Policy Research Working Paper 8522

Tax Evasion in Africa and Latin America

The Role of Distortionary Infrastructures and Policies

Wilfried A. Kouamé Jonathan Goyette

Africa RegionOffice of the Chief EconomistJuly 2018

WPS8522P

ublic

Dis

clos

ure

Aut

horiz

edP

ublic

Dis

clos

ure

Aut

horiz

edP

ublic

Dis

clos

ure

Aut

horiz

edP

ublic

Dis

clos

ure

Aut

horiz

ed

Produced by the Research Support Team

Abstract

The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent.

Policy Research Working Paper 8522

This paper examines the impact of the quality of the busi-ness environment as well as the monitoring capacity of the tax agency on firms’ tax evasion and production decisions. First, the paper uses firm-level data for 30 African and Latin American countries to show that tax evasion and distortions stemming from the business environment are positively and significantly correlated, while sales not reported for tax purposes and institutional quality are negatively and sig-nificantly correlated. Second, the paper develops a general equilibrium model where heterogeneous firms make tax evasion decisions based on their assessment of the quality of their business environment as well as the monitoring

capacity of the tax agency. The model simulations for each country in the African and Latin American sample show that the model can explain 35 percent of the variation in tax evasion and more than 49 percent of the dispersion in output per worker across the sample countries. Finally, a series of counterfactual experiments shows that, at the cur-rent level of deterrence, governments could decrease sales not reported for tax purposes by 21 percent, by reducing distortions stemming from the business environment by half. The paper presents empirical supporting evidence consistent with testable predictions of the model.

This paper is a product of the Office of the Chief Economist, Africa Region. It is part of a larger effort by the World Bank to provide open access to its research and make a contribution to development policy discussions around the world. Policy Research Working Papers are also posted on the Web at http://www.worldbank.org/research. The authors may be contacted at wkouame@worldbank.org.

Tax Evasion in Africa and Latin America: The Role of

Distortionary Infrastructures and Policies

Wilfried A. Kouamé∗ and Jonathan Goyette†

Keywords: Distortions, tax evasion, business environment, infrastructure, institutional quality, Africa and Latin American.

JEL codes: H2; H3; H5; O1

∗Corresponding author: Wilfried A. Kouamé - The World Bank and Economics Department, Universitéde Sherbrooke: wkouame@worldbank.org/wilfried.kouame@usherbrooke.ca. Authors are grateful to DiegoRestuccia, Théophile Azomahou, Moussa Blimpo, Jean-François Rouillard, Cesar Calderon, and PunamChulan-Pole for their comments. The paper has also benefited from presentations at Oxford UniversityCSAE 2017 Conference, Canadian Economics Association and Public Economic Theory conferences in 2016.Earlier versions of the paper circulated under the titles "Distortions, Policy Ineffectiveness and Tax Evasion"and "Distortions, Social Infrastructures and Tax Evasion".

†Economics Department, Université de Sherbrooke

I Introduction

A sound tax system and a healthy business environment are crucial for the welfare of an

economy (Easterly and Rebelo, 1993; Aterido et al., 2011). However, many African and

Latin American economies face major challenges on both fronts. Indeed, high levels of tax

evasion are ubiquitous in these countries. On average, 25% of total sales are not reported

for tax purpose in African and Latin American countries compared with only 7% in OECD

countries.1 This shortfall has important social and economic consequences, as evaded taxes

reduce a government’s ability to invest in productive infrastructures and institutional quality,

and to provide public goods and services. Moreover, insufficient government resources result

in an inefficient business environment as various distortions (power outages, corruption, etc.)

hinder firms’ performance and their ability to create jobs (Besley and Burgess, 2004; Aterido

et al., 2011; Buera et al., 2013; Goyette and Gallipoli, 2015). As such, African and Latin

American economies fare poorly in terms of ease of doing business.2

The coexistence of tax evasion and inefficient business environments thus raises the ques-

tion about a potential link between these variables in developing countries. This paper aims

to examine how distortions stemming from the business environment and institutional qual-

ity, i.e., a tax agency’s monitoring capacity, affect firms’ production and tax evasion decisions.

We focus on two specific mechanisms. First, entrepreneurs, knowing that institutional quality

is low, also anticipate low monitoring of taxes and have, therefore, more incentive to evade

their taxes. Second, we argue that a poor business environment generates costs for firms and

that this creates a wedge between firms’ potential and realized profits. As a result, firms

have an incentive to under-report their sales for tax purposes in order to compensate for the

losses incurred due to the distortions in their business environment.

Using firm-level data from the World Bank Enterprise Surveys (WBES) as well as data

from the World Governance Indicators (WGI), we provide supporting evidence consistent1See Table A.1 in Appendix A.2See Appendix A.

2

with the mechanisms described above. Tax evasion is proxied by the amount of sales not

declared for tax purposes. Distortions from the business environment are calculated based on

the losses in annual sales due to power outages, corruption, etc. Finally, institutional quality

is based on various variables related to the perceptions of the quality of public services and

the credibility of the government. We find a positive and significant relationship between

losses due to distortions from the business environment and tax evasion. On the contrary,

institutional quality is negatively related to firms’ tax evasion.

However, these relationships are subject to omitted variable bias, potential measurement

error, and reverse causality. As a result, we are not able to identify, using standard econo-

metrics, the mechanisms through which distortions and institutional quality affect firms’ tax

evasion behavior. Instead, we develop a general equilibrium model and use simulations to

match the empirical evidence. Building on Restuccia and Rogerson (2008), the economic

environment consists of (i) one representative household which maximizes his inter-temporal

utility, (ii) a government which balances its budget and, (iii) heterogeneous firms which max-

imize their profits and make tax evasion decisions based on their idiosyncratic productivity

level as well as their idiosyncratic level of distortions stemming from the business environ-

ment. We calibrate the model to the United States and treat that country as an economy

with no distortions as is standard in the literature. This benchmark economy allows nor-

malizing GDP per worker in both the model and the data. The model is then simulated

for each of the 30 African and Latin American countries from the WBES sample, using the

benchmark distribution of productivity and the idiosyncratic distribution of losses due to the

distortions from the business environment of a specific country.

The model explains 35 percent of the variation of tax evasion and 49 percent of the

dispersion of output per worker in the data. The simulation for each region provides similar

findings. As robustness checks, we calibrate the model to the Chilean economy, which exhibits

the lowest combination of distortions and tax evasion in the data. Moreover, the model

is simulated using an alternative measure of the deterrence probability. These additional

3

robustness checks show that the findings remain robust, and the model explains at least 20

percent of the variation of tax evasion across countries in all cases.

Having established the ability of the model to replicate some relevant moments in the data

related to tax evasion and output across African and Latin American countries, we conduct

a set of tax neutral counterfactual experiments to examine various policy implications. First,

we examine what happens in Guinea, the country with the highest level of tax evasion in the

data, when distortions are reduced to the average level of the sample. Such improvements

generate a drop between 18.34 and 40.16 percent in tax evasion while output per worker

increases by between one- to six-fold. Second, we examine what happens when governments

could reduce distortions stemming from the business environment by half. Such reduction

could reduce sales not reported for tax purposes by about 21%.

This paper is closely related to Restuccia and Rogerson (2008) and Bah and Fang (2015).

The authors argue in the former paper that a country’s policies and institutions can create

taxes or subsidies (distortions) on establishment output. These distortions reduce aggregate

total factor productivity (TFP) and can explain up to 50% of the cross-country differences

in output, capital accumulation, and TFP. Bah and Fang (2015) introduce distortions as an

idiosyncratic tax on output in the general equilibrium model of Amaral and Quintin (2010).

In addition to a distribution of productivity approximated with firms’ size, Bah and Fang

(2015) use the distribution of distortions from the data and collateral constraints to explain

some of the variations in output in Africa. This paper differs from Restuccia and Rogerson

(2008) and Bah and Fang (2015) by focusing on the effect of distortions and monitoring

capacity on firms’ tax evasion behavior. The paper highlights two mechanisms explaining

the role of distortions stemming from the business environment and institutional quality on

tax evasion in African and Latin American countries.

Well-developed infrastructures and institutions are essential to promoting economic growth

by reducing transaction cost for firms as well as for households. Firms’ performance is af-

fected by the quality of infrastructures such as transport, energy, water, and sanitation as

4

those infrastructures and services are used in the production processes and delivery of goods

and services (Bah and Fang, 2015). However, in developing countries, the cost of trans-

portation, logistics, telecommunication, water, electricity, security, and bribes are high, and

firms suffer great losses due to the poor quality of public infrastructures and services (Eifert

et al., 2006). The latter increases transaction costs and makes firms less competitive and

productive than their international counterparts (Bah and Fang, 2015). Similarly, inefficient

institutions in developing countries create barriers to opportunities and increase costs and

risks for microenterprises as well as multinationals (World Bank, 2005; Botero et al., 2004).

Inefficient institutions and policies limit market access and increase the size of the unoffi-

cial economy (Botero et al., 2004; López de Silanes et al., 2002). Also, recent literature on

policy distortions unanimously demonstrates that ineffective public policies lower aggregate

total factor productivity (TFP) and explain an important share of TFP dispersion across

countries (Hseih and Klenow, 2009; Restuccia and Rogerson, 2013; Wu, 2018). The bene-

fits of improving the institutional quality are not limited to developing countries as Prado

(2011) shows on a sample of OECD countries that policies reducing regulation costs have a

significant positive impact on the supply of both private and publicly produced goods, and

effectively reduce the size of the informal sector. We show in this paper that distortions

stemming from the business environment and institutional quality affect firms’ tax evasion

decisions and production.

