IB&M Master Thesis What is the effect of trade liberalization on inequality in the context of Latin American countries? How does corruption matter?

IB&M Master Thesis

What is the effect of trade liberalization on inequality

in the context of Latin American countries? How does

corruption matter?

First Supervisor: Dr. Sathyajit Gubbi

Co-assessor: L. Gee

Abstract

How trade liberalization policies have been affecting inequality levels is a topic which has generated intense academic debate and discussion. The current relationship might be at some extent influenced by the corruption levels within the countries. This corruption levels will negatively influence the hypothetical redistributive effects that trade liberalization will generate and consequently, directly passing on the inequality levels. This MSc thesis examines if corruption levels have a significant in the relationship between trade liberalization and inequality using a panel data approach. The findings confirm that the moderation effect is not significant when examining the relationship between trade liberalization and inequality in the interaction effect. Despite this, some interesting findings, some of them following a trend but not significant, have been found and further discuss

Table of Contents

1. Introduction ………..6

2. Theoretical background………9

2.1 Literature review………...9

2.1.1 Classic trade theories………..9

2.1.2 Trade liberalization ………...10

2.1.3 Trade liberalization in developing countries: The Latin America case……….12

2.2 Theory and hypothesis development ………16

2.2.1 The impact of trade liberalization on inequality levels………..16

2.2.2 Specific patterns in developing countries: Latin America……….18

2.2.3 The moderating role of corruption……….20

3. Data and methodology ……….22

3.1 Research model……….22

3.2 Empirical setting………...23

3.3 Sample and data………23

4. Results………..33

4.1 Regression model………..33

5. Concluding remarks………39

5.1 Conclusion……….39

5.2 Limitation and future research………..40

References………42

Appendix………..50

TABLES

Table 1: Descriptive statistics………..31

Table 2: Correlations………...34

Table 3: Model summary……….35

Table 4: ANOVA……….38

Table 5: Coefficients………39

GRAPHS

Graph 2: Normal P-P Plot……….50

Graph 3: Scatterplot analysis (trade-inequality) ………..51

Graph 4: Scatterplot analysis (corruption-inequality) ………..52

FIGURES

1.INTRODUCTION

The discussion on the distributive implications of globalization, and in particular of trade liberalization has gained relevance in recent years in the academic and political context. Although the evidence on the relationship between trade liberalization and poverty is diverse, the general consensus is that in the long term, the insertion of developing countries in the world economy and international trade offers great opportunities to improve growth and alleviate poverty ( Edwards, 1993; Rodrik, 1995; Barro & Sala-i-Martin, 1995; Wacziarg & Welch, 2008; Parikh & Stirbu, 2004). However, the effects in underdeveloped countries as in the case of Latin American countries, the results may not follow the expected theoretical line. Empirical evidence is shown that in the short term, Latin American countries have been facing different patterns compared with other regions in the world (Wood, 1997; Beyer, Rojas & Vergara, 1999; Goldberg & Pavnick, 2004; Buitrago, 2009).

classic theory states, is expected that more openness and better consolidate trade policies will lead a better economic growth and hence, inequality reduction. However, it might not be the case of Latin American region (Murakami, 2018).

In addition, due to the little evidence existed in the literature, the present paper introduces the moderating role of corruption, which prevails in a worrying way in the Latin American region (Ostry, 1991; Weyland, 1998). Authors have argued that liberalization and openness will lead to a reduction of corruption (Ades & Di Tella, 1997; Dincer & Gunalp, 2008). However, in Latin American, where the legal institutions are not well developed, might not be the same (Johnson, Kaufmann & Shleifer, 1997). On the other hand, regarding the relationship between corruption and inequality, the results are still open to debate due to different empirical pieces of evidence found. In the case of the relationship between corruption and inequality, some authors justify that corruption negatively affects inequality levels, arguing that more corruption will lead to higher inequality levels (Ades & Di Tella, 1997; Dincer et al, 2008). However, on the other hand, some authors showed evidence to the contrary, especially in countries where the informal market prevails and the legal institutions are not well established, as the case of Latin America (Dobson & Dobson, 2010, Andres & Dobson, 2011).

Therefore, the present research of the following MSc thesis is focused on a conceptual model that examines the relationship between trade liberalization and inequality where corruption is playing a moderator role in the model. For that purpose, in the first part, the literature review, the classic theories and trade liberalization policies are explained., with special emphasis in Latin American countries. In the second part, the theory is focused on trying to analyze the relationship and the impact that trade liberalization has on inequality levels. In order to analyze the relationship the flowing hypothesis 1 has been developed: In

the long term (1996-2016), there is a negative relationship between trade and inequality in Latin American countries. On the other hand, the second part of the theory is based on the

moderating role of corruption and its implications regarding the relationship between trade liberalization and inequality. For such purpose, hypothesis 2 is studied: : There is an

negative relationship with inequality for low corruption and less negative relationship for high corruption values.

In order to carry out the analysis, the study is focused on Latin American countries. In particular Argentina, Bolivia, Brazil, Chile, Colombia, Costa Rica, Ecuador, Honduras, Mexico, Panama, Peru, Paraguay, El Salvador and Uruguay. The context in which the countries are texted goes through the years 1996 and 2016. Panel data is implemented in order to conduct the linear regression. The proxy for texting the model is based on macroeconomic indicators such as trade in % GDP, Gini coefficient or Corruption Perception Index. Moreover, control variables (economic growth, FDI and inflation) are texted in order to minimize the effects of un-observed factor on inequality levels.

2.LITERATURE REVIEW

2.1.1 Classical trade theories

Adam Smith was the pioneer who firstly illustrates that two nation can equally benefit from trade even when one is more efficient in the production of one good and the other nation is more efficient at producing another one. Accordingly, this fact will enable each nation to specialize their production in the specific goods or services which have absolute advantage, increasing overall production (Smith, 1937; Krist, 2013). Few years later, Ricardo exposed some limitations and introduced new contributions. A country has a comparative advantage in the production of a good if the relative opportunity cost in the production of that good is lower than in other countries. Each country will specialize in a good in which it has a comparative advantage, both countries can gain benefits for trade (Ricardo, 1817).

classic theories claim that more trade and openness will lead to inequality reductions in developing countries (Wood, 1997).

