4
.34)
1.9(0
.21)
1.9(0
.35)
79.8
(7.85
)1.5
(0.31
)
TOTA
L NON
-LAB
OUR
EVEN
TS7.9
(0.37
)73
.2(2.
07)
5.8(0.
32)
3.9(0.
29)
47.8
(3.83
)1.9
(0.21
)19
.5(1.
07)
87.8
(1.93
)17
.1(1.
02)
6Gr
owth
in the
inco
me fr
om pe
nsion
s 6.5
(0.34
)81
.7(1
.98)
5.4(0
.31)
2.6(0
.24)
60.5
(4.58
)1.5
(0.19
)18
.0(1
.04)
90.3
(1.84
)16
.3(1
.00)
8Gr
owth
in ot
her i
ncom
e1.4
(0.15
)32
.7(5
.34)
0.5(0
.09)
1.3(0
.17)
23.6
(5.64
)0.3
(0.09
)1.5
(0.32
)56
.6(1
0.57)
0.8(0
.25)
III - L
abor
and n
on-
labor
inco
me ev
ents
9Gr
owth
in la
bor a
nd no
n-lab
or in
come
s6.2
(0.32
)78
.2(2
.13)
4.8(0
.29)
5.1(0
.34)
67.4
(3.11
)3.4
(0.28
)9.3
(0.79
)95
.2(1
.71)
8.9(0
.78)
IV E
xclus
ively
demo
grap
hic ev
ents
10Re
ducti
on in
the t
otal n
umbe
r of h
ouse
hold
memb
ers;
the to
tal
nomi
nal in
come
rema
ins co
nstan
t5.3
(0.28
)12
.2(1
.85)
0.6(0
.11)
6.2(0
.35)
10.8
(1.87
)0.7
(0.12
)2.5
(0.38
)22
.1(6
.67)
0.6(0
.20)
V -
Demo
grap
hic
even
ts lea
ding t
o inc
ome c
hang
es11
Grow
th in
the n
umbe
r of la
bor o
r non
-labo
r inc
ome e
arne
rs
due t
o the
entra
nce o
f mem
bers
to the
hous
ehold
12.7
(0.44
)52
.2(1
.84)
6.6(0
.33)
12.7
(0.51
)45
.3(2
.15)
5.8(0
.36)
12.6
(0.83
)72
.4(2
.95)
9.1(0
.75)
VI -
Comb
inatio
n of
demo
grap
hic an
d inc
ome e
vents
12Gr
owth
in th
e tota
l nom
inal in
come
and g
rowt
h in t
he nu
mber
of
hous
ehold
mem
bers
.8.1
(0.35
)71
.9(1
.94)
5.8(0
.31)
8.5(0
.41)
67.3
(2.29
)5.7
(0.35
)6.9
(0.67
)87
.9(2
.97)
6.1(0
.64)
21.1
(0.52
)56
.4(1
.37)
11.9
(0.42
)21
.8(0
.61)
49.4
(1.58
)10
.7(0
.46)
19.2
(1.00
)79
.5(2
.16)
15.3
(0.92
)TO
TAL H
OUSE
HOLD
S W
ITH E
VENT
S79
.20.0
053
.2(0
.73)
42.1
(0.65
)77
.0(0
.64)
43.7
(0.86
)33
.6(0
.72)
85.7
(0.89
)77
.8(1
.12)
66.6
(1.20
)TO
TAL H
OUSE
HOLD
S W
ITHOU
T EVE
NTS
20.8
(0.53
)0.0
0.00
.0.0
023
.0(0
.64)
0.00.0
0.
0.00
14.3
(0.89
)0.0
0.00
.0.0
0TO
TAL H
OUSE
HOLD
S10
0.0(0
.53)
0.00.0
0.
0.00
100.0
0.00
0.00.0
0.
0.00
100.0
0.00
0.00.0
0.
0.00
TOTA
L NUM
BER
of H
OUSE
HOLD
S W
ITH E
VENT
S2,6
13,66
70.0
00
0.00
00.0
01,8
84,05
30.0
00
0.00
00.0
072
9,614
0.00
00.0
00
0.00
TOTA
L NUM
BER
of H
OUSE
HOLD
S W
ITHOU
T EVE
NTS
685,8
000.0
00
0.00
00.0
056
3,608
0.00
00.0
00
0.00
122,1
920.0
00
0.00
00.0
0TO
TAL N
UMBE
R of
HOU
SEHO
LDS
3,299
,467
0.00
00.0
00
0.00
2,447
,661
0.00
00.0
00
0.00
851,8
060.0
00
0.00
00.0
0
1/ De
comp
ositio
n ba
sed
on eq
uatio
n [2]
2/ Me
tropo
litan a
reas
of R
ecife
, Salv
ador
, Belo
Hor
izonte
, Rio
de Ja
neiro
, São
Pau
lo an
d Por
to Al
egre
.
D E M O G R A P H I
C &
C O M B I N E D
E V E N T S
Exit
Even
ts
N O N D E M O G R A P H I C
I - E
xclus
ively
labor
inc
ome e
vents
P(S/
E)Ex
itP(
even
t)P(
S/E)
P(ev
ent)
Table
9.B.
Dec
ompo
sition
of th
e exit
rate
from
pove
rty. S
ensit
ivity
analy
sis to
pov
erty
line (
Redu
ction
of 10
% in
pove
ry lin
e) 1/
Braz
il. Ur
ban a
reas
. Per
iod 20
03-20
06. 2
/
Total
hous
ehold
sHo
useh
olds w
ith ch
ildre
nHo
useh
olds w
ithou
t chil
dren
VII -
Eve
nts no
t clas
sified
P(ev
ent)
P(S/
E)Ex
it
II - E
xclus
ively
non-
labor
inco
me ev
ents
112 113
N°(1
)SE
(2)
SE(1
)*(2)
=(3)
SE(1
)SE
(2)
SE(1
)*(2)
=(3)
SE(1
)SE
(2)
SE(1
)*(2)
=(3)
SETO
TAL
LABO
R EV
ENTS
39.09
(2.35
)68
.76(3
.49)
26.87
(2.11
)43
.91(2
.91)
63.71
(4.30
)27
.98(2
.64)
30.20
(3.81
)82
.27(5
.36)
24.84
(3.54
)1
Grow
th in
the
num
ber o
f em
ploy
ed m
embe
rs11
.3(1
.52)
78.3
(5.7
9)8.
8(1
.38)
10.4
(1.8
4)74
.9(8
.09)
7.8
(1.6
4)13
.0(2
.64)
83.1
(7.7
9)10
.8(2
.51)
1.1
Grow
th in
the
num
ber o
f reg
ister
ed w
age e
arne
rs4.
3(1
.00)
100.
00.
004.
3(1
.00)
3.0
(1.0
4)10
0.0
0.00
3.0
(1.0
4)6.
6(2
.06)
100.
00.
006.
6(2
.06)
1.2
Grow
th in
the
num
ber o
f non
-regi
ster
ed w
age e
arne
rs3.
2(0
.82)
68.8
(11.
73)
2.2
(0.6
9)3.
5(1
.10)
69.9
(14.
09)
2.5
(0.9
4)2.
7(1
.14)
66.1
(20.
90)
1.8
(0.9
0)1.
3Gr
owth
in th
e nu
mbe
r of n
on-w
age e
arne
rs3.
8(0
.92)
61.8
(12.
09)
2.3
(0.7
4)3.
8(1
.18)
59.8
(15.
52)
2.3
(0.9
2)3.
7(1
.47)
65.7
(18.
67)
2.4
(1.2
5)
2Gr
owth
in to
tal h
ourly
wag
e of m
embe
rs e
mplo
yed
in bo
th
obse
rvat
ions
8.0
(1.2
8)54
.4(7
.98)
4.4
(0.9
6)11
.2(1
.84)
49.7
(8.3
4)5.
6(1
.31)
2.1
(1.2
2)10
0.0
0.00
2.1
(1.2
2)
3Gr
owth
in th
e nu
mbe
r of w
orkin
g ho
urs
of m
embe
rs e
mplo
yed
in bo
th o
bser
vatio
ns3.
0(0
.79)
39.5
(12.
82)
1.2
(0.4
9)3.
4(1
.01)
27.6
(13.
17)
0.9
(0.5
1)2.
4(1
.24)
70.4
(24.
44)
1.7
(1.0
2)
4Gr
owth
in th
e nu
mbe
r of w
orkin
g ho
urs
and
in th
e to
tal h
ourly
wa
ge o
f mem
bers
em
ploye
d in
both
obs
erva
tions
8.1
(1.2
9)54
.8(8
.33)
4.5
(0.9
7)10
.1(1
.76)
55.2
(9.2
1)5.
6(1
.34)
4.5
(1.7
1)53
.1(1
9.36
)2.
4(1
.23)
5Gr
owth
in th
e to
tal m
onth
ly wa
ge o
f mem
bers
em
ploye
d in
both
ob
serv
ation
s and
in th
e nu
mbe
r of e
mplo
yed
mem
bers
8.6
(1.3
3)93
.0(4
.08)
8.0
(1.2
9)8.
9(1
.65)
91.5
(5.7
2)8.
2(1
.58)
8.1
(2.2
7)96
.0(3
.97)
7.8
(2.2
5)
TOTA
L NO
N-LA
BOR
EVEN
TS8.0
(1.32
)15
.2(5
.95)
1.2(0
.51)
3.2(1
.02)
18.0
(11.7
4)0.6
(0.41
)17
.0(3
.10)
14.2
(6.89
)2.4
(1.24
)6
Grow
th in
the
inco
me f
rom
pen
sions
4.
9(1
.08)
19.7
(8.3
5)1.
0(0
.45)
0.6
(0.4
1)10
0.0
0.00
0.6
(0.4
1)13
.0(2
.84)
13.1
(7.5
2)1.
7(1
.03)
7Gr
owth
in p
ublic
mon
etar
y tra
nsfe
rs (s
ocial
poli
cy)
2.1
(0.6
7)0.
00.
00.
0.00
2.6
(0.9
4)0.
00.
00.
0.00
1.0
(0.7
7)0.
00.
00.
0.00
8Gr
owth
in o
ther
non
-labo
r inc
omes
1.
1(0
.49)
23.9
(20.
60)
0.3
(0.2
5)0.
00.
000.
00.
000.
00.
003.
0(1
.38)
23.9
(20.
60)
0.7
(0.7
1)
III -
Labo
r and
non
-lab
or in
com
e ev
ents
9Gr
owth
in la
bor a
nd n
on-la
bor i
ncom
es9.
5(1
.39)
72.9
(6.2
9)6.
9(1
.18)
10.2
(1.7
1)66
.1(8
.26)
6.7
(1.4
4)8.
1(2
.38)
88.4
(7.7
6)7.
2(2
.06)
IV E
xclus
ively
dem
ogra
phic
even
ts10
Redu
ction
in th
e to
tal n
umbe
r of h
ouse
hold
mem
bers
; the
tota
l no
mina
l inco
me
rem
ains c
onsta
nt2.
8(0
.78)
28.2
(12.
65)
0.8
(0.4
1)3.
1(1
.05)
34.3
(16.
24)
1.1
(0.6
2)2.
2(1
.07)
12.1
(12.
10)
0.3
(0.2
6)
V -
Dem
ogra
phic
even
ts lea
ding
to
incom
e ch
ange
s11
Grow
th in
the
num
ber o
f lab
or o
r non
-labo
r inc
ome e
arne
rs
due
to th
e en
tranc
e of
mem
bers
to th
e ho
useh
old0.
8(0
.44)
70.2
(24.
54)
0.6
(0.3
6)0.
8(0
.54)
49.9
(35.
36)
0.4
(0.3
8)1.
0(0
.74)
100.
00.
001.
0(0
.74)
VI -
Com
binat
ion o
f de
mog
raph
ic an
d inc
ome
even
ts12
Grow
th in
the
tota
l nom
inal
inco
me
and
grow
th in
the
num
ber o
f ho
useh
old
mem
bers
.6.
8(1
.18)
78.3
(7.7
2)5.
3(1
.04)
8.5
(1.6
3)77
.8(8
.63)
6.6
(1.4
3)3.
7(1
.47)
80.6
(17.
15)
3.0
(1.3
0)
11.3
(1.4
8)88
.0(4
.40)
9.9
(1.4
1)13
.3(1
.98)
88.3
(4.9
8)11
.7(1
.87)
7.7
(2.1
2)87
.0(9
.39)
6.7
(2.0
2)TO
TAL
HOUS
EHOL
DS W
ITH
EVE
NTS
78.3
(1.9
1)65
.9(2
.49)
51.7
(2.3
6)83
.0(2
.13)
66.4
(3.0
3)55
.1(2
.92)
69.8
(3.6
7)64
.9(4
.49)
45.3
(4.0
1)TO
TAL
HOUS
EHOL
DS W
ITHO
UT E
VENT
S21
.7(1
.91)
0.0
0.00
.0.
0017
.0(2
.13)
0.0
0.00
.0.
0030
.2(3
.67)
0.0
0.00
.0.
00TO
TAL
HOUS
EHOL
DS10
0.0
0.00
0.0
0.00
51.7
0.00
100.
00.
000.
00.
0055
.10.
0010
0.0
0.00
0.0
0.00
45.3
0.00
TOTA
L NU
MBE
R of
HOU
SEHO
LDS
WIT
H E
VENT
S56
,805
0.00
00.
000
0.00
38,9
990.
000
0.00
00.
0017
,806
0.00
00.
000
0.00
TOTA
L NU
MBE
R of
HOU
SEHO
LDS
WIT
HOUT
EVE
NTS
15,7
050.
000
0.00
00.
008,
000
0.00
00.
000
0.00
7,70
50.
000
0.00
00.
00TO
TAL
NUM
BER
of H
OUSE
HOLD
S72
,510
0.00
00.
000
0.00
46,9
990.
000
0.00
00.
0025
,511
0.00
00.
000
0.00
1/ D
ecom
posit
ion b
ased
on
equa
tion
[2]
Sour
ce: A
utho
rs’ e
labor
ation
bas
ed o
n da
ta fr
om E
ncue
sta d
e Ho
gare
s de
Prop
ósito
s Múlt
iples
(INE
C).
D E M O G R A P H I C &
C O M B I N E D
E V E N T S
VII
- Eve
nts n
ot cl
assif
ied
Exit
Even
ts
N O N D E M O G R A P H I C
I - E
xclus
ively
labor
inc
ome
even
ts
II - E
xclus
ively
non-
labor
inco
me
even
ts
P(S/
E)Ex
itP(
even
t)P(
S/E)
Tabl
e 9.C
. Dec
ompo
sitio
n of
the e
xit ra
te fr
om p
over
ty. S
ensit
ivity
analy
sis to
pov
erty
line (
Redu
ctio
n of
10%
in p
over
y lin
e) 1/
Cost
a Rica
.Urb
an ar
eas.
Perio
d 20
06-2
008.
Tota
l hou
seho
lds
Hous
ehol
ds w
ith ch
ildre
nHo
useh
olds
with
out c
hild
ren
P(ev
ent)
P(S/
E)Ex
itP(
even
t)
114
N°(1)
SE(2)
SE(1)
*(2)=
(3)SE
(1)SE
(2)SE
(1)*(2
)=(3)
SE(1)
SE(2)
SE(1)
*(2)=
(3)SE
TOTA
L LAB
OR E
VENT
S37
.35(0.
69)
49.72
(1.16
)18
.57(0.
56)
38.79
(0.77
)47
.78(1.
27)
18.53
(0.62
)30
.83(1.
54)
60.80
(2.86
)18
.75(1.
30)
1Gr
owth
in th
e num
ber o
f emp
loyed
mem
bers
5.0(0
.30)
44.7
(3.09
)2.3
(0.21
)4.6
(0.32
)40
.3(3
.50)
1.9(0
.21)
6.8(0
.82)
58.4
(6.14
)4.0
(0.65
)1.1
Grow
th in
the n
umbe
r of r
egist
ered
wag
e ear
ners
0.5(0
.09)
69.3
(9.09
)0.3
(0.08
)0.4
(0.09
)58
.3(1
1.72)
0.2(0
.07)
0.9(0
.31)
91.4
(8.43
)0.8
(0.30
)1.2
Grow
th in
the n
umbe
r of n
on-re
gister
ed w
age e
arne
rs2.2
(0.20
)44
.3(4
.67)
1.0(0
.14)
2.2(0
.22)
38.4
(5.01
)0.8
(0.14
)2.2
(0.49
)71
.6(1
0.23)
1.6(0
.42)
1.3Gr
owth
in th
e num
ber o
f non
-wag
e ear
ners
2.4(0
.21)
40.1
(4.48
)0.9
(0.13
)2.1
(0.22
)38
.9(5
.33)
0.8(0
.14)
3.7(0
.61)
42.9
(8.22
)1.6
(0.41
)
2Gr
owth
in to
tal ho
urly
wage
of m
embe
rs em
ploye
d in b
oth
obse
rvatio
ns11
.8(0
.45)
42.1
(2.00
)5.0
(0.30
)12
.5(0
.51)
40.5
(2.14
)5.0
(0.34
)8.8
(0.92
)52
.2(5
.46)
4.6(0
.67)
3Gr
owth
in th
e num
ber o
f wor
king h
ours
of m
embe
rs em
ploye
d in
both
obse
rvatio
ns4.2
(0.28
)25
.5(3
.03)
1.1(0
.15)
4.4(0
.31)
22.1
(3.14
)1.0
(0.16
)3.4
(0.59
)45
.7(8
.85)
1.5(0
.41)
4Gr
owth
in th
e num
ber o
f wor
king h
ours
and i
n the
total
hour
ly wa
ge of
mem
bers
emplo
yed i
n both
obse
rvatio
ns9.2
(0.40
)56
.8(2
.29)
5.2(0
.31)
9.6(0
.45)
54.8
(2.50
)5.3
(0.35
)7.3
(0.85
)69
.0(5
.58)
5.0(0
.72)
5Gr
owth
in th
e tota
l mon
thly
wage
of m
embe
rs em
ploye
d in b
oth
obse
rvatio
ns an
d in t
he nu
mber
of em
ploye
d mem
bers
7.1(0
.36)
71.0
(2.35
)5.1
(0.31
)7.7
(0.41
)70
.0(2
.52)
5.4(0
.35)
4.6(0
.68)
79.2
(6.33
)3.6
(0.61
)
TOTA
L NON
-LAB
OR E
VENT
S5.9
(0.35
)40
.2(2.
88)
2.4(0.
22)
3.9(0.
31)
28.8
(3.71
)1.1
(0.17
)14
.9(1.
23)
53.8
(4.19
)8.0
(0.92
)6
Grow
th in
the i
ncom
e fro
m pe
nsion
s 0.7
(0.12
)54
.5(8
.06)
0.4(0
.09)
0.4(0
.10)
46.1
(11.4
6)0.2
(0.07
)2.2
(0.50
)62
.2(1
1.41)
1.3(0
.39)
7Gr
owth
in pu
blic m
oneta
ry tr
ansfe
rs (s
ocial
polic
y) 1.0
(0.14
)0.0
0.00
.0.0
00.9
(0.15
)0.0
0.00
.0.0
01.4
(0.38
)0.0
0.00
.0.0
08
Grow
th in
othe
r non
-labo
r inco
mes
4.1(0
.30)
47.5
(3.50
)2.0
(0.20
)2.6
(0.26
)36
.3(4
.92)
0.9(0
.16)
11.3
(1.11
)58
.9(4
.67)
6.7(0
.85)
III - L
abor
and n
on-
labor
inco
me ev
ents
9Gr
owth
in la
bor a
nd no
n-lab
or in
come
s12
.7(0
.47)
58.9
(1.94
)7.5
(0.37
)12
.8(0
.53)
54.0
(2.18
)6.9
(0.40
)12
.1(1
.07)
82.4
(3.34
)10
.0(0
.98)
IV E
xclus
ively
demo
grap
hic ev
ents
10Re
ducti
on in
the t
otal n
umbe
r of h
ouse
hold
memb
ers;
the to
tal
nomi
nal in
come
rema
ins co
nstan
t5.5
(0.32
)16
.4(2
.10)
0.9(0
.12)
5.9(0
.37)
16.5
(2.26
)1.0
(0.14
)3.6
(0.61
)16
.0(5
.62)
0.6(0
.21)
V -
Demo
grap
hic
even
ts lea
ding t
o inc
ome c
hang
es11
Grow
th in
the n
umbe
r of la
bor o
r non
-labo
r inc
ome e
arne
rs
due t
o the
entra
nce o
f mem
bers
to the
hous
ehold
2.5(0
.22)
53.5
(4.53
)1.3
(0.17
)2.3
(0.23
)51
.0(5
.17)
1.2(0
.17)
3.3(0
.62)
61.3
(9.31
)2.0
(0.48
)
VI -
Comb
inatio
n of
demo
grap
hic an
d inc
ome e
vents
12Gr
owth
in th
e tota
l nom
inal in
come
and g
rowt
h in t
he nu
mber
of
hous
ehold
mem
bers
.7.8
(0.38
)72
.7(2
.24)
5.7(0
.33)
8.2(0
.43)
71.6
(2.46
)5.9
(0.37
)6.1
(0.81
)79
.5(5
.06)
4.9(0
.74)
7.3(0
.37)
57.1
(2.60
)4.1
(0.28
)6.3
(0.39
)55
.0(3
.12)
3.5(0
.29)
11.5
(1.02
)62
.3(4
.62)
7.2(0
.82)
TOTA
L HOU
SEHO
LDS
WIT
H E
VENT
S79
.0(0
.58)
51.2
(0.81
)40
.5(0
.71)
78.3
(0.65
)48
.7(0
.90)
38.1
(0.77
)82
.3(1
.26)
62.4
(1.76
)51
.3(1
.67)
TOTA
L HOU
SEHO
LDS
WIT
HOUT
EVE
NTS
21.0
(0.58
)0.3
(0.16
)0.1
(0.03
)21
.7(0
.65)
0.3(0
.18)
0.1(0
.04)
17.7
(1.26
)0.2
(0.21
)0.0
(0.04
)TO
TAL H
OUSE
HOLD
S10
0.040
.510
0.038
.110
0.051
.4TO
TAL N
UMBE
R of
HOU
SEHO
LDS
WIT
H E
VENT
S1,6
12,57
91,3
08,68
030
3,899
TOTA
L NUM
BER
of H
OUSE
HOLD
S W
ITHO
UT E
VENT
S42
8,718
363,2
2065
,498
TOTA
L NUM
BER
of H
OUSE
HOLD
S2,0
41,29
71,6
71,90
036
9,397
1/ De
comp
ositio
n ba
sed
on eq
uatio
n [2]
Sour
ce: A
uthor
s’ ela
bora
tion b
ased
on da
ta fro
m En
cues
ta de
Emp
leo, D
esem
pleo y
Sub
emple
o (IN
EC).
