1
The effects of duration and degree of
democracy on income inequality in
Chile, Costa Rica and Uruguay
Vincy Wemmenhove 1361597
Bachelor International Relations and Organizations Bachelor project 9
Supervisor: Brenda van Coppenolle Word count: 8064
2
Introduction
Since the eighteenth century an idea has emerged that progress in political arena could lead to economic development. During the eighteenth century the first democracies appeared and the Industrial Revolution was at its peak in Western Europe. Soon after a belief developed among scholars (Greig, Hulme & Turner, 2007, pp. 53-57) from these recent Western democracies that democracies increase economic growth and overall economic development. This notion has remained prominent until this day. Many people believe that there is a positive relationship between democracy and economic development and often this theory is used to explain why developing countries have such low levels of economic development. Recent papers (such as Przeworski & Limongi, 1993 and Haggard & Kaufman, 2012) have started to question this relationship and have begun analyzing this relationship to see if it exists or if authoritarian regimes are more likely to increase economic development as opposed to democratic regimes.
In recent times the focus is no longer solely on the relationship between democracies and economic growth, but there is an increasing amount of scholars interested in other aspects of economic development such as income inequality. The underlying theory behind this relationship is that political equality should also lead to income equality. Since citizens are in charge in democracies, they will choose what will benefit the greatest amount of people. It is believed that lower levels of income inequality will benefit the most amount of people and thus that there will be a certain decrease depending on the electoral system in income inequality levels (Iversen & Soskice, 2006, p. 165).
This relationship has frequently been applied to Latin America to explain why this region has the highest income inequality levels in the world, especially when compared to Western countries. Many countries in Latin America have had long histories of authoritarian leadership with few democratic experiences and thus it is often assumed that this authoritarian tradition can be an important factor in explaining the high levels of income inequality in this region (Wiarda & Kline, 2014, p. 64). On the other hand, others (Sirowy & Inkeles, 1990 and Gradstein & Milanovic, 2004) have mentioned that authoritarian regimes can create lower levels of income inequality. Especially in developing regions like Latin America, authoritarian regimes can lead to political stability necessary for economic stability and can impose harsh economic policies that will lead to lower levels of income inequality.
3
For those reasons I would like to test whether or not this idea can be used to explain the reason why there are such high levels of income inequality in Latin America. Since a lack of democratic tradition is often used to explain high levels of income inequality in developing regions, it is useful to compare a couple of these developing countries to determine if a lack of a democratic tradition is the sole reason for high levels of income inequality in those regions. For my research I will focus on three countries in Latin America: Chile, Costa Rica and Uruguay. These countries are similar in many respects such as voting systems, ruling political parties and a history of social welfare systems. They are also considered to be the most democratic in Latin America, yet they differ in the degree and duration of their democratic experiences and all three have different levels of income inequality (Wiarda & Kline, 2014, p. 127, 237 & 261). Thus my focus will be on those three cases to explain the following question: how does the degree and duration of democracy influence income inequality in Chile, Costa Rica and Uruguay?
In order to adequately answer this question, I will first explain the theoretical debate surrounding this broad topic. Then I will proceed to explain how I will test and analyze these cases, why I chose these cases and potential problems surrounding my research. Next I will test these cases and then I will finally conclude that for these three countries there is not strong evidence that there is a relationship between the degree and duration of democracy and income inequality, however my evidence also does not refute the theory either. First it is necessary to explore previous literature surrounding this topic.
Theoretical debate
This debate started mostly with the relationship between democracy and economic growth after the introduction of democracies and Industrial Revolution in the Western world. Many previous studies done during the 1950s until around the 1980s (see Sirowy & Inkeles, 1990 for a list of these studies) have shown that there is not one clear conclusive result on how democracies influence economic growth. Some studies have shown that there is a positive relationship, while others have shown that there is negative relationship and some could not detect a significant relationship between the two variables. Overall those studies have demonstrated that different types of relationships have been found between the two and therefore it is difficult to determine how regime types influence income inequality based on those studies.