2.1.2 Trade liberalization

In the last decades, the world economy has experienced rapid growth rates in general terms as a consequence of technological and diverse integration policies across countries. Trade liberalization policies marked a turning point in terms of economic consolidation within countries and regions worldwide, reducing trade barriers and promoting free markets with common policies and legislations (Helpman & Krugman, 1985; Kawai, 1994; Perry & Olarreaga, 2006; Krugman, Obstfeld & Melitz, 2006; Breinlich & Cuñat, 2015). In the literature, trade liberalization is often defined as "a situation without tariff barriers or with a reduction of tariff and non-tariff barriers imposed on the inflow and outflow of goods and services" (Sinaga, 2010). Also, Dijkastra (2000) defined trade liberalization as "the policies in charge of reducing restrictions on the free trade of goods and services. Especially, comprising the reduction of import quota and import tariffs, export restrictions reduction and lowering export taxes".

liberalization based on Ricardo´s comparative advantage argument. As long as the relative prices before trade differ, the specialization will promote and increase welfare in both countries. Secondly, countries might face dynamic effects, such as technological change, learning, and economic growth, which change the production function in profitable directions. Third, the economies that trade more will adapt better to external structural shocks. Trade liberalization will decrease the waste stemming from rent-seeking activities (Rodrick, 1995).

Moreover, classic models expose that trade liberalization causes an increase in the economy real income, improving in that way, the consumption possibilities in the aggregate . As a consequence, prices will adjust to prevailing levels in the world market and, therefore, the necessary incentives are created for economies to transform their productive structure and specialize in activities in which their cost are relatively lower. In this way, the positive impact of trade opening on growth is due to more efficient productive resources allocation and economies specialization (Cho et al, 2000; Krugman, Obstfeld & Melitz, 2006). However, some authors are not taking into an account that greater capital accumulation and more specialization do not imply better redistributive policies and mechanism among the population (Taylor, 1998; Tavares & Wacziarg, 2001).

On the other hand and despite the results obtained, some authors argue that trade liberalization might not be a realizable and viable option for all countries. Especially, developing countries, which may face structural issues when they are opening their economies, being detrimental for the subsequent economic growth and social development in the short term (Baldwin & Krugman, 1989; Devarajan et al, 1989, Ostry, 1991; Dornbusch, 1992; Goldberg & et al, 2004). In particular, for countries with a limited industrial base and imperfect competition, the expected theoretical trade liberalization effects should not consider unequivocal positive (Harris, 1984). A country without a solid industrial based, which is crucial for the long-run growth, will face a long-run deterioration in the trade terms because it’s basing its commercial sector on primary exports with low value added (Rodrick, 1995; Dijkstra, 2000).

As an example, Dean Baker (2008), carries out an analysis which is not aligned with the theory that states that more openness and liberalization will lead to better resource allocation and better gains redistribution. The first concept is based on the acceptance that there are winners and losers in trade, leaving behind theories that argued that everyone can benefit from trade. The commercial opening carries out changes in the prices of resources, benefiting, therefore, the one that is more intensive in a given territory. When we refer to developed countries, an increase in the salary of skilled labor would be expected, to the detriment of the unskilled, and income inequality consequently increased. The theory of trade argues that despite there being losers, the possible benefits of the winners would compensate for the losses. In this way, it is understood that those who are injured must sacrifice themselves in favor of greater economic efficiency.

The second premise that Baker (2008) points out is based on the possible redistribution from the winners to the losers of commerce. However, the trade is seeking greater economic efficiency, redistribution would be carried out by collecting taxes, reducing economic efficiency. Finally, the third aspect is based on "trade protectionism", that is, the presence of barriers that maintain better conditions for the most qualified workers in developed countries.

2.1.3 Trade Liberalization in developing countries: the Latin America case

As mentioned in the existing literature, trade liberalization and subsequent economic growth have not been manifested in the countries equally (Edwards, 1993; Rodrik, 1995 Parikh & Joshi, 2005). In particular, Latin American countries have experimented different patterns compared with other regions as a result of different trade policies throughout the last decades (Dornbusch, 1992; Edwards, 1993; Rodrik, 1995).

institutional environment (corruption, lack of transparency and inefficient bureaucracy) and a series of systematic imbalances referring to public finances, trade balance, price level and the existence of national monopolies (De Gregorio, 1992). Apart from this, a constant danger in the commercial exchange restrictions is also present due to political situations and high uncertainty levels, both social and economic (Prebich, 1981).

During the 80s, many semi-industrialized countries began to adopt market-oriented reforms (Edwards, 1993), characterized by the liberalization of international trade and the attraction of foreign investment flows, attributing the export sector as the main growth engine (Holland, Vieira & Canuto, 2004). In particular, Latin American countries have been adopting external oriented growth policies with the main purpose of consolidating macroeconomic stability and position the exports manufacturing sector as sustainable growth nuclei (Edwards, 1993; Perry et al, 2006; Holland et al, 2004). In addition, they have been seeking to reduce the vulnerability that exports may face regarding external shocks, encourage competition among industries and increase technology transfer and productive efficiency (IMF, 1997; OECD, 1999).

As a consequence of the adopted policies, Latin American economies registered significant growth rates during initial periods, being able, in the short term, to effectively contain the fiscal deficit and the inflation rate, presenting better insertion in the international market and higher foreign direct investment flow (Holland et al, 2004). Nonetheless, It was also a period of system imbalances that manifested in various ways across the region (Perry et al, 2006; Kalyoncu, Ozturk, Artan & Kalyoncu, 2009; Holland et al, 2004).

fact occurs due to the commercial pattern that prevails in the Latin American economies is based mainly on following a commercial dynamic oriented to the import of high technology goods and a net exporter of goods based on natural resources and with a low-medium technological content (Holland et al, 2004). It may be one of the arguments which explains why the results of the liberalization have not been the same in Southeast Asian countries compared with Latin American countries (Prebisch, 1981; Perry et al, 2006).

Imagen 1: Growth and Exports in Latin America and East Asia: 1965-1989. Retrieved from Edwards (1993)

Secondly, the trade liberalization policies adopted in Latin America in the 80s and early 90s became a major liberalization of financial markets, generating contradictory impacts within the capital flows from foreign countries. The massive inflow of capital caused the real exchange rate to appreciate, parallel to the stability of the nominal exchange rate. As a result, the trade balance turned negative in the short and medium run (Ostry, 1991, Parikh, 2005, Edwards, 1993; Reinhart, 1995; Holland et al, 2004). Latin America maintains a synchronized movement between the exchange rate and the competitiveness of the external sector. During the phases of better commercial position, and therefore of greater export dynamism, the exchange rate becomes a relevant factor, an insertion in the world market is created in the function of relative prices, in the place of productivity and technological innovation. It is important to analyze to what extent the economic opening process, based on liberalization and flexible exchange rate, created possible restrictions to economic growth in the long term, linked to the position of the trade balance and the characteristics of the pattern of productive specialization (Reinhart, 1995).