D E M O G R A P H I C &
C O M B I N E D
E V E N T S
VII -
Eve
nts no
t clas
sified
Exit
Even
ts
N O N D E M O G R A P H I C
I - E
xclus
ively
labor
inc
ome e
vents
II - E
xclus
ively
non-
labor
inco
me ev
ents
P(S/
E)Ex
itP(
even
t)P(
S/E)
Table
9.D.
Dec
ompo
sition
of th
e exit
rate
from
pove
rty. S
ensit
ivity
analy
sis to
pov
erty
line (
Redu
ction
of 10
% in
pove
ry lin
e) 1/
Ecua
dor.
Urba
n are
as. P
eriod
2004
-2008
.
Total
hous
ehold
sHo
useh
olds w
ith ch
ildre
nHo
useh
olds w
ithou
t chil
dren
P(ev
ent)
P(S/
E)Ex
itP(
even
t)
114 115
N°(1)
SE(2)
SE(1)
*(2)=(
3)SE
(1)SE
(2)SE
(1)*(2
)=(3)
SE(1)
SE(2)
SE(1)
*(2)=(
3)SE
TOTA
L LAB
OR E
VENT
S28
.47(1.
21)
49.88
(2.57
)14
.20(0.
96)
29.09
(1.30
)46
.75(2.
76)
13.60
(1.03
)24
.38(3.
24)
74.34
(6.43
)18
.13(2.
77)
1Gr
owth
in the
numb
er of
emplo
yed m
embe
rs6.4
(0.64
)51
.3(5.
24)
3.3(0.
47)
6.3(0.
68)
46.0
(5.76
)2.9
(0.49
)7.0
(1.90
)82
.1(9.
45)
5.8(1.
62)
1.1Gr
owth
in the
numb
er of
regis
tered
wag
e ear
ners
0.9(0.
24)
63.1
(13.94
)0.5
(0.19
)0.9
(0.27
)58
.3(15
.25)
0.5(0.
20)
0.7(0.
52)
100.0
0.00
0.7(0.
52)
1.2Gr
owth
in the
numb
er of
non-
regis
tered
wag
e ear
ners
2.9(0.
44)
49.2
(7.57
)1.4
(0.32
)2.7
(0.44
)43
.1(8.
61)
1.1(0.
31)
4.3(1.
64)
74.4
(13.83
)3.2
(1.29
)1.3
Grow
th in
the nu
mber
of no
n-wa
ge ea
rner
s2.6
(0.42
)49
.6(8.
23)
1.3(0.
31)
2.7(0.
46)
44.9
(8.75
)1.2
(0.32
)2.0
(0.94
)91
.8(8.
50)
1.8(0.
93)
2Gr
owth
in tot
al ho
urly
wage
of m
embe
rs em
ploye
d in b
oth
obse
rvatio
ns5.7
(0.61
)42
.3(5.
66)
2.4(0.
42)
5.9(0.
67)
37.9
(5.90
)2.2
(0.43
)4.4
(1.45
)80
.1(13
.16)
3.5(1.
34)
3Gr
owth
in the
numb
er of
work
ing ho
urs o
f mem
bers
emplo
yed i
n bo
th ob
serva
tions
4.0(0.
53)
30.8
(6.38
)1.2
(0.30
)4.2
(0.58
)29
.4(6.
54)
1.2(0.
33)
2.4(1.
16)
47.2
(25.20
)1.1
(0.80
)
4Gr
owth
in the
numb
er of
work
ing ho
urs a
nd in
the t
otal h
ourly
wa
ge of
mem
bers
emplo
yed i
n both
obse
rvatio
ns4.7
(0.54
)58
.3(5.
81)
2.7(0.
43)
4.8(0.
58)
58.3
(6.10
)2.8
(0.47
)3.8
(1.44
)58
.3(18
.89)
2.2(1.
15)
5Gr
owth
in the
total
mon
thly
wage
of m
embe
rs em
ploye
d in b
oth
obse
rvatio
ns an
d in t
he nu
mber
of em
ploye
d mem
bers
7.8(0.
70)
59.1
(4.61
)4.6
(0.57
)7.9
(0.76
)56
.2(4.
99)
4.4(0.
61)
6.8(1.
68)
81.1
(9.33
)5.5
(1.58
)
TOTA
L NON
-LAB
OR E
VENT
S4.0
(0.53
)32
.5(6.
20)
1.3(0.
30)
3.0(0.
46)
34.8
(7.60
)1.1
(0.29
)10
.1(2.
39)
27.8
(10.33
)2.8
(1.27
)6
Grow
th in
the in
come
from
pens
ions
0.2(0.
13)
57.0
(34.68
)0.1
(0.11
)0.1
(0.09
)0.0
0.00
0.00.0
00.8
(0.80
)10
0.00.0
00.8
(0.80
)7
Grow
th in
publi
c mon
etary
trans
fers (
socia
l poli
cy)
0.1(0.
09)
0.00.0
0.
0.00
0.1(0.
10)
0.00.0
0.
0.00
0.00.0
00.0
0.00
0.00.0
08
Grow
th in
other
non-l
abor
incom
es
3.7(0.
51)
32.2
(6.33
)1.2
(0.28
)2.8
(0.44
)37
.6(7.
98)
1.1(0.
29)
9.3(2.
28)
21.6
(9.22
)2.0
(1.01
)
III - L
abor
and n
on-
labor
incom
e eve
nts9
Grow
th in
labor
and n
on-la
bor in
come
s13
.3(0.
89)
64.8
(3.29
)8.6
(0.75
)13
.6(0.
97)
63.5
(3.53
)8.6
(0.81
)11
.2(2.
11)
74.5
(8.12
)8.3
(1.92
)
IV E
xclus
ively
demo
graph
ic ev
ents
10Re
ducti
on in
the t
otal n
umbe
r of h
ouse
hold
memb
ers;
the to
tal
nomi
nal in
come
rema
ins co
nstan
t5.7
(0.58
)21
.7(4.
38)
1.2(0.
29)
6.0(0.
63)
20.9
(4.54
)1.3
(0.31
)4.1
(1.30
)29
.6(16
.72)
1.2(0.
85)
V - D
emog
raphic
ev
ents
leadin
g to
incom
e cha
nges
11Gr
owth
in the
numb
er of
labor
or no
n-lab
or inc
ome e
arners
due t
o the
entra
nce o
f mem
bers
to the
hous
ehold
2.1(0.
34)
47.4
(8.41
)1.0
(0.25
)2.0
(0.36
)43
.8(9.
20)
0.9(0.
26)
2.8(1.
13)
64.0
(19.54
)1.8
(0.93
)
VI - C
ombin
ation
of
demo
graph
ic an
d inc
ome e
vents
12Gr
owth
in the
total
nomi
nal in
come
and g
rowth
in the
numb
er of
ho
useh
old m
embe
rs.
8.6(0.
72)
71.8
(3.83
)6.2
(0.63
)8.8
(0.78
)69
.3(4.
15)
6.1(0.
67)
7.6(1.
96)
91.2
(8.32
)7.0
(1.87
)
11.1
(0.84
)67
.5(3.
61)
7.5(0.
74)
10.2
(0.85
)67
.0(3.
99)
6.9(0.
74)
16.7
(2.88
)69
.5(8.
41)
11.6
(2.51
)TO
TAL H
OUSE
HOLD
S W
ITH E
VENT
S73
.3(1.
11)
54.6
(1.58
)40
.0(1.
37)
72.7
(1.21
)52
.8(1.
70)
38.4
(1.47
)76
.8(2.
87)
66.1
(3.89
)50
.8(3.
58)
TOTA
L HOU
SEHO
LDS
WITH
OUT E
VENT
S26
.7(1.
11)
4.8(1.
08)
1.3(0.
29)
27.3
(1.21
)5.1
(1.17
)1.4
(0.32
)23
.2(2.
87)
2.7(2.
70)
0.6(0.
63)
TOTA
L HOU
SEHO
LDS
100.0
0.00
0.00.0
041
.30.0
010
0.00.0
00.0
0.00
39.7
0.00
100.0
0.00
0.00.0
051
.40.0
0TO
TAL N
UMBE
R of
HOU
SEHO
LDS
WITH
EVE
NTS
2,879
,829
0.00
00.0
00
0.00
2,480
,408
0.00
00.0
00
0.00
399,4
210.0
00
0.00
00.0
0TO
TAL N
UMBE
R of
HOU
SEHO
LDS
WITH
OUT E
VENT
S1,0
51,64
90.0
00
0.00
00.0
093
1,188
0.00
00.0
00
0.00
120,4
610.0
00
0.00
00.0
0TO
TAL N
UMBE
R of
HOU
SEHO
LDS
3,931
,478
0.00
00.0
00
0.00
3,411
,596
0.00
00.0
00
0.00
519,8
820.0
00
0.00
00.0
0
1/ De
comp
ositio
n ba
sed
on eq
uatio
n [2]
Sour
ce: A
uthors
’ elab
oratio
n bas
ed on
data
from
Encu
esta
Nacio
nal d
e Hog
ares (
INEI
)
P(S/
E)
Table
9.E.
Dec
ompo
sition
of th
e exit
rate
from
pove
rty. S
ensit
ivity
analy
sis to
pov
erty
line (
Redu
ction
of 10
% in
pove
ry lin
e) 1/
Peru
. Urb
an ar
eas.
Perio
d 200
2-200
6.
Total
hous
ehold
sHo
useh
olds w
ith ch
ildre
nHo
useh
olds w
ithou
t chil
dren
P(ev
ent)
P(S/
E)Ex
itP(
even
t)
D E M O G R A P H I C &
C O M B I N E D
E V E N T S
VII -
Eve
nts no
t clas
sified
Exit
Even
ts
N O N D E M O G R A P H I C
I - Ex
clusiv
ely la
bor
incom
e eve
nts
II - E
xclus
ively
non-
labor
incom
e eve
nts
P(S/
E)Ex
itP(
even
t)
116
N°(1)
SE(2)
SE(1)
*(2)=(
3)SE
(1)SE
(2)SE
(1)*(2
)=(3)
SE(1)
SE(2)
SE(1)
*(2)=(
3)SE
TOTA
L LAB
OR E
VENT
S36
.91(0.
67)
36.20
(1.09
)13
.33(0.
45)
40.91
(0.81
)31
.12(1.
17)
12.71
(0.52
)26
.24(1.
10)
57.38
(2.41
)14
.97(0.
89)
1Gr
owth
in th
e num
ber o
f emp
loyed
mem
bers
6.5(0
.35)
31.3
(2.48
)2.0
(0.19
)6.2
(0.41
)20
.7(2
.62)
1.3(0
.18)
7.4(0
.66)
54.8
(4.57
)4.0
(0.49
)1.1
Grow
th in
the n
umbe
r of r
egist
ered
wag
e ear
ners
1.2(0
.17)
55.5
(6.96
)0.7
(0.11
)1.2
(0.20
)42
.7(8
.06)
0.5(0
.11)
1.4(0
.33)
83.0
(8.68
)1.2
(0.30
)1.2
Grow
th in
the n
umbe
r of n
on-re
gister
ed w
age e
arne
rs3.5
(0.26
)22
.1(2
.94)
0.8(0
.11)
3.6(0
.32)
13.0
(2.75
)0.5
(0.10
)3.3
(0.45
)49
.0(6
.96)
1.6(0
.31)
1.3Gr
owth
in th
e num
ber o
f non
-wag
e ear
ners
1.8(0
.17)
32.8
(4.66
)0.6
(0.10
)1.4
(0.19
)22
.6(5
.94)
0.3(0
.10)
2.7(0
.38)
46.7
(7.10
)1.3
(0.26
)
2Gr
owth
in to
tal ho
urly
wage
of m
embe
rs em
ploye
d in b
oth
obse
rvatio
ns11
.1(0
.42)
28.5
(1.77
)3.2
(0.23
)13
.2(0
.53)
26.3
(1.89
)3.5
(0.29
)5.6
(0.58
)42
.3(5
.04)
2.4(0
.34)
3Gr
owth
in th
e num
ber o
f wor
king h
ours
of m
embe
rs em
ploye
d in
both
obse
rvatio
ns3.5
(0.25
)16
.7(2
.41)
0.6(0
.09)
3.9(0
.31)
10.7
(2.28
)0.4
(0.09
)2.6
(0.38
)40
.2(7
.00)
1.1(0
.23)
4Gr
owth
in th
e num
ber o
f wor
king h
ours
and i
n the
total
hour
ly wa
ge of
mem
bers
emplo
yed i
n both
obse
rvatio
ns8.6
(0.44
)38
.5(2
.50)
3.3(0
.24)
9.9(0
.57)
32.9
(2.63
)3.3
(0.28
)5.0
(0.52
)67
.7(4
.61)
3.4(0
.46)
5Gr
owth
in th
e tota
l mon
thly
wage
of m
embe
rs em
ploye
d in b
oth
obse
rvatio
ns an
d in t
he nu
mber
of em
ploye
d mem
bers
7.1(0
.34)
59.6
(2.44
)4.2
(0.27
)7.7
(0.42
)55
.5(2
.77)
4.3(0
.32)
5.5(0
.59)
75.5
(4.81
)4.1
(0.52
)
TOTA
L NON
-LAB
OR E
VENT
S9.6
(0.37
)33
.3(1.
90)
3.2(0.
22)
4.5(0.
27)
8.5(1.
66)
0.4(0.
08)
23.3
(1.07
)45
.9(2.
60)
10.7
(0.75
)6
Grow
th in
the i
ncom
e fro
m pe
nsion
s 6.2
(0.31
)37
.3(2
.48)
2.3(0
.19)
1.7(0
.16)
10.3
(2.07
)0.2
(0.03
)18
.1(0
.99)
44.0
(2.96
)8.0
(0.66
)7
Grow
th in
publi
c mon
etary
tran
sfers
(soc
ial po
licy)
1.4(0
.14)
3.7(1
.92)
0.1(0
.03)
1.7(0
.17)
0.4(0
.37)
0.0(0
.01)
0.8(0
.20)
22.3
(11.0
6)0.2
(0.10
)8
Grow
th in
othe
r non
-labo
r inco
mes
2.0(0
.18)
41.9
(4.33
)0.8
(0.11
)1.1
(0.15
)18
.1(5
.45)
0.2(0
.07)
4.4(0
.51)
58.0
(5.90
)2.5
(0.37
)
III - L
abor
and n
on-
labor
inco
me ev
ents
9Gr
owth
in la
bor a
nd no
n-lab
or in
come
s12
.5(0
.44)
50.9
(1.86
)6.3
(0.33
)13
.3(0
.53)
43.8
(2.14
)5.8
(0.38
)10
.4(0
.75)
74.9
(3.27
)7.7
(0.67
)
IV E
xclus
ively
demo
grap
hic ev
ents
10Re
ducti
on in
the t
otal n
umbe
r of h
ouse
hold
memb
ers;
the to
tal
nomi
nal in
come
rema
ins co
nstan
t5.9
(0.34
)17
.1(2
.44)
0.7(0
.11)
6.3(0
.42)
12.8
(2.29
)0.6
(0.11
)4.9
(0.54
)33
.5(6
.93)
1.1(0
.29)
V -
Demo
grap
hic
even
ts lea
ding t
o inc
ome c
hang
es11
Grow
th in
the n
umbe
r of la
bor o
r non
-labo
r inc
ome e
arne
rs
due t
o the
entra
nce o
f mem
bers
to the
hous
ehold
1.6(0
.18)
25.2
(4.53
)0.4
(0.07
)1.3
(0.18
)13
.1(3
.20)
0.2(0
.04)
2.5(0
.42)
42.6
(8.75
)0.9
(0.25
)
VI -
Comb
inatio
n of
demo
grap
hic an
d inc
ome e
vents
12Gr
owth
in th
e tota
l nom
inal in
come
and g
rowt
h in t
he nu
mber
of
hous
ehold
mem
bers
.6.2
(0.31
)52
.4(2
.89)
2.6(0
.20)
6.0(0
.36)
45.9
(3.35
)2.3
(0.23
)6.7
(0.63
)69
.9(5
.44)
3.5(0
.43)
5.0(0
.33)
46.0
(3.43
)2.2
(0.20
)4.4
(0.39
)34
.5(4
.13)
1.5(0
.20)
6.4(0
.62)
66.6
(4.68
)4.2
(0.52
)TO
TAL H
OUSE
HOLD
S W
ITH E
VENT
S77
.7(0
.56)
38.7
(0.77
)28
.8(0
.61)
76.7
(0.68
)31
.8(0
.87)
23.4
(0.67
)80
.4(1
.02)
56.4
(1.43
)43
.1(1
.24)
TOTA
L HOU
SEHO
LDS
WITH
OUT E
VENT
S22
.3(0
.56)
0.0(0
.04)
0.0(0
.01)
23.3
(0.68
)0.0
0.00
0.00.0
019
.6(1
.02)
0.2(0
.18)
0.0(0
.04)
TOTA
L HOU
SEHO
LDS
100.0
0.00
0.00.0
028
.80.0
010
0.00.0
00.0
0.00
23.4
0.00
100.0
0.00
0.00.0
043
.20.0
0TO
TAL N
UMBE
R of
HOU
SEHO
LDS
WITH
EVE
NTS
2,614
,721
0.00
00.0
00
0.00
1,876
,736
0.00
00.0
00
0.00
737,9
850.0
00
0.00
00.0
0TO
TAL N
UMBE
R of
HOU
SEHO
LDS
WITH
OUT E
VENT
S74
9,895
0.00
00.0
00
0.00
570,5
200.0
00
0.00
00.0
017
9,375
0.00
00.0
00
0.00
TOTA
L NUM
BER
of H
OUSE
HOLD
S3,3
64,61
60.0
00
0.00
00.0
02,4
47,25
60.0
00
0.00
00.0
091
7,360
0.00
00.0
00
0.00
1/ De
comp
ositio
n ba
sed
on eq
uatio
n [2]
Sour
ce: A
uthor
s’ ela
bora
tion b
ased
on da
ta fro
m En
cues
ta Pe
rman
ente
de H
ogar
es (I
NDEC
).
D E M O G R A P H I C &
C O M B I N E D
E V E N T S
VII -
Eve
nts no
t clas
sified
Exit
Even
ts
N O N D E M O G R A P H I C
I - E
xclus
ively
labor
inc
ome e
vents
II - E
xclus
ively
non-
labor
inco
me ev
ents
P(S/
E)Ex
itP(
even
t)P(
S/E)
Table
10.A
. Dec
ompo
sition
of th
e exit
rate
from
pove
rty. S
ensit
ivity
analy
sis to
pov
erty
line (
Incre
ase o
f 10%
in po
very
line)
1/Ar
gent
ina. U
rban
area
s. Pe
riod 2
003 2
006.