The remainder of the paper is structured as follows. Section II describes the data and

examines the relationship between distortions and firms’ tax evasion empirically. Section

III presents the theoretical model. In section IV, we calibrate the model using the United

States as a benchmark economy; we then describe our quantitative analysis and the results

of the counterfactual experiments. Section V assesses the sensitivity of the findings. The

concluding remarks and policy implications are discussed in section VI.

5

II Empirical Evidence

This section provides empirical evidence consistent with the testable predictions of the gen-

eral equilibrium model in the next section. We first provide descriptive evidence before using

a multilevel mixed model to examine the impact of distortionary infrastructures and policies

on firms’ tax evasion. The multilevel mixed model allows dealing with the hierarchic structure

of the data and takes into account the potential dependence between firms of a given coun-

try. Moreover, this methodology allows including both microeconomic and macroeconomic

variables while accounting for country fixed effects.

II.1 Data and descriptive statistics

We use the World Bank Enterprise Surveys (WBES) data which are a collection of a firm-

level surveys of a representative sample (random stratified sampling) of firms mainly in

developing countries. Questionnaires cover a wide range of business environment topics

like infrastructure, performance measures, crime, corruption, competition, access to finance.

The surveys are conducted within a framework of common guidelines in the design and

implementation. A module of identical questions included in all questionnaires is used for

assembling the data set, which allows cross-country comparisons. The analysis focuses on

African and Latin American countries having at least 200 establishments as well as tax

evasion and distortions data.3 The distributions of firms used in the structural model below

are representative at the country level. The sample comprises 19,490 firms in 30 African

and Latin American countries during the period 2002-2006.4 The appendix provides a list of3The sample also excludes countries that were involved in armed conflicts over the period of the survey.

We do so to ensure that the measure of institutional quality, as well as the costs stemming from the businessenvironment, are not tainted by armed conflicts. All conclusions of the paper remain the same without theserestrictions and using countries with at least 100 establishments.

4The dataset employed in this paper does not include informal firms due to data issues. However, weacknowledge that informality is pervasive in developing countries and might be connected with low domesticresource mobilization. As discussed by Besley and Persson (2014), the informal sector is inherently hardto tax because transactions are not recorded, and incomes from informal firms are difficult to measure.Moreover, informality is a source of misallocation (D’Erasmo and Boedo, 2012) affecting both productivityand tax collection (Ordóñez, 2014). Consequently, we expect in this paper that the estimated effects ofdistortionary infrastructures and policies provide only lower bounds for tax evasion in African and Latin

6

countries.

Tax evasion is captured by the proportion of total sales not reported for tax purposes.5

Distortions are measured as the sum of the losses (in percentage of total sales) due to power

outage or surges from the public grid, insufficient water supply, unavailable main line tele-

phone service, transport failures, crime (loss due to theft, robbery, vandalism or arson) and

gifts or informal payment to public officials to "get things done".6

Institutional quality is proxied using a measure of the effectiveness of public policies from

the World Governance Indicators (WGI). More particularly, institutional quality captures

the perceptions of the quality of public services, the quality of the civil service and the

degree of its independence from political pressures, the quality of policy formulation and

implementation, and the credibility of the government’s commitment to such policies. We

use the percentile distribution of these public policies effectiveness variables.

In the regressions below, we also account for a set of firm individual characteristics. Firms’

access to finance is measured as the share of working capital financed by commercial banks.

Regulatory burden is measured by the percentage of time the senior management spends

dealing with requirements imposed by government regulation. We include the percentage of

the firm owned by foreign interests and the government, firm’s age captured by three cate-

gorical variables: young (1-5 years old), mature (6-15 years old), and older (more than 15

years old), and the percentage of the establishment’s sales exported. Finally, we account for

the size of the firms using four dummy variables capturing firms’ size categories. Microen-

terprises have fewer than 10 permanent employees. Small and medium have between 11 and

American countries, as the paper does not account for the distortions generated by informal firms and theirimpacts on tax collection.

5The question is: "Recognizing the difficulties many enterprises face in fully complying with taxes andregulations, what percentage of total sales would you estimate the typical establishment in your area ofactivity reports for tax purposes?"

6The questions related to the components of distortionary infrastructures are stated as follows. (i) "Pleaseestimate the losses (as a percentage of total sales) of theft, robbery, vandalism or arson against your estab-lishment in the last year." (ii) "What percentage of your total sales value was lost last year due to poweroutages, insufficient water supply, unavailable mainline telephone service, and transport failures?" (iii) "Wehave heard that establishments are sometimes required to make gifts or informal payments to public officialsto "get things done" about customs, taxes, licenses, regulations, services, etc. On average, what percentageof annual sales value would such expenses cost a typical firm like yours?"

7

50 or between 51 and 200 permanent employees, respectively. Firms with more than 200

permanent employees are classified as large firms. Older firms and larger firms are used as

the reference categories in the regression analysis.

Table 1 presents the descriptive statistics. On average, about 23.84 percent of total sales

are not reported for tax purposes (tax evasion) and 4.69 percent of annual sales are lost due

to the poor quality of infrastructures, crime, and informal payment to public officials (dis-

tortionary infrastructures). Approximately, 75.77 percent of establishments’ working capital

is financed by commercial banks. On average 12.16 percent of senior management’s time is

spent dealing with requirements imposed by government regulation each week (taxes, cus-

toms, labor regulations, licensing and registration), and only 9.80 percent of establishments’

total sales are exported. Firms are on average 20 years old. Finally, the sample contains 27.2

percent of microenterprises, 43.4 percent of small firms, 20 percent of medium firms and 9.4

percent of large firms.

Figure 1 plots the correlation between tax evasion and distortionary infrastructures across

countries. Each circle represents one country. As it can be seen, there is a positive correlation

between tax evasion and the losses stemming from the business environment. Conversely,

Figure 2 presents the evidence of a negative relationship between institutional quality and

tax evasion. This evidence supports that high distortionary infrastructures and policies are

correlated with the proportions of sales not reported for tax purposes.

Table 1. Summary statistics

Variable Mean Std. Dev. Min. Max.Tax evasion 23.84 32.53 0 100Distortionary infrastructures 4.74 10.53 0 100Access to finance 75.77 31.79 0 100Management time 12.16 18.21 0 100Foreign share 10.33 28.70 0 100Government share 0.52 6.55 0 100Sales exported 11.76 24.02 0 100Age 19.95 17.88 0 196Notes. All variables are expressed in %, except age.

8

ARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARG

BOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOL

BRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRA

CHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHL

COLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOL

CRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRIECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECU

SLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLV

GTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTM

HNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKEN

MDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDG

MEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEX

MARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMAR

NICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNIC

PANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPAN

PRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRY

PERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAF

URYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURY

AGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGO

BWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWA

GINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGIN

MRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRT

NAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAM

RWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWA

SWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZ

TZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZA UGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGA0

20

40

60

Ta

x E

va

sio

n (

% o

f to

tal sa

les)

0 5 10 15 20Distortions (% of total sales)

Figure 1. Distortionary infrastructures and Tax evasion

ARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARGARG

BOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOLBOL

BRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRABRA

CHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHLCHL

COLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOLCOL

CRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRICRIECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECUECU

SLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLVSLV

GTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTMGTM

HNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDHNDKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKENKEN

MDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDGMDG

MEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEXMEX

MARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMARMAR

NICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNICNIC

PANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPANPAN

PRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRYPRY

PERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERPERZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAFZAF

URYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURYURY

AGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGOAGO

BWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWABWA

GINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGINGIN

MRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRTMRT

NAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAMNAM

RWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWARWA

SWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZSWZ

TZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZATZAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGAUGA

020

40

60

Ta

x E

va

sio

n (

% o

f to

tal sa

les)

0 .2 .4 .6 .8Policy effectiveness

Figure 2. Distortionary policies and Tax evasion

9

II.2 Methodology: Multilevel mixed model

This section examines the role of distortionary infrastructure and policies on firms’ tax eva-

sion using a multilevel mixed model. This model takes into account the fact that firms in

a given country share similar contextual characteristics such as institutional environment,

macroeconomic framework, and policies which affect their tax evasion behavior. Standard

estimation methods ignore such clustering effects which generate biased standard errors. As

firms in a given country may not be independent, standard errors in standard estimation

methods may be underestimated. The model allows the intercept to vary across countries

(Hox et al., 2010). Moreover, this methodology allows to include both microeconomic and

macroeconomic variables as well as country fixed effects. The approach considers a two-level

model where the highest level is the country, and the lowest level is the firms such that:

Level 1:

Tax_Evasionic = α0c + βDic + ηXic + γIQc + δc + µs + ηt + εic, εic ∼ N(0, σ2) (1)

Level 2:

α0c = α00 + ϑc, ϑc ∼ N(0, σ2), ϑc⊥εic (2)

Combining the two previous equations, the model can be written as follows:

Tax_Evasion_ic = α00 + βDic + γIQc + ηXic + δc + µs + ηt + ϑc + εic (3)

where, ϑc + εic is the error term of the model with ϑc the country-specific error term and εic

the firm-level error term. Tax_Evasion_ic refers to the proportion of sales not reported for

tax purposes of the firm i in the country c. Dic denotes distortions, IQc institutional quality,

and Xic defines firm individual characteristics. All these variables are the same as defined

previously. Finally, δc, µs, and ηt refer to country, sector and year fixed effects respectively.

