De facto property rights recognition, labor supply, and investment of the

poor in Brazil

Rafael Santos DantasMaria Tannuri-Pianto

(University of Brasilia)

World Bank Conference on Land and Poverty Washington DC, March 24-27 2014 1

2

The Brasilia context

• Almost a quarter of the population of the Federal District was residing in non-regularized properties in 2006.

• 17 out of its 31 administrative regions have part or all of their territory irregularly occupied.

• The biggest slum in Brazil, with 80,000 people: Sol Nascente.

• History of evictions and relocations by the government.– Some of them marked by violence.

• Grilagem: Organized crime of land grabbing

3

4

5

Land Tenure Regularization Program

• Legal Framework–Federal Government• Constitution: defines urban policy

guidelines – director plan and “social role of property”• Federal Law nº 10.257/01, City Statue: – LTR as urban policy guideline; – Special Zones of Social Interest (ZEIS);–Concession of Real Rights of Use (CDRU).

6

Land Tenure Regularization Program

• Legal Framework–Federal District Government• Directive Plan of Land Arrangement

(2009)– LTR Strategy: Delimitation of areas

subject to regularization

• Decree nº 23.592/03 : eligibility criteria

7

Land Tenure Regularization Program

• Program Timing:1. Definition of areas subject to

regularization• ARIS (Regularization Areas of Social Interest)

2. Delivery of documents proving occupation

3. Formalization

• Our Approach: the definition of areas subject to regularization is a de facto recognition of property rights.

8

Land Tenure Regularization Program

Background

• Investment– Besley (1995); Alston, Libecap, Schneider (1996); Field

(2005); Galiani e Schargrodsky (2010)

• Credit Market– Field & Torero (2006); Galiani & Schargrodsky (2010);

Moura & Piza (2011)

• Labor Market– Field (2007); Moura et al (2011)

• Land Market– Lanjouw & Levy (2002); Alston & Mueller (2010);

Deininger et al. (2011); Macours et al (2010)

• Income– Andrade (2004); Moura & Bueno (2009)

9

10

Conceptual Framework

• Extension of Field’s (2003) model of household labor allocation in a incomplete property rights setting.

• 2 adults, both contributing to property security and home production

• Wages are higher than home production marginal value

• Results: the burden of protecting the property will rely on the adult that has comparative advantage in providing security, even if receives higher wages.

11

Empirical Investigation

• Socioeconomic Survey in Territories of Social Vulnerability (2010)

• Not designed to investigate the program.

• 8,395 households– 1,338 household belonging to ARIS

12

Empirical Investigation

• Propensity Score Matching• Identification Strategy:– Households within an ARIS– Households not under tenancy contracts– Living in for at least 5 years

13

Empirical Investigation

• Definition of Treatment & Control Groups:– Treatment: Households in ARIS with a property

document – not the deed – and that perceives the property as being regularized (239 households)

– Control: Households in ARIS with no property document and that does not perceives the property as regularized (277 households)

• Perception of regularization– ARIS not regularized by the time data was

gathered;– Associated with a positive answer to the question

about property land being settled.

14

Empirical InvestigationVariables Treatment

GroupControl Group Difference p-value

Sex head of household (=1 if female) 0,2795 0.3369 -0.0575 0.32Age head of household 43.7128 43.1130 0.5998 0.77Working (=1 if hh employed) 0.6466 0.7097 -0.0631 0.31Race (=1 if hh black or mixed race) 0.6485 0.7323 -0.0838 0.20Years of education head of household 6.2186 6.0250 0.1936 0.73Spouse (=1 if hh has a partner) 0.7195 0.7086 0.0109 0.84Spouse working 0.3346 0.3962 -0.0616 0.23Spouse years of education 5.6518 4.6980 0.9538 0.10*Number of other adults living in household 2.0764 2.0926 -0.0162 0.92Years of education other adults in household 4.1454 4.8861 -0.7407 0.33Proportion of other adults employed 0.2040 0.2496 -0.0457 0.22Number of kids in household 0.7898 0.6656 0.1242 0.30Elderly living in household 0.1001 0.0998 0.0002 0.99Number of inactive adults in household 0.2000 0.2400 -0.0400 0.43Years living in residence 9.6809 9.2437 0.4372 0.62Wealth accumulation index 12.7210 13.8400 -1.1190 0.12Property taxes (=1 if hh is billed) 0.9478 0.9529 -0.0051 0.84

15

Propensity Score Model

• Include a set of region indicators that are strongly significant;

• Spouse years of education, years living in residence , property taxes are significant. All increase the probability of being in treatment group.