Total
hous
ehold
sHo
useh
olds w
ith ch
ildre
nHo
useh
olds w
ithou
t chil
dren
P(ev
ent)
P(S/
E)Ex
itP(
even
t)
116 117
N°(1)
SE(2)
SE(1)
*(2)=(
3)SE
(1)SE
(2)SE
(1)*(2
)=(3)
SE(1)
SE(2)
SE(1)
*(2)=(
3)SE
TOTA
L LAB
OR EV
ENTS
18.02
(0.46
)30
.00(1.
31)
5.41
(0.28
)18
.84(0.
55)
24.41
(1.41
)4.6
0(0.
30)
15.73
(0.85
)48
.80(2.
95)
7.67
(0.63
)1
Grow
th in
the n
umbe
r of e
mploy
ed m
embe
rs4.2
(0.24
)39
.9(2.
91)
1.7(0.
16)
3.6(0.
27)
25.5
(3.32
)0.9
(0.14
)6.0
(0.56
)63
.8(4.
51)
3.9(0.
46)
1.1Gr
owth
in th
e num
ber o
f regis
tered
wag
e earn
ers1.7
(0.15
)50
.5(4.
66)
0.8(0.
11)
1.4(0.
17)
38.4
(5.87
)0.5
(0.11
)2.4
(0.35
)70
.5(6.
62)
1.7(0.
30)
1.2Gr
owth
in th
e num
ber o
f non
-regis
tered
wag
e earn
ers1.5
(0.15
)26
.4(4.
59)
0.4(0.
08)
1.4(0.
17)
17.6
(4.75
)0.2
(0.07
)1.6
(0.30
)48
.0(9.
24)
0.8(0.
22)
1.3Gr
owth
in th
e num
ber o
f non
-wag
e earn
ers1.1
(0.13
)41
.9(5.
69)
0.5(0.
08)
0.8(0.
12)
16.7
(5.72
)0.1
(0.05
)2.0
(0.34
)68
.4(7.
77)
1.4(0.
28)
2Gr
owth
in to
tal ho
urly w
age o
f mem
bers
emplo
yed i
n both
ob
serva
tions
7.0(0.
31)
15.9
(1.67
)1.1
(0.13
)7.7
(0.38
)13
.8(1.
76)
1.1(0.
15)
4.9(0.
50)
25.2
(4.53
)1.2
(0.25
)
3Gr
owth
in th
e num
ber o
f work
ing ho
urs of
mem
bers
emplo
yed i
n bo
th ob
serva
tions
1.3(0.
14)
16.4
(4.11
)0.2
(0.06
)1.4
(0.17
)13
.8(4.
20)
0.2(0.
07)
1.0(0.
24)
26.6
(11.10
)0.3
(0.14
)
4Gr
owth
in th
e num
ber o
f work
ing ho
urs an
d in t
he to
tal ho
urly
wage
of m
embe
rs em
ploye
d in b
oth ob
serva
tions
2.6(0.
19)
34.5
(3.62
)0.9
(0.12
)3.0
(0.24
)33
.2(3.
94)
1.0(0.
15)
1.6(0.
29)
41.6
(9.20
)0.7
(0.18
)
5Gr
owth
in th
e tota
l mon
thly w
age o
f mem
bers
emplo
yed i
n both
ob
serva
tions
and i
n the
numb
er of
emplo
yed m
embe
rs2.9
(0.20
)51
.9(3.
56)
1.5(0.
14)
3.1(0.
24)
46.0
(3.99
)1.4
(0.17
)2.2
(0.34
)74
.7(6.
89)
1.7(0.
29)
TOTA
L NON
-LABO
UR EV
ENTS
7.4(0.
32)
66.7
(2.05
)5.0
(0.27
)3.7
(0.26
)38
.6(3.
49)
1.4(0.
17)
17.9
(0.91
)82
.9(2.
08)
14.9
(0.85
)6
Grow
th in
the in
come
from
pens
ions
6.1(0.
29)
74.6
(2.10
)4.6
(0.26
)2.5
(0.21
)49
.5(4.
38)
1.2(0.
16)
16.4
(0.88
)85
.3(2.
07)
14.0
(0.83
)8
Grow
th in
other
incom
e1.3
(0.13
)29
.1(4.
69)
0.4(0.
07)
1.2(0.
15)
16.5
(4.54
)0.2
(0.06
)1.5
(0.28
)57
.5(9.
04)
0.9(0.
22)
III - L
abor
and n
on-
labor
incom
e eve
nts9
Grow
th in
labo
r and
non-l
abor
incom
es5.9
(0.28
)72
.7(2.
11)
4.3(0.
25)
4.9(0.
30)
59.9
(3.02
)2.9
(0.24
)9.0
(0.69
)92
.0(1.
95)
8.3(0.
67)
IV Ex
clusiv
ely
demo
graph
ic ev
ents
10Re
ducti
on in
the t
otal n
umbe
r of h
ouse
hold
memb
ers; th
e tota
l no
mina
l inco
me re
mains
cons
tant
5.6(0.
26)
14.6
(1.77
)0.8
(0.11
)6.4
(0.32
)13
.5(1.
86)
0.9(0.
13)
3.4(0.
40)
20.4
(5.14
)0.7
(0.20
)
V - D
emog
raphic
ev
ents
leadin
g to
incom
e cha
nges
11Gr
owth
in th
e num
ber o
f labo
r or n
on-la
bor in
come
earne
rs du
e to t
he en
tranc
e of m
embe
rs to
the ho
useh
old11
.8(0.
38)
45.9
(1.70
)5.4
(0.27
)11
.6(0.
44)
38.6
(1.98
)4.5
(0.29
)12
.6(0.
74)
64.7
(2.93
)8.2
(0.63
)
VI - C
ombin
ation
of
demo
graph
ic an
d inc
ome e
vents
12Gr
owth
in th
e tota
l nom
inal in
come
and g
rowth
in the
numb
er of
hous
ehold
mem
bers.
7.7(0.
31)
66.8
(1.90
)5.1
(0.26
)8.1
(0.36
)62
.0(2.
21)
5.0(0.
30)
6.5(0.
57)
83.4
(3.34
)5.4
(0.52
)
20.9
(0.47
)50
.1(1.
26)
10.4
(0.36
)21
.4(0.
55)
41.8
(1.44
)9.0
(0.39
)19
.3(0.
89)
76.1
(2.06
)14
.6(0.
81)
TOTA
L HOU
SEHO
LDS W
ITH E
VENT
S77
.40.0
047
.2(0.
67)
36.5
(0.57
)74
.9(0.
60)
37.7
(0.77
)28
.2(0.
62)
84.4
(0.81
)70
.8(1.
12)
59.7
(1.11
)TO
TAL H
OUSE
HOLD
S WITH
OUT E
VENT
S22
.6(0.
49)
0.0(0.
03)
0.0(0.
01)
25.1
(0.60
)0.0
(0.04
)0.0
(0.01
)15
.6(0.
81)
0.00.0
00.0
0.00
TOTA
L HOU
SEHO
LDS
100.0
(0.49
)0.0
0.00
36.5
0.00
100.0
0.00
0.00.0
028
.30.0
010
0.00.0
00.0
0.00
59.7
0.00
TOTA
L NUM
BER o
f HOU
SEHO
LDS W
ITH E
VENT
S3,2
12,36
90.0
00
0.00
00.0
02,2
92,38
60.0
00
0.00
00.0
091
9,983
0.00
00.0
00
0.00
TOTA
L NUM
BER o
f HOU
SEHO
LDS W
ITHOU
T EVE
NTS
938,8
550.0
00
0.00
00.0
076
8,425
0.00
00.0
00
0.00
170,4
300.0
00
0.00
00.0
0TO
TAL N
UMBE
R of H
OUSE
HOLD
S4,1
51,22
40.0
00
0.00
00.0
03,0
60,81
10.0
00
0.00
00.0
01,0
90,41
30.0
00
0.00
00.0
0
1/ De
comp
ositio
n ba
sed
on eq
uatio
n [2]
2/ Me
tropo
litan a
reas o
f Rec
ife, S
alvad
or, Be
lo Ho
rizon
te, R
io de
Jane
iro, S
ão Pa
ulo an
d Port
o Aleg
re.
Table
10.B.
Dec
ompo
sition
of th
e exit
rate
from
pove
rty. S
ensit
ivity
analy
sis to
pov
erty l
ine (In
creas
e of 1
0% in
pove
ry lin
e) 1/
Braz
il. Ur
ban a
reas.
Perio
d 200
3-200
6. 2/
Total
hous
ehold
sHo
useh
olds w
ith ch
ildren
Hous
ehold
s with
out c
hildre
nEx
itEv
ents
N O N D E M O G R A P H I C
I - Ex
clusiv
ely la
bor
incom
e eve
nts
P(S/E)
Exit
P(eve
nt)P(S
/E)P(e
vent)
P(S/E)
II - Ex
clusiv
ely no
n-lab
or inc
ome e
vents
VII -
Even
ts no
t clas
sified
P(eve
nt)
D E M O G R A P H I
C &
C O M B I N E D
E V E N T S
Exit
118
N°(1)
SE(2)
SE(1)
*(2)=(
3)SE
(1)SE
(2)SE
(1)*(2
)=(3)
SE(1)
SE(2)
SE(1)
*(2)=(
3)SE
TOTA
L LAB
OR EV
ENTS
38.61
(2.00
)60
.95(3.
20)
23.53
(1.73
)43
.36(2.
44)
58.25
(3.77
)25
.26(2.
15)
29.11
(3.33
)69
.01(5.
91)
20.09
(2.90
)1
Grow
th in
the n
umbe
r of e
mploy
ed m
embe
rs9.2
(1.18
)73
.5(5.
85)
6.7(1.
04)
8.2(1.
39)
76.2
(7.42
)6.3
(1.24
)11
.0(2.
17)
69.4
(9.45
)7.6
(1.89
)1.1
Grow
th in
the n
umbe
r of r
egist
ered w
age e
arners
3.2(0.
74)
86.4
(7.76
)2.8
(0.69
)2.3
(0.75
)95
.0(4.
99)
2.1(0.
74)
5.1(1.
61)
78.7
(13.32
)4.0
(1.43
)1.2
Grow
th in
the n
umbe
r of n
on-re
gister
ed w
age e
arners
2.6(0.
63)
68.6
(11.22
)1.7
(0.53
)2.5
(0.79
)69
.9(14
.09)
1.8(0.
67)
2.6(1.
02)
66.2
(18.44
)1.7
(0.85
)1.3
Grow
th in
the n
umbe
r of n
on-w
age e
arners
3.4(0.
74)
64.9
(10.56
)2.2
(0.61
)3.5
(0.93
)68
.4(12
.94)
2.4(0.
77)
3.3(1.
21)
57.5
(18.35
)1.9
(0.97
)
2Gr
owth
in to
tal ho
urly
wage
of m
embe
rs em
ploye
d in b
oth
obse
rvatio
ns9.0
(1.16
)41
.8(6.
65)
3.8(0.
77)
11.8
(1.59
)36
.1(6.
82)
4.3(0.
98)
3.3(1.
32)
83.3
(15.24
)2.7
(1.21
)
3Gr
owth
in th
e num
ber o
f wor
king h
ours
of m
embe
rs em
ploye
d in
both
obse
rvatio
ns2.8
(0.66
)15
.3(8.
57)
0.4(0.
27)
2.7(0.
77)
0.00.0
00.0
0.00
3.2(1.
25)
40.6
(19.34
)1.3
(0.79
)
4Gr
owth
in th
e num
ber o
f wor
king h
ours
and i
n the
total
hour
ly wa
ge of
mem
bers
emplo
yed i
n both
obse
rvatio
ns7.7
(1.08
)54
.4(7.
27)
4.2(0.
81)
9.5(1.
44)
54.0
(8.03
)5.1
(1.08
)4.3
(1.44
)56
.1(17
.16)
2.4(1.
09)
5Gr
owth
in th
e tota
l mon
thly
wage
of m
embe
rs em
ploye
d in b
oth
obse
rvatio
ns an
d in t
he nu
mber
of em
ploye
d mem
bers
9.9(1.
25)
85.0
(4.54
)8.4
(1.15
)11
.2(1.
61)
86.1
(5.06
)9.6
(1.48
)7.4
(1.91
)81
.9(10
.01)
6.0(1.
76)
TOTA
L NON
-LAB
OR EV
ENTS
8.7(1.
20)
32.0
(6.44
)2.8
(0.69
)3.6
(0.90
)19
.6(9.
26)
0.7(0.
36)
19.0
(2.94
)36
.7(7.
95)
7.0(1.
89)
6Gr
owth
in th
e inc
ome f
rom
pens
ions
5.6(1.
01)
38.4
(8.41
)2.2
(0.61
)0.8
(0.40
)38
.5(24
.67)
0.3(0.
22)
15.4
(2.77
)38
.4(8.
91)
5.9(1.
75)
7Gr
owth
in pu
blic m
oneta
ry tra
nsfer
s (so
cial p
olicy
) 2.2
(0.57
)12
.7(8.
32)
0.3(0.
19)
2.8(0.
81)
14.5
(9.41
)0.4
(0.29
)0.8
(0.60
)0.0
0.00
0.00.0
08
Grow
th in
othe
r non
-labo
r inco
mes
0.9(0.
40)
38.1
(20.91
)0.4
(0.26
)0.0
0.00
0.00.0
00.0
0.00
2.8(1.
19)
38.1
(20.91
)1.1
(0.76
)
III - L
abor
and n
on-
labor
incom
e eve
nts9
Grow
th in
labo
r and
non-l
abor
incom
es9.2
(1.19
)58
.0(6.
49)
5.3(0.
89)
9.9(1.
47)
49.0
(7.39
)4.8
(1.02
)7.8
(2.00
)80
.7(12
.58)
6.3(1.
70)
IV E
xclus
ively
demo
graph
ic ev
ents
10Re
ducti
on in
the t
otal n
umbe
r of h
ouse
hold
memb
ers; th
e tota
l no
mina
l inco
me re
mains
cons
tant
3.1(0.
69)
29.3
(9.98
)0.9
(0.36
)3.0
(0.86
)28
.2(12
.64)
0.9(0.
45)
3.2(1.
16)
31.4
(16.14
)1.0
(0.61
)
V - D
emog
raphic
ev
ents
leadin
g to
incom
e cha
nges
11Gr
owth
in th
e num
ber o
f labo
r or n
on-la
bor in
come
earn
ers
due t
o the
entra
nce o
f mem
bers
to the
hous
ehold
0.6(0.
32)
40.5
(25.09
)0.2
(0.19
)0.5
(0.39
)0.0
0.00
0.00.0
00.7
(0.57
)10
0.00.0
00.7
(0.57
)
VI - C
ombin
ation
of
demo
graph
ic an
d inc
ome e
vents
12Gr
owth
in th
e tota
l nom
inal in
come
and g
rowth
in the
numb
er of
ho
useh
old m
embe
rs.7.0
(1.03
)69
.6(7.
16)
4.9(0.
86)
8.3(1.
36)
69.1
(8.12
)5.7
(1.12
)4.5
(1.46
)71
.4(15
.09)
3.2(1.
24)
10.1
(1.20
)66
.8(5.
93)
6.7(1.
01)
11.6
(1.55
)69
.3(6.
56)
8.0(1.
34)
7.1(1.
79)
58.7
(13.03
)4.2
(1.37
)TO
TAL H
OUSE
HOLD
S WITH
EVE
NTS
77.3
(1.68
)57
.4(2.
24)
44.4
(1.99
)80
.3(1.
90)
56.5
(2.68
)45
.4(2.
40)
71.5
(3.26
)59
.4(4.
09)
42.5
(3.53
)TO
TAL H
OUSE
HOLD
S WITH
OUT E
VENT
S22
.7(1.
68)
0.00.0
0.
0.00
19.7
(1.90
)0.0
0.00
.0.0
028
.5(3.
26)
0.00.0
0.
0.00
TOTA
L HOU
SEHO
LDS
100.0
0.00
0.00.0
044
.40.0
010
0.00.0
00.0
0.00
45.4
0.00
100.0
0.00
0.00.0
042
.50.0
0TO
TAL N
UMBE
R of
HOU
SEHO
LDS W
ITH E
VENT
S76
,567
0.00
00.0
00
0.00
52,98
50.0
00
0.00
00.0
023
,582
0.00
00.0
00
0.00
TOTA
L NUM
BER
of H
OUSE
HOLD
S WITH
OUT E
VENT
S22
,448
0.00
00.0
00
0.00
13,02
70.0
00
0.00
00.0
09,4
210.0
00
0.00
00.0
0TO
TAL N
UMBE
R of
HOU
SEHO
LDS
99,01
50.0
00
0.00
00.0
066
,012
0.00
00.0
00
0.00
33,00
30.0
00
0.00
00.0
0
1/ De
comp
ositio
n ba
sed
on eq
uatio
n [2]
Sour
ce: A
uthors
’ elab
oratio
n bas
ed on
data
from
Encu
esta
Perm
anen
te de
Hog
ares (
EPH-
INDE
C).
Sour
ce: A
uthors
’ elab
oratio
n bas
ed on
data
from
Encu
esta
de H
ogare
s de P
ropós
itos M
últipl
es (IN
EC).
P(S/E
)
Table
10.C
. Dec
ompo
sition
of th
e exit
rate
from
pove
rty. S
ensit
ivity
analy
sis to
pov
erty l
ine (In
creas
e of 1
0% in
pove
ry lin
e) 1/
Costa
Rica
.Urb
an ar
eas.
Perio
d 200
6-200
8.
Total
hous
ehold
sHo
useh
olds w
ith ch
ildren
Hous
ehold
s with
out c
hildr
enP(
even
t)P(
S/E)
Exit
P(ev
ent)
D E M O G R A P H I C &
C O M B I N E D
E V E N T S
VII -
Eve
nts no
t clas
sified
Exit
Even
ts
N O N D E M O G R A P H I C
I - Ex
clusiv
ely la
bor
incom
e eve
nts
II - E
xclus
ively
non-
labor
incom
e eve
nts
P(S/E
)Ex
itP(
even
t)
118 119
N°(1)
SE(2)
SE(1)
*(2)=(
3)SE
(1)SE
(2)SE
(1)*(2
)=(3)
SE(1)
SE(2)
SE(1)
*(2)=(
3)SE
TOTA
L LAB
OR EV
ENTS
36.45
(0.61
)41
.29(1.
03)
15.05
(0.46
)37
.68(0.
69)
38.63
(1.12
)14
.55(0.
50)
31.19
(1.34
)54
.99(2.
56)
17.15
(1.09
)1
Grow
th in
the n
umbe
r of e
mploy
ed m
embe
rs4.7
(0.26
)34
.7(2.
77)
1.6(0.
16)
4.4(0.
29)
30.2
(3.11
)1.3
(0.16
)6.1
(0.67
)48
.6(5.
71)
3.0(0.
48)
1.1Gr
owth
in th
e num
ber o
f reg
istere
d wag
e earn
ers0.4
(0.08
)57
.1(9.
11)
0.2(0.
06)
0.3(0.
08)
38.1
(10.16
)0.1
(0.04
)0.8
(0.26
)92
.8(7.
04)
0.7(0.
25)
1.2Gr
owth
in th
e num
ber o
f non
-regis
tered
wag
e earn
ers2.2
(0.18
)31
.7(3.
87)
0.7(0.
10)
2.2(0.
20)
25.3
(3.98
)0.6
(0.10
)2.3
(0.43
)58
.2(9.
46)
1.3(0.
33)
1.3Gr
owth
in th
e num
ber o
f non
-wag
e earn
ers2.0
(0.18
)33
.3(4.
24)
0.7(0.
11)
1.8(0.
19)
34.7
(5.19
)0.6
(0.12
)3.0
(0.46
)29
.6(7.
11)
0.9(0.
25)
2Gr
owth
in to
tal ho
urly
wage
of m
embe
rs em
ploye
d in b
oth
obse
rvatio
ns11
.9(0.
40)
35.7
(1.71
)4.3
(0.25
)12
.6(0.
46)
33.4
(1.83
)4.2
(0.28
)9.0
(0.80
)49
.1(4.
67)
4.4(0.
57)
3Gr
owth
in th
e num
ber o
f wor
king h
ours
of me
mbers
emplo
yed i
n bo
th ob
serva
tions
4.4(0.
26)
21.7
(2.52
)1.0
(0.13
)4.6
(0.29
)18
.6(2.
60)
0.8(0.
13)
3.7(0.
55)
37.9
(7.39
)1.4
(0.35
)
4Gr
owth
in th
e num
ber o
f wor
king h
ours
and i
n the
total
hour
ly wa
ge of
mem
bers
emplo
yed i
n both
obse
rvatio
ns8.8
(0.35
)46
.8(2.
08)
4.1(0.
25)
9.1(0.
39)
43.7
(2.27
)4.0
(0.27
)7.9
(0.79
)62
.3(4.
92)
4.9(0.
63)
5Gr
owth
in th
e tota
l mon
thly
wage
of m
embe
rs em
ploye
d in b
oth
obse
rvatio
ns an
d in t
he nu
mber
of em
ploye
d mem
bers
6.6(0.