These fixed effects control for potentially important omitted variables at the country, sector

and year level while accounting for differences in demand conditions, productive structure

and culture of opportunistic behaviors.

10

II.3 Results

Table 2 reports the findings of the estimation of the equation (3). As it can be seen from

the first row, distortions are positively related to firms’ tax evasion. These coefficients are

statistically significant at the 1 percent level and suggest that a 1 percent loss in total sales

due to distortionary infrastructures increases firms’ tax evasion from 0.11 to 0.14 percent.

Conversely, sales not reported for tax purposes decrease with institutional quality (last row).

A one-unit increase in the index of the public policies effectiveness decreases firms’ tax evasion

from 62.79 to 70.71 percentage points.

Moreover, as it can be seen in columns (3) to (8) having a higher working capital financed

by commercial banks decreases firms’ tax evasion, suggesting that access to finance reduces

the likelihood that an establishment under-reports its total sales for tax purposes. Financially

constrained establishments seem to use tax evasion as an alternative source of financing or as

a way to survive in a highly distorted environment. Columns (7) and (8) shows that larger

and medium establishments are less likely to underreport sales for tax purposes (large firms

are the omitted category).7 Finally, foreign ownership status is negatively associated with

firms’ tax evasion. The coefficients are statistically significant at the 1 and 5 percent levels.

All estimates include country, sector and year fixed effects.

Although these results corroborate similar findings in the literature, one should treat

them with caution as we do not tackle potential endogeneity issues such as omitted variable

bias, potential measurement error, and reverse causality. Moreover, the empirical approach

limits the possibility to identify the mechanisms through which the losses stemming from

the poor business environment and public policies ineffectiveness affect firms’ tax evasion.

That is why we develop a theoretical model in the next section, which allows examining two

mechanisms relating distortions, institutional quality, and tax evasion behavior.

7The findings remain the same capturing the size and age of the firms by continuous variables.

11

Table2.

Impa

ctsof

distortion

aryinfrastructuresan

dpo

licieson

firm’s

taxevasion

(1)

(2)

(3)

(4)

(5)

(6)

(7)

(8)

Dep

endent

variab

le:Ta

xevasion

Distortions

0.13

7***

0.13

2***

0.12

7***

0.122*

**0.12

2***

0.12

1***

0.10

9**

0.10

7**

(0.040

6)(0.044

5)(0.044

8)(0.045

5)(0.045

5)(0.045

7)(0.045

7)(0.045

1)Man

agem

entTim

e-0.019

3-0.019

8-0.0201

-0.020

1-0.019

0-0.015

1-0.014

6(0.024

0)(0.024

4)(0.024

3)(0.024

3)(0.024

6)(0.023

7)(0.023

7)Accessto

finan

ce-0.023

0*-0.023

5*-0.023

6*-0.023

0*-0.022

6*-0.022

1*(0.012

8)(0.012

8)(0.012

8)(0.012

9)(0.012

5)(0.012

4)Fo

reignshare

-0.045

7***

-0.0459*

**-0.041

0***

-0.026

7**

-0.027

9**

(0.014

0)(0.014

0)(0.013

0)(0.0121)

(0.012

0)Governm

entshare

-0.033

9-0.034

0-0.003

91-0.002

09(0.037

4)(0.037

1)(0.035

3)(0.035

2)Sa

lesexpo

rted

-0.030

0*-0.007

92-0.009

15(0.017

0)(0.015

7)(0.016

1)Age

-0.038

4-0.019

9(0.025

9)(0.019

3)Microenterprise

8.40

2***

7.91

7***

(1.623

)(1.586

)Sm

all

4.36

9***

4.10

3***

(1.466

)(1.414

)Medium

1.35

11.246

(1.167

)(1.145

)You

ng2.21

9(2.044

)Mature

0.69

6(0.868

)Institutiona

lQ.

-62.79

***

-70.71

***

-69.15

***

-68.11

***

-68.03

***

-69.04

***

-66.46

***

-67.08

***

(16.34

)(14.38

)(14.91

)(15.63

)(15.69

)(16.19

)(16.53

)(16.30)

Observation

s19

,490

17,250

17,100

17,089

17,089

17,068

17,068

17,068

Not

es.The

tablepresents

theestimates

oftheeff

ects

ofdistortion

aryinfrastructuresan

dpo

licieson

firms’

taxevasionusingamultilevelmixed

effects

mod

el.Allregression

sinclud

ecoun

try,

sector,a

ndyear

fixed

effects.Rob

uststan

dard

errors

clusteredat

thecoun

try-sector

levelinpa

rentheses.

***,

**,*

deno

tesign

ificanc

eat

the1,

5,an

d10

percentlevel.

12

III The model

In this section, we develop a general equilibrium model to analyze potential mechanisms

linking distortions and institutional quality to firms’ tax evasion decision. We identify two

mechanisms that can explain the relationship between these variables. First, entrepreneurs,

knowing that institutional quality is low, also anticipate low monitoring of their taxes and

have, therefore, more incentives to evade their taxes. The second mechanism is when dis-

tortionary infrastructures and policies generate losses for firms by creating a wedge between

firms’ potential and realized profits. We assume that firms compensate for some of these

losses by evading their taxes.

The economic environment consists of (i) firms which maximize their profits and are het-

erogeneous along two dimensions: productivity and the level of distortions they face due

to unsound business infrastructures and policies; (ii) one representative household which

maximizes its inter-temporal utility, and (iii) a government which balances its budget.

III.1 Representative household

Consumers are aggregated through a representative household with preferences described by

the following utility function:

∞∑t=0

βtu(ct)

where ct is consumption at date t and β ∈ (0, 1) is the discount factor. The household is

endowed with one unit of productive time in each period and K0 > 0 units of the capital

stock at date 0. We assume that u(ct) satisfies the usual Inada conditions. The representative

household maximizes its lifetime utility subject to the following budget constraint:

∞∑t=0

(ct +Kt+1 + (1− δ)Kt) =∞∑t=0

(rtKt + wtNt + Πt)

13

where, wt and rt are, respectively, the rental price of labor lt and capital Kt at t, and Nt

the total labor supply to the market. We assume that the representative consumer does not

value leisure, i.e. Nt = 1. Finally, Πt is total profits from the operations of all firms.

III.2 Firms

Entrepreneurs are risk-neutral and produce a homogenous consumption good. Each estab-

lishment i makes production decisions to maximize profits based on an idiosyncratic produc-

tivity level zi, which is constant over time and varies across establishments. Values of the

parameter zi, are drawn from a probability density function of g(z). The production function

F (zi, A, ki, li) takes as input capital ki and labor li. We include a variable A capturing insti-

tutional quality. This variable is the same for all the establishments in the same country and

captures the quality of public policies formulation and implementation as well as the quality

of public services. The production function is assumed to exhibit decreasing returns to scale

in both capital and labor, and to satisfy the usual Inada conditions:

F (zi, A, ki, li) = ziAkηi lγi , 0 < η + γ < 1.

There is a proportional tax τ on establishments’ total sales, which is assumed to have two

components, τ c and τ di . The first component τ c represents a tax collected by the government

to finance institutional quality, which can be assimilated public goods and services. The

second component τ di refers to the fraction of output which is lost due to distortionary

infrastructures. To summarize τ = τ di + τ c. We assume that establishments differ in their

idiosyncratic levels of productivity and distortions.

In each period an establishment reports a proportion (1−ϑi) of its total sales for tax purposes

based on its idiosyncratic level of distortions. Hence, each establishment evades a proportion

ϑi of its total sales. We assume there is a probability p to be detected for tax evasion. In

this case, the evading firm must pay a fine. The expected fine is assumed to be a convex

14

function of the establishment’s tax evasion so that the marginal cost of tax evasion is positive

and increasing in tax evasion: B = pϑθiF (zi, A, ki, li), with θ > 2. The expected profit for an

establishment with productivity zi is given by:

π(zi, τi) = (1− τi)(1− ϑi)F (zi, A, ki, li) + ϑiF (zi, A, ki, li)︸ ︷︷ ︸T

−wli − rki − pϑθiF (zi, A, ki, li)︸ ︷︷ ︸B

−cf (4)

where cf is a fixed cost of operation for an incumbent establishment, p is the probability of

detection, w and r are, respectively, the rental prices of labor and capital. T is the amount of

total sales evaded. We assume that this amount is beneficial for the establishment (private

benefit) but is a dead loss at the country level. By rearranging equation (4) we have

π(zi, τi) = [1− (1− ϑi)τi − pϑθi ]F (zi, A, ki, li)− wli − rki − cf (5)

For the sake of simplicity, we abandon the index i; however, it is understood that z, k ,l, τ ,

ϑ are different for each establishment in what follows.

III.3 Government

The government provides public goods and services with institutional quality A by balancing

its budget in each period:

∫ zmax

1

τ c(1− ϑi)F (zi, A, ki, li)d(τ, z) +

∫ zmax

1

pϑθiF (zi, A, ki, li)d(τ, z) ≥ C(A, p)

The government revenues appear on the left hand side (LHS) of this equation. The first

component is the revenues from proportional taxation on establishments’ total sales. The

second component is the revenues from detection activities. The right hand side (RHS) is the

cost associated with the provision of institutional quality A and detection activities p which

is assumed to be a convex cost function in both A and p.