16

Results – Labor Market

• Treated household supply between 18% and 24% more weekly labor hours than control.

• Men in treated household supply 20% more weekly labor hours than men in control group.

• Women in treated household supply 10% more weekly labor hours than women in control group, although this result is not statistically significant.

17

Results – Household Earnings

• Program has large and significant effect on per capita labor earnings: Individuals in treated group earn around 30% more than in control;

• Employed men in treated group earn 55% more than employed men in control;

• Men’s hourly earnings are 50% higher in treated households than in control;

• Possible mechanism: expansion of exchange circles to outside community boundaries;

– Likelihood that a treated household has at least one adult man with commuting time to work greater than 90 minutes is 44%, while it falls to 33% for a control group household.

18

Results – Investment

• Effects on quality of external wall, roof material, and size of property not economically relevant.

• Impact on likelihood of future investments: 14 p.p higher in treated households than in control ones (0.51/0.37).

• Not just an “income effect”• Impact on likelihood of future investments

controlling for household earnings: 10 p.p (0.51/0.41)

19

Threats to Internal Validity

1. Difference between groups

2. Non-observables

3. Endogeneity

20

Going Forward

• Anticipation bias• Specifically designed questionnaire• Explore discontinuity in ARIS

delimitation– Boundaries may be biased

• Interference of other Housing Policy: “Programa Morar Bem”– Revealed preference

21

Tables

22

Labor MarketHours worked

household Radius(0.01) Radius(0.005) Kernel(0.06) Kernel(0.01)

Treatment 61.95567 61.21687 60.81333 61.95567Control 49.96729 50.38259 51.18076 49.70243ATT 11.98837 10.83428 9.63257 12.25324T-stat 2.14** 1.97** 1.81* 2.17**ATE 10.67428 9.38546 8.54985 10.47978Obs(Treat/Control) 203/220 166/182 225/240 203/220

Hours worked by women Radius(0.01) Radius(0.005) Kernel(0.06) Kernel(0.01)

Treatment 20.42036 20.23092 20.12148 20.42036Control 18.49923 18.49219 18.07289 18.68868ATT 1.92114 1.73873 2.04859 1.73168T-stat 0.66 0.60 0.74 0.59ATE 2.64988 2.38711 2.74359 2.29938Obs (Treat./Control) 203/220 166/182 225/240 203/220

23

Labor MarketHours worked by

men Radius(0.01) Radius(0.005) Kernel(0.06) Kernel(0.01)

Treatment 31.28325 31.73494 30.96222 31.28325Control 25.23452 26.11479 26.83680 25.03516ATT 6.04873 5.62015 4.12542 6.24809T-stat 2.08** 1.95* 1.50 2.14**ATE 4.72680 4.00330 22.84689 4.88659Obs (Treat./Control) 203/220 166/182 225/240 203/220

24

Household EarningsHousehold

Earnings Radius(0.01) Radius(0.005) Kernel(0.06) Kernel(0.01)

Treatment 1044.68 1026.9 1077.73 1044.68Control 705.8 751.83 746.24 708.61ATT 338.88 275.07 331.49 336.07T-stat 2.67*** 2.18** 2.68*** 2.63***ATE 282.98 254.31 258.06 272.29Obs (Treat./Control) 203/220 166/182 225/240 203/220

Per Capita LaborEarnings Radius(0.01) Radius(0.005) Kernel(0.06) Kernel(0.01)