Mexico (Cragg & Epelbaum, 1996), documenting a speedy increase in the professional occupational reward and administrators. In contrast, analyses in other countries, such as Colombia (Attanasio, Goldberg & Pavcnik, 2004) do not document any significant variance.

As mentioned before, the present study is focused on how trade liberalization affects inequality levels in Latin America, with corruption as a moderator. Most of the exited literature is focus on the short-term affects during the 80s and 90s. However, the present master thesis is focus on the sample analysis throughout a longer period of time, specifically from 1996 to 2016. It is interesting to determine if the patterns that characterize these countries have remained in force or if, in the long term, they have managed to achieve stable growth rates, consolidate macroeconomic stability and the imposition of efficient redistributive policies that accomplish inequality and poverty levels reduction. At the same time, the study investigates the role of corruption as a moderator between trade liberalization and inequality levels and analyzes if countries with higher corruption index and lower institutional quality are those that show stagnation or lower growth levels, either economic and inequality. Therefore, this paper seeks to answer the following research question: How does trade liberalization affect inequality levels in countries with high levels

of corruption?

2.2 Literature review and hypothesis development

2.2.1 The impact of trade liberalization on inequality levels

the benefits generated from this process are not equally shared among the population. Thus, it will not be beneficial for an effective distribution of resources and the reduction of the inequality levels among the citizens (Kuznets, 1995; Winters, McCulloch & McKay, 2004, Goldber et al, 2004; Meschi & Vivarelli, 2007; IMF, 2007). In fact, the existed literature attributes the inequality increases since 1980s to the opening of international trade and the technology improvement (Bogliaccini, 2013).

Apart from that, empirical studies suggest that income inequality in a country is costly, not only in the social sphere but because it prevents the benefits of globalization from being completely exploited (Birsdall, 2007). These costs that must be paid in the short term, affecting the population with the least resources and generating negative effects in the income distribution (Winters et al, 2004). As mentioned above, technological diffusion is another important aspect that some authors have been pointed out. Being of vital importance the development of internal capacities and mechanisms that allow the real use and the greater appropriation of the innovation to which a certain country could be exposed. If the assumption of technological equality between different countries is relaxed, an increase in inequality would be brought about by the difference in the use of technology in some sectors as opposed to others, and those that employ skilled workers are generally benefited to the detriment of sector which are employing un-skilled labor force (Meschi & Vivarelli. 2007).

2.2.2 Specific patterns in developing countries : Latin America

As mentioned above, trade theory states that a commercial opening would generate benefits in any scenario for the economy in general and it is whole (Cho et al, 2000; Wacziarg, 2001; Krugman et al, 2006). Nonetheless, although the empirical evidence is quite ambiguous, taking into account the existing literature we can suppose that, at least for Latin America, trade openness might increase inequality in the short term in many cases, in turn widening the gap between skilled and unskilled workers (Leamer, Maul, rodriguez & Schott, 1999; Beyer et al, 1999; Weisbrot & Baker, 2003; Goldberg et al, 2004). Wood (1997) in his study reports that in Latin American region, the inequality levels have been steadily growing since the mid-1980s, coinciding with trade liberalization episodes. In the case of this region, some studies state that the increasing demand for skilled-workers might be one of the main consequence for the increasing inequality (Murphy & Welch, 1993 ; Goldberg et al, 2004). Also, Perry et al (2006) observed how during the liberalization process, Latin American countries faced wage inequality and skills premiums, being in many countries an increase in overall income inequality.

Graph: 1 Comparison between poor population income before and after liberalization. Own elaboration based on Dollar and Kraay (2002) and Wacziarg and Welch (2003)

In addition, these countries have institutional limitations, both in the administrative and fiscal areas, so the adoption of efficient redistributive policies seems fundamental for the future (Jansen & Lee, 2007).

To sum up, the existed literature still remains quite ambiguous when analyzing the effects that liberalization has on inequality levels, especially in Latin American countries. There are empirical evidence throughout the literature which expose that, in Latin America in the short term, it may cause negative redistributive effects (Wood, 1997; Winters et al, 2004; Gasparini & Lustig, 2011), affecting the labor force demand (Goldberg et al, 2004, Meschi et al, 2007; Bogliaccini, 2013), households consumption (McCulloch, 2003) and consequently, inequality levels. However, other authors support the classic theory assumptions, arguing that trade liberalization will promote economic growth, better resource allocation and inequality levels reduction in the long term (Cho et al, 2000; Wacziarg, 2001; Reinikka et al, 2001; Rodriguez et al, 2001; Edwards et al, 2002; Krugman et al, 2006; Perry et al, 2006). An analysis is carrying out following the line of the classic theory and existing literature and taking into an account the long term perspective. It is expected expect a stagnation or progressive overall inequality decreasing in the long term in Latin American countries. For that purpose, the flowing hypothesis is been proposed :

0 500 1000 1500 2000

Brazil Chile Colombia Mexico Peru Costa

Rica SalvadorEl Ecuador

Income of the poor before and after trade episodes

Hypothesis 1: In the long term (1996-2016), there is a negative relationship between trade and inequality in Latin American countries.

2.2.3 The moderating role of corruption

Latin American widely seems like a region with one of the highest corruption levels in the world (Ostry, 1991; Weyland, 1998). As mentioned before, those countries have been adopting both political and economic opening policies, implementing democratized and neoliberal economic reforms. However, despite this, the region still suffers from weak institutional system and high corruption levels in most of their public organisms (Weyland, 1998; Tulchin, Espach & Espach, 2000).

In previous years, some authors have been argued that trade liberalization and more openness will lead to a corruption levels reduction (Ades et al, 1997; Dincer et al, 2008). Bureaucrats are not able to collect as many bribes as before the openness, partly, thanks to the liberalization process (Leiken, 1996). The empirical analysis which argued that market reforms are sources of corruption often fail to take into an account the amount of corruption that would otherwise be carried out in the absence of trade openness and economic reforms. Less regulatory and trade interventions from the government, stability, and moderate tax impositions will result beneficial for reducing corruption ( Ades et al, 1999).