15

III.4 Equilibrium

We consider the steady-state competitive equilibrium of the model in which the decision

problems are described as follows.

III.4.1 Consumer’s problem

Using the first order conditions, we find a solution to the consumer’s problem with the rental

price of capital r and the consumption c being constant, that is:

r =1

β− (1− δ) (6)

III.4.2 Incumbent establishment’s problem

Solving for the first order conditions, the optimal tax evasion and factor demands are:

ϑ∗ = [1

θpτ ]

1θ−1 (7)

k∗ = [zA(1− (1 +1− θθ

(1

pθτ)

1θ−1 )τ)]

11−α−γ [

γ

w]

γ1−α−γ [

α

r]

1−γ1−α−γ (8)

l∗ = [zA(1− (1 +1− θθ

(1

pθτ)

1θ−1 )τ)]

11−α−γ [

γ

w]

1−α1−α−γ [

α

r]

α1−α−γ (9)

Given that the establishment-level productivity and tax rate are constant over time, the

discounted present value of an incumbent establishment is given by :

V (z, τ) =π(z, τ)

(1− ρ)(10)

16

where ρ = 1−λ1+r−δ is the discount rate for the establishment and λ the probability of death

which is assumed to be constant. Substituting ρ in equation (10) we have

V (z, τ) =(1− δ) + r

(λ− δ) + rπ(z, τ) (11)

III.4.3 Entry

Establishments’ entry decision is made on the basis of the distribution over potential draws

for the pair (z, τ). Let x̄(z, τ) denote the optimal entry decision with the convention that the

establishment enters and remains in operation if x̄(z, τ) = 1. The actual discounted value of

a potential entrant Ve is given by:

Ve(z, τ) =∑(z,τ)

maxx̄∈[0,1]

[x̄(z, τ)V (z, τ)d(z, τ)− ce]

where, ce is the entry cost paid by a new establishment, and g(z, τ) is the probability density

function. In an equilibrium with entry, the free entry condition is fulfilled for Ve(z, τ) = 0. In

the steady state and according to equations (6) and (11), V (z, τ) is determined only by the

endogenous variable w. Hence, there is a unique value of the wage rate w for which Ve = 0

as in Restuccia and Rogerson (2008).

III.5 Invariant distribution of establishment

Let E and µ(z, τ) denote, respectively, the mass of entrants and the distribution of firms in

period t. Given the decision rule for production of entering establishments x̄(z, τ), the next

period distribution of firms µ′ over the pair (z, τ) satisfies the following condition:

µ′(z, τ) = (1− λ)µ(z, τ) + x̄(z, τ)d(z, τ)E

where (1 − λ)µ(z, τ) refers to the mass of incumbent establishments that have survived,

and x̄(z, τ)d(z, τ)E represents the mass of entering establishments that enter and remain in

17

operation. The unique invariant distribution of establishment µ̂(z, τ) is characterized by a

constant distribution of µ over time and a death rate bounded away from 0. This invariant

distribution of establishments is given by:

µ̂(z, τ) = Ex̄(z, τ)

λd(z, τ)

III.5.1 Labor market clearing

Given values for w and r, the steady state of this model is characterized by the functions

ϑ̄(z, τ), k̄(z, τ), l̄(z, τ), x̄(z, τ) and the associated invariant distribution of establishments

µ̂(z, τ). Using these functions, the aggregate labor demand and the steady-state equilibrium

level of entry are given by:

N(r, w) = E∑(z,τ)

l̄(z, τ)µ̂(z, τ)

E =N(r, w)∑

(z,τ) l̄(z, τ)µ̂(z, τ)

III.5.2 Definition of a competitive equilibrium

The steady-state competitive equilibrium with entry is a set of prices {w∗, r∗}, a set of

decision rules {k∗, l∗, ϑ∗; c∗, K ′∗}, a distribution of establishments µ(z, τ), value functions

π(z, τ), V (z, τ), Ve(z, τ), and a mass of entry E such that :

1. Given prices (w, r) and preferences, the pair {c∗, K ′∗} maximizes the consumer lifetime

utility;

2. Given prices (w, r), the functions π(z, τ), V (z, τ), and Ve(z, τ) solve incumbent and

entering establishment’s problems, with {k∗, l∗, ϑ∗} the optimal policy functions;

3. The free-entry and invariant distribution conditions are satisfied i.e.

Ve = 0,

18

µ(z, τ) = Ex̄(z, τ)

λd(z, τ), ∀ z, τ

4. Market clearing conditions are satisfied :

c+ δK =∑(z,τ)

[f(z, A, k̄, l̄)− cf ]µ(z, τ)− ceE

K =∑(z,τ)

k̄(z, τ)µ(z, τ)

1 =∑(z,τ)

l̄(z, τ)µ(z, τ)

III.6 Theoretical predictions

From the optimal tax evasion equation (7), we can draw two predictions:

Prediction 1: ∂ϑ∗

∂τd> 0, ∀ τ ∈ ]0, 1]. An increase in distortions stemming from busi-

ness environment increases firms’ tax evasion.

Prediction 2: ∂ϑ∗∂p

< 0, ∀ p ∈ ]0, 1]. A higher probability of detection reduces firms’

tax evasion.

19

IV Quantitative model

In this section, we calibrate the model using the United States economy and then simulate

it on a sample of 30 African and Latin American countries.

IV.1 Calibration

The model is calibrated on the United States, which is considered as an economy with no

distortions as is standard in the literature. Several of the parameter values are assigned

following the literature. The period in the model corresponds to one year in the data.

Preferences. The yearly interest rate is targeted to 4% as in Restuccia and Rogerson (2008)

and Bah and Fang (2015). This implies a value of β = 1(1+0.04)

= 0.96.

Technology. We follow the literature by using 0.85 as the returns to scale of the production

function.8 Parameter values η and γ are attributed to match capital and labor shares of

income, respectively, 13and 2

3of the returns to scale of the production function. We choose

δ so that the investment to output ratio is equal to 20%, implying δ = 0.08 as in Restuccia

and Rogerson (2008). The range of employment across establishments in the data determines

the range of establishment-level productivity. According to the United States data from the

Census Bureau,9 the number of employees at the establishment level ranges from 1 to 10,000.

Hence, the minimum and maximum levels of productivity are chosen to obtain the range of

employment, as in the data. Normalizing the lowest firm-level productivity to 1, the highest

level of productivity is chosen to obtain the maximum number of employees of 10,000, as in

the data. We approximate the distribution of establishment-level productivity with 100 grid

points. We choose a log-spaced grid so that the invariant distribution of establishment size

across employment level matches the data. Descriptive statistics about firms highlight one

important stylized fact. Establishments with fewer than 20 employees represent 86.1 percent

of all the establishments. However, these establishments account only for 24.9 percent of the8See for instance Restuccia and Rogerson (2008), Atkeson and Kehoe (2005), and Pavcnik (2002).9The data can be downloaded from the website of the US Census Bureau: http://www.census.gov/

econ/susb/data/susb2007.html in the table "U.S. & states, totals".

20

employment.

Distortions We use the distortions from WBES in the same way as we did in the empir-

ical section as a proxy of distortionary infrastructures. As mentioned previously, we sum

up the percentage of sales lost due to poor infrastructure and services (losses due to power

outage or surges from public grid, insufficient water supply, unavailable mainline telephone

service, transport failures); crime (losses due to theft, robbery, vandalism or arson against

the establishment); and corruption (informal payments to "get things done"). For a given

country, we compute losses for each establishment and use the distribution of distortionary

infrastructures across establishments in the simulation.

Institutional Quality We calibrate institutional quality A using the measure of public

policies effectiveness from the Worldwide Governance Indicators (WGI) as described above.

Recall that this variable is defined as perceptions of the quality of public services, the quality

of the civil service and the degree of its independence from political pressures, the quality of

policy formulation and implementation, and the credibility of the government’s commitment

to such policies. We use the percentile distribution of policy effectiveness variable ranged

from 0 to 1.10 In the calibration, we approximate the detection probability using this in-

dex of public policies effectiveness. We argue that effective government in formulating and

implementing public policies will also be effective in detecting illegal activities such as tax

evasion. In the sensitivity analysis, we relax this assumption using an alternative measure

of detection probability and show that the predictions of the paper remain the same and are

not driven by this assumption.

In the model, the cost of tax evasion for an establishment is assumed to be a convex function,

with a convexity parameter θ > 2. This assumption guarantees that the marginal cost of

tax evasion is positive and increasing with the share of sales not reported for tax purposes.

The convexity parameter θ is calibrated to 3. Sensitivity analysis shows that the findings are10To the best of our knowledge, the proportional tax rate on sales is not publicly available for each country

in the sample. Calibrating institutional quality using the data allows us minimizing errors arising from a notaccurate calibration of the proportional tax rate on sales τ c for each country. The proportional tax rate onsales τ c is therefore normalized to 0 in the quantitative model.

21

robust using alternative values of θ. Table 3 summarizes the parameter values.