31)
61.7
(2.32
)4.1
(0.25
)7.1
(0.36
)59
.5(2.
52)
4.2(0.
29)
4.5(0.
59)
76.6
(5.66
)3.4
(0.52
)
TOTA
L NON
-LAB
OR EV
ENTS
5.7(0.
30)
36.3
(2.53
)2.1
(0.18
)3.7
(0.27
)22
.6(3.
07)
0.8(0.
13)
14.2
(1.05
)51
.7(3.
74)
7.3(0.
76)
6Gr
owth
in th
e inc
ome f
rom
pens
ions
0.8(0.
11)
49.9
(6.86
)0.4
(0.08
)0.4
(0.09
)29
.5(9.
21)
0.1(0.
05)
2.3(0.
45)
66.5
(9.17
)1.5
(0.36
)7
Grow
th in
publi
c mon
etary
trans
fers (
socia
l poli
cy)
0.9(0.
11)
0.00.0
0.
0.00
0.8(0.
12)
0.00.0
0.
0.00
1.0(0.
28)
0.00.0
0.
0.00
8Gr
owth
in ot
her n
on-la
bor in
come
s 4.0
(0.26
)41
.3(3.
08)
1.7(0.
16)
2.5(0.
23)
28.7
(4.11
)0.7
(0.12
)10
.8(0.
94)
53.5
(4.24
)5.8
(0.68
)
III - L
abor
and n
on-
labor
incom
e eve
nts9
Grow
th in
labo
r and
non-l
abor
incom
es11
.9(0.
41)
52.5
(1.82
)6.2
(0.31
)12
.0(0.
46)
47.2
(2.03
)5.6
(0.33
)11
.7(0.
91)
75.8
(3.44
)8.9
(0.81
)
IV E
xclus
ively
demo
graph
ic ev
ents
10Re
ducti
on in
the t
otal n
umbe
r of h
ouse
hold
memb
ers; th
e tota
l no
mina
l inco
me re
mains
cons
tant
5.8(0.
30)
16.4
(1.83
)1.0
(0.12
)6.2
(0.34
)16
.4(1.
98)
1.0(0.
13)
4.4(0.
60)
16.4
(4.73
)0.7
(0.23
)
V - D
emog
raphic
ev
ents
leadin
g to
incom
e cha
nges
11Gr
owth
in th
e num
ber o
f labo
r or n
on-la
bor in
come
earn
ers du
e to
the en
tranc
e of m
embe
rs to
the ho
useh
old2.3
(0.19
)44
.3(4.
06)
1.0(0.
13)
2.2(0.
20)
41.8
(4.69
)0.9
(0.14
)2.9
(0.49
)52
.4(8.
07)
1.5(0.
34)
VI - C
ombin
ation
of
demo
graph
ic an
d inc
ome e
vents
12Gr
owth
in th
e tota
l nom
inal in
come
and g
rowth
in the
numb
er of
ho
useh
old m
embe
rs.7.3
(0.33
)66
.7(2.
18)
4.9(0.
27)
7.6(0.
37)
64.4
(2.41
)4.9
(0.30
)5.9
(0.70
)79
.4(4.
48)
4.7(0.
65)
7.2(0.
33)
50.4
(2.33
)3.6
(0.23
)6.3
(0.34
)47
.1(2.
79)
3.0(0.
24)
11.3
(0.89
)58
.2(4.
11)
6.6(0.
70)
TOTA
L HOU
SEHO
LDS W
ITH E
VENT
S76
.7(0.
53)
44.1
(0.72
)33
.9(0.
61)
75.6
(0.60
)40
.8(0.
80)
30.8
(0.66
)81
.5(1.
10)
57.4
(1.56
)46
.8(1.
43)
TOTA
L HOU
SEHO
LDS W
ITHOU
T EVE
NTS
23.3
(0.53
)0.0
0.00
0.00.0
024
.4(0.
60)
0.00.0
00.0
0.00
18.5
(1.10
)0.0
0.00
0.00.0
0TO
TAL H
OUSE
HOLD
S10
0.033
.910
0.030
.810
0.046
.8TO
TAL N
UMBE
R of
HOU
SEHO
LDS W
ITH E
VENT
S2,0
05,29
01,6
00,20
340
5,087
TOTA
L NUM
BER
of H
OUSE
HOLD
S WITH
OUT E
VENT
S60
8,817
517,1
6191
,656
TOTA
L NUM
BER
of H
OUSE
HOLD
S2,6
14,10
72,1
17,36
449
6,743
1/ De
comp
ositio
n ba
sed
on eq
uatio
n [2]
Sour
ce: A
uthors
’ elab
oratio
n bas
ed on
data
from
Encu
esta
Perm
anen
te de
Hog
ares (
EPH-
INDE
C).
Sour
ce: A
uthors
’ elab
oratio
n bas
ed on
data
from
Encu
esta
de E
mpleo
, Des
emple
o y S
ubem
pleo (
INEC
).
P(S/E
)
Table
10.D
. Dec
ompo
sition
of th
e exit
rate
from
pove
rty. S
ensit
ivity
analy
sis to
pov
erty l
ine (In
creas
e of 1
0% in
pove
ry lin
e) 1/
Ecua
dor. U
rban
area
s. Pe
riod 2
004-2
008.
Total
hous
ehold
sHo
useh
olds w
ith ch
ildren
Hous
ehold
s with
out c
hildr
enP(
even
t)P(
S/E)
Exit
P(ev
ent)
D E M O G R A P H I C &
C O M B I N E D
E V E N T S
VII -
Eve
nts no
t clas
sified
Exit
Even
ts
N O N D E M O G R A P H I C
I - Ex
clusiv
ely la
bor
incom
e eve
nts
II - E
xclus
ively
non-
labor
incom
e eve
nts
P(S/E
)Ex
itP(
even
t)
120
N°(1)
SE(2)
SE(1)
*(2)=(
3)SE
(1)SE
(2)SE
(1)*(2
)=(3)
SE(1)
SE(2)
SE(1)
*(2)=(
3)SE
TOTA
L LAB
OR E
VENT
S27
.64(1.
11)
34.56
(2.40
)8.5
2(0.
69)
28.11
(1.20
)33
.57(2.
56)
8.51
(0.76
)24
.70(2.
90)
41.49
(6.59
)8.5
4(1.
65)
1Gr
owth
in the
numb
er of
emplo
yed m
embe
rs6.0
(0.58
)32
.4(4.
69)
1.7(0.
30)
5.9(0.
62)
29.9
(4.96
)1.6
(0.31
)6.8
(1.72
)45
.3(13
.17)
2.6(0.
97)
1.1Gr
owth
in the
numb
er of
regis
tered
wag
e ear
ners
0.8(0.
22)
55.4
(13.62
)0.4
(0.15
)0.8
(0.23
)44
.9(15
.29)
0.3(0.
14)
1.1(0.
65)
100.0
0.00
0.9(0.
55)
1.2Gr
owth
in the
numb
er of
non-
regis
tered
wag
e ear
ners
2.8(0.
40)
26.9
(6.57
)0.7
(0.19
)2.6
(0.40
)26
.2(7.
08)
0.6(0.
19)
4.1(1.
47)
29.6
(16.57
)1.0
(0.66
)1.3
Grow
th in
the nu
mber
of no
n-wa
ge ea
rner
s2.4
(0.37
)31
.0(7.
30)
0.7(0.
19)
2.5(0.
42)
29.3
(7.59
)0.7
(0.20
)1.7
(0.77
)47
.5(23
.86)
0.7(0.
47)
2Gr
owth
in tot
al ho
urly
wage
of m
embe
rs em
ploye
d in b
oth
obse
rvatio
ns5.8
(0.56
)24
.8(4.
58)
1.3(0.
27)
5.9(0.
61)
23.1
(4.79
)1.2
(0.29
)5.3
(1.40
)35
.9(13
.98)
1.6(0.
79)
3Gr
owth
in the
numb
er of
work
ing ho
urs o
f mem
bers
emplo
yed i
n bo
th ob
serva
tions
3.8(0.
48)
22.0
(5.46
)0.7
(0.21
)4.0
(0.53
)20
.7(5.
64)
0.7(0.
23)
2.8(1.
12)
32.8
(19.41
)0.8
(0.55
)
4Gr
owth
in the
numb
er of
work
ing ho
urs a
nd in
the t
otal h
ourly
wa
ge of
mem
bers
emplo
yed i
n both
obse
rvatio
ns4.9
(0.51
)38
.8(5.
37)
1.7(0.
30)
5.1(0.
55)
41.5
(5.73
)1.9
(0.35
)4.0
(1.31
)17
.3(11
.98)
0.6(0.
42)
5Gr
owth
in the
total
mon
thly
wage
of m
embe
rs em
ploye
d in b
oth
obse
rvatio
ns an
d in t
he nu
mber
of em
ploye
d mem
bers
7.1(0.
62)
48.2
(4.59
)3.1
(0.41
)7.3
(0.68
)46
.4(4.
90)
3.1(0.
45)
5.8(1.
39)
63.0
(12.13
)3.0
(0.99
)
TOTA
L NON
-LAB
OR E
VENT
S3.6
(0.47
)26
.2(5.
87)
0.8(0.
22)
2.7(0.
41)
25.8
(6.91
)0.6
(0.20
)9.2
(2.10
)26
.9(10
.59)
2.1(0.
93)
6Gr
owth
in the
inco
me fr
om pe
nsion
s 0.2
(0.12
)57
.0(34
.68)
0.1(0.
08)
0.1(0.
08)
0.00.0
00.0
0.00
0.7(0.
66)
100.0
0.00
0.5(0.
55)
7Gr
owth
in pu
blic m
oneta
ry tra
nsfer
s (so
cial p
olicy
) 0.1
(0.08
)0.0
0.00
.0.0
00.1
(0.09
)0.0
0.00
.0.0
00.0
0.00
0.00.0
00.0
0.00
8Gr
owth
in oth
er no
n-lab
or inc
omes
3.3
(0.45
)25
.4(5.
93)
0.8(0.
20)
2.5(0.
39)
27.8
(7.33
)0.6
(0.20
)8.6
(2.01
)21
.2(9.
69)
1.5(0.
76)
III - L
abor
and n
on-
labor
incom
e eve
nts9
Grow
th in
labor
and n
on-la
bor in
come
s13
.0(0.
83)
57.1
(3.29
)6.6
(0.59
)13
.3(0.
91)
55.3
(3.53
)6.6
(0.64
)10
.7(1.
87)
70.6
(8.00
)6.3
(1.41
)
IV E
xclus
ively
demo
graph
ic ev
ents
10Re
ducti
on in
the t
otal n
umbe
r of h
ouse
hold
memb
ers;
the to
tal
nomi
nal in
come
rema
ins co
nstan
t6.2
(0.57
)16
.8(3.
75)
0.9(0.
23)
6.4(0.
62)
14.8
(3.67
)0.8
(0.23
)5.0
(1.36
)32
.8(14
.49)
1.4(0.
78)
V - D
emog
raphic
ev
ents
leadin
g to
incom
e cha
nges
11Gr
owth
in the
numb
er of
labor
or no
n-lab
or inc
ome e
arners
due t
o the
entra
nce o
f mem
bers
to the
hous
ehold
2.1(0.
33)
31.8
(7.50
)0.6
(0.18
)2.0
(0.33
)26
.8(7.
96)
0.5(0.
17)
3.2(1.
15)
50.4
(18.05
)1.4
(0.70
)
VI - C
ombin
ation
of
demo
graph
ic an
d inc
ome e
vents
12Gr
owth
in the
total
nomi
nal in
come
and g
rowth
in the
numb
er of
ho
useh
old m
embe
rs.
8.5(0.
67)
59.1
(4.04
)4.5
(0.49
)8.8
(0.73
)57
.1(4.
30)
4.5(0.
53)
7.0(1.
68)
74.4
(11.14
)4.3
(1.24
)
10.7
(0.77
)54
.4(3.
70)
5.2(0.
55)
9.8(0.
78)
52.2
(4.09
)4.6
(0.55
)15
.9(2.
58)
62.8
(8.25
)8.3
(1.83
)TO
TAL H
OUSE
HOLD
S W
ITH E
VENT
S71
.6(1.
03)
42.5
(1.52
)27
.1(1.
09)
71.0
(1.13
)41
.0(1.
61)
26.2
(1.15
)75
.8(2.
57)
51.1
(3.97
)32
.3(2.
96)
TOTA
L HOU
SEHO
LDS
WITH
OUT E
VENT
S28
.4(1.
03)
2.2(0.
66)
0.6(0.
17)
29.0
(1.13
)2.3
(0.70
)0.6
(0.18
)24
.2(2.
57)
2.0(2.
01)
0.4(0.
41)
TOTA
L HOU
SEHO
LDS
100.0
0.00
0.00.0
027
.70.0
010
0.00.0
00.0
0.00
26.8
0.00
100.0
0.00
0.00.0
032
.70.0
0TO
TAL N
UMBE
R of
HOU
SEHO
LDS
WITH
EVE
NTS
3,237
,279
0.00
00.0
00
0.00
2,759
,411
0.00
00.0
00
0.00
477,8
680.0
00
0.00
00.0
0TO
TAL N
UMBE
R of
HOU
SEHO
LDS
WITH
OUT E
VENT
S1,2
80,93
20.0
00
0.00
00.0
01,1
27,98
60.0
00
0.00
00.0
015
2,946
0.00
00.0
00
0.00
TOTA
L NUM
BER
of H
OUSE
HOLD
S4,5
18,21
10.0
00
0.00
00.0
03,8
87,39
70.0
00
0.00
00.0
063
0,814
0.00
00.0
00
0.00
1/ De
comp
ositio
n ba
sed
on eq
uatio
n [2]
Sour
ce: A
uthors
’ elab
oratio
n bas
ed on
data
from
Encu
esta
Nacio
nal d
e Hog
ares (
INEI
)
D E M O G R A P H I C &
C O M B I N E D
E V E N T S
VII -
Eve
nts no
t clas
sified
Exit
Even
ts
N O N D E M O G R A P H I C
I - Ex
clusiv
ely la
bor
incom
e eve
nts
II - E
xclus
ively
non-
labor
incom
e eve
nts
P(S/
E)Ex
itP(
even
t)P(
S/E)
Table
10.E
. Dec
ompo
sition
of th
e exit
rate
from
pove
rty. S
ensit
ivity
analy
sis to
pov
erty
line (
Incre
ase o
f 10%
in po
very
line)
1/Pe
ru. U
rban
area
s. Pe
riod 2
002-2
006.
Total
hous
ehold
sHo
useh
olds w
ith ch
ildre
nHo
useh
olds w
ithou
t chil
dren
P(ev
ent)
P(S/
E)Ex
itP(
even
t)
120 121
N°(1)
SE(2)
SE(1)
*(2)=(
3)SE
(1)SE
(2)SE
(1)*(2
)=(3)
SE(1)
SE(2)
SE(1)
*(2)=(
3)SE
TOTA
L LAB
OR E
VENT
S13
.95(0.
34)
21.12
(1.07
)2.9
2(0.
16)
18.20
(0.62
)26
.63(1.
64)
4.81
(0.33
)11
.45(0.
40)
15.94
(1.39
)1.8
0(0.
17)
1Re
ducti
on in
the n
umbe
r of e
mploy
ed m
embe
rs3.2
(0.17
)34
.0(2.
74)
1.0(0.
11)
3.2(0.
29)
46.8
(4.65
)1.5
(0.21
)3.1
(0.22
)26
.1(3.
25)
0.8(0.
12)
1.1Re
ducti
on in
the n
umbe
r of r
egist
ered
wag
e ear
ners
0.9(0.
10)
33.9
(5.92
)0.3
(0.07
)1.0
(0.17
)47
.5(8.
56)
0.5(0.
12)
0.8(0.
13)
24.0
(7.84
)0.2
(0.07
)1.2
Redu
ction
in th
e num
ber o
f non
-regis
tered
wag
e ear
ners
1.2(0.
11)
28.2
(4.06
)0.3
(0.06
)1.4
(0.21
)34
.2(6.
80)
0.5(0.
12)
1.1(0.
13)
23.6
(4.86
)0.3
(0.06
)1.3
Redu
ction
in th
e num
ber o
f non
-wag
e ear
ners
1.0(0.
09)
41.1
(4.51
)0.4
(0.06
)0.8
(0.13
)69
.6(7.
18)
0.5(0.
11)
1.2(0.
13)
30.0
(4.90
)0.3
(0.07
)
2Re
ducti
on in
total
hour
ly wa
ge of
mem
bers
emplo
yed i
n both
ob
serva
tions
4.5(0.
21)
13.3
(1.52
)0.6
(0.07
)6.3
(0.39
)19
.1(2.
49)
1.2(0.
17)
3.5(0.
23)
7.1(1.
53)
0.2(0.
05)
3Re
ducti
on in
the n
umbe
r of w
orkin
g hou
rs of
mem
bers
emplo
yed
in bo
th ob
serva
tions
2.3(0.
15)
10.8
(1.65
)0.2
(0.04
)3.1
(0.28
)14
.0(2.
61)
0.4(0.
08)
1.9(0.
16)
7.6(2.
03)
0.1(0.
04)
4Re
ducti
on in
the n
umbe
r of w
orkin
g hou
rs an
d in t
he to
tal ho
urly
wage
of m
embe
rs em
ploye
d in b
oth ob
serva
tions
2.6(0.
15)
19.5
(2.28
)0.5
(0.06
)3.8
(0.30
)24
.1(3.
22)
0.9(0.
14)
1.8(0.
16)
13.9
(3.18
)0.3
(0.06
)
5Re
ducti
on in
the t
otal m
onth
ly wa
ge of
mem
bers
emplo
yed i
n both
ob
serva
tions
and i
n the
numb
er of
emplo
yed m
embe
rs1.4
(0.12
)38
.4(4.
36)
0.5(0.
07)
1.7(0.
21)
45.5
(6.05
)0.8
(0.14
)1.1
(0.15
)32
.1(6.
22)
0.4(0.
09)
TOTA
L NON
-LAB
OR E
VENT
S7.3
(0.24
)12
.7(1.
09)
0.9(0.
08)
3.0(0.
24)
36.7
(3.95
)1.1
(0.15
)9.8
(0.35
)8.5
(1.00
)0.8
(0.10
)6
Redu
ction
in th
e inc
ome f
rom
pens
ions
3.9(0.
18)
8.1(1.
22)
0.3(0.
05)
0.7(0.
11)
29.0
(7.26
)0.2
(0.06
)5.8
(0.27
)6.6
(1.17
)0.4
(0.07
)7
Redu
ction
in pu
blic m
oneta
ry tra
nsfer
s (so
cial p
olicy
) 0.4
(0.06
)36
.5(6.
68)
0.1(0.
03)
0.8(0.
14)
44.4
(8.46
)0.4
(0.09
)0.2
(0.04
)15
.0(6.
14)
0.0(0.
01)
8Re
ducti
on in
othe
r non
-labo
r inco
mes
3.0(0.
16)
15.5
(1.85
)0.5
(0.06
)1.5
(0.17
)36
.0(5.
45)
0.5(0.
10)
3.9(0.
23)
10.9
(1.81
)0.4
(0.07
)
III - L
abor
and n
on-
labor
incom
e eve
nts9
Redu
ction
in la
bor a
nd no
n-lab
or inc
omes
2.2(0.
14)
29.0
(2.74
)0.6
(0.07
)2.7
(0.24
)42
.6(4.
41)
1.1(0.
15)
1.9(0.
17)
17.9
(3.20
)0.3
(0.07
)
IV E
xclus
ively
demo
graph
ic ev
ents
10Gr
owth
in the
total
numb
er of
hous
ehold
mem
bers
; the t
otal
nomi
nal in
come
rema
ins co
nstan
t5.9
(0.21
)5.7
(0.76
)0.3
(0.05
)6.5
(0.38
)5.4
(1.13
)0.3
(0.08
)5.5
(0.26
)5.9
(1.03
)0.3
(0.06
)
V - D
emog
raphic
ev
ents
leadin
g to
incom
e cha
nges
11Re
ducti
on in
the n
umbe
r of la
bor o
r non
-labo
r inc
ome e
arne
rs
due t
o the
entra
nce o
f mem
bers
to the
hous
ehold
3.3(0.
17)
9.5(1.
74)
0.2(0.
04)
2.9(0.
25)
13.6
(3.46
)0.3
(0.08
)3.6
(0.22
)7.2
(1.90
)0.2
(0.04
)
VI - C
ombin
ation
of
demo
graph
ic an
d inc
ome e
vents
12Re
ducti
on in
the t
otal
nomi
nal in
come
and g
rowth
in the
numb
er
of ho
useh
old m
embe
rs.
1.8(0.
12)
47.1
(3.55
)0.8
(0.09
)2.4
(0.25
)61
.4(4.
96)
1.5(0.
20)
1.4(0.