On the other hand, some authors have expressed that corruption causes negative effects on inequality, more corruption will lead to increase inequality levels (Gupta, Davoodi, & Alonso-Terme, 2002; Gyimah-Brempong & Muñoz de Camacho, 2006; Li, Xu, & Zou, 2000). However, conventional believe that corruption negatively affects inequality has been recently questioned, especially in Latin America. In particular, Dobson & Dobson (2010) and Andres & Dobson (2011) provide evidence that supports the affirmation with the explanation of the large informal sector that prevails in the region. The individuals with the lowest income do not have the required characteristics that achieve a job in the formal economy, work opportunities are restricted by discrimination and institutional barriers. Therefore, the informal sector is willing to provide job and opportunities to the poorest, while policies imposing to reduce corruption through labor market and regulations could be highly detrimental for the welfare and employment in the informal sector.

Hypothesis 2: There is an interaction effect between trade liberalization and corruption on inequality. Trade has a negative relationship with inequality for low corruption and less negative relationship for high corruption values.

Additionally, is very important for the present thesis to take into an account of how the variables are measured, especially inequality and corruption. For that reason, in the following section, a descriptive analysis will be conducted related to the data, sample and the variable measurement

3. Data and Methodology

3.1 Research model H2 + H1 -

Figure 1: Research conceptual model

Figure 1 represents the model of this research, indicating that corruption would moderate the relationship between trade liberalization and inequality levels. Regarding the first hypotheses, trade liberalization will negatively affect inequality levels, assuming in that way that more trade liberalization would decrease the levels of inequality in Latin

IV: Trade Liberalization

American countries. On the other hand, corruption moderator effect will play a positive role in the relationship between trade liberalization and inequality levels. In particular, less corruption will lead to fewer inequality levels. In that sense, it is important to take into an account of how the corruption and inequality variables are measured. For that reason, in the following section, more details about the variables and the data analysis will be conducted.

3.2 Empirical setting

Latin American countries are a suitable empirical setting for the proposed research. The region during the last decades has been adopting trade liberalization policies, in order to consolidate their economies in the global market and achieve sustainable growth throughout the years (Edwards, 1993). However, in opposition to what the classical theory of commerce affirms, some of the Latin American did not face the expected liberalization benefits (Dornbusch, 1992; Rodrik, 1993). In addition, these countries, as well documented by some authors, have been registered significant inequality levels increase and stagnation after the openness episodes (Goldberg et al, 2004), becoming in one of the regions in the world with highest levels of inequality (Alvaredo & Gasparini, 2015). Finally, corruption has been present throughout the years in Latin American countries. In particular, the chronic corruption that reigns in most of the areas could be one of the determinants of slow and volatile growth of the region after liberalization policies (Ades et al, 1999).

3.3 Sample and data

El Salvador, Uruguay. Finally, Guatemala, Nicaragua, and Venezuela are excluded because of the lack of reliable data, especially related to the Gini coefficient.

Therefore, the database implemented in this research is based on macroeconomic indicators of the previously mentioned countries covering the period between the years 1996 and 2016. I specifically choose a long period of time in order to examine the particular liberalization effects in the long term perspective, rather than just on the short-term perspective. Consequently, I have been trying to identify under the relation between trade liberalization and inequality levels, a factor that could give a clear answer to the formulated hypotheses.

On the other hand, the data used in this MSc thesis research comes from different international organizations. First, the World Trade Organization (WTO) Data portal which provides quantitative data regarding economic and trade policy aspects. The data provided broadcast trade in services statistics and merchandise trade, market access indicators, non-tariff information, and other indicators. To be more concrete, the panel database contains 240 indicators, 289 reporting economies all over the world and throughout 72 years, from 1948 to 2019.

Second, World Bank Open Data, which is more oriented into world development indicators, has also been employed to get the data to conduct the analysis. In particular, it employees 76 different databases across 264 countries, using 1.599 series throughout 59 years. Third, in order to measure the GINI coefficient, the current version of the World Income Inequality Database (WIID4) has been used. It comprehends over 11.000 data points in total, 3.500 unique country years observations and covers 189 countries worldwide and including until the year 2017. The WIID4 has measured the GINI coefficient in a scale from 0-100 and other indicators regarding income inequality, either developing, developed or transition economies (UNU-WIDER, 2017).

support victims and witnesses of corruption. They work together with citizens, governments, and organizations to prevent bribery, secret deals and abuse of power. They provide different databases related to corruption, effectiveness and institutional quality around the globe, including information between the years 1995 and 2018.

3.4 Variable measurement

3.4.1 Dependent variable

Inequality levels were derived from the World Income Inequality Database (WIID4). In particular, to measure the distribution inequality the GINI coefficient has been used. It measures and shows to what extent people or households income distribution in a specific country be distinguished from the ideal or perfect equal income distribution. The range varies from 0 to 100, where 0 indicates equity in the distribution of income and while 100 represents a total inequality.

However, multiple sources have been used to establish the observations of the Gini coefficient in the WIID4 database. In particular, the database is based on a wide variety of assets and associate to a broad variety of population concepts and income, statistical approaches and sample sizes. Moreover, it can be seen a difference between income measurement (consumption, gross income, net income …) and the receptor of that income (family, household, person and tax unit ) (UNU-WINDER, 2018). In this way, the Gini coefficients will face quality issues, particularly in developing countries. The income concept and beneficiary units will be different among the countries and throughout the successive surveys, resulting in not stable and comparable within the years (Deninger & Squire, 1996; Coudouel, Hentschel & Wodon, 2002). In order to consider reliable the data available, Deninger et al (1996), developed three minimal standards of reliability for making the Gini variable solid and comparable throughout the years.

or general assumptions regarding redistributive patterns. Second, exhaustive coverage of the population is also required. Using not representative sample can easily lead to biased estimations, therefore, the sample must be represented and cover all of the population. Lastly, it is required that the inequality index may be based on representative coverage of the diverse income sources and different population segments (Deninger et al, 1996).

3.4.2 Independent variable

Trade liberalization plays the role of an independent variable in this MSc thesis. The data has been resulting from the World Trade Organization (WTO) and the World Bank Open Database. Trade liberalization is measured by taking the trade openness ratio, which is calculated from the sum of import and exports of goods and services measured as a share of gross domestic product (in %) (The World Bank, 2019). This proxy has been used in the existed literature (Dollar, 1992; Agosin & Ffrench-Davis, 1995; Vamvakidis, 2002). Nonetheless, some authors have distinguished two main disadvantages or limitations that the use of this indicator can imply. In particular, Frankel & Romer (1999), argue that when examining the relationship between trade liberalization and inequality using Trade (%) GDP, some aspects are not included in the analysis. For example, they argue that countries which are adopting free-trade policies will normally adopt free-market domestic policies and also domestic stable fiscal and monetary policies. These policies are likely to affect income, however, the benefits that trade would bring are omitted from the income equation. The second aspect is related with countries’ geographic characteristics, which are considered as an influential determinant of trade (Frankel, Stein & Wei; 1997), arguing that trade does not take into consideration the particular countries features, in this case, geographical aspects (Frankel et al, 1999).