Table 3. Benchmark calibration

Parameter Value Description Sourceτ d [0; 1] Distortionary infrastructures WBESA [0; 1] Public policies Effectiveness WGIp [0; 1] Detection probability WGIβ 0.96 Real rate of return Literatureη 0.283 Capital income share Literatureγ 0.567 Labor income share Literatureδ 0.08 Investment to output ratio Literatureλ 0.1 Annual exit rate Literatureθ 3 Evasion cost parameter Assumptionce 1 Entry costs Literaturecf 0 Fixed costs Literaturez [1; 3.98] Distribution of productivity Relative labor demandNotes. See Bah and Fang (2015), Restuccia and Rogerson (2008), Atkeson and Kehoe (2005)

and Pavcnik (2002) for parameter values from the literature.

22

IV.2 Quantitative analysis

The quantitative analysis consists in validating the model and conducting counterfactual ex-

periments to assess the implications of changes in the distortions stemming from the business

environment and the detection probability.

IV.2.1 Validation of the model

We simulate the model for 30 African and Latin American countries using the calibrated

parameters described above. For each country, the model predicts the level of tax evasion

and output per worker. Figure 3 plots GDP per worker from the model against GDP per

worker from the World Bank’s World Development Indicators (WDI). The reported value of

the GDP per worker is normalized by the United States levels in both the model and the

data. Each circle represents one country and corresponds to the correlation of output per

worker between the simulated data and the WDI data. The straight line is obtained from

an OLS regression between the model and the data. As it can be seen, the predicted values

of output per worker in the model are positively correlated with the data. The coefficient

is statistically significant at the 1% level, and the regression coefficient is 4.58. Similarly,

Figure 4 plots the level of tax evasion from the model and the data. The predicted values of

tax evasion are highly correlated with the WBES data. The regression coefficient is 0.92 and

is statistically significant at the 1% level. Focusing on the R-squared, the model explains

49% of the variation of the GDP per worker and 35% of the dispersion of tax evasion among

African and Latin American countries.

23

ARG

BOL

BRA

CHL

COL CRI

ECU

EGY

SLVGTM

HND

KEN

MDG

MEX

MAR

NIC

PAN

PRY

PER

ZAF

URY

AGO

BWA

GIN

MRT

NAM

RWA

SWZ

TZAUGA

0.1

.2.3

.4.5

Mo

de

l

0 .02 .04 .06 .08Data

R−squared=.4931813103211073

Figure 3. GDP per worker - data vs the model predictions

ARG

BOL

BRA

CHL

COL

CRIECU

EGYSLV

GTM

HNDKEN

MDG

MEX

MAR

NIC

PAN

PRY

PERZAF

URY

AGO

BWA

GIN

MRT

NAM

RWA

SWZ

TZA UGA

0.2

.4.6

Mo

de

l

.1 .2 .3 .4 .5Data

R−squared=.3505665524954171

Figure 4. Tax evasion - data vs the model predictions

24

IV.2.2 Counterfactual experiments

Having established the ability of the model to replicate some of the variations of tax evasion

and output per worker across African and Latin American countries, we conduct in this sec-

tion three tax neutral counterfactual experiments to examine the implications of changes in

the distortions stemming from the business environment and the detection probability. Tax

neutral experiments guarantee that the tax revenues remain the same as in the benchmark.

In those experiments, the tax revenues from the decrease in tax evasion are used as a subsidy

to production, i.e., the price of the final good is now a function of this subsidy.

The counterfactual experiments focus on Guinea, the country with the highest level of tax

evasion in the data. We first examine what happens regarding tax evasion, ceteris paribus,

when the losses stemming from the business environment in Guinea are reduced to the aver-

age level observed in the sample. In particular, we replace the distribution of losses stemming

from business environment across size in Guinea by the average distribution of losses across

firms’ size from the sample. All the parameters of the model remain the same as previously,

except the losses stemming from the business environment. In the second experiment, we

explore the second mechanism by examining the impact of a change in Guinea’s deterrence

probability. In particular, we increase the deterrence probability by 50 percent in Guinea. All

other parameters remain the same. Finally, we combine in a third experiment both changes,

i.e., a distribution of distortions similar to the average level of the sample and a detection

probability 50% larger. Table 5 reports the change on key variables relative to the baseline

findings on Guinea.

Experiment 1. In the first scenario, we conduct a tax neutral experiment analysis and

set the distribution of losses stemming from business environment to the average level by size

category in the sample. As described in Table 4, the distribution of losses by size in Guinea

is between 14.97 and 20.60 percent of sales, whereas in the sample the latter is between 5.18

and 7.36 percent of their sales. On average, counterfactual experiment 1 corresponds to a

25

65.25 percent drop in the losses stemming from the business environment in Guinea.

Table 4. Distribution of losses in Guinea vs. the sample

Establishment size Average losses Average losses(number of employees) in Guinea in the sample< 10 19.59 7.3610 to 19 19.24 7.0020 to 49 14.97 6.3950 to 99 20.60 5.49≥ 100 17.00 5.18

The results show that a drop to the average level by size category generates a 26.72 percent

drop in sales not reported for tax purposes in Guinea. Guinean firms’ sales not reported

for tax purposes decline from 53.74 percent to 39.38 percent. As discussed above, losses

stemming from the business environment are a proportional tax on firms’ sales and create

therefore a wedge between potential and realized profits. In such environment, a drop in

additional costs stemming from business environment decreases sales not reported for tax

purposes as witnessed in experiment 1. Using the proceeds from increased tax revenues to

subsidize production, a drop in distortionary infrastructures in Guinea to the average level

of the sample generates a 1.06-fold increase in output per worker. The aggregate TFP is

one of the channels through which the output per worker is stimulated. As Table 5 shows,

the aggregate TFP increases slightly by 0.16 percent. Also, a drop in distortions stemming

from the business environment stimulates the entry of firms in the market and the aggregate

capital as both increase by 22 percent.

This experiment seems to benefit more to medium and large firms, which represent firms

having between 20-100 and more than 100 employees. The share of firms in both groups

increases by about 8.23 and 2.12 percent respectively while the share of small firms in the

distribution of establishments decreases by 0.98 percent. Moreover, this experiment gener-

ates a scale effect as the average size of firms increases by 21.52 percent. This is a direct

consequence of the increase in the share of medium and large firms in the distribution of

establishments. Finally, this experiment implies a reallocation of the labor force towards

26

large firms. The share of total employment in large firms grows by around 3 percent, while

both medium and small firms face a drop in their employment share.

Experiment 2. In the second counterfactual experiment, we examine what happens in

Guinea when the detection probability can be improved by 50 percent while the distribution

of losses stemming from the business environment is left untouched. When the detection

probability is increased from 0.0634 to 0.0951, sales not reported for tax purposes decline

from 53.73 percent to 43.88 percent, corresponding to a 18.34 percent drop in firms’ tax eva-

sion in Guinea. The improvement of the institutional quality raises the expected costs related

to tax evasion activities as monitoring and auditing increase. Consequently, the incentive for

firms to underreport their sales for tax purposes declines.

In this tax neutral experiment, the output per worker is nearly 6.33-fold as large as the

initial output per worker of the country.11 Compared to the baseline findings on Guinea,

a 50 percent improvement in the institutional quality causes the aggregate TFP to rise by

208 percent and boosts the entry of firms by 173 percent. Similarly, the aggregate capital

increases by 173 percent. These important changes in aggregate TFP and the entry of firms

might explain the substantive growth in output per worker. This experiment facilitates the

emergence of medium firms as their share in the distribution of firms increases by 8.71 percent

while the share of small firms decreases by 0.98 percent for a constant share of large firms.

The evidence is consistent with the literature on the missing middle in the size distribution

of firms in developing countries, where the quality of the business environment is identified as11Although the change in the output per worker might appear high, they are in line with similar evidence in

the literature. IMF (2003) shows for instance that improving the quality of institutions in Cameroon (-0.72)to the average level of institutions in the sample (0.13) generates almost a 5-fold increase in income per capitain Cameroon. Rodrik et al. (2004) find that one standard deviation increase in institutional quality generatesa 6.4-fold difference in income per capita. Similarly, Acemoglu et al. (2001) show that improving the qualityof institutions in Nigeria to the level of Chile could lead to a 7-fold increase in Nigeria’s income. Also, Halland Jones (1999) show in a cross-country regression that difference in institutions between Niger and theUnited States which is only 3.78-fold is more than enough to explain the 35-fold difference in output perworker. The contribution of Hall and Jones (1999), one of the most important in the literature, highlightshow a small change in institutions can generate an exponential variation in output per worker.

27

a critical factor explaining the limited number of medium-sized firms.12 The counterfactual

experiment shows that improving the quality of business environment favors the distribution

of firms toward medium-sized firms. Moreover, the experiment shows no scale effect. The

average size of firms decreases slightly by 4.69 percent. Finally, the improvement in the

institutional quality generates a reallocation of the labor force towards medium firms, which

account for an 8.86 percent increase in employment share.