13)
31.9
(4.43
)0.4
(0.07
)
3.1(0.
18)
27.8
(2.80
)0.7
(0.08
)3.3
(0.29
)46
.9(5.
01)
1.2(0.
18)
3.0(0.
22)
16.6
(2.88
)0.4
(0.08
)TO
TAL H
OUSE
HOLD
S W
ITH E
VENT
S37
.5(0.
46)
18.4
(0.61
)6.5
(0.23
)39
.0(0.
77)
27.8
(1.13
)10
.4(0.
47)
36.6
(0.58
)12
.5(0.
67)
4.3(0.
24)
TOTA
L HOU
SEHO
LDS
WITH
OUT E
VENT
S62
.5(0.
46)
0.6(0.
11)
0.4(0.
07)
61.0
(0.77
)1.0
(0.24
)0.6
(0.14
)63
.4(0.
58)
0.3(0.
11)
0.2(0.
07)
TOTA
L HOU
SEHO
LDS
100.0
0.00
0.00.0
06.9
0.00
100.0
0.00
0.00.0
010
.90.0
010
0.00.0
00.0
0.00
4.50.0
0TO
TAL N
UMBE
R of
HOU
SEHO
LDS
WITH
EVE
NTS
2,615
,636
0.00
00.0
00
0.00
1,008
,235
0.00
00.0
00
0.00
1,607
,401
0.00
00.0
00
0.00
TOTA
L NUM
BER
of H
OUSE
HOLD
S W
ITHOU
T EVE
NTS
4,360
,766
0.00
00.0
00
0.00
1,574
,667
0.00
00.0
00
0.00
2,786
,099
0.00
00.0
00
0.00
TOTA
L NUM
BER
of H
OUSE
HOLD
S6,9
76,40
20.0
00
0.00
00.0
02,5
82,90
20.0
00
0.00
00.0
04,3
93,50
00.0
00
0.00
00.0
0
1/ De
comp
ositio
n ba
sed
on eq
uatio
n [2]
Sour
ce: A
uthors
’ elab
oratio
n bas
ed on
data
from
Encu
esta
Perm
anen
te de
Hog
ares (
INDE
C).
D E M O G R A P H I C &
C O M B I N E D
E V E N T S
VII -
Eve
nts no
t clas
sified
Entry
Even
ts
N O N D E M O G R A P H I C
I - Ex
clusiv
ely la
bor
incom
e eve
nts
II - E
xclus
ively
non-
labor
incom
e eve
nts
P(S/
E)En
tryP(
even
t)P(
S/E)
Table
11.A
. Dec
ompo
sition
of th
e ent
ry ra
te fro
m po
verty
. Sen
sitivi
ty an
alysis
to p
over
ty lin
e (Re
ducti
on of
10%
in po
very
line)
1/Ar
gent
ina. U
rban
area
s. Pe
riod 2
003 2
006.
Total
hous
ehold
sHo
useh
olds w
ith ch
ildre
nHo
useh
olds w
ithou
t chil
dren
P(ev
ent)
P(S/
E)En
tryP(
even
t)
122
N°(1)
SE(2)
SE(1)
*(2)=
(3)SE
(1)SE
(2)SE
(1)*(2
)=(3)
SE(1)
SE(2)
SE(1)
*(2)=
(3)SE
TOTA
L lab
or E
VENT
S7.6
0(0.
23)
22.92
(1.29
)1.7
4(0.
11)
9.95
(0.38
)26
.59(1.
74)
2.65
(0.20
)5.6
0(0.
27)
17.39
(1.85
)0.9
7(0.
11)
1Re
ducti
on in
the n
umbe
r of e
mploy
ed m
embe
rs1.6
(0.11
)42
.3(3
.36)
0.7(0
.07)
1.8(0
.17)
41.0
(4.58
)0.8
(0.11
)1.3
(0.13
)43
.9(4
.92)
0.6(0
.09)
1.1Re
ducti
on in
the n
umbe
r of r
egist
ered
wag
e ear
ners
0.8(0
.07)
43.5
(4.77
)0.3
(0.05
)1.0
(0.12
)45
.0(6
.29)
0.4(0
.08)
0.6(0
.09)
41.2
(7.32
)0.2
(0.06
)1.2
Redu
ction
in th
e num
ber o
f non
-regis
tered
wag
e ear
ners
0.5(0
.06)
39.9
(6.50
)0.2
(0.04
)0.6
(0.11
)31
.3(7
.92)
0.2(0
.06)
0.3(0
.07)
54.3
(10.2
7)0.2
(0.05
)1.3
Redu
ction
in th
e num
ber o
f non
-wag
e ear
ners
0.4(0
.05)
42.9
(6.84
)0.2
(0.03
)0.3
(0.06
)48
.8(1
1.53)
0.1(0
.04)
0.4(0
.07)
39.9
(8.47
)0.2
(0.05
)
2Re
ducti
on in
total
hour
ly wa
ge of
mem
bers
emplo
yed i
n both
ob
serva
tions
2.8(0
.14)
13.9
(1.72
)0.4
(0.05
)3.6
(0.24
)19
.1(2
.53)
0.7(0
.10)
2.1(0
.17)
6.4(1
.94)
0.1(0
.04)
3Re
ducti
on in
the n
umbe
r of w
orkin
g hou
rs of
mem
bers
emplo
yed
in bo
th ob
serva
tions
1.0(0
.08)
13.4
(3.05
)0.1
(0.03
)1.3
(0.14
)18
.7(4
.46)
0.2(0
.07)
0.7(0
.09)
4.5(2
.70)
0.0(0
.02)
4Re
ducti
on in
the n
umbe
r of w
orkin
g hou
rs an
d in t
he to
tal ho
urly
wage
of m
embe
rs em
ploye
d in b
oth ob
serva
tions
1.5(0
.11)
17.4
(2.60
)0.3
(0.04
)2.1
(0.19
)18
.5(3
.32)
0.4(0
.08)
1.1(0
.12)
15.5
(4.18
)0.2
(0.05
)
5Re
ducti
on in
the t
otal m
onth
ly wa
ge of
mem
bers
emplo
yed i
n both
ob
serva
tions
and i
n the
numb
er of
emplo
yed m
embe
rs0.7
(0.07
)40
.0(4
.77)
0.3(0
.04)
1.1(0
.13)
52.0
(6.10
)0.6
(0.09
)0.4
(0.07
)13
.7(5
.14)
0.1(0
.02)
TOTA
L NON
-LAB
OR E
VENT
S11
.0(0.
27)
20.6
(1.03
)2.3
(0.13
)6.1
(0.30
)27
.5(2.
25)
1.7(0.
16)
15.2
(0.42
)18
.3(1.
14)
2.8(0.
19)
6Re
ducti
on in
the i
ncom
e fro
m pe
nsion
s 9.2
(0.25
)22
.2(1
.15)
2.0(0
.12)
4.5(0
.26)
30.0
(2.67
)1.3
(0.14
)13
.2(0
.39)
20.0
(1.26
)2.6
(0.18
)8
Redu
ction
in ot
her n
on-la
bour
inco
mes
1.8(0
.12)
12.8
(2.13
)0.2
(0.04
)1.6
(0.16
)20
.9(4
.07)
0.3(0
.08)
2.0(0
.17)
7.2(2
.11)
0.1(0
.04)
III - L
abor
and n
on-
labor
inco
me ev
ents
9Re
ducti
on in
labo
r and
non-
labor
inco
mes
2.8(0
.14)
27.7
(2.24
)0.8
(0.07
)2.4
(0.19
)33
.9(3
.75)
0.8(0
.11)
3.1(0
.20)
23.6
(2.74
)0.7
(0.10
)
IV E
xclus
ively
demo
grap
hic ev
ents
10Gr
owth
in the
total
numb
er of
hous
ehold
mem
bers
; the t
otal
nomi
nal in
come
rema
ins co
nstan
t5.8
(0.20
)7.7
(1.04
)0.4
(0.06
)5.0
(0.28
)8.5
(1.63
)0.4
(0.09
)6.4
(0.28
)7.2
(1.35
)0.5
(0.09
)
V -
Demo
grap
hic
even
ts lea
ding t
o inc
ome c
hang
es11
Redu
ction
in th
e num
ber o
f labo
r or n
on-la
bor i
ncom
e ear
ners
du
e to t
he en
tranc
e of m
embe
rs to
the ho
useh
old9.6
(0.25
)29
.8(1
.23)
2.8(0
.14)
10.6
(0.38
)37
.5(1
.81)
4.0(0
.24)
8.7(0
.33)
21.8
(1.60
)1.9
(0.16
)
VI -
Comb
inatio
n of
demo
grap
hic an
d inc
ome e
vents
12Re
ducti
on in
the t
otal
nomi
nal in
come
and g
rowt
h in t
he nu
mber
of
hous
ehold
mem
bers
.3.3
(0.15
)41
.6(2
.21)
1.4(0
.09)
3.4(0
.22)
54.0
(3.33
)1.8
(0.16
)3.3
(0.20
)30
.7(2
.75)
1.0(0
.10)
12.0
(0.27
)27
.3(1
.04)
3.3(0
.14)
13.7
(0.42
)33
.3(1
.50)
4.6(0
.24)
10.5
(0.35
)20
.5(1
.39)
2.1(0
.16)
TOTA
L HOU
SEHO
LDS
WITH
EVE
NTS
52.1
(0.43
)24
.5(0
.49)
12.7
(0.28
)51
.2(0
.63)
31.1
(0.79
)15
.9(0
.44)
0.0(0
.58)
19.0
(0.61
)10
.0(0
.34)
TOTA
L HOU
SEHO
LDS
WITH
OUT E
VENT
S47
.9(0
.43)
0.9(0
.12)
0.4(0
.06)
48.8
(0.63
)1.7
(0.23
)0.8
(0.11
)47
.2(0
.58)
0.2(0
.08)
0.1(0
.04)
TOTA
L HOU
SEHO
LDS
100.0
0.00
0.00.0
013
.20.0
054
.90.0
00.0
0.00
2.50.0
047
.20.0
00.0
0.00
10.1
0.00
TOTA
L NUM
BER
of H
OUSE
HOLD
S W
ITH E
VENT
S4,4
41,54
50.0
00
0.00
00.0
02,0
06,86
50.0
00
0.00
00.0
00
0.00
00.0
00
0.00
TOTA
L NUM
BER
of H
OUSE
HOLD
S W
ITHOU
T EVE
NTS
4,091
,644
0.00
00.0
00
0.00
1,912
,022
0.00
00.0
00
0.00
2,179
,622
0.00
00.0
00
0.00
TOTA
L NUM
BER
of H
OUSE
HOLD
S8,5
33,18
90.0
00
0.00
00.0
03,9
18,88
70.0
00
0.00
00.0
04,6
14,30
20.0
00
0.00
00.0
0
1/ De
comp
ositio
n ba
sed
on eq
uatio
n [2]
2/ Me
tropo
litan a
reas
of R
ecife
, Salv
ador
, Belo
Hor
izonte
, Rio
de Ja
neiro
, São
Pau
lo an
d Por
to Al
egre
.
P(S/
E)
Table
11.B
. Dec
ompo
sition
of th
e ent
ry ra
te to
pove
rty. S
ensit
ivity
analy
sis to
pov
erty
line (
Redu
ction
of 10
% in
pove
ry lin
e) 1/
Braz
il. Ur
ban a
reas
. Per
iod 20
03-20
06. 2
/
Total
hous
ehold
sHo
useh
olds w
ith ch
ildre
nHo
useh
olds w
ithou
t chil
dren
P(ev
ent)
P(S/
E)En
tryP(
even
t)
D E M O G R A P H I C
& C O M B I N E D
E V E N T S
VII -
Eve
nts no
t clas
sified
Entry
Even
ts
N O N D E M O G R A P H I C
I - E
xclus
ively
labor
inc
ome e
vents
II - E
xclus
ively
non-
labor
inco
me ev
ents
P(S/
E)En
tryP(
even
t)
122 123
N°(1)
SE(2)
SE(1)
*(2)=(
3)SE
(1)SE
(2)SE
(1)*(2
)=(3)
SE(1)
SE(2)
SE(1)
*(2)=(
3)SE
TOTA
L LAB
OR E
VENT
S24
.10(1.
03)
19.17
(1.91
)4.6
2(0.
50)
27.35
(1.42
)22
.21(2.
57)
6.08
(0.77
)20
.12(1.
46)
14.12
(2.72
)2.8
4(0.
58)
1Re
ducti
on in
the n
umbe
r of e
mploy
ed m
embe
rs6.3
(0.59
)32
.3(4.
49)
2.0(0.
34)
5.3(0.
73)
43.2
(6.95
)2.3
(0.49
)7.4
(0.95
)22
.7(5.
48)
1.7(0.
46)
1.1Re
ducti
on in
the n
umbe
r of r
egist
ered
wag
e ear
ners
3.1(0.
42)
31.0
(6.35
)1.0
(0.24
)2.4
(0.50
)42
.8(10
.60)
1.0(0.
34)
4.1(0.
72)
22.6
(7.29
)0.9
(0.33
)1.2
Redu
ction
in th
e num
ber o
f non
-regis
tered
wag
e ear
ners
1.7(0.
32)
28.7
(8.04
)0.5
(0.16
)1.7
(0.42
)33
.9(11
.45)
0.6(0.
23)
1.7(0.
47)
22.4
(10.92
)0.4
(0.21
)1.3
Redu
ction
in th
e num
ber o
f non
-wag
e ear
ners
1.4(0.
29)
39.5
(9.94
)0.6
(0.18
)1.3
(0.36
)56
.6(14
.07)
0.7(0.
27)
1.6(0.
47)
23.2
(12.63
)0.4
(0.24
)
2Re
ducti
on in
total
hour
ly wa
ge of
mem
bers
emplo
yed i
n both
ob
serva
tions
8.3(0.
68)
8.6(2.
28)
0.7(0.
20)
10.7
(1.02
)11
.7(3.
14)
1.2(0.
35)
5.4(0.
83)
1.2(1.
20)
0.1(0.
07)
3Re
ducti
on in
the n
umbe
r of w
orkin
g hou
rs of
mem
bers
emplo
yed
in bo
th ob
serva
tions
2.4(0.
37)
18.4
(5.95
)0.4
(0.16
)3.0
(0.56
)22
.0(7.
76)
0.7(0.
26)
1.7(0.
47)
10.7
(8.05
)0.2
(0.14
)
4Re
ducti
on in
the n
umbe
r of w
orkin
g hou
rs an
d in t
he to
tal ho
urly
wage
of m
embe
rs em
ploye
d in b
oth ob
serva
tions
3.0(0.
41)
13.0
(4.35
)0.4
(0.14
)3.5
(0.59
)16
.3(6.
33)
0.6(0.
24)
2.4(0.
53)
7.2(4.
10)
0.2(0.
10)
5Re
ducti
on in
the t
otal m
onth
ly wa
ge of
mem
bers
emplo
yed i
n both
ob
serva
tions
and i
n the
numb
er of
emplo
yed m
embe
rs4.1
(0.48
)25
.4(5.
14)
1.0(0.
24)
4.9(0.
70)
26.7
(6.42
)1.3
(0.36
)3.2
(0.65
)23
.0(8.
58)
0.7(0.
31)
TOTA
L NON
-LAB
OR E
VENT
S3.5
(0.47
)7.9
(3.43
)0.3
(0.13
)2.2
(0.49
)5.3
(5.18
)0.1
(0.12
)5.2
(0.84
)9.2
(4.43
)0.5
(0.25
)6
Redu
ction
in th
e inc
ome f
rom
pens
ions
1.5(0.
30)
10.1
(5.83
)0.2
(0.09
)0.3
(0.19
)0.0
0.00
0.00.0
02.9
(0.61
)11
.6(6.
59)
0.3(0.
20)
7Re
ducti
on in
publi
c mon
etary
trans
fers (
socia
l poli
cy)
0.1(0.
08)
0.00.0
0.
0.00
0.3(0.
15)
0.00.0
0.
0.00
0.00.0
00.0
0.00
0.00.0
08
Redu
ction
in ot
her n
on-la
bor in
come
s 1.9
(0.36
)6.7
(4.42
)0.1
(0.09
)1.6
(0.43
)7.3
(7.04
)0.1
(0.12
)2.3
(0.61
)6.2
(5.61
)0.1
(0.14
)
III - L
abor
and n
on-
labor
incom
e eve
nts9
Redu
ction
in la
bor a
nd no
n-lab
or inc
omes
3.6(0.
46)
24.0
(5.43
)0.9
(0.23
)2.7
(0.51
)22
.6(8.
24)
0.6(0.
26)
4.6(0.
81)
25.0
(7.20
)1.2
(0.39
)
IV E
xclus
ively
demo
graph
ic ev
ents
10Gr
owth
in the
total
numb
er of
hous
ehold
mem
bers
; the t
otal
nomi
nal in
come
rema
ins co
nstan
t7.1
(0.61
)3.0
(1.32
)0.2
(0.09
)6.2
(0.78
)0.0
0.00
0.00.0
08.2
(0.96
)5.8
(2.51
)0.5
(0.21
)
V - D
emog
raphic
ev
ents
leadin
g to
incom
e cha
nges
11Re
ducti
on in
the n
umbe
r of la
bor o
r non
-labo
r inc
ome e
arne
rs
due t
o the
entra
nce o
f mem
bers
to the
hous
ehold
0.3(0.
12)
0.00.0
0.
0.00
0.1(0.
12)
0.00.0
0.
0.00
0.5(0.
22)
0.00.0
0.
0.00
VI - C
ombin
ation
of
demo
graph
ic an
d inc
ome e
vents
12Re
ducti
on in
the t
otal
nomi
nal in
come
and g
rowth
in the
numb
er
of ho
useh
old m
embe
rs.
2.8(0.
40)
30.2
(6.57
)0.9
(0.22
)2.2
(0.47
)33
.6(10
.45)
0.7(0.
28)
3.6(0.
67)
27.7
(8.36
)1.0
(0.35
)
6.1(0.
58)
42.1
(4.90
)2.6
(0.39
)7.1
(0.85
)44
.8(6.
18)
3.2(0.
58)
4.8(0.
77)
37.3
(7.99
)1.8
(0.49
)TO
TAL H
OUSE
HOLD
S W
ITH E
VENT
S47
.5(1.
20)
19.8
(1.38
)9.4
(0.70
)47
.8(1.
61)
22.4
(1.95
)10
.7(1.
01)
47.1
(1.79
)16
.5(1.
91)
7.8(0.
95)
TOTA
L HOU
SEHO
LDS
WITH
OUT E
VENT
S52
.5(1.
20)
0.6(0.
26)
0.3(0.
14)
52.2
(1.61
)0.9
(0.43
)0.5
(0.23
)52
.9(1.
79)
0.3(0.
27)
0.1(0.
14)
TOTA
L HOU
SEHO
LDS
100.0
0.00
0.00.0
09.7
0.00
100.0
0.00
0.00.0
011
.20.0
010
0.00.0
00.0
0.00
7.90.0
0TO
TAL N
UMBE
R of
HOU
SEHO
LDS
WITH
EVE
NTS
136,6
830.0
00
0.00
00.0
075
,670
0.00
00.0
00
0.00
61,01
30.0
00
0.00
00.0
0TO
TAL N
UMBE
R of
HOU
SEHO
LDS
WITH
OUT E
VENT
S15
1,055
0.00
00.0
00
0.00
82,51
20.0
00
0.00
00.0
068
,543
0.00
00.0
00
0.00
TOTA
L NUM
BER
of H
OUSE
HOLD
S28
7,738
0.00
00.0
00
0.00
158,1
820.0
00
0.00
00.0
012
9,556
0.00
00.0
00
0.00
1/ De
comp
ositio
n ba
sed
on eq
uatio
n [2]
Sour
ce: A
uthors
’ elab
oratio
n bas
ed on
data
from
Encu
esta
de H
ogare
s de P
ropós
itos M
últipl
es (IN
EC).
P(S/
E)
Table
11.C
. Dec
ompo
sition
of th
e ent
ry ra
te to
pove
rty. S
ensit
ivity
analy
sis to
pov
erty
line (
Redu
ction
of 10
% in
pove
ry lin
e) 1/
Costa
Rica
. Urb
an ar
eas.
Perio
d 200
6-200
8.
Total
hous
ehold
sHo
useh
olds w
ith ch
ildre
nHo
useh
olds w
ithou
t chil
dren
P(ev
ent)
P(S/
E)En
tryP(
even
t)
D E M O G R A P H I C &
C O M B I N E D
E V E N T S
VII -
Eve
nts no
t clas
sified
Entry
Even
ts
N O N D E M O G R A P H I C
I - Ex
clusiv
ely la
bor
incom
e eve
nts
II - E
xclus
ively
non-
labor
incom
e eve
nts
P(S/
E)En
tryP(
even
t)
124
N°(1)
SE(2)
SE(1)
*(2)=(
3)SE
(1)SE
(2)SE
(1)*(2
)=(3)
SE(1)
SE(2)
SE(1)
*(2)=(
3)SE
TOTA
L LAB
OR E
VENT
S21
.62(0.