3.4.3 Moderator variable

ranks 180 countries regarding public sector corruption levels based on surveys, experts and business people. For such purpose, it uses a scale from 0 to 10, where 0 indicates an extremely corrupt index while 10 signifies the total absence of corruption in the public sector. Nonetheless, it has been documented by some authors that the CPI shows some limitations, which will be discussed at the end of the present study.

3.4.4 Control variables

The present thesis adds economic growth, foreign direct investment and inflation as control variables. The first control variable, economic growth, it is measured with the annual percentage growth rate of GDP at constant US$. It encompasses data from 264 countries and regions worldwide from 1961 to 2017. Economic growth (CV 1) has been chosen because of its ambiguous relationship with trade liberalization and inequality levels. The classic trade theories have shown the importance of trade liberalization for economic growth (Devarajan et al, 1989; Edwards, 1993; Barro et al, 1995; Wacziarg et al, 2000 Parikh, 2006). Nonetheless, trade liberalization did not bring the expected benefits in terms of economic growth and better resource allocation (Balassa, 1985, Buitrago, 2009, Rodrick, 1993). Related with income inequality, Kuznets (1955) states that income inequality is reduced in advanced nations in the last phases of the economic growth process. However, as mentioned before, in developing countries such as Latin American ones, the economic growth does not necessarily imply a reduction in the inequality levels (Goldberg et al, 2004). Therefore, it has been considered that the control variable economic growth may have ambiguous repercussion on inequality levels.

taxes and institutional protection (Bogliaccini, 2013). However, it is consider that the control variable FDI may have a positive repercussion in the reduction of the inequality levels.

To conclude, the last control variable which is conducted in the model is inflation (CV3). It is measured reflecting consumer prices cost change in annual percentage, studying 292 countries from 1960 to 2017.Part of the existing literature shows a positive correlation between inflation and income across the countries (Albanesi, 2007; Thalassinos, Ugurlu & Muratoglu, 2012). Growing inflation may affect relative prices, whereas prices are getting lower when talking about chronic inflation. As a consequence of these price distortions, low income population is going to face more significant effects on their consumption (Bogliaccini, 2013). Therefore, according with the theoretical evidence, a positive relationship is expected when analyzing inflation and inequality.

3.5 Econometric model

As a consequence, the econometric model which is implemented in the present study will allow us to partially solve this data inconvenience. Therefore, a panel data set, also known as longitudinal, will be implemented. The data set comprises a group of cross-sectional units (countries) observed across time (years). Linear regression with fixed-effects analyzes the relationship between the outcome and predictor variables within a unit (country). However, each country can involve different characteristics that may or may not have a direct influence in the other variables (dependent, moderator and control). Therefore, the fixed-effect model assumes that something within each country, which is not observable, may impact the outcomes variables. The effect of those time-invariant characteristics will be removed by the model so we can measure the predictor's net effect on the outcome variable (Torres-Reyna, 2007). Consequently, the net effect of the IV (trade liberalization), CV (economic growth) and MV (corruption) on the DV (inequality) can be analyzed.

Another important characteristic related to the fixed-effects model is that the mentioned time-invariant characteristics are exclusive to each entity (country) and should not be correlated with other individual features. Fixed-effects model is designed to conduct a study regarding the changes within a person or entity (country). As a consequence of the differences between entities, the entity´s error terms and the constant should not be correlated with the others in order to guarantee that the model is suitable (Torres-Reyna, 2007).

The following equation will be conducted or analyzing and testing the hypothesis :

𝑌 = 𝛽%+ 𝛽& . 𝜒& + 𝛽). 𝜒) + 𝛽*. 𝜒&. 𝜒) + 𝔲

,-• 𝛽% unknown intercept.

• 𝑌_,- is the dependent variable (DV) where i=entity and t=time. • 𝑋_,- represents one independent variable.

• 𝛽_- is the coefficient for the IV. • 𝔲,- represents the error term.

3.5.1 Descriptive statistics

First of all, the number of observations fluctuate as a consequence of the incomplete dataset in some of the countries in which the study will be focused on. Some difficulties have been found regarding the Gini coefficient available data, especially in Nicaragua and Guatemala. The following table (table 1) represent the descriptive statistics. It can be seen that the dependent variable (inequality) shows the lowest number of observations with 262. Furthermore, it should be taken into an account that the four variables are measured using different scales. For example, Inequality is analyzed using a scale from 0-100 (0: no inequality; 100: total inequality), whereas the moderator Corruption uses a scale from 0 to 10 (0: high corruption; 10: no corruption). Moreover, both Trade liberalization and

Economic Growth are expressed in percentages (%) without a specific range.

The table 1 also indicates a moderately high Gini coefficient mean of 50,54. Bolivia is the country with the lowest inequality index with 61,6 in 2000, whereas the lowest Gini coefficient, 39,7, can be seen in Uruguay in 2012. Moreover, regarding the independent variable, trade liberalization, it can be observed that the highest value (165,34) is registered in Panama in 1997. On the other hand, the lowest (15,64) belongs to Brazil in 1996. To conclude, it is important to mention that corruption presents a mean (3,88) which is closest to the minimum value (1,41) than the maximum one (7,91). This fact means that the countries have on average high levels of corruption perception index (3,88). Latin American countries show high levels of corruption, where informal mechanisms and lack of transparency prevail (Ades et al, 1999).

3.5.2 Scatterplot analysis (trade-inequality and trade-corruption )

more trade liberalization will lead to better resource allocation and therefore, the reduction of inequality levels (Krugman et al, 2006). On the other hand, it might be partly supported by the theory developed by Goldberg and Pavnick (2004) in Latin American countries, which claim that trade liberalization, apart from other aspects, can lead to stagnation or increasing of inequality levels.

4. RESULTS

4.1 Regression analysis

The normality assumption on linear regression assumes that the residuals of the DV must be normally distributed. The normal P-P Plot of regression (graph 1) proves the linear regression assumption. It can be seen how the data is uniformly following the line, presenting no dispersion.