Experiment 3. Finally, we combine both changes, i.e., a reduction in distortions by 65.25%

on average and an increase in detection by 50% and observe a decrease in tax evasion by

40.16 percent, i.e., sales not reported for tax purposes declines from 53.74 percent to 32.16

percent, while the output per worker is now nearly 6.26-fold larger than its initial level. As

it can be seen, the combination of a drop in losses stemming from the business environment

with an improvement of the institutional quality generates a more substantive decline in

sales not reported for tax purposes relative to previous experiments. In this experiment,

both changes contribute to improving the overall business environment. Qualitatively the

channels are similar to those described above: as distortions in infrastructures drop and in-

stitutional quality increases, the aggregate TFP, the mass of firms entering on the market,

and the aggregate capital increase by 185.86 and 200.22 percent respectively. Moreover, we

observe a reallocation of firms towards medium and large firms and a scale effect. As Table 5

shows, the shares of medium and large firms increase 8.23 and 2.12 percent respectively. The

average size of firms grows by 20.99 percent suggesting thereby a scale effect. Finally, this

experiment generates a reallocation of the labor force towards large firms. The total employ-

ment share of large firms increases by about 3 percent, while the employment shares of small

and medium firms decrease by 16.43 and 6.53 percent respectively.

12See for instance Sleuwaegen and Goedhuys (2002) and Tybout (2000) for extensive discussion on themissing middle in developing countries.

28

Table 5. Effects from counterfactual experiments

Counterfactual experiments

Variables Experiment 1 Experiment 2 Experiment 3Tax evasion -26.72 -18.34 -40.16Output per worker 1.06 6.33 6.26Aggregate TFP 0.16 208.07 184.86Aggregate capital 22.92 173.24 200.22Mass of entry 22.92 173.24 200.22Share of small firms -0.98 -0.98 -0.98Share of medium firms 8.23 8.71 8.23Share of large firms 2.12 0 2.12Employment share - Small -16.75 - 0.09 -16.43Employment share - Medium -6.94 8.86 -6.53Employment share - Large 3.04 -1.04 2.95Average size of firms 21.52 -4.69 20.99Notes. Except for the change in output per worker that is reported in fold, all effects arereported in percentage change. Here small, medium and large firms refer to firms having

between 1-19, 20-99, and 100 and more employees.

In summary, the counterfactual experiments show that a drop in losses stemming from the

business environment along with an improvement in detection probability in Guinea could

decrease sales not reported for tax purposes by between 17.78 and 32.86 percent. The drop in

tax evasion combined with the improvement in institutional quality generate between a 1.06-

and 6.33-fold increase output per worker in Guinea through their effects on the aggregate

TFP, the mass entering on the market, and the reallocation of firms and employment.

29

V Sensitivity analysis

This section assesses the sensitivity of the findings simulating the model for each region,

employing alternative detection probability and a recalibrating the model on the Chilean

economy.

V.1 Simulations for each region

So far, the model shows the role of distortionary infrastructures and policies in explaining the

variation in tax evasion and output per worker across African and Latin American countries.

In this section, we simulate the model for the two groups of countries separately. The intuition

of this exercise is twofold. First, we evaluate how the model fares in explaining the variation

of tax evasion and GDP per worker within each region. Second, we show that compositional

effects across the continents do not drive the findings of the paper. The simulation procedure

is the same as previously, except that we are now focusing on each region. Figure 5 plots

tax evasion and output per worker from the model against the data for African countries.

As previously, each circle represents one country and the straight line is obtained from the

OLS regression between the model and the data. The simulated data are highly correlated

with the actual data, with a regression coefficient of around 0.88 for tax evasion and 5.58

for GDP per worker. Based on R-squared values, the model explains 31% of the variation

of tax evasion and 43% of the dispersion in GDP per worker across African countries. Sim-

ilarly, Figure 6 plots tax evasion and output per worker from the model against the data

for Latin American countries. For this group of countries, the regression coefficient is 0.77

for tax evasion and 3.98 for GDP per worker. Using the R-squared, the model explains 23%

of the dispersion of tax evasion and 71% of the variation of the output per worker in Latin

American countries. All regression coefficients are statistically significant. These findings are

comparable to those obtained in the overall sample.

30

EGY

KEN

MDGMAR

ZAF

AGO

BWA

GIN

MRT

NAM

RWA

SWZ

TZA UGA

0.2

.4.6

Model

.1 .2 .3 .4 .5Data

R−squared=.3079162075672508

(a) Tax evasion

EGY

KEN

MDG

MAR

ZAF

AGO

BWA

GIN

MRT

NAM

RWA

SWZ

TZAUGA

0.1

.2.3

.4M

odel

0 .01 .02 .03 .04Data

R−squared=.4296512555503219

(b) GDP per worker

Figure 5. Tax evasion and GDP per worker - data vs the model predictions across Africancountries

ARG

BOL

BRA

CHL

COL

CRI

ECU

SLV

GTM

HND

MEX

NIC

PAN

PRY

PER

URY

.1.2

.3.4

Model

.05 .1 .15 .2 .25Data

R−squared=.2279314032419751

(a) Tax evasion

ARG

BOL

BRA

CHL

COLCRI

ECU SLVGTM

HND

MEX

NIC

PAN

PRY

PER

URY

VEN

.1.2

.3.4

.5M

odel

0 .02 .04 .06 .08Data

R−squared=.4735910912144733

(b) GDP per worker

Figure 6. Tax evasion and GDP per worker - data vs the model predictions across LatinAmerican countries

31

V.2 Alternative detection probability

In the baseline calibration, we use the same variable to capture both institutional quality and

the detection probability. As argued previously, the level of institutional quality may send an

equivalent signal of the level of monitoring and auditing, i.e., the detection probability. How-

ever, this assumption can raise questions as the latter may drive our findings. In this section,

we check the sensitivity of the results using the perception of the control of corruption as an

alternative measure of the detection probability. The latter is from the World Governance

Indicators’ database and is defined as the perceptions of the extent to which public power

is exercised for private gain, including both petty and grand forms of corruption, as well as

"capture" of the state by elites and private interests. We use the percentile ranking of the

index which is ranged between 0 and 1. The simulation procedure is the same as previously,

and the model is calibrated to the United States economy. Figure 7 plots tax evasion and

output per worker from the model against the data for African and Latin American countries.

As previously, the model’s prediction is strongly correlated with the data. The model still

explains around 20% of the variation in tax evasion and 49% of the dispersion of GDP per

worker across African and Latin American countries.

32

ARG

BOL

BRA

CHL

COL

CRIECU

EGYSLV

GTM

HNDKEN

MDG

MEX

MAR

NIC

PAN

PRY

PERZAF

URY

AGO

BWA

GIN

MRT

NAM

RWA

SWZ

TZA UGA

0.2

.4.6

Mo

de

l

.1 .2 .3 .4Data

R−squared=.1977976014267405

(a) Tax evasion

ARG

BOL

BRA

CHL

COL CRI

ECU

EGY

SLVGTM

HND

KEN

MDG

MEX

MAR

NIC

PAN

PRY

PER

ZAF

URY

AGO

BWA

GIN

MRT

NAM

RWA

SWZ

TZAUGA

0.1

.2.3

.4.5

Mo

de

l

0 .02 .04 .06 .08Data

R−squared=.4941246242786512

(b) GDP per worker

Figure 7. Tax evasion and GDP per worker - Data vs the model predictions - Alternativedetection probability

Counterfactual experiments. With the calibrated model, we first conduct three tax

neutral counterfactual experiments on Guinea as in the quantitative analysis before exploring

two additional counterfactual analyses. In the first round of experiments, we examine what

happens in Guinea regarding tax evasion if the level of distortionary infrastructures and

deterrence probability can be improved to the average level of the sample and by 50 percent

respectively. These changes lead to a decline in sales not reported for tax purposes by

26.72 percent and 18.34 percent if the level of distortionary infrastructures and deterrence

probability can be improved to the average level of the sample and by 50 percent respectively.

Regarding the output per worker, it grows by 1.22- and 6.46-fold respectively. The third

experiment combines both changes in the level of distortionary infrastructures and deterrence

probability and leads to a 40.16 percent decline in sales not reported for tax purposes, while

the output per worker grows by 6.62-fold. These counterfactual experiments provide the same

conclusion as in the quantitative analysis, where an improvement in the level of distortionary

33

infrastructures and institutional quality leads to a decline in tax evasion between 18.34 and

40.16 percent.

We also conduct two additional counterfactual experiments that are consistent with the

previous findings. Firstly, using the current level of control of corruption as a proxy of the

detection probability, we conduct a counterfactual experiment that examines what happens

in Africa and Latin America if the losses stemming from the business environment are cut

down by half. The deterrence probability is proxied by the index of the control of corruption,

while the other parameters remain the same, except the level of distortionary infrastruc-

tures. The latter is assumed to be cut down by half in each country. This counterfactual

experiment shows that cutting down losses stemming from the poor business environment by

half, governments could reduce firms’ tax evasion by 20.62% on average in African and Latin

American countries. Firms’ tax evasion declines from 17.84% to 14.16%.

Secondly, we use so far the difference in both the detection probability and the level of distor-

tionary infrastructures to explain 35% of the variation of tax evasion across African and Latin

American countries. Using such approach, we are unable to distinguish between the effects of

deterrence policies and distortionary infrastructures on tax evasion, while the distinction can

be highly relevant regarding policy implications. Alternatively, we conduct a counterfactual

experiment consisting of setting the same level of detection probability for all the countries.

The latter corresponds to the average level of the detection probability in the sample. We

further examine how much the difference in losses stemming from the business environment

can explain the variation of tax evasion across African and Latin American countries. The

calibration and simulation procedures are the same as previously, except that the detection

probability is the same for all the countries. Figure 8 plots tax evasion and output per worker

from the model against the data for African and Latin America countries. As it can be seen,

the model’s predicted value remains highly correlated with the data, and the model explains

14% of the variation of tax evasion across African and Latin America countries. Also, the

model explains 50% of the variation of GDP per worker across countries.