41)
26.30
(0.92
)5.6
9(0.
22)
24.75
(0.55
)30
.68(1.
16)
7.59
(0.33
)16
.89(0.
59)
16.57
(1.38
)2.8
0(0.
25)
1Re
ducti
on in
the n
umbe
r of e
mploy
ed m
embe
rs3.0
(0.17
)33
.1(2.
62)
1.0(0.
09)
3.1(0.
22)
35.7
(3.37
)1.1
(0.13
)2.8
(0.26
)28
.7(4.
14)
0.8(0.
14)
1.1Re
ducti
on in
the n
umbe
r of r
egist
ered
wag
e ear
ners
0.6(0.
08)
22.0
(5.54
)0.1
(0.04
)0.6
(0.11
)29
.1(7.
71)
0.2(0.
06)
0.5(0.
12)
9.4(5.
89)
0.1(0.
03)
1.2Re
ducti
on in
the n
umbe
r of n
on-re
gister
ed w
age e
arne
rs1.0
(0.10
)34
.3(4.
47)
0.3(0.
06)
1.1(0.
13)
35.1
(5.47
)0.4
(0.07
)0.9
(0.15
)32
.7(7.
76)
0.3(0.
08)
1.3Re
ducti
on in
the n
umbe
r of n
on-w
age e
arne
rs1.4
(0.11
)36
.9(3.
93)
0.5(0.
07)
1.4(0.
15)
39.1
(5.13
)0.5
(0.09
)1.4
(0.18
)33
.6(6.
07)
0.5(0.
10)
2Re
ducti
on in
total
hour
ly wa
ge of
mem
bers
emplo
yed i
n both
ob
serva
tions
8.6(0.
28)
21.8
(1.37
)1.9
(0.13
)9.9
(0.38
)26
.4(1.
76)
2.6(0.
20)
6.6(0.
39)
11.2
(1.89
)0.7
(0.13
)
3Re
ducti
on in
the n
umbe
r of w
orkin
g hou
rs of
mem
bers
emplo
yed
in bo
th ob
serva
tions
2.8(0.
16)
19.1
(2.22
)0.5
(0.07
)3.2
(0.22
)20
.9(2.
79)
0.7(0.
10)
2.3(0.
23)
15.1
(3.56
)0.3
(0.09
)
4Re
ducti
on in
the n
umbe
r of w
orkin
g hou
rs an
d in t
he to
tal ho
urly
wage
of m
embe
rs em
ploye
d in b
oth ob
serva
tions
4.6(0.
20)
26.0
(1.92
)1.2
(0.10
)5.2
(0.27
)30
.3(2.
44)
1.6(0.
15)
3.7(0.
30)
17.0
(2.86
)0.6
(0.12
)
5Re
ducti
on in
the t
otal m
onth
ly wa
ge of
mem
bers
emplo
yed i
n both
ob
serva
tions
and i
n the
numb
er of
emplo
yed m
embe
rs2.6
(0.15
)42
.1(3.
02)
1.1(0.
10)
3.3(0.
23)
48.6
(3.47
)1.6
(0.16
)1.4
(0.18
)18
.6(4.
57)
0.3(0.
07)
TOTA
L NON
-LAB
OR E
VENT
S5.0
(0.22
)19
.3(1.
68)
1.0(0.
09)
3.5(0.
23)
21.5
(2.68
)0.8
(0.11
)7.2
(0.41
)17
.7(2.
14)
1.3(0.
17)
6Re
ducti
on in
the i
ncom
e fro
m pe
nsion
s 1.0
(0.10
)14
.4(3.
44)
0.1(0.
04)
0.5(0.
09)
27.2
(7.99
)0.1
(0.05
)1.7
(0.21
)8.8
(3.17
)0.1
(0.06
)7
Redu
ction
in pu
blic m
oneta
ry tra
nsfer
s (so
cial p
olicy
) 0.1
(0.03
)0.0
0.00
.0.0
00.1
(0.04
)0.0
0.00
.0.0
00.1
(0.04
)0.0
0.00
.0.0
08
Redu
ction
in ot
her n
on-la
bor in
come
s 3.9
(0.19
)20
.9(1.
95)
0.8(0.
09)
2.9(0.
21)
21.2
(2.91
)0.6
(0.09
)5.5
(0.36
)20
.7(2.
62)
1.1(0.
16)
III - L
abor
and n
on-
labor
incom
e eve
nts9
Redu
ction
in la
bor a
nd no
n-lab
or inc
omes
6.3(0.
24)
34.2
(1.85
)2.1
(0.14
)6.6
(0.31
)39
.6(2.
37)
2.6(0.
20)
5.7(0.
37)
24.8
(2.85
)1.4
(0.18
)
IV E
xclus
ively
demo
graph
ic ev
ents
10Gr
owth
in the
total
numb
er of
hous
ehold
mem
bers
; the t
otal
nomi
nal in
come
rema
ins co
nstan
t8.5
(0.28
)7.6
(0.90
)0.6
(0.08
)7.0
(0.33
)7.6
(1.27
)0.5
(0.09
)10
.7(0.
50)
7.7(1.
27)
0.8(0.
14)
V - D
emog
raphic
ev
ents
leadin
g to
incom
e cha
nges
11Re
ducti
on in
the n
umbe
r of la
bor o
r non
-labo
r inc
ome e
arne
rs
due t
o the
entra
nce o
f mem
bers
to the
hous
ehold
2.9(0.
17)
23.7
(2.42
)0.7
(0.08
)2.7
(0.21
)30
.3(3.
44)
0.8(0.
11)
3.1(0.
28)
14.9
(3.17
)0.5
(0.11
)
VI - C
ombin
ation
of
demo
graph
ic an
d inc
ome e
vents
12Re
ducti
on in
the t
otal
nomi
nal in
come
and g
rowth
in the
numb
er
of ho
useh
old m
embe
rs.
5.3(0.
23)
41.9
(2.10
)2.2
(0.15
)5.4
(0.29
)45
.3(2.
72)
2.4(0.
20)
5.2(0.
36)
36.4
(3.29
)1.9
(0.21
)
6.6(0.
25)
22.8
(1.57
)1.5
(0.12
)6.9
(0.32
)26
.1(2.
07)
1.8(0.
17)
6.2(0.
39)
17.2
(2.32
)1.1
(0.16
)TO
TAL H
OUSE
HOLD
S W
ITH E
VENT
S56
.2(0.
50)
24.7
(0.57
)13
.9(0.
34)
56.9
(0.64
)29
.1(0.
76)
16.6
(0.47
)55
.1(0.
81)
17.7
(0.81
)9.7
(0.47
)TO
TAL H
OUSE
HOLD
S W
ITHOU
T EVE
NTS
43.8
(0.50
)0.3
(0.09
)0.1
(0.04
)43
.1(0.
64)
0.5(0.
14)
0.2(0.
06)
44.9
(0.81
)0.1
(0.06
)0.0
(0.03
)TO
TAL H
OUSE
HOLD
S10
0.014
.010
0.016
.810
0.09.8
TOTA
L NUM
BER
of H
OUSE
HOLD
S W
ITH E
VENT
S2,4
90,78
01,5
19,68
397
1,097
TOTA
L NUM
BER
of H
OUSE
HOLD
S W
ITHOU
T EVE
NTS
1,944
,316
1,151
,519
792,7
97TO
TAL N
UMBE
R of
HOU
SEHO
LDS
4,435
,096
2,671
,202
1,763
,894
1/ De
comp
ositio
n ba
sed
on eq
uatio
n [2]
Sour
ce: A
uthors
’ elab
oratio
n bas
ed on
data
from
Encu
esta
de E
mpleo
, Des
emple
o y S
ubem
pleo (
INEC
).
P(S/
E)
Table
11.D
. Dec
ompo
sition
of th
e ent
ry ra
te to
pove
rty. S
ensit
ivity
analy
sis to
pov
erty
line (
Redu
ction
of 10
% in
pove
ry lin
e) 1/
Ecua
dor.
Urba
n are
as. P
eriod
2004
-2008
.
Total
hous
ehold
sHo
useh
olds w
ith ch
ildre
nHo
useh
olds w
ithou
t chil
dren
P(ev
ent)
P(S/
E)En
tryP(
even
t)
D E M O G R A P H I C &
C O M B I N E D
E V E N T S
VII -
Eve
nts no
t clas
sified
Entry
Even
ts
N O N D E M O G R A P H I C
I - Ex
clusiv
ely la
bor
incom
e eve
nts
II - E
xclus
ively
non-
labor
incom
e eve
nts
P(S/
E)En
tryP(
even
t)
124 125
N°(1)
SE(2)
SE(1)
*(2)=(
3)SE
(1)SE
(2)SE
(1)*(2
)=(3)
SE(1)
SE(2)
SE(1)
*(2)=(
3)SE
TOTA
L LAB
OR EV
ENTS
18.91
(0.85
)24
.89(2.
10)
4.31
(0.41
)20
.15(1.
09)
31.59
(2.72
)5.7
0(0.
58)
16.73
(1.35
)10
.73(2.
43)
1.71
(0.40
)1
Redu
ction
in th
e num
ber o
f emp
loyed
mem
bers
4.0(0.
41)
42.7
(5.28
)1.5
(0.24
)3.8
(0.49
)53
.7(6.
65)
1.8(0.
33)
4.2(0.
75)
25.1
(7.08
)1.0
(0.30
)1.1
Redu
ction
in th
e num
ber o
f reg
istere
d wag
e earn
ers1.2
(0.23
)40
.1(9.
29)
0.4(0.
12)
1.1(0.
27)
60.4
(12.49
)0.6
(0.18
)1.4
(0.41
)12
.3(8.
54)
0.2(0.
12)
1.2Re
ducti
on in
the n
umbe
r of n
on-re
gister
ed w
age e
arners
1.2(0.
23)
48.9
(9.36
)0.6
(0.14
)1.2
(0.27
)55
.5(11
.53)
0.6(0.
19)
1.4(0.
42)
38.9
(14.33
)0.5
(0.21
)1.3
Redu
ction
in th
e num
ber o
f non
-wag
e earn
ers1.5
(0.26
)39
.8(8.
36)
0.6(0.
15)
1.6(0.
33)
47.9
(10.61
)0.7
(0.21
)1.5
(0.42
)24
.3(11
.96)
0.3(0.
19)
2Re
ducti
on in
total
hour
ly wa
ge of
mem
bers
emplo
yed i
n both
ob
serva
tions
5.7(0.
52)
12.7
(2.83
)0.7
(0.16
)6.1
(0.65
)16
.8(3.
91)
0.9(0.
23)
5.1(0.
84)
4.3(2.
56)
0.2(0.
13)
3Re
ducti
on in
the n
umbe
r of w
orkin
g hou
rs of
mem
bers
emplo
yed
in bo
th ob
serva
tions
1.9(0.
28)
16.3
(5.48
)0.3
(0.10
)2.0
(0.35
)24
.5(7.
82)
0.4(0.
16)
1.8(0.
45)
0.00.0
00.0
0.00
4Re
ducti
on in
the n
umbe
r of w
orkin
g hou
rs an
d in t
he to
tal ho
urly
wage
of m
embe
rs em
ploye
d in b
oth ob
serva
tions
3.7(0.
40)
21.9
(4.90
)0.8
(0.19
)4.3
(0.54
)27
.1(6.
25)
1.0(0.
29)
2.8(0.
55)
7.7(4.
75)
0.2(0.
13)
5Re
ducti
on in
the t
otal m
onth
ly wa
ge of
mem
bers
emplo
yed i
n both
ob
serva
tions
and i
n the
numb
er of
emplo
yed m
embe
rs3.6
(0.39
)32
.4(5.
07)
1.1(0.
20)
4.0(0.
52)
41.4
(6.49
)1.5
(0.29
)2.9
(0.56
)10
.7(6.
12)
0.3(0.
18)
TOTA
L NON
-LAB
OR EV
ENTS
4.8(0.
49)
9.0(2.
80)
0.4(0.
13)
4.1(0.
56)
10.2
(3.89
)0.4
(0.15
)5.9
(0.86
)7.6
(3.97
)0.4
(0.23
)6
Redu
ction
in th
e inc
ome f
rom
pens
ions
0.8(0.
18)
5.4(3.
93)
0.0(0.
03)
0.4(0.
16)
10.3
(10.12
)0.0
(0.04
)1.4
(0.40
)2.9
(2.96
)0.0
(0.04
)7
Redu
ction
in pu
blic m
oneta
ry tra
nsfer
s (so
cial p
olicy
) 0.0
0.00
0.00.0
00.0
0.00
0.00.0
00.0
0.00
0.00.0
00.0
(0.76
)0.0
0.00
0.00.0
08
Redu
ction
in ot
her n
on-la
bor in
come
s 4.0
(0.44
)9.7
(3.21
)0.4
(0.12
)3.7
(0.51
)10
.1(4.
13)
0.3(0.
15)
4.5(0.
76)
9.1(5.
06)
0.4(0.
23)
III - L
abor
and n
on-
labor
incom
e eve
nts9
Redu
ction
in la
bor a
nd no
n-lab
or inc
omes
11.0
(0.68
)34
.4(3.
13)
3.5(0.
37)
12.1
(0.86
)38
.8(3.
80)
4.2(0.
51)
9.1(1.
07)
24.0
(5.07
)2.1
(0.49
)
IV E
xclus
ively
demo
graph
ic ev
ents
10Gr
owth
in the
total
numb
er of
hous
ehold
mem
bers
; the t
otal
nomi
nal in
come
rema
ins co
nstan
t9.4
(0.61
)4.7
(1.39
)0.4
(0.12
)8.1
(0.72
)3.4
(1.65
)0.2
(0.12
)11
.6(1.
15)
6.3(2.
33)
0.7(0.
27)
V - D
emog
raphic
ev
ents
leadin
g to
incom
e cha
nges
11Re
ducti
on in
the n
umbe
r of la
bor o
r non
-labo
r inco
me ea
rners
due
to the
entra
nce o
f mem
bers
to the
hous
ehold
2.4(0.
31)
19.4
(5.19
)0.4
(0.13
)2.2
(0.38
)28
.2(7.
68)
0.6(0.
19)
2.7(0.
55)
6.9(4.
48)
0.2(0.
12)
VI - C
ombin
ation
of
demo
graph
ic an
d inc
ome e
vents
12Re
ducti
on in
the t
otal
nomi
nal in
come
and g
rowth
in the
numb
er of
hous
ehold
mem
bers.
11.9
(0.70
)42
.4(3.
01)
4.6(0.
44)
12.7
(0.89
)45
.5(3.
65)
5.2(0.
57)
10.5
(1.14
)35
.9(5.
11)
3.6(0.
66)
8.2(0.
58)
36.5
(3.52
)2.7
(0.33
)8.6
(0.73
)42
.5(4.
22)
3.3(0.
44)
7.3(0.
99)
24.0
(5.97
)1.7
(0.48
)TO
TAL H
OUSE
HOLD
S WITH
EVE
NTS
66.5
(0.91
)26
.8(1.
15)
16.4
(0.76
)68
.0(1.
18)
32.1
(1.48
)19
.6(1.
00)
63.8
(1.43
)17
.0(1.
64)
10.4
(1.06
)TO
TAL H
OUSE
HOLD
S WITH
OUT E
VENT
S33
.5(0.
91)
0.6(0.
29)
0.2(0.
09)
32.0
(1.18
)0.3
(0.26
)0.1
(0.07
)36
.2(1.
43)
1.2(0.
63)
0.4(0.
22)
TOTA
L HOU
SEHO
LDS
100.0
0.00
0.00.0
016
.60.0
010
0.00.0
00.0
0.00
19.6
0.00
100.0
0.00
0.00.0
010
.80.0
0TO
TAL N
UMBE
R of
HOU
SEHO
LDS W
ITH E
VENT
S4,2
91,02
00.0
00
0.00
00.0
02,7
95,66
60.0
00
0.00
00.0
01,4
95,35
40.0
00
0.00
00.0
0TO
TAL N
UMBE
R of
HOU
SEHO
LDS W
ITHOU
T EVE
NTS
2,161
,061
0.00
00.0
00
0.00
1,313
,840
0.00
00.0
00
0.00
847,2
210.0
00
0.00
00.0
0TO
TAL N
UMBE
R of
HOU
SEHO
LDS
6,452
,081
0.00
00.0
00
0.00
4,109
,506
0.00
00.0
00
0.00
2,342
,575
0.00
00.0
00
0.00
1/ De
comp
ositio
n ba
sed
on eq
uatio
n [2]
Sour
ce: A
uthors
’ elab
oratio
n bas
ed on
data
from
Encu
esta
Nacio
nal d
e Hog
ares (
ENEI
)
D E M O G R A P H I C &
C O M B I N E D
E V E N T S
VII -
Eve
nts no
t clas
sified
Entry
Even
ts
N O N D E M O G R A P H I C
I - Ex
clusiv
ely la
bor
incom
e eve
nts
II - E
xclus
ively
non-
labor
incom
e eve
nts
P(S/E
)En
tryP(
even
t)P(
S/E)
Table
11.E.
Dec
ompo
sition
of th
e ent
ry rat
e fro
m po
verty
. Sen
sitivi
ty an
alysis
to p
overt
y line
(Red
uctio
n of 1
0% in
pove
ry lin
e) 1/
Peru
. Urb
an ar
eas.
Perio
d 200
2-200
6.