The mentioned variance can be also observed in the ANOVA (table 4). The variable economic growth, FDI and inflation themselves are consider significant in the model (p-value=.035). Moreover, when adding trade liberalization and corruption we can also observe a significant variation in the model (p-value=0,00). It can be stated that if the significance is below .05, (𝑝-value <.05), it means that the model explains a significant amount of variance in inequality. Following the same patterns than R-Square, the first model (model 1) does explain a significant amount of variance in inequality (𝑝 = 0,035). Moreover, when adding the two other independent variables in the second model ( trade liberalization and corruption) it also explains significant variance within the model ( 𝑝 = 0,000). Finally, it does not change for the interaction effect (model 3) because it is already zero.

Turning into the table 5, it can be observed in column B that a country with mean value of trade liberalization (60,68) and corruption (3,83) will present an inequality value of 49 (𝐵 = 48,950). Moreover, we can state that the relationship between the economic growth and inequality is negative and significant (𝐵 = −0,196, 𝑆𝐸 = 0,091, 𝑝 − 𝑣𝑎𝑙𝑢𝑒 = 0,033). Moreover, the relationship between FDI (% GDP) and inequality is also positive and significant (𝐵 = 0,524, 𝑆𝐸 = 0,112 , 𝑝 − 𝑣𝑎𝑙𝑢𝑒 = 0,00). Instead, Inflation does not have a significant effect in the model (𝐵 = 0,042 , 𝑆𝐸 = 0,035, 𝑝 − 𝑣𝑎𝑙𝑢𝑒 = 0,240) Trade liberalization does not have a significant effect in the model (𝐵 = 0,156, 𝑆𝐸 = 0,330, 𝑝 − 𝑣𝑎𝑙𝑢𝑒 = 0,638). However, corruption itself seems to have a significant and negative effect on inequality (𝐵 = −2,152, 𝑆𝐸 = 0,322, 𝑝 − 𝑣𝑎𝑙𝑢𝑒 = 0,00), whereas the corruption moderation effect is not significant ( 𝑝 − 𝑣𝑎𝑙𝑢𝑒 =0,863).

5. Concluding remarks 5.1 Conclusion

The present study has the main object of examined whether the relationship between trade liberalization and inequality levels is influenced by corruption in Latin American countries. In particular, the study was focused on fourteen countries from 1996 to 2016. In order to study this moderation model (see figure 1), the following research question has been conducted: What is the effect of trade liberalization on inequality in the context of Latin

American countries? How does corruption matter ?.

A panel database has been implemented to analyze the relationship between trade liberalization and inequality, with the moderation effect of corruption. After executing the regression model, some conclusion can be argued. First of all, the classic models state that more openness and trade liberalization will lead to a better resource allocation and inequality levels reduction (Deverajan et al, 1898; Edwards, 1993). However, after conducted the regression model, trade does not have a significant effect in the inequality levels reduction, therefore, it cannot state that more trade will lead to a reduction in the inequality levels in Latin America from 1996 to 2016. Theoretically, we cannot say that it is close, it is either significant or not significant but, a trend can be seen that probably it would be more significant if it was more data. In the model the coefficient between trade and inequality was positive. Therefore, in the case that the relationship was significant, we could affirm that more trade would lead to more inequality. This finding is more in line with the theory that claims that, especially in Latin America, trade liberalization episodes could lead to stagnation or even the increase in inequality levels (Rodrik, 1995; Perry et al, 2006; Buitrago, 2009; Díaz et al, 2017). Therefore, the first hypothesis cannot be either accepted or rejected because of the non-significant data. However, if the data were significant, we would be closer to rejecting it rather than accepting it.

inequality cannot be affirmed in the present model with the data implemented. However, we could find some results regarding the corruption variable itself, not within the interaction effect. Firstly, there is an interaction effect between trade liberalization and corruption on inequality. Accordingly, there is a positive trend in Latin American countries which partially supports the theory explained by Ades et al (1997) and Dincer et al (2008), in where they claim that trade liberalization and more openness will provoke a reduction in the levels of corruption. Secondly, as pointed out before, corruption does not have a significant effect in the interaction model, however, corruption itself has a positive significant impact on the reduction of inequality levels in Latin American countries. If the corruption perception index (CPI) goes up, which means less corruption and better institutional environment, the inequality levels will go down. In particular, taking into an account the B=-1,199 (table 5), we can affirm that for every unit that corruption increases, the inequality variable will go down by 1,199. This finding, support the theory presented by Li et al (2000), Gupta et al (2002) and Gyimah-Brempong et al (2006), where they empirically prove that corruption will generate a negative effect on inequality, stating that more corruption will increase inequality levels.

To conclude, it can be observed that we can observe how the control variables (economic growth, FDI and inflation) used have a significant relationship in the model. However, when we turn into the interaction effect between trade liberalization , corruption and inequality, the results are not the same.

5.2 Limitation and future research.

Several limitations could be found in the present master thesis which may be further developed and improve in future research.

Moreover, the GINI coefficient is based on a wide variety of assets and a broad variety of population, income and sample sizes. The GINI coefficient will face problems, especially in developing countries due to the lack of data and measurement instruments. For future research, the inequality data should be focus on a more reliable data source, such as household surveys or individuals as observed units, and representative sample. In that way, the sample will be bigger and the collected data will reflect in a better way the inequality perception and effects. Also, the CPI is mainly based on perceptions that business people and governments are giving related to the levels of corruption and institutional quality. Either inequality and corruption are really abstract variables which are not easy to measure and interpreted. For example, the perception of corruption might be different across the countries, what is considered in a country as corruption does not have to be considered in another one. In addition, it is impossible to determine and measure all the causes that could cause inequality. The mentioned correlation was conducted using control variables, as Bogliaccini (2013), in order to minimize the unobserved effects on inequality. However, despite this, the interaction effect still remains not significant within the model, partly as a consequence of the data limitation.

Another important limitation which can easily provoke interpretation misunderstanding is related to the measurement of the variables. Corruption perception index is measured on a scale from 0 to 10, where 10 express absent in terms of corruption whereas 0 shows high levels of corruption. Turning into inequality, the GINI coefficient is measured with a scale from 0 to 100, where 100 reflects high levels of inequality while 0 express total equity. Both variables are measure with indicators using different scales and where the higher levels express different patterns. Therefore, future research should consider the possibility to adopt

References

Ades, A., & Di Tella, R. (1997). The new economics of corruption: a survey and some new results. Political Studies, 45(3), 496-515.

Ades, A., & Di Tella, R. (1999). Rents, competition, and corruption. American economic

review, 89(4), 982-993.

Ades, A., & Tella, R. D. (1997). National champions and corruption: some unpleasant interventionist arithmetic. The Economic Journal, 107(443), 1023-1042.