34

ARG

BOL

BRA

CHL

COL

CRIECU

EGYSLV

GTM

HNDKEN

MDG

MEX

MAR

NIC

PAN

PRY

PERZAF

URY

AGO

BWA

GIN

MRT

NAM

RWA

SWZ

TZA UGA

0.2

.4.6

Mod

el

.1 .15 .2 .25Data

R−squared=.1407139537033489

(a) Tax evasion

ARG

BOL

BRA

CHL

COL CRI

ECU

EGY

SLVGTM

HND

KEN

MDG

MEX

MAR

NIC

PAN

PRY

PER

ZAF

URY

AGO

BWA

GIN

MRT

NAM

RWA

SWZ

TZAUGA

0.1

.2.3

.4.5

Mod

el

0 .02 .04 .06 .08Data

R−squared=.4995535949815552

(b) GDP per worker

Figure 8. Tax evasion and GDP per worker - data vs the model predictions using the samedetection probability for all the countries

V.3 Alternative calibration

The analysis so far relied on the model that is calibrated to the United States’ economy. How-

ever, the latter can raise criticism the parameters for the United States and those for African

and Latina American economies might be different. As a sensitivity check, we recalibrate the

model to the Chilean economy, the country with the lowest combination of distortions and

tax evasion in our data and one of the largest sample size.

The parameter value for the labor share in Chile is set to the average share of labor income

in the GDP from the ILOSTAT.13 We set labor share to be 0.414 which is the average labor’s

share in the GDP.14 As in the baseline calibration, we assume a decreasing return in the

firm-level production function. The extent of the returns is obtained under the assumption13The ILOSTAT is the world leading data source on labor statistics from the International Labour Orga-

nization (ILO). The data can be downloaded from the following website: http://www.ilo.org/ilostat.14The average labor’s share of the GDP is available for 2009, 2011, and 2013.

35

that the average share of labor income in the GDP corresponds to 23of the returns to scale.

The returns to scale obtained is 0.62, very close to the calibration in Gourio and Kashyap

(2007) of the returns to scale for Chile.15 We set the interest rate to be 17.6 percent which

is the average discount rate for African and Latin American economies from the IMF Inter-

national Financial Statistics (IFS) dataset. Using this information, the parameter β is set to

0.850. We choose the depreciation rate of capital to be δ = 0.06 which is the calibrated value

from Gourio and Kashyap (2007). Since there are no studies on the exit rate parameter in

developing countries, we set the annual exit rate to the same value as previously, i.e., λ = 0.1.

The other parameters of the model remain the same as previously. Table 6 below summarizes

the parameters in the alternative calibration.

Table 6. Alternative calibration

Parameter Value Description Sourceτ d [0; 1] Distortionary infrastructures WBESA [0; 1] Public policies effectiveness WGIp [0; 1] Detection probability WGIβ 0.850 Real rate of return IMF International

Financial Statistics (IFS)η 0.207 Capital income share ILOSTAT and Literatureγ 0.414 Labor income share ILOSTATδ 0.06 Investment to output ratio Gourio and Kashyap (2007)λ 0.1 Annual exit rate Literatureθ 3 Evasion cost parameter Assumptionce 1 Entry costs Literaturecf 0 Fixed costs Literaturez [1; 3.98] Distribution of productivity Relative labor demand

Notes. See Restuccia and Rogerson (2008) for parameter values from the literature.

15Gourio and Kashyap (2007) calibrated the returns to scale to 0.6 in Chile. Our findings remain the sameusing this returns to scale and 1

3 and 23 as the share of capital and labor respectively.

36

Figure 9 plots tax evasion and output per worker from the model against the data for

African and Latin American countries. Here, the output per worker is normalized by the

output per worker of Chile. As previously, the data on GDP per worker are from the World

Bank’s World Development Indicators. The reported value of the GDP per worker is nor-

malized by the levels in Chile in both the model and the data. As previously, each circle

represents one country and corresponds to a combination of the output per worker from both

the model and the data. As it can be seen, the model’s predicted values are highly corre-

lated with the data, with a regression coefficient of around 0.92 for tax evasion and 0.20 for

GDP per worker. These coefficients are statistically significant at the 1 and 10 percent level

respectively. Using Chile as a benchmark economy, the model explains 35% of the variation

of tax evasion and around 10% of the dispersion in GDP per worker across African and Latin

American countries. These findings are comparable to those obtained from the calibration

to the United-States, where the model explains 35% of the variation of tax evasion and 49%

of the variation of GDP per worker across African and Latin American countries.

With the calibrated model on the Chile economy, we conduct three tax neutral counterfac-

tual experiments using Guinea as in the quantitative analysis. Especially, we examine what

happens in terms of tax evasion if both the level of distortionary infrastructures and policies

can be improved to the average level of the sample and by 50 percent respectively. The

approach consists of examining the effects of the change separately in experiments 1 and 2

before combining them in experiment 3. Experiments 1 and 2 lead respectively to a decline

in sales not reported for tax purposes by 26.72 and 18.34 percent. In these experiments, the

output per worker increases up to 13.20-fold respectively. The third experiment leads to a

decrease in firms’ tax evasion by 40.16 percent, while the output per worker is nearly 5-fold

as large as in the benchmark situation of the country. The counterfactual experiments are

in line with the findings from the calibration to the U.S. economy.

37

ARG

BOL

BRA

CHL

COL

CRIECU

EGYSLV

GTM

HNDKEN

MDG

MEX

MAR

NIC

PAN

PRY

PERZAF

URY

AGO

BWA

GIN

MRT

NAM

RWA

SWZ

TZA UGA

0.2

.4.6

Mo

de

l

.1 .2 .3 .4 .5Data

R−squared=.3505665524954171

(a) Tax evasion

ARG

BOL

BRA

CHL

COL CRI

ECU

EGY

SLVGTM

HND

KEN

MDG

MEX

MAR

NIC

PAN

PRY

PER

ZAF

URY

AGO

BWA

GIN

MRT

NAM

RWA

SWZ

TZAUGA

0.2

.4.6

.81

Mo

de

l

0 .5 1 1.5 2Data

R−squared=.1007588272129979

(b) GDP per worker

Figure 9. Tax evasion and GDP per worker - Data vs the model predictions across - Alter-native calibration

38

VI Conclusion

Firms in developing countries face considerable constraints in their development due to dis-

tortionary infrastructures and policies, and weak institutions. In parallel, tax evasion is

widespread and pervasive in these countries, generating a substantial negative effect on do-

mestic resource mobilization as well as on government’s ability to provide infrastructure and

public goods. African and Latin American countries rank poorly in terms of ease of doing

business, with around one-quarter of annual sales not reported for tax purposes. As a novelty,

this paper examines empirically and theoretically the impacts of distortionary infrastructures

and institutional quality on firms’ tax evasion in African and Latin American countries.

The paper sheds light on two potential mechanisms driving firms’ tax evasion which have

not been given a lot of attention in the literature. First, entrepreneurs, knowing that govern-

ment effectiveness is weak, also anticipate low monitoring and have, therefore, more incentive

to evade their taxes. In the second mechanism, distortions and poor institutional quality gen-

erate losses for firms by creating a wedge between potential and realized profits. Tax evasion

is a way for firms to counterbalance the negative effects of distortionary infrastructures and

policies. This paper provides supporting evidence of the negative impact of distortionary

infrastructures and policies on firms’ tax evasion. In particular, using WBES data, we show

that losses stemming from the business environment are positively correlated with firms’ tax

evasion. Using WGI data, we show that institutional quality is negatively related to the

proportion of sales not reported for tax purposes.

To understand the mechanisms through which distortionary infrastructures and policies

affect firms’ tax evasion, we develop a general equilibrium model with heterogeneous firms.

The model is calibrated using the United States as a benchmark economy. We simulate tax

evasion and output data for firms in a sample of 30 African and Latin American countries.

The simulated output per worker and tax evasion are strongly correlated with the data, as

the model explains 35 percent of the variation in tax evasion and 49 percent of the dispersion

in output per worker across African and Latin American countries.

39

Having established the ability of the model to replicate some of the variations of tax eva-

sion across countries, we conduct counterfactual experiments to examine the implications of

changes in distortionary infrastructures and policies. The counterfactual experiments show

that African and Latin American countries could reduce tax evasion by about 21 percent by

cutting down the losses stemming from the business environment by half. Finally, setting

the same detection probability for all the countries, the last counterfactual experiment shows

that distortionary infrastructures explain 14 percent of the variation of tax evasion across

African and Latin American countries.

The paper highlights that distortionary infrastructures and policies can generate a vicious

circle of underdevelopment in which tax evasion and distortions interact positively. Firms’

tax evasion shrinks public revenues and this, in turn, reduces the government’s capacity

to curtail tax evasion, invest in productive infrastructures, and thereby reduce distortions

in the business environment. Hence, improving the quality of the business environment in

Africa and Latin America would generate a double-dividend. On the one hand, it would

boost firms’ productivity through better public infrastructure, better access to finance, and

a lighter regulatory environment. On the other hand, improving the quality of the business

environment could reduce firms’ tax evasion. African and Latin American countries could

generate additional domestic revenues and boost the aggregate productivity simultaneously

by pushing the agenda of reforms that help to improve the quality of infrastructures and

policies related to the business environment. The impacts of distortionary infrastructures

and policies examined in this paper might represent lower bounds for tax evasion at the

country level as the standardized dataset employed does not include informal firms. Those

firms are widespread in African and Latin American countries, are inherently hard to tax,

and might represent an additional source of distortion and resource misallocation.