4
Recent studies (Acemoglu, Johnson, Robinson & Yared, 2009 and Gerring, Bond, Barndt & Moreno, 2005) have looked at different aspects of this relationship. Acemoglu, Johnson, Robinson and Yared (2009) look at the affect economic growth has on regime type. They approach this debate from a different angle, one in which economic growth influences regime type. Their study does acknowledge that democratic countries with lower GDP per capita incomes are more likely to experience coups. Nonetheless the conclusion was that there was no relationship between countries with high GDP per capita incomes and democracy. Thus the reverse relationship of economic growth on democracy was not significant according to their data. However as Gerring, Bond, Barndt and Moreno (2005) point out, the richest countries in the world are most often democracies, yet it is difficult to find a correlation between democracies and economic growth. Therefore they believed that it is important to study how duration of democracy influences economic growth. Their study counted the amount of democratic stock, using Polity IV scores as their measurement system, a country built over time. They assumed that the more democratic stock a country has, the higher the level of economic growth will be. According to their study there was a significant relationship between the two.
The underlying theory behind this relationship is that countries with more democratic stock have been able to gain more physical, human, social and political capital (Gerring, Bond, Barndt & Moreno, 2005, p. 326 & p. 343). Thus a democracy will have built up more institutions, increased education for more people, improved health and created laws that promote economic growth. Solid institutions and educated healthy adults are essential for economic growth. In addition as in the case for certain countries like Chile, they discovered that the previous democratic stock the country had built up helped it sustain its economic growth during times of dictatorship and crises. Przeworski and Limongi (1993, pp. 62-64) also point out that a certain level of development is needed for democracies to be stable and for economic growth to increase. Instability in the political system is unattractive for investors and can lead to a decrease in investments which leads to a decrease in economic growth. They believed that this reason was the most important factor to explain the positive relationship between democracy and economic growth. With all the research done on this topic, it is not surprising that many of these scholars have suggested that this theory could also be applied to income inequality. Many of these scholars like Przeworski and Limongi (1993) have mentioned that a positive relationship should
5
exist between the two. Democracy creates political equality, so it would be natural that it also creates income equality.
Like earlier studies done with the focus on the economic growth, most studies done on the relationship between democracy and income inequality (see Gradstein & Milanovic, 2004 and Sirowy & Inkeles, 1990 for a summary of these studies) have shown that there is not one conclusive result. Most of the these studies did agree that there is some type of relationship between democracy and income inequality, however it was unclear what type of relationship there was between the two and how strong it was. It is generally assumed that democracies lead to lower levels of income inequality, but some studies like Reuveny and Li (2003) have found that democracies lead to higher levels of income inequality. Most of these previous studies like the studies done on economic growth did not take in account the influence of previous democratic experience or democratic stock can have on income inequality levels. The only study that perhaps comes close is the one done by Timmons (2010). This study does not look at one point in time and tries to incorporate the idea of lagged results, however it is done over an overall short period of time. The conclusion of the study is that there is not a systematic relationship between income inequality and democracy.
Many studies have also been done pertaining to the effects the type of political party in government and voting systems have on income inequality levels. For those studies there are more conclusive results that certain voting system such as proportional representation and political parties such as ideologically left parties lead to lower levels of income inequality (see Cusack, Iversen & Soskice, 2007 and Iversen & Soskice, 2006). In some of these papers it is also mentioned that democracies when compared to non-democracies can lead to lower levels of income inequality due to the fact that the people can vote and will want to vote in favor of the best system for the majority of the people. In addition, citizens have more property rights under democracies and thus have more control in the economy. Once again these property rights should lead to lower levels of income inequality. Non-political explanations are given as well that explain those low levels of income inequality for democracies such as economic factors and technological shocks that also occurred around the times of these political changes. (Boix, 2010, Iversen & Soskice, 2006, Przeworski & Limongi, 1993). It should be noted that these studies were mostly focused on voting systems and political parties and did not study the effects democracy on its own had on income inequality levels.