Total
hous
ehold
sHo
useh
olds w
ith ch
ildren
Hous
ehold
s with
out c
hildr
enP(
even
t)P(
S/E)
Entry
P(ev
ent)
126
N°(1
)SE
(2)
SE(1
)*(2)
=(3)
SE(1
)SE
(2)
SE(1
)*(2)
=(3)
SE(1
)SE
(2)
SE(1
)*(2)
=(3)
SETO
TAL
LABO
R EV
ENTS
14.31
(0.37
)25
.44(1
.21)
3.60
(0.19
)19
.18(0
.70)
31.44
(1.86
)5.9
8(0
.41)
11.78
(0.42
)20
.33(1
.57)
2.36
(0.20
)1
Redu
ction
in th
e num
ber o
f em
ploy
ed m
embe
rs3.2
(0.19
)39
.0(2
.98)
1.2(0
.12)
3.2(0
.32)
50.0
(5.24
)1.5
(0.23
)3.2
(0.23
)33
.3(3
.56)
1.0(0
.14)
1.1Re
ducti
on in
the n
umbe
r of r
egist
ered
wag
e ear
ners
1.0(0
.11)
42.7
(6.09
)0.4
(0.08
)1.1
(0.19
)60
.6(8
.32)
0.6(0
.15)
0.9(0
.14)
31.0
(8.02
)0.3
(0.08
)1.2
Redu
ction
in th
e num
ber o
f non
-regi
ster
ed w
age e
arne
rs1.2
(0.12
)35
.1(4
.65)
0.4(0
.07)
1.4(0
.23)
38.4
(7.81
)0.5
(0.14
)1.1
(0.13
)32
.8(5
.75)
0.4(0
.08)
1.3Re
ducti
on in
the n
umbe
r of n
on-w
age e
arne
rs1.0
(0.10
)40
.5(4
.81)
0.4(0
.06)
0.6(0
.13)
58.7
(9.64
)0.4
(0.10
)1.2
(0.13
)35
.4(5
.28)
0.4(0
.07)
2Re
ducti
on in
total
hou
rly w
age o
f mem
bers
emplo
yed i
n both
ob
serva
tions
4.7(0
.22)
17.1
(1.77
)0.8
(0.09
)6.7
(0.44
)22
.4(2
.73)
1.5(0
.20)
3.6(0
.24)
12.0
(2.25
)0.4
(0.09
)
3Re
ducti
on in
the n
umbe
r of w
orkin
g ho
urs o
f mem
bers
emplo
yed
in bo
th ob
serva
tions
2.4(0
.16)
15.9
(2.22
)0.4
(0.06
)3.4
(0.33
)22
.7(3
.78)
0.8(0
.14)
1.9(0
.17)
9.5(2
.34)
0.2(0
.05)
4Re
ducti
on in
the n
umbe
r of w
orkin
g ho
urs a
nd in
the t
otal h
ourly
wa
ge of
mem
bers
emplo
yed i
n both
obse
rvatio
ns2.7
(0.16
)25
.7(2
.67)
0.7(0
.08)
4.2(0
.35)
32.9
(3.89
)1.4
(0.19
)1.9
(0.17
)17
.6(3
.55)
0.3(0
.07)
5Re
ducti
on in
the t
otal m
onth
ly wa
ge of
mem
bers
emplo
yed i
n both
ob
serva
tions
and i
n the
num
ber o
f em
ploy
ed m
embe
rs1.4
(0.13
)40
.0(4
.73)
0.5(0
.08)
1.7(0
.23)
47.7
(6.85
)0.8
(0.16
)1.2
(0.16
)34
.1(6
.42)
0.4(0
.09)
TOTA
L NO
N-LA
BOR
EVEN
TS7.6
(0.26
)15
.7(1
.23)
1.2(0
.10)
2.8(0
.25)
34.1
(4.30
)0.9
(0.14
)10
.1(0
.36)
13.1
(1.23
)1.3
(0.13
)6
Redu
ction
in th
e inc
ome f
rom
pen
sions
4.1
(0.19
)11
.3(1
.37)
0.5(0
.06)
0.6(0
.11)
32.9
(7.82
)0.2
(0.06
)5.9
(0.28
)10
.1(1
.36)
0.6(0
.08)
7Re
ducti
on in
pub
lic m
onet
ary t
rans
fers
(soc
ial po
licy)
0.3(0
.05)
34.1
(7.21
)0.1
(0.03
)0.7
(0.13
)41
.9(9
.85)
0.3(0
.08)
0.2(0
.04)
18.9
(7.13
)0.0
(0.01
)8
Redu
ction
in ot
her n
on-la
bor in
com
es
3.1(0
.17)
19.6
(2.18
)0.6
(0.07
)1.5
(0.19
)31
.2(5
.83)
0.4(0
.10)
4.0(0
.25)
17.4
(2.32
)0.7
(0.10
)
III - L
abor
and n
on-
labor
inco
me e
vents
9Re
ducti
on in
labo
r and
non-
labor
inco
mes
2.3(0
.15)
34.1
(3.01
)0.8
(0.08
)2.8
(0.27
)42
.4(4
.59)
1.2(0
.15)
2.0(0
.18)
28.0
(4.00
)0.6
(0.09
)
IV E
xclus
ively
dem
ogra
phic
even
ts10
Grow
th in
the to
tal n
umbe
r of h
ouse
hold
mem
bers
; the t
otal
nom
inal in
com
e rem
ains c
onsta
nt5.6
(0.22
)10
.3(1
.24)
0.6(0
.07)
6.1(0
.41)
10.9
(2.35
)0.7
(0.15
)5.3
(0.26
)10
.0(1
.39)
0.5(0
.08)
V -
Dem
ogra
phic
even
ts lea
ding t
o inc
ome c
hang
es11
Redu
ction
in th
e num
ber o
f lab
or o
r non
-labo
r inc
ome e
arne
rs
due t
o the
entra
nce o
f mem
bers
to th
e hou
seho
ld3.3
(0.18
)10
.0(1
.86)
0.2(0
.04)
2.9(0
.29)
17.1
(4.28
)0.4
(0.10
)3.5
(0.23
)6.5
(1.78
)0.1
(0.04
)
VI -
Com
binati
on of
de
mog
raph
ic an
d inc
ome e
vents
12Re
ducti
on in
the t
otal
nom
inal
inco
me a
nd gr
owth
in the
num
ber
of h
ouse
hold
mem
bers
.1.7
(0.13
)50
.8(3
.79)
0.9(0
.09)
2.3(0
.26)
66.6
(5.52
)1.5
(0.21
)1.4
(0.14
)36
.9(4
.76)
0.5(0
.08)
3.1(0
.18)
32.6
(3.10
)0.8
(0.09
)3.2
(0.31
)56
.6(5
.43)
1.4(0
.21)
3.0(0
.23)
20.5
(3.19
)0.5
(0.09
)TO
TAL
HOUS
EHOL
DS W
ITH
EVE
NTS
37.9
(0.49
)22
.4(0
.69)
8.0(0
.27)
39.3
(0.85
)32
.2(1
.30)
12.1
(0.55
)37
.1(0
.60)
16.9
(0.78
)5.9
(0.29
)TO
TAL
HOUS
EHOL
DS W
ITHO
UT E
VENT
S62
.1(0
.49)
0.8(0
.19)
0.5(0
.12)
60.7
(0.85
)1.8
(0.54
)1.1
(0.32
)62
.9(0
.60)
0.3(0
.10)
0.2(0
.06)
TOTA
L HO
USEH
OLDS
100.0
0.00
0.00.0
08.5
0.00
100.0
0.00
0.00.0
013
.10.0
010
0.00.0
00.0
0.00
6.10.0
0TO
TAL
NUMB
ER o
f HOU
SEHO
LDS
WIT
H E
VENT
S2,3
82,46
30.0
00
0.00
00.0
084
6,569
0.00
00.0
00
0.00
1,535
,894
0.00
00.0
00
0.00
TOTA
L NU
MBER
of H
OUSE
HOLD
S W
ITHO
UT E
VENT
S3,9
11,38
60.0
00
0.00
00.0
01,3
08,16
30.0
00
0.00
00.0
02,6
03,22
30.0
00
0.00
00.0
0TO
TAL
NUMB
ER o
f HOU
SEHO
LDS
6,293
,849
0.00
00.0
00
0.00
2,154
,732
0.00
00.0
00
0.00
4,139
,117
0.00
00.0
00
0.00
1/ De
com
posit
ion b
ased
on e
quati
on [2
]So
urce
: Auth
ors’
elabo
ratio
n bas
ed on
data
from
Enc
uesta
Per
man
ente
de H
ogar
es (I
NDEC
).
P(S/
E)
Tabl
e 12.A
. Dec
ompo
sitio
n of
the e
ntry
rate
from
pov
erty
. Sen
sitivi
ty an
alysis
to p
over
ty lin
e (In
crea
se o
f 10%
in p
over
y lin
e) 1/
Arge
ntin
a. Ur
ban
area
s. Pe
riod
2003
2006
.
Tota
l hou
seho
lds
Hous
ehol
ds w
ith ch
ildre
nHo
useh
olds
with
out c
hild
ren
P(ev
ent)
P(S/
E)En
tryP(
even
t)
D E M O G R A P H I C &
C O M B I N E D
E V E N T S
VII -
Eve
nts no
t clas
sified
Entry
Even
ts
N O N D E M O G R A P H I C
I - E
xclus
ively
labor
inc
ome e
vents
II - E
xclus
ively
non-
labor
inco
me e
vents
P(S/
E)En
tryP(
even
t)
126 127
N°(1)
SE(2)
SE(1)
*(2)=(
3)SE
(1)SE
(2)SE
(1)*(2
)=(3)
SE(1)
SE(2)
SE(1)
*(2)=(
3)SE
TOTA
L lab
or EV
ENTS
7.75
(0.24
)23
.74(1.
36)
1.84
(0.12
)10
.41(0.
42)
25.87
(1.84
)2.6
9(0.
22)
5.74
(0.28
)20
.82(1.
99)
1.19
(0.13
)1
Redu
ction
in th
e num
ber o
f emp
loyed
mem
bers
1.5(0.
11)
46.7
(3.62
)0.7
(0.08
)1.8
(0.19
)42
.4(5.
08)
0.8(0.
12)
1.3(0.
14)
51.0
(5.08
)0.7
(0.10
)1.1
Redu
ction
in th
e num
ber o
f regis
tered
wag
e earn
ers0.8
(0.08
)47
.8(5.
05)
0.4(0.
05)
1.0(0.
14)
42.8
(6.77
)0.4
(0.09
)0.6
(0.09
)53
.9(7.
47)
0.3(0.
07)
1.2Re
ducti
on in
the n
umbe
r of n
on-re
gister
ed w
age e
arners
0.4(0.
06)
43.4
(7.28
)0.2
(0.04
)0.6
(0.11
)33
.6(9.
24)
0.2(0.
06)
0.3(0.
07)
57.6
(10.79
)0.2
(0.05
)1.3
Redu
ction
in th
e num
ber o
f non
-wag
e earn
ers0.3
(0.05
)47
.9(7.
32)
0.2(0.
04)
0.2(0.
06)
61.2
(12.31
)0.2
(0.05
)0.4
(0.08
)42
.0(8.
96)
0.2(0.
05)
2Re
ducti
on in
total
hour
ly wa
ge of
mem
bers
emplo
yed i
n both
ob
serva
tions
2.9(0.
15)
13.6
(1.75
)0.4
(0.05
)3.8
(0.26
)16
.0(2.
48)
0.6(0.
10)
2.2(0.
18)
10.4
(2.37
)0.2
(0.06
)
3Re
ducti
on in
the n
umbe
r of w
orkin
g hou
rs of
memb
ers em
ploye
d in
both
obse
rvatio
ns1.0
(0.09
)15
.8(3.
44)
0.2(0.
04)
1.4(0.
16)
23.2
(5.13
)0.3
(0.08
)0.7
(0.10
)3.9
(2.63
)0.0
(0.02
)
4Re
ducti
on in
the n
umbe
r of w
orkin
g hou
rs an
d in t
he to
tal ho
urly
wage
of m
embe
rs em
ploye
d in b
oth ob
serva
tions
1.6(0.
12)
18.5
(2.69
)0.3
(0.05
)2.3
(0.21
)19
.1(3.
44)
0.4(0.
09)
1.1(0.
13)
17.5
(4.29
)0.2
(0.05
)
5Re
ducti
on in
the t
otal m
onthl
y wag
e of m
embe
rs em
ploye
d in b
oth
obse
rvatio
ns an
d in t
he nu
mber
of em
ploye
d mem
bers
0.7(0.
08)
38.4
(5.04
)0.3
(0.05
)1.1
(0.14
)51
.2(6.
76)
0.6(0.
10)
0.4(0.
08)
15.1
(5.30
)0.1
(0.03
)
TOTA
L NON
-LABO
R EV
ENTS
11.9
(0.29
)24
.9(1.
12)
3.0(0.
15)
6.7(0.
35)
32.8
(2.48
)2.2
(0.20
)15
.9(0.
44)
22.4
(1.24
)3.6
(0.22
)6
Redu
ction
in th
e inc
ome f
rom
pens
ions
10.0
(0.27
)26
.7(1.
25)
2.7(0.
14)
4.9(0.
30)
35.3
(2.93
)1.7
(0.17
)13
.8(0.
41)
24.4
(1.37
)3.4
(0.21
)8
Redu
ction
in oth
er no
n-lab
our in
come
s 2.0
(0.13
)15
.7(2.
35)
0.3(0.
05)
1.8(0.
18)
26.0
(4.59
)0.5
(0.09
)2.1
(0.18
)9.2
(2.33
)0.2
(0.05
)
III - L
abor
and n
on-
labor
incom
e eve
nts9
Redu
ction
in la
bor a
nd no
n-lab
or inc
omes
3.0(0.
15)
30.4
(2.34
)0.9
(0.08
)2.7
(0.22
)35
.3(3.
91)
1.0(0.
13)
3.2(0.
21)
27.2
(2.91
)0.9
(0.11
)
IV Ex
clusiv
ely
demo
graph
ic ev
ents
10Gr
owth
in the
total
numb
er of
hous
ehold
mem
bers;
the t
otal
nomi
nal in
come
rema
ins co
nstan
t5.5
(0.21
)10
.0(1.
26)
0.5(0.
07)
4.5(0.
28)
10.2
(2.12
)0.5
(0.10
)6.2
(0.29
)9.9
(1.56
)0.6
(0.10
)
V - D
emog
raphic
ev
ents
leadin
g to
incom
e cha
nges
11Re
ducti
on in
the n
umbe
r of la
bor o
r non
-labo
r inco
me ea
rners
du
e to t
he en
tranc
e of m
embe
rs to
the ho
useh
old9.6
(0.26
)34
.5(1.
36)
3.3(0.
16)
10.8
(0.41
)43
.7(2.
01)
4.7(0.
28)
8.7(0.
34)
25.9
(1.76
)2.3
(0.18
)
VI - C
ombin
ation
of
demo
graph
ic an
d inc
ome e
vents
12Re
ducti
on in
the t
otal n
omina
l inco
me an
d grow
th in
the nu
mber
of ho
useh
old m
embe
rs.3.4
(0.16
)47
.6(2.
34)
1.6(0.
11)
3.5(0.
25)
58.9
(3.50
)2.1
(0.19
)3.3
(0.21
)38
.6(3.
01)
1.3(0.
12)
12.5
(0.29
)32
.6(1.
15)
4.1(0.
17)
14.7
(0.47
)40
.1(1.
69)
5.9(0.
31)
10.8
(0.37
)24
.9(1.
51)
2.7(0.
18)
TOTA
L HOU
SEHO
LDS W
ITH E
VENT
S53
.6(0.
45)
28.5
(0.54
)15
.3(0.
32)
53.4
(0.68
)35
.6(0.
88)
19.0
(0.53
)0.0
(0.60
)23
.2(0.
68)
12.5
(0.39
)TO
TAL H
OUSE
HOLD
S WITH
OUT E
VENT
S46
.4(0.
45)
1.0(0.
13)
0.5(0.
06)
46.6
(0.68
)1.7
(0.25
)0.8
(0.12
)46
.2(0.
60)
0.5(0.
12)
0.2(0.
05)
TOTA
L HOU
SEHO
LDS
100.0
0.00
0.00.0
015
.70.0
053
.30.0
00.0
0.00
3.00.0
046
.20.0
00.0
0.00
12.7
0.00
TOTA
L NUM
BER
of HO
USEH
OLDS
WITH
EVE
NTS
4,120
,010
0.00
00.0
00
0.00
1,763
,740
0.00
00.0
00
0.00
00.0
00
0.00
00.0
0TO
TAL N
UMBE
R of
HOUS
EHOL
DS W
ITHOU
T EVE
NTS
3,561
,422
0.00
00.0
00
0.00
1,541
,997
0.00
00.0
00
0.00
2,019
,425
0.00
00.0
00
0.00
TOTA
L NUM
BER
of HO
USEH
OLDS
7,681
,432
0.00
00.0
00
0.00
3,305
,737
0.00
00.0
00
0.00
4,375
,695
0.00
00.0
00
0.00
1/ De
comp
ositio
n ba
sed
on eq
uatio
n [2]
2/ Me
tropo
litan a
reas o
f Rec
ife, S
alvad
or, Be
lo Ho
rizon
te, R
io de
Jane
iro, S
ão Pa
ulo an
d Port
o Aleg
re.
D E M O G R A P H I C
& C O M B I N E D
E V E N T S
VII -
Even
ts no
t clas
sified
Entry
Even
ts
N O N D E M O G R A P H I C
I - Ex
clusiv
ely la
bor
incom
e eve
nts
II - Ex
clusiv
ely no
n-lab
or inc
ome e
vents
P(S/E
)En
tryP(
even
t)P(
S/E)
Table
12.B
. Dec
ompo
sition
of th
e entr
y rate
to po
verty
. Sen
sitivi
ty an
alysis
to p
overt
y line
(Incre
ase o
f 10%
in po
very
line)
1/Br
azil.
Urba
n area
s. Pe
riod 2
003-2
006.
2/
Total
hous
ehold
sHo
useh
olds w
ith ch
ildren
Hous
ehold
s with
out c
hildr
enP(
even
t)P(
S/E)
Entry
P(ev
ent)
128
N°(1)
SE(2)
SE(1)
*(2)=(
3)SE
(1)SE
(2)SE
(1)*(2
)=(3)
SE(1)
SE(2)
SE(1)
*(2)=(
3)SE
TOTA
L LAB
OR EV
ENTS
24.83
(1.09
)24
.58(2.
19)
6.10
(0.61
)28
.30(1.
54)
30.50
(3.03
)8.6
3(0.
98)
20.88
(1.52
)15
.42(2.
86)
3.22
(0.64
)1
Redu
ction
in th
e num
ber o
f emp
loyed
mem
bers
6.4(0.
63)
34.6
(4.81
)2.2
(0.38
)5.3
(0.78
)50
.5(7.
54)
2.7(0.
57)
7.6(0.
99)
22.0
(5.59
)1.7
(0.48
)1.1
Redu
ction
in th
e num
ber o
f regis
tered
wag
e earn
ers3.3
(0.45
)40
.6(6.
93)
1.3(0.
29)
2.5(0.
55)
65.8
(10.26
)1.6
(0.45
)4.2
(0.75
)23
.4(7.
51)
1.0(0.
35)
1.2Re
ducti
on in
the n
umbe
r of n
on-re
gister
ed w
age e
arners
1.7(0.
33)
24.5
(8.12
)0.4
(0.16
)1.6
(0.44
)28
.8(11
.90)
0.5(0.
22)
1.8(0.
50)
20.1
(10.96
)0.4
(0.22
)1.3
Redu
ction
in th
e num
ber o
f non
-wag
e earn
ers1.4
(0.31
)33
.1(10
.38)
0.5(0.
18)
1.2(0.
38)
48.3
(15.76
)0.6
(0.27
)1.7
(0.49
)20
.5(12
.83)
0.3(0.
24)
2Re
ducti
on in
total
hour
ly wa
ge of
mem
bers
emplo
yed i
n both
ob
serva
tions
8.9(0.
73)
17.1
(3.21
)1.5
(0.31
)11
.7(1.
13)
23.9
(4.38
)2.8
(0.58
)5.7
(0.88
)1.2
(1.20
)0.1
(0.07
)
3Re
ducti
on in
the n
umbe
r of w
orkin
g hou
rs of
memb
ers em
ploye
d in
both
obse
rvatio
ns2.2
(0.38
)10
.8(4.
74)
0.2(0.
11)
2.6(0.
55)
9.1(5.
57)
0.2(0.
15)
1.8(0.
50)
13.7
(8.55
)0.2
(0.16
)
4Re
ducti
on in
the n
umbe
r of w
orkin
g hou
rs an
d in t
he to
tal ho
urly
wage
of m
embe
rs em
ploye
d in b
oth ob
serva
tions
3.0(0.
43)
15.7
(4.99
)0.5
(0.16
)3.5
(0.64
)19
.5(7.
39)
0.7(0.
28)
2.5(0.
55)
9.5(4.
82)
0.2(0.
12)
5Re
ducti
on in
the t
otal m
onthl
y wag
e of m
embe
rs em
ploye
d in b
oth
obse
rvatio
ns an
d in t
he nu
mber
of em
ploye
d mem
bers
4.3(0.
52)
38.6
(5.98
)1.7
(0.32
)5.2
(0.77
)43
.0(7.
55)
2.2(0.
51)
3.3(0.
67)
30.5
(9.53
)1.0
(0.37
)
TOTA
L NON
-LABO
R EV
ENTS
3.9(0.
51)
16.9
(4.87
)0.7
(0.21
)2.4
(0.55
)5.5
(5.39
)0.1
(0.13
)5.5
(0.89
)22
.5(6.
54)
1.2(0.
42)
6Re
ducti
on in
the i
ncom
e fro
m pe
nsion
s 1.7
(0.33
)26
.6(8.
78)
0.4(0.
17)
0.4(0.
22)
0.00.0
00.0
0.00
3.1(0.
64)
30.3
(9.73
)0.9
(0.36
)7
Redu
ction
in pu
blic m
oneta
ry tra
nsfer
s (so
cial p
olicy
) 0.1
(0.08
)0.0
0.00
.0.0
00.2
(0.14
)0.0
0.00
.0.0
00.0
0.00
0.00.0
00.0
0.00
8Re
ducti
on in
othe
r non
-labo
r inco
mes
2.1(0.
40)
10.0
(5.32
)0.2
(0.12
)1.8
(0.48
)7.3
(7.04
)0.1
(0.13
)2.4
(0.64
)12
.3(7.
68)
0.3(0.
21)
III - L
abor
and n
on-
labor
incom
e eve
nts.
(Incre
ase o
f 10%
in
pove
ry lin
e) 1/
9Re
ducti
on in
labo
r and
non-l
abor
incom
es3.7
(0.49
)25
.9(5.
72)
1.0(0.
25)
2.6(0.
53)
17.3
(7.92
)0.5
(0.23
)4.9
(0.86
)31
.1(7.
74)
1.5(0.
47)
IV E
xclus
ively
demo
graph
ic ev
ents
10Gr
owth
in the
total
numb
er of
hous
ehold
mem
bers;
the t
otal
nomi
nal in
come
rema
ins co
nstan
t6.9
(0.63
)5.1
(1.93
)0.3
(0.14
)5.8
(0.81
)0.0
0.00
0.00.0
08.1
(0.99
)9.2
(3.44
)0.7
(0.29
)
V - D
emog
raphic
ev
ents
leadin
g to
incom
e cha
nges
11Re
ducti
on in
the n
umbe
r of la
bor o
r non
-labo
r inco
me ea
rners
du
e to t
he en
tranc
e of m
embe
rs to
the ho
useh
old0.3
(0.13
)18
.5(13
.04)
0.1(0.
04)
0.1(0.
13)
0.00.0
00.0
0.00
0.5(0.
24)
23.9
(16.56
)0.1
(0.09
)
VI - C
ombin
ation
of
demo
graph
ic an
d inc
ome e
vents
12Re
ducti
on in
the t
otal n
omina
l inco
me an
d grow
th in
the nu
mber
of ho
useh
old m
embe
rs.3.0
(0.43
)35
.4(6.