40

References

Acemoglu, D., Johnson, S., and Robinson, J. (2001). The colonial origins of comparative

development: An empirical investigation. American Economic Review, 91(5):1369–1401.

Amaral, P. and Quintin, E. (2010). Limited enforcement, financial intermediation, and

economic development: A quantitative assessment. International Economic Review,

51(3):785–811.

Aterido, R., Hallward-Driemeier, M., and Pagés, C. (2011). Big constraints to small firms’

growth? business environment and employment growth across firms. Economic Develop-

ment and Cultural Change, 59(3):609–647.

Atkeson, A. and Kehoe, P. J. (2005). Modeling and measuring organization capital. Journal

of Political Economy, 113(5):1026–1053.

Bah, E.-H. and Fang, L. (2015). Impact of the business environment on output and produc-

tivity in africa. Journal of Development Economics, 114(C):159–171.

Besley, T. and Burgess, R. (2004). Can labor regulation hinder economic performance?

evidence from india. The Quarterly Journal of Economics, 119(1):91–134.

Besley, T. and Persson, T. (2014). Why do developing countries tax so little? Journal of

Economic Perspectives, 28(4):99–120.

Botero, J. C., Djankov, S., La-Porta, R., de Silanes, F. L., and Shleifer, A. (2004). The

regulation of labor. Quarterly Journal of Economics, 119(4):1339–1382.

Buera, F. J., Moll, B., and Shin, Y. (2013). Financial frictions and the persistence of history:

A quantitative exploration. Journal of Political Economy, 121(2):221–272.

D’Erasmo, P. and Boedo, H. (2012). Financial structure, informality and development.

Journal of Monetary Economics, 59(3):286–302.

41

Easterly, W. and Rebelo, S. (1993). Fiscal policy and economic growth. Journal of Monetary

Economics, 32(3):417–458.

Eifert, B., Gelb, A., and Ramachandran, V. (2006). Business environment and comparative

advantage in africa: Evidence from the investment climate data. In Annual World Bank

Conference on Development Economics (ABCDE) 2006: Growth and Integration, pages

195–233. The World Bank.

Gourio, F. and Kashyap, A. (2007). Investment spikes: New facts and a general equilibrium

exploration. Journal of Monetary Economics, 54(1):1–22.

Goyette, J. and Gallipoli, G. (2015). Distortions, efficiency and the size distribution of firms.

Journal of Macroeconomics, 45:202–221.

Hall, R. E. and Jones, C. I. (1999). Why do some countries produce so much more output

per worker than others? The Quarterly Journal of Economics, 114(1):83–116.

Hox, J. J., Moerbeek, M., and van de Schoot, R. (2010). Multilevel analysis: Techniques and

Applications. Routledge.

Hseih, C.-T. and Klenow, P. (2009). Misallocation and manufacturing tfp in china and india.

The Quarterly Journal of Economics, 124(4):1403–1448.

IMF (2003). World Economic Outlook, April 2003 : Growth and Institutions. IMF.

López de Silanes, F., Djankov, S., La Porta, R., and Shleifer, A. (2002). The regulation of

entry. Quarterly Journal of Economics, 117(1):1–37.

Ordóñez, J. (2014). Tax collection, the informal sector, and productivity. Review of Economic

Dynamics, 17(2):262–286.

Pavcnik, N. (2002). Trade liberalization, exit, and productivity improvements: Evidence

from chilean plants. The Review of Economic Studies, 69(1):245–276.

42

Prado, M. (2011). Government policy in the formal and informal sectors. European Economic

Review, 55(8):1120–1136.

Restuccia, D. and Rogerson, R. (2008). Policy distortions and aggregate productivity with

heterogeneous establishments. Review of Economic Dynamics, 11(4):707–720.

Restuccia, D. and Rogerson, R. (2013). Misallocation and productivity. Review of Economic

Dynamics, 16(1):1–10.

Rodrik, D., Subramanian, A., and Trebbi, F. (2004). Institutions rule: the primacy of

institutions over geography and integration in economic development. Journal of Economic

Growth, 9(2):131–165.

Sleuwaegen, L. and Goedhuys, M. (2002). Growth of firms in developing countries, evidence

from cote d’ivoire. Journal of development Economics, 68(1):117–135.

Tybout, J. (2000). Manufacturing firms in developing countries: How well do they do, and

why? Journal of Economic Literature, 38(1):11–44.

World Bank (2005). World Development Report 2005 : A better Investment Climate for

Everyone. Oxford University Press.

Wu, G. L. (2018). Capital misallocation in china: Financial frictions or policy distortions?

Journal of Development Economics, 130(1):203–223.

43

Appendices

A Investment climate indicators

The World Bank’s Doing Business index measures aspects of regulation that enable or pre-

vent private sector businesses from starting, operating and growing. These regulations are

measured using 11 indicator sets: opening a business, dealing with construction permits,

getting electricity, registering property, getting credit, protecting minority investors, paying

taxes, trading across borders, enforcing contracts, resolving insolvency and labor market reg-

ulation. The annual report provides a ranking at the country level in term of the ease of

doing business. In the last Doing Business Report in 2017, the average regional rankings of

African and Latin American countries were 143 and 107,16 respectively, constraining firms’

creation and the long-term performance of incumbent firms.

Moreover, comparing different indicators of investment climate variables from the World

Bank Enterprise Surveys (WBES) for the high-income countries of the Organisation for Eco-

nomic Co-operation and Development (OECD) and African and Latin American countries,

it can be seen that firms in these countries face a poor investment climate compared to their

counterparts in OECD countries (cf. Table A.1 below). For example, respectively, 22.4 per-

cent and 19.3 percent of firms in African and Latin American countries identify corruption

as a major or severe obstacle to business, while the corresponding number is 7.2 percent in

the OECD countries. Similarly, the average percentile rank of policy effectiveness is 45 in

African and Latin American countries against 85 in the OECD sample. The high-income

OECD countries include Germany, Greece, Ireland, Portugal, The Republic of South Korea

and Spain.

16See the 2017 Doing Business Report: http://www.doingbusiness.org/~/media/WBG/DoingBusiness/Documents/Annual-Reports/English/DB17-Report.pdf

44

Table A.1. African and Latin American countries vs OECD counterparts

AFR LAC OECDElectricity as an obstacle to business (% of firms) 17 9.3 5.2Costs of power outages (%) 7 4 2.3Corruption as an obstacle to business (% of firms) 22.4 19.3 7.2Costs of corruption (% of sales) 2.7 1.9 0.27Crime as an obstacle to business (% of firms) 14 17 6Costs of crime (% of sales) 1.5 1 0.42Regulation (% of time spent) 8.60 13.9 1.69Tax evasion (%) 28.29 22.51 6.80Institutional Quality 39 48 85Notes. An investment climate variable is considered as an obstacle to businessif a firm states that the latter is a major or a very severe obstacle to business.

45

B List of countries

Table B.1. Business environment in Africa and Latin American countries

Country Year Sample size Tax evasion (%) Distortions (%) InstitutionalQuality

Angola 2006 409 50.95 7.10 0.06Argentina 2006 913 17.21 1.76 0.55Bolivia 2006 544 20.34 3.72 0.29Botswana 2006 309 47.80 3.14 0.70Brazil 2003 1506 32.70 2.82 0.61Chile 2004 894 2.97 2.20 0.88

2006 925 13.20 2.13 0.83Colombia 2006 918 17.10 2.47 0.53Costa Rica 2005 284 28.49 7.04 0.59Ecuador 2003 356 20.47 8.71 0.20

2006 603 26.24 3.31 0.17Egypt, Arab Rep. 2004 946 16.50 7.36 0.48El Salvador 2003 409 23.21 3.93 0.43

2006 546 18.93 3.90 0.50Guatemala 2003 437 22.59 6.49 0.39

2006 491 26.83 4.76 0.31Guinea 2006 213 64.25 19.70 0.06Honduras 2003 333 31.66 5.97 0.33

2006 355 15.85 4.69 0.30Kenya 2003 218 13.32 11.94 0.31Madagascar 2005 271 6.41 11.07 0.42Mauritania 2006 225 47.13 6.22 0.26Mexico 2006 1295 23.67 2.02 0.60Morocco 2004 833 3.91 0.49 0.56Namibia 2006 301 24.68 2.29 0.59Nicaragua 2003 346 32.58 9.03 0.25

2006 415 40.85 10.51 0.21Panama 2006 545 36.99 5.50 0.57Paraguay 2006 452 19.34 6.02 0.20Peru 2006 607 10.53 1.00 0.32Rwanda 2006 209 19.00 8.89 0.45South Africa 2003 560 9.16 1.43 0.74Swaziland 2006 290 58.13 4.04 0.21Tanzania 2003 229 30.06 9.95 0.41

2006 416 47.03 12.53 0.43Uganda 2006 546 46.50 13.59 0.37Uruguay 2006 341 14.85 0.83 0.66

46