Agosin, M. R., & Ffrench-Davis, R. (1995). Trade liberalization and growth: Recent experiences in Latin America. Journal of Interamerican Studies and World Affairs, 37(3), 9-58.

Albanesi, S. (2007). Inflation and inequality. Journal of Monetary Economics, 54(4), 1088-1114.

Alvaredo, F., & Gasparini, L. (2015). Recent trends in inequality and poverty in developing countries. In Handbook of income distribution (Vol. 2, pp. 697-805). Elsevier.

Andres, A. R., & Ramlogan-Dobson, C. (2011). Is corruption really bad for inequality? Evidence from Latin America. Journal of Development Studies, 47(7), 959-976.

Attanasio, O., Goldberg, P. K., & Pavcnik, N. (2004). Trade reforms and wage inequality in Colombia. Journal of development Economics, 74(2), 331-366.

Atkinson, Anthony.B .(1970) . “On the measurement of inequality,” in Journal of

Economic Theory (San Diego, CA), Vol. 2; reprinted in Atkinson, 1983.

Baker, D. (2008). Trade and inequality: The role of economists. Real world economic

review, 45.

Balassa, B. (2013). The theory of economic integration (Routledge revivals). Routledge. Baldwin, R., & Krugman, P. (1989). Persistent trade effects of large exchange rate shocks. The Quarterly Journal of Economics, 104(4), 635-654.

Behrman, J., Birdsall, N., & Szekely, M. (2003). Economic policy and wage differentials in Latin America.

Beyer, H., Rojas, P., & Vergara, R. (1999). Trade liberalization and wage inequality. Journal of Development Economics, 59(1), 103-123.

Birdsall, N. (2007). Income distribution: effects on growth and development. Center for

Global Development Working Paper, (118).

Bogliaccini, J. A. (2013). Trade liberalization, deindustrialization, and inequality: Evidence from middle-income Latin American countries. Latin American Research Review, 79-105. Breinlich, H., & Cuñat, A. (2015). Tariffs, trade and productivity: A quantitative evaluation of heterogeneous firm models. The Economic Journal, 126(595), 1660-1702.

Calderón, C., & Chong, A. (2001). External sector and income inequality in interdependent economies using a dynamic panel data approach. Economics Letters, 71(2), 225-231. Cho, D. S., & Moon, H. C. (2000). From Adam Smith to Michael Porter: evolution of

competitiveness theory (Vol. 2). World Scientific.

Cornia, G. A. (2014). Inequality trends and their determinants: Latin America over the period 1990–2010. Falling Inequality in Latin America: Policy Changes and Lessons, 24-49.

Coudouel, A., Hentschel, J. S., & Wodon, Q. T. (2002). Poverty measurement and analysis. A Sourcebook for poverty reduction strategies, 1, 27-74.

Cragg, M. I., & Epelbaum, M. (1996). Why has wage dispersion grown in Mexico? Is it the incidence of reforms or the growing demand for skills?. Journal of development

Economics, 51(1), 99-116.

Ricardo, D. (1817). On the Principles of Political Economy and Taxation: London.

Deardorff, A. V. (1994). Overview of the Samuelson theorem. The

Stolper-Samuelson theorem: A golden jubilee, (344), 7.

De Gregorio, J. (1992). Economic growth in latin america. Journal of development

Deninger, K., & Squire, L. (1996). Measuring Income Inequality: A New Database, sThe World Bank Economic Review.

Devarajan, S., & Rodrik, D. (1989). Trade liberalization in developing countries: do imperfect competition and scale economies matter?. The American Economic

Review, 79(2), 283-287.

Dijkstra, A. Geske. "Trade liberalization and industrial development in Latin America." World Development 28.9 (2000): 1567-1582.

Dincer, O. C., & Gunalp, B. (2008). Corruption, income inequality, and poverty in the United States.

Dobson, S., & Ramlogan-Dobson, C. (2010). Is there a trade-off between income inequality and corruption? Evidence from Latin America. Economics letters, 107(2), 102-104.

Dobson, S., & Ramlogan-Dobson, C. (2012). Why is corruption less harmful to income inequality in Latin America?. World Development, 40(8), 1534-1545.

Dollar, D. (1992). Outward-oriented developing economies really do grow more rapidly: evidence from 95 LDCs, 1976-1985. Economic development and cultural change, 40(3), 523-544.

Dollar, D., & Kraay, A. (2002). Growth is Good for the Poor. Journal of economic

growth, 7(3), 195-225.

Dornbusch, R. (1992). The case for trade liberalization in developing countries. Journal of

Economic Perspectives, 6(1), 69-85.

Edwards, S. (1993). Openness, trade liberalization, and growth in developing countries. Journal of economic Literature, 31(3), 1358-1393.

Edwards, S., & Lederman, D. (1998). The political economy of unilateral trade

liberalization: the case of Chile (No. w6510). National Bureau of Economic Research.

Feenstra, Robert C. (2015). Advanced international trade: theory and evidence. Princeton university

Feenstra, R. C., & Hanson, G. H. (1997). Foreign direct investment and relative wages: Evidence from Mexico's maquiladoras. Journal of international economics, 42(3-4), 371-393.

Frankel, J. A., & Romer, D. H. (1999). Does trade cause growth?. American economic

review, 89(3), 379-399.

Frankel, J. A., Stein, E., & Wei, S. J. (1997). Regional trading blocs in the world economic

system. Peterson Institute.

Galiani, S., & Sanguinetti, P. (2003). The impact of trade liberalization on wage inequality: evidence from Argentina. Journal of development Economics, 72(2), 497-513.

Gasparini, L., & Lustig, N. (2011). The rise and fall of income inequality in Latin

America (No. 118).

scaGoldberg, P. K., & Pavcnik, N. (2004). Trade, inequality, and poverty: What do we

know? Evidence from recent trade liberalization episodes in developing countries (No.

w10593). National Bureau of Economic Research.

Gupta, S., Davoodi, H., & Alonso-Terme, R. (2002). Does corruption affect income inequality and poverty?. Economics of governance, 3(1), 23-45.

Gyimah-Brempong, K. and S. Muñoz de Camacho, 2006, Corruption, growth and income distribution: are there regional differences?, Economics of Governance 7, 245-269

Harris, R. (1984). Applied general equilibrium analysis of small open economies with scale economies and imperfect competition. The American Economic Review, 74(5), 1016-1032. Helpman, E. (1981). International trade in the presence of product differentiation, economies of scale and monopolistic competition: A Chamberlin-Heckscher-Ohlin approach. Journal of international economics, 11(3), 305-340.