90)
1.1(0.
25)
2.3(0.
52)
40.2
(10.97
)0.9
(0.33
)3.7
(0.70
)32
.0(8.
76)
1.2(0.
38)
6.3(0.
63)
48.1
(5.13
)3.0
(0.44
)7.7
(0.94
)51
.6(6.
44)
4.0(0.
69)
4.8(0.
79)
41.5
(8.43
)2.0
(0.52
)TO
TAL H
OUSE
HOLD
S WITH
EVE
NTS
48.9
(1.27
)25
.0(1.
56)
12.2
(0.83
)49
.2(1.
74)
28.7
(2.24
)14
.1(1.
22)
48.4
(1.85
)20
.7(2.
11)
10.0
(1.10
)TO
TAL H
OUSE
HOLD
S WITH
OUT E
VENT
S51
.1(1.
27)
0.4(0.
21)
0.2(0.
11)
50.8
(1.74
)0.7
(0.39
)0.4
(0.20
)51
.6(1.
85)
0.00.0
00.0
0.00
TOTA
L HOU
SEHO
LDS
100.0
0.00
0.00.0
012
.40.0
010
0.00.0
00.0
0.00
14.5
0.00
100.0
0.00
0.00.0
010
.00.0
0TO
TAL N
UMBE
R of
HOUS
EHOL
DS W
ITH E
VENT
S12
7,616
0.00
00.0
00
0.00
68,51
50.0
00
0.00
00.0
059
,101
0.00
00.0
00
0.00
TOTA
L NUM
BER
of HO
USEH
OLDS
WITH
OUT E
VENT
S13
3,617
0.00
00.0
00
0.00
70,65
40.0
00
0.00
00.0
062
,963
0.00
00.0
00
0.00
TOTA
L NUM
BER
of HO
USEH
OLDS
261,2
330.0
00
0.00
00.0
013
9,169
0.00
00.0
00
0.00
122,0
640.0
00
0.00
00.0
0
1/ De
comp
ositio
n ba
sed
on eq
uatio
n [2]
Sour
ce: A
uthors
’ elab
oratio
n bas
ed on
data
from
Encu
esta
de H
ogare
s de P
ropós
itos M
últipl
es (IN
EC).
D E M O G R A P H I C &
C O M B I N E D
E V E N T S
VII -
Eve
nts no
t clas
sified
Entry
Even
ts
N O N D E M O G R A P H I C
I - Ex
clusiv
ely la
bor
incom
e eve
nts
II - E
xclus
ively
non-
labor
incom
e eve
nts
P(S/E
)En
tryP(
even
t)P(
S/E)
Table
12.C
. Dec
ompo
sition
of th
e entr
y rate
to po
verty
. Sen
sitivi
ty an
alysis
to p
overt
y line
(Incre
ase o
f 10%
in po
very
line)
1/Co
sta R
ica. U
rban
area
s. Pe
riod 2
006 -
2008
.
Total
hous
ehold
sHo
useh
olds w
ith ch
ildren
Hous
ehold
s with
out c
hildr
enP(
even
t)P(
S/E)
Entry
P(ev
ent)
128 129
N°(1)
SE(2)
SE(1)
*(2)=(
3)SE
(1)SE
(2)SE
(1)*(2
)=(3)
SE(1)
SE(2)
SE(1)
*(2)=(
3)SE
TOTA
L LAB
OR E
VENT
S22
.09(0.
44)
31.62
(1.04
)6.9
8(0.
27)
25.65
(0.61
)36
.82(1.
32)
9.44
(0.41
)17
.24(0.
62)
21.09
(1.56
)3.6
4(0.
30)
1Re
ducti
on in
the n
umbe
r of e
mploy
ed m
embe
rs3.1
(0.18
)40
.7(2.
96)
1.3(0.
12)
3.3(0.
25)
45.7
(3.86
)1.5
(0.17
)2.8
(0.27
)32
.8(4.
48)
0.9(0.
15)
1.1Re
ducti
on in
the n
umbe
r of r
egist
ered
wag
e ear
ners
0.7(0.
09)
37.3
(6.65
)0.2
(0.06
)0.7
(0.12
)43
.6(8.
64)
0.3(0.
08)
0.6(0.
13)
26.3
(9.89
)0.1
(0.07
)1.2
Redu
ction
in th
e num
ber o
f non
-regis
tered
wag
e ear
ners
1.0(0.
11)
43.0
(5.07
)0.4
(0.07
)1.1
(0.15
)46
.2(6.
43)
0.5(0.
10)
0.9(0.
16)
37.7
(8.17
)0.3
(0.09
)1.3
Redu
ction
in th
e num
ber o
f non
-wag
e ear
ners
1.4(0.
12)
40.7
(4.28
)0.6
(0.08
)1.5
(0.16
)46
.3(5.
65)
0.7(0.
11)
1.3(0.
18)
32.3
(6.29
)0.4
(0.10
)
2Re
ducti
on in
total
hour
ly wa
ge of
mem
bers
emplo
yed i
n both
ob
serva
tions
8.8(0.
30)
26.4
(1.57
)2.3
(0.16
)10
.1(0.
42)
32.1
(2.04
)3.2
(0.25
)6.9
(0.41
)15
.1(2.
14)
1.0(0.
16)
3Re
ducti
on in
the n
umbe
r of w
orkin
g hou
rs of
mem
bers
emplo
yed
in bo
th ob
serva
tions
2.7(0.
17)
19.6
(2.47
)0.5
(0.07
)3.1
(0.24
)21
.1(3.
15)
0.6(0.
11)
2.2(0.
23)
16.7
(3.94
)0.4
(0.09
)
4Re
ducti
on in
the n
umbe
r of w
orkin
g hou
rs an
d in t
he to
tal ho
urly
wage
of m
embe
rs em
ploye
d in b
oth ob
serva
tions
4.8(0.
22)
30.8
(2.14
)1.5
(0.12
)5.5
(0.31
)35
.4(2.
69)
2.0(0.
18)
3.9(0.
32)
21.9
(3.36
)0.8
(0.15
)
5Re
ducti
on in
the t
otal m
onth
ly wa
ge of
mem
bers
emplo
yed i
n both
ob
serva
tions
and i
n the
numb
er of
emplo
yed m
embe
rs2.7
(0.17
)51
.5(3.
19)
1.4(0.
12)
3.6(0.
26)
57.6
(3.61
)2.1
(0.20
)1.5
(0.19
)31
.2(5.
97)
0.5(0.
10)
TOTA
L NON
-LAB
OR E
VENT
S5.3
(0.24
)21
.0(1.
82)
1.1(0.
11)
3.7(0.
26)
20.9
(2.83
)0.8
(0.11
)7.6
(0.44
)21
.0(2.
37)
1.6(0.
20)
6Re
ducti
on in
the i
ncom
e fro
m pe
nsion
s 1.1
(0.11
)15
.7(3.
73)
0.2(0.
04)
0.5(0.
10)
34.1
(9.14
)0.2
(0.06
)1.8
(0.23
)8.5
(3.14
)0.2
(0.06
)7
Redu
ction
in pu
blic m
oneta
ry tra
nsfer
s (so
cial p
olicy
) 0.1
(0.03
)22
.1(14
.15)
0.0(0.
01)
0.1(0.
05)
30.2
(18.57
)0.0
(0.02
)0.1
(0.04
)0.0
0.00
0.00.0
08
Redu
ction
in ot
her n
on-la
bor in
come
s 4.2
(0.21
)22
.3(2.
09)
0.9(0.
10)
3.0(0.
23)
18.3
(2.87
)0.6
(0.09
)5.7
(0.38
)25
.2(2.
92)
1.4(0.
19)
III - L
abor
and n
on-
labor
incom
e eve
nts9
Redu
ction
in la
bor a
nd no
n-lab
or inc
omes
6.7(0.
27)
40.1
(2.00
)2.7
(0.17
)7.2
(0.36
)46
.1(2.
55)
3.3(0.
25)
6.0(0.
39)
30.1
(3.08
)1.8
(0.22
)
IV E
xclus
ively
demo
graph
ic ev
ents
10Gr
owth
in the
total
numb
er of
hous
ehold
mem
bers
; the t
otal
nomi
nal in
come
rema
ins co
nstan
t8.3
(0.30
)9.4
(1.06
)0.8
(0.09
)6.6
(0.35
)6.8
(1.38
)0.4
(0.09
)10
.5(0.
51)
11.6
(1.57
)1.2
(0.17
)
V - D
emog
raphic
ev
ents
leadin
g to
incom
e cha
nges
11Re
ducti
on in
the n
umbe
r of la
bor o
r non
-labo
r inc
ome e
arne
rs
due t
o the
entra
nce o
f mem
bers
to the
hous
ehold
2.8(0.
18)
23.9
(2.62
)0.7
(0.08
)2.5
(0.22
)30
.9(3.
90)
0.8(0.
12)
3.2(0.
29)
16.4
(3.35
)0.5
(0.12
)
VI - C
ombin
ation
of
demo
graph
ic an
d inc
ome e
vents
12Re
ducti
on in
the t
otal
nomi
nal in
come
and g
rowth
in the
numb
er
of ho
useh
old m
embe
rs.
5.6(0.
25)
47.3
(2.26
)2.6
(0.17
)5.5
(0.32
)47
.9(2.
94)
2.7(0.
22)
5.6(0.
38)
46.4
(3.49
)2.6
(0.26
)
6.9(0.
27)
28.0
(1.79
)1.9
(0.14
)7.4
(0.36
)32
.9(2.
38)
2.4(0.
21)
6.3(0.
41)
20.3
(2.56
)1.3
(0.18
)TO
TAL H
OUSE
HOLD
S W
ITH E
VENT
S57
.6(0.
54)
29.1
(0.64
)16
.8(0.
40)
58.6
(0.69
)33
.9(0.
86)
19.9
(0.56
)56
.4(0.
84)
22.4
(0.91
)12
.6(0.
55)
TOTA
L HOU
SEHO
LDS
WITH
OUT E
VENT
S42
.4(0.
54)
0.4(0.
11)
0.2(0.
05)
41.4
(0.69
)0.6
(0.17
)0.2
(0.07
)43
.6(0.
84)
0.3(0.
15)
0.1(0.
06)
TOTA
L HOU
SEHO
LDS
100.0
17.0
100.0
20.1
100.0
12.8
TOTA
L NUM
BER
of H
OUSE
HOLD
S W
ITH E
VENT
S2,2
26,41
51,3
03,70
792
2,708
TOTA
L NUM
BER
of H
OUSE
HOLD
S W
ITHOU
T EVE
NTS
1,635
,871
922,0
3171
3,840
TOTA
L NUM
BER
of H
OUSE
HOLD
S3,8
62,28
62,2
25,73
81,6
36,54
8
1/ De
comp
ositio
n ba
sed
on eq
uatio
n [2]
Sour
ce: A
uthors
’ elab
oratio
n bas
ed on
data
from
Encu
esta
de E
mpleo
, Des
emple
o y S
ubem
pleo (
INEC
).
D E M O G R A P H I C &
C O M B I N E D
E V E N T S
VII -
Eve
nts no
t clas
sified
Entry
Even
ts
N O N D E M O G R A P H I C
I - Ex
clusiv
ely la
bor
incom
e eve
nts
II - E
xclus
ively
non-
labor
incom
e eve
nts
P(S/
E)En
tryP(
even
t)P(
S/E)
Table
12.D
. Dec
ompo
sition
of th
e ent
ry ra
te to
pove
rty. S
ensit
ivity
analy
sis to
pov
erty
line (
Incre
ase o
f 10%
in po
very
line)
1/Ec
uado
r. Ur
ban a
reas
. Per
iod 20
04-20
08.
Total
hous
ehold
sHo
useh
olds w
ith ch
ildre
nHo
useh
olds w
ithou
t chil
dren
P(ev
ent)
P(S/
E)En
tryP(
even
t)
130
N°(1)
SE(2)
SE(1)
*(2)=(
3)SE
(1)SE
(2)SE
(1)*(2
)=(3)
SE(1)
SE(2)
SE(1)
*(2)=(
3)SE
TOTA
L LAB
OR E
VENT
S19
.14(0.
90)
36.12
(2.39
)6.9
1(0.
55)
20.43
(1.17
)43
.88(3.
05)
8.97
(0.78
)17
.00(1.
38)
20.58
(3.32
)3.5
0(0.
62)
1Re
ducti
on in
the n
umbe
r of e
mploy
ed m
embe
rs4.1
(0.44
)54
.8(5.
33)
2.3(0.
32)
4.0(0.
54)
63.1
(6.69
)2.5
(0.43
)4.3
(0.79
)41
.9(8.
09)
1.8(0.
46)
1.1Re
ducti
on in
the n
umbe
r of r
egist
ered
wag
e ear
ners
1.2(0.
24)
45.1
(9.88
)0.6
(0.16
)1.1
(0.29
)63
.3(13
.08)
0.7(0.
22)
1.4(0.
44)
22.2
(11.98
)0.3
(0.19
)1.2
Redu
ction
in th
e num
ber o
f non
-regis
tered
wag
e ear
ners
1.3(0.
25)
64.6
(9.56
)0.8
(0.19
)1.3
(0.30
)63
.5(11
.53)
0.8(0.
24)
1.3(0.
43)
66.4
(16.76
)0.9
(0.32
)1.3
Redu
ction
in th
e num
ber o
f non
-wag
e ear
ners
1.6(0.
28)
54.3
(8.76
)0.9
(0.21
)1.7
(0.36
)62
.5(10
.53)
1.0(0.
29)
1.5(0.
45)
39.4
(14.22
)0.6
(0.27
)
2Re
ducti
on in
total
hour
ly wa
ge of
mem
bers
emplo
yed i
n both
ob
serva
tions
5.6(0.
54)
25.3
(3.99
)1.4
(0.25
)6.0
(0.68
)30
.4(5.
25)
1.8(0.
35)
4.9(0.
86)
14.9
(5.29
)0.7
(0.28
)
3Re
ducti
on in
the n
umbe
r of w
orkin
g hou
rs of
mem
bers
emplo
yed
in bo
th ob
serva
tions
1.7(0.
28)
18.5
(6.39
)0.3
(0.12
)1.7
(0.34
)26
.2(9.
27)
0.4(0.
18)
1.9(0.
48)
7.1(6.
86)
0.1(0.
13)
4Re
ducti
on in
the n
umbe
r of w
orkin
g hou
rs an
d in t
he to
tal ho
urly
wage
of m
embe
rs em
ploye
d in b
oth ob
serva
tions
4.0(0.
43)
31.0
(5.21
)1.2
(0.25
)4.7
(0.59
)40
.9(6.
52)
1.9(0.
39)
2.8(0.
57)
3.8(2.
77)
0.1(0.
08)
5Re
ducti
on in
the t
otal m
onth
ly wa
ge of
mem
bers
emplo
yed i
n both
ob
serva
tions
and i
n the
numb
er of
emplo
yed m
embe
rs3.7
(0.42
)45
.4(5.
62)
1.7(0.
27)
4.1(0.
56)
55.1
(6.92
)2.3
(0.41
)3.1
(0.60
)23
.6(8.
12)
0.7(0.
28)
TOTA
L NON
-LAB
OR E
VENT
S4.8
(0.52
)16
.6(3.
80)
0.8(0.
20)
4.1(0.
60)
19.5
(5.53
)0.8
(0.24
)6.0
(0.90
)13
.4(5.
09)
0.8(0.
33)
6Re
ducti
on in
the i
ncom
e fro
m pe
nsion
s 0.7
(0.18
)10
.3(7.
19)
0.1(0.
06)
0.4(0.
17)
11.7
(11.45
)0.0
(0.05
)1.3
(0.39
)9.7
(9.18
)0.1
(0.12
)7
Redu
ction
in pu
blic m
oneta
ry tra
nsfer
s (so
cial p
olicy
) 0.0
0.00
0.00.0
00.0
0.00
0.00.0
00.0
0.00
0.00.0
00.0
0.00
0.00.0
00.0
0.00
8Re
ducti
on in
othe
r non
-labo
r inco
mes
4.1(0.
46)
17.7
(4.27
)0.7
(0.19
)12
.6(0.
93)
20.3
(5.86
)0.8
(0.24
)4.8
(0.80
)14
.4(6.
07)
0.7(0.
31)
III - L
abor
and n
on-
labor
incom
e eve
nts9
Redu
ction
in la
bor a
nd no
n-lab
or inc
omes
11.4
(0.72
)44
.7(3.
27)
5.1(0.
48)
8.2(0.
77)
49.5
(3.93
)6.3
(0.67
)9.4
(1.13
)34
.0(5.
72)
3.2(0.
63)
IV E
xclus
ively
demo
graph
ic ev
ents
10Gr
owth
in the
total
numb
er of
hous
ehold
mem
bers
; the t
otal
nomi
nal in
come
rema
ins co
nstan
t9.4
(0.64
)13
.0(2.
37)
1.2(0.
24)
2.3(0.
40)
8.1(2.
47)
0.7(0.
21)
11.4
(1.19
)18
.9(4.
16)
2.2(0.
52)
V - D
emog
raphic
ev
ents
leadin
g to
incom
e cha
nges
11Re
ducti
on in
the n
umbe
r of la
bor o
r non
-labo
r inco
me ea
rners
due
to the
entra
nce o
f mem
bers
to the
hous
ehold
2.5(0.
33)
24.5
(5.85
)0.6
(0.17
)12
.7(0.
93)
37.7
(8.76
)0.9
(0.25
)2.8
(0.58
)6.9
(4.48
)0.2
(0.13
)
VI - C
ombin
ation
of
demo
graph
ic an
d inc
ome e
vents
12Re
ducti
on in
the t
otal
nomi
nal in
come
and g
rowth
in the
numb
er
of ho
useh
old m
embe
rs.
12.0
(0.73
)46
.3(3.
11)
5.6(0.
52)
8.7(0.
76)
49.4
(3.76
)6.3
(0.69
)10
.9(1.
19)
40.3
(5.30
)4.4
(0.76
)
8.2(0.
60)
41.3
(3.82
)3.4
(0.40
)26
.3(1.
24)
47.7
(4.58
)4.1
(0.54
)7.3
(1.01
)28
.8(6.
47)
2.1(0.
56)
TOTA
L HOU
SEHO
LDS
WITH
EVE
NTS
67.5
(0.96
)35
.0(1.
30)
23.6
(0.98
)69
.1(1.
25)
40.5
(1.63
)28
.0(1.
29)
64.9
(1.49
)25
.2(1.
96)
16.3
(1.38
)TO
TAL H
OUSE
HOLD
S W
ITHOU
T EVE
NTS
32.5
(0.96
)1.7
(0.56
)0.6
(0.18
)30
.9(1.
25)
2.0(0.
81)
0.6(0.
25)
35.1
(1.49
)1.4
(0.71
)0.5
(0.25
)TO
TAL H
OUSE
HOLD
S10
0.00.0
00.0
0.00
24.2
0.00
100.0
0.00
0.00.0
028
.60.0
010
0.00.0
00.0
0.00
16.8
0.00
TOTA
L NUM
BER
of H
OUSE
HOLD
S W
ITH E
VENT
S3,9
85,05
00.0
00
0.00
00.0
02,5
46,38
50.0
00
0.00
00.0
01,4
38,66
50.0
00
0.00
00.0
0TO
TAL N
UMBE
R of
HOU
SEHO
LDS
WITH
OUT E
VENT
S1,9
18,10
20.0
00
0.00
00.0
01,1
40,64
10.0
00
0.00
00.0
077
7,461
0.00
00.0
00
0.00
TOTA
L NUM
BER
of H
OUSE
HOLD
S5,9
03,15
20.0
00
0.00
00.0
03,6
87,02
60.0
00
0.00
00.0
02,2
16,12
60.0
00
0.00
00.0
0
1/ De
comp
ositio
n ba
sed
on eq
uatio
n [2]
Sour
ce: A
uthors
’ elab
oratio
n bas
ed on
data
from
Encu
esta
Nacio
nal d
e Hog
ares (
INEI
)
P(S/
E)
Table
12.E
. Dec
ompo
sition
of th
e ent
ry ra
te fro
m po
verty
. Sen
sitivi
ty an
alysis
to p
over
ty lin
e (Inc
reas
e of 1
0% in
pove
ry lin
e) 1/
Peru
. Urb
an ar
eas.
Perio
d 200
2-200
6.
Total
hous
ehold
sHo
useh
olds w
ith ch
ildre
nHo
useh
olds w
ithou
t chil
dren
P(ev
ent)
P(S/
E)En
tryP(
even
t)
D E M O G R A P H I C &
C O M B I N E D
E V E N T S
VII -
Eve
nts no
t clas
sified
Entry
Even
ts
N O N D E M O G R A P H I C
I - Ex
clusiv
ely la
bor
incom
e eve
nts
II - E
xclus
ively
non-
labor
incom
e eve
nts
P(S/
E)En
tryP(
even
t)