6
In addition, it should be taken into taken for consideration that for this paper I will focus on how regime types influence income inequality, but there have been studies (Haggard & Kaufman, 2012) done on that have suggested that income inequality influences regime type. Thus a reverse relationship could also be present, but I have decided to focus on how regime type influences income inequality, since for the countries that I selected regime type is often suspected to explain the high levels of income inequality in this region. In addition Acemoglu, Johnson, Robinson and Yared (2009, p. 1055) determined that economic factors are not the most important factors in explaining democratic transitions in countries with colonial pasts. More often factors such as indigenous population density before colonization and constraints on the executive are more important in determining which countries transition to a democracy. Thus it does not seem likely that income inequality would influence regime types in my countries. Nonetheless it should be noted that there is a possibility that a reverse relationship exists for my three countries.
It has been suggested that the high levels of income inequality in Latin America and in other developing countries can be explained by their history of strong authoritarian leaders and their lack of democratic traditions (Hoffman & Centeno, 2003, p. 370, De Ferranti, Perry, Ferreira & Walton, 2004, p. 110 and Wiarda & Kline, 2014, p. 64). There have also been papers (Haggard & Kaufman, 2012) that have suggested that authoritarian regimes lead to lower levels of income inequality. This perspective is known as the conflict perspective. Thus Latin America’s authoritarian past might not be able to explain its high income inequality levels according to this theory. There are also many other factors that can explain the high income inequality levels in this region such as economic policies, social welfare systems, ethnic problems and clientelism (Wesson, 1982, p. 125). In general, there is much confusion about what exactly has led to the high levels of income inequality in this region (Tsounta & Osueke, 2014) yet one of the main reasons that is repeatedly mentioned is the authoritarian tradition in this region and it is the theory that is adhered the most to in the literature for this topic.
It is strange that few studies have been done to see if this factor can explain the high income inequality levels in Latin America. The only comprehensive study that I could find on the topic was done by Huber, Nielsen, Pribble and Stephens (2006). Their study looked if the length of democracy and the prominence of left parties in the government lowered income inequality levels in countries in Latin America. Their conclusion was that both length of
7
democracy and prominence of left parties were important determinants for differences in income inequality levels in that region. Their focus was solely on length while it is also has been shown by other studies focused on economic growth that the degree of democracy such as the human, social and physical capital a country has gained is also an important factor that could be taken into consideration for studies done on income inequality. Although there have been few studies with the focus on Latin America, there have been many large-n studies mentioned earlier that do include Latin American countries. They do not always include all Latin American countries, especially not Costa Rica, but Chile and Uruguay have been studied in these past papers (such as Murtin & Wacziarg, 2014) and they have confirmed that more democratic experience has an influence on income inequality levels.
Although multiple studies have been done on the subject of the relationship between democracy and economic growth and the relationship between democracy and income inequality, limited research has been done on the topic of the influence of the degree and duration of democracy on income inequality. Therefore I would like to make a start in analyzing this relationship with a simple qualitative research done in three Latin American countries, since it has been shown that most often this theory has been applied to explain the high levels of income inequality in Latin America.
Research methodology
To explain how I am going to conduct this research, it is important to give a brief outline on how I will conduct my research before explaining my reasoning why I chose to focus on these specific aspects. I will analyze Chile, Costa Rica and Uruguay over a period of time from 1960 until current times. Each of these three countries has been a democracy for different amounts of time and at different degrees. Therefore I hope that it is possible to explain the variation of their income inequality levels by their democratic experiences.