Helpman, E., & Krugman, P. R. (1985). Market structure and foreign trade: Increasing

returns, imperfect competition, and the international economy. MIT press.

Holland, M., Vieira Vilela, F., & Canuto, O. (2004). Economic growth and the balance-of-payments constraint in Latin America. Investigación económica, 63(247).

World Bank (1998). Retrieved from : https://www.worldbank.org

IMF (1997). Retrieved from: https://www.imf.org/external/pubs/weomay/chapter1.pdf Jansen, M., & Lee, E. (2007). Trade and employment: challenges for policy research: a

joint study of the International Labour Office and the Secretariat of the World Trade Organization. International Labour Organization.

Johnson, S., Kaufmann, D., & Shleifer, A. (1997). The unofficial economy in transition. Kalyoncu, H., Ozturk, I., Artan, S., & Kalyoncu, K. (2009). Devaluation and trade balance in Latin American countries. Zbornik radova Ekonomskog fakulteta u Rijeci, časopis za

ekonomsku teoriju i praksu-Proceedings of Rijeka Faculty of Economics, Journal of Economics and Business, 27(1), 115-128.

Kawai, H. (1994). International comparative analysis of economic growth: trade liberalization and productivity. Developing Economies, 32(4), 373-397.

Krist, W. (2013). Globalization and America's Trade Agreements. Woodrow Wilson Center Press.

Krugman, P. R., Obstfeld, M., & Melitz, M. (2006). International economics: Trade and policy.

Kuznets, S. (1955). Economic growth and income inequality. The American economic

review, 1-28.

Leamer, E. E., Maul, H., Rodriguez, S., & Schott, P. K. (1999). Does natural resource abundance increase Latin American income inequality?. Journal of development

Economics, 59(1), 3-42.

Leiken, R. (1996). Controlling the Global Corruption Epidemic. Foreign Policy, (105), 55-73. doi:10.2307/1148973

Lustig, N., Lopez-Calva, L. F., & Ortiz-Juarez, E. (2012). Declining inequality in Latin

America in the 2000s: the cases of Argentina, Brazil, and Mexico. The World Bank.

McCulloch, A. (2003). An examination of social capital and social disorganization in neighbourhoods in the British household panel study. Social Science & Medicine, 56(7), 1425-1438.

Meschi, E. F., & Vivarelli, M. (2007). Globalization and income inequality.

Meschi, E., & Vivarelli, M. (2007). Trade openness and income inequality in developing countries.

Murakami, Y. (2018). Globalization and Income Inequality in Latin America: A Review of

Theoretical Developments and Recent Evidence (No. DP2018-16).

Murphy, K. M., & Welch, F. (1993). Occupational change and the demand for skill, 1940-1990. The American Economic Review, 83(2), 122-126.

OECD (1999), Development Co-operation Report 1998: Efforts and Policies of the

Members of the Development Assistance Committee, OECD Publishing,

Paris, https://doi.org/10.1787/dcr-1998-en.

Ostry, J. D. (1991). Trade liberalization in developing countries: Initial trade distortions and imported intermediate inputs. Staff Papers, 38(3), 447-479.

Parikh, A., & Stirbu, C. (2004). Relationship between trade liberalisation, economic growth and trade balance: an econometric investigation.

Parikh, M. A., & Joshi, K. (2005). Purchasing process transformation: restructuring for small purchases. International Journal of Operations & Production Management, 25(11), 1042-1061.

Perry, G., & Olarreaga, M. (2006). Trade liberalization, inequality and poverty reduction in Latin America. ABCDE, St. Petersburg.

Prebisch, R. (1981). Raul Prebisch on Latin American Development. Population and

Development Review, 7(3), 563-568. doi:10.2307/1972587

Reinhart, C. M. (1995). Devaluation, relative prices, and international trade: evidence from developing countries. Staff papers, 42(2), 290-312.

Reinikka, R., & Collier, P. (Eds.). (2001). Uganda's recovery: the role of farms, firms, and

government. The World Bank.

Rodriguez, F., & Rodrik, D. (2000). Trade policy and economic growth: a skeptic's guide to the cross-national evidence. NBER macroeconomics annual, 15, 261-325.

Rodrik, D. (1993). Trade and industrial policy reform in developing countries: a review of

recent theory and evidence(No. w4417). National Bureau of Economic Research.

Rodrik, D. (1995). Political economy of trade policy. Handbook of international

economics, 3, 1457-1494.

Roemer, M., & Gugerty, M. K. (1997). Does economic growth reduce poverty?. CAER II Discussion Paper 4, Cambridge, MA: Harvard Institute for International Development Salas, S., Chuenpagdee, R., Charles, A. T., & Seijo, J. C. (Eds.). (2011). Coastal fisheries

of Latin America and the Caribbean(Vol. 544). Rome: Food and Agriculture Organization

of the United Nations.

Smith, A. (1937). The Wealth of Nations. Modern Library. New York, 423.

Tavares, J., & Wacziarg, R. (2001). How democracy affects growth. European economic

review, 45(8), 1341-1378.

Taylor, A. M. (1998). On the costs of inward-looking development: price distortions, growth, and divergence in Latin America. The Journal of Economic History, 58(1), 1-28. Thalassinos, E., Ugurlu, E., & Muratoglu, Y. (2012). Income Inequality and Inflation in the EU. European Research Studies, 15(1), 127.

Torres-Reyna, O. (2007). Panel data analysis fixed and random effects using Stata (v. 4.2). Data & Statistical Services, Princeton University.

Tulchin, J. S., Espach, R. H., & Espach, R. (Eds.). (2000). Combating Corruption in Latin

Vamvakidis, A. (2002). How robust is the growth-openness connection? Historical evidence. Journal of Economic Growth, 7(1), 57-80.

Wacziarg, R., & Welch, K. H. (2008). Trade liberalization and growth: New evidence. The

World Bank Economic Review, 22(2), 187-231.

Wacziarg, R. (1999). Measuring the dynamic gains from trade. The World Bank.

Weisbrot, M., & Baker, D. (2003). The relative impact of trade liberalization on developing countries. 62(244).

Weyland, K. G. (1998). The politics of corruption in Latin America. Journal of

Democracy, 9(2), 108-121.

Winters, L. A., McCulloch, N., & McKay, A. (2004). Trade liberalization and poverty: the evidence so far. Journal of economic literature, 42(1), 72-115.

Wood, A. (1997). Openness and wage inequality in developing countries: the Latin American challenge to East Asian conventional wisdom. The World Bank Economic

Review, 11(1), 33-57.

APPENDIX