Treatment 289.40 287.45 293.10 289.40Control 215.98 224.18 218.57 217.62ATT 73.42 63.27 74.53 71.78T-stat 2.08** 1.83* 2.23** 2.02**ATE 61.31 55.63 54.3 58.04Obs. (Treat./Control) 203/220 166/182 225/240 203/220

25

Household EarningsEmployed Men Labor Earnings Radius(0.01) Radius(0.005) Kernel(0.06) Kernel(0.01)

Treatment 546.65 536.11 559.16 546.65Control 329.68 363.07 361.82 321.26ATT 216.97 173.04 197.34 225.39T-stat 2.8*** 2.3** 2.69*** 2.9***ATE 183.49 162.38 145.43 184.49Obs. (Treat./Control) 203/220 166/182 225/240 203/220

Employed MenHourly Earnings Radius(0.01) Radius(0.005) Kernel(0.06) Kernel(0.01)

Treatment 4.53 4.41 4.64 4.53Control 3.05 3.29 2.88 3.05ATT 1.48 1.12 1.76 1.48T-stat 2.49** 1.75* 3.05*** 2.44**ATE 1.53 1.21 1.48 1.51Obs. (Treat./Control) 115/154 96/117 158/175 115/154

26

Household EarningsIndicator of at least one men in household with commuting time > 1h30min

Radius(0.01) Radius(0.005) Kernel(0.06) Kernel(0.01)

Treatment 0.44 0.45 0.44 0.44Control 0.3 0.33 0.32 0.29ATT 0.14 0.12 0.12 0.15T-stat 2.23** 1.92* 2.02** 2.32**ATE 0.15 0.13 0.12 0.17Obs. (Treat./Control) 203/220 166/182 225/240 203/220

27

Actual InvestmentExternal

Masonry Wall Radius(0.01) Radius(0.005) Kernel(0.06) Kernel(0.01)

Treatment 0.9812 0.9838 0.978 0.9812Control 0.9281 0.9399 0.9078 0.9291ATT 0.0531 0.0439 0.0702 0.0521T-stat 2.42** 1.99** 3.28*** 2.35**ATE 0.041 0.037 0.049 0.17Obs. (Treat./Control) 213/232 186/195 228/245 213/232

Shingled Roof Radius(0.01) Radius(0.005) Kernel(0.06) Kernel(0.01)Treatment 0.9718 0.9785 0.9736 0.9718Control 0.9329 0.9449 0.9205 0.9321ATT 0.0389 0.0336 0.0531 0.0397T-stat 1.27 1.2 1.82* 1.29ATE 0.026 0.025 0.013 0.03Obs. (Treat./Control) 213/232 186/195 228/245 213/232

28

Actual InvestmentNumber of Bedrooms Radius(0.01) Radius(0.005) Kernel(0.06) Kernel(0.01)

Treatment 2.289 2.309 2.309 2.289Control 2.101 2.111 2.127 2.099ATT 0.188 0.198 0.182 0.19T-stat 1.65* 1.69* 1.66* 1.66*ATE 0.106 0.128 0.13 0.11Obs. (Treat./Control) 213/232 186/195 228/245 213/232

29

Future InvestmentPropensityto Invest Radius(0.01) Radius(0.005) Kernel(0.06) Kernel(0.01)

Treatment 0.516 0.532 0.500 0.516Control 0.376 0.363 0.399 0.371ATT 0.14 0.169 0.101 0.145T-stat 2.19** 2.63*** 1.59 2.24**ATE 0.111 0.173 0.064 0.123Obs. (Treat./Control) 213/232 186/195 228/245 213/232

Propensityto Invest – HH

incomeRadius(0.01) Radius(0.005) Kernel(0.06) Kernel(0.01)

Treatment 0.515 0.511 0.509 0.515Control 0.41 0.42 0.398 0.412ATT 0.105 0.091 0.111 0.103T-stat 1.71* 1.43 1.82* 1.66*ATE 0.083 0.103 0.078 0.088Obs. (Treat./Control) 213/232 186/195 228/245 213/232