First I will discuss how I will measure duration of democracy and what my definition of democracy is. For measuring democracy I will use the Polity IV index made by the Integrated Network for Societal Conflict Research. Most researchers (see Gerring, Bond, Barndt & Moreno, 2005, Haggard & Kaufman, 2012 and Papaioannou & Siourounis, 2008) on this topic use the Polity IV index to measure democracy. The index is made by looking at how the executive is
8
elected, the checks on the executive and the type of electoral competition present. Thus my definition of democracy follows those guidelines. That definition might be limited because there are other aspects of democracy that this index does not take into account such as how free the elections are. It is impossible to find an index that measures all aspects of democracy. Still this definition of democracy is considered to be an open and broad one by some (Gerring, Bond, Barndt & Moreno, 2005, p. 338). For my research, I believe that the Polity IV index definition covers the most important aspects of a democracy. According to that index, a country is considered to be a democracy with score of 6 and above on a 21-point scale from -10 to 10. Thus I will consider them to be a democracy when these countries have a score of 6 or above. Along with analyzing their scores from 1960 to 2015 and comparing them with the income inequality scores, I will also look at the amount of democratic stock a country has accumulated over time. Democratic stock will be measured by adding all Polity IV scores to demonstrate how much experience a country has gained in democracy. This way of measuring was also done by Gerring, Bond, Brandt and Moreno (2005) although my method slightly differs since I did not add a yearly 1% depreciation rate and I started measuring the scores earlier than they did. These slight differences should not affect my research, but my democratic stock scores for Uruguay and Chile will be different than theirs. The Polity IV index has scores for Chile from 1818, for Costa Rica from 1838 and for Uruguay from 1830. Since they all have different starting dates, I will only look at the scores from 1838 and onwards so that no country has an unfair advantage or disadvantage. There are some problems with this index outside of its definitions as those previous researchers have also stated. These problems mostly involve the measurement error index that it uses, however most scales have some measurement error problem so it should not be a huge issue that will affect my research. There is one downside for this index regarding the polity scores for Costa Rica. For Costa Rica it does not take in account its brief coup that started in 1948 and ended a year later (Wiarda & Kline, 2014, p. 367). It should not have a great effect on the results since I mostly look at the results from the 1960s and onward, but with democratic stock it should be noted that Costa Rica’s score should be a little lower than it truly is. This slight increase should not impose a problem overall.
The next important concept in my research is degree of democracy. It was defined earlier as the political, social, human and physical capital a country gains once it transitions to a democracy. Degree of democracy ranges from a broad number of topics such as education,
9
freedom of citizens, health and corruption. Therefore it can be measured by various variables, but I have chosen two variables to focus on. The first measure that will be used is Freedom House. Freedom House is quite different from the Polity IV index, since Polity IV index mostly focuses if power is properly distributed and how the elections are held. The Freedom House focuses on the rights that citizens and non-governmental organizations have rather than look at if the government and electoral institutions function in a democratic manner. Many other studies (Acemoglu, Johnson, Robinson & Yared, 2009, Gerring, Bond, Barndt & Moreno, 2005 and Murtin & Wacziarg, 2014) have used Freedom House scores as an additional measure to the Polity IV index. Papaioannou and Siourounis (2008, p. 1538) used it as a primary measure to measure when democratic transitions occurred. They acknowledged that Freedom House was not the best measure for democracy since it took a while to notice a transition based on those results. They used the traditional statuses which only has three categories and makes it more difficult to detect small changes. Therefore I have decided to not to use those statuses. Instead the focus will be on the civil liberties and political rights scores that decide those statuses. These two factors should be able to measure the degree of democracy since they should increase after democratic institutions come into place. There can be a short time lag to see the effects of a democratic transition on the Freedom House scores but it should not be as severe as the time lag for income inequality.
Many researchers (Gradstein & Milanovic, 2004 and Murtin & Wacziarg, 2014) have claimed that education, specifically enrollment rates in colleges and university, is another important factor in determining how democratic a country is. It is considered to fall under the human and social capital a country has gained. Therefore it is important to include these measurements to determine the degree of democracy. I will look at tertiary educational enrollment levels in these three countries since there is the most data on these numbers as opposed to attainment levels. It is said that the degree of democracy can be measured by the amount of people that enroll into tertiary education. It is believed that once a country becomes more democratic, it will produce more tertiary education graduates. More people are willing to go to university and can go to university since less of their rights are restrained. There are some studies (Stasavage, 2005 and Timmons, 2010) that are unsure if there is a relationship between democracy and education and their results were inconclusive. There has also been limited research done on the topic especially for the relationship between democracy and tertiary
10
education. The debate has mostly been philosophical. Thus I will also look if there is a relationship between the tertiary education enrollment and democracy in my three countries. The United Nations Educational, Scientific and Cultural Organization (UNESCO) has a great dataset on education in general and therefore on tertiary education enrollment levels so I will be using that as my measurement for education. The effects of tertiary education can have a time lag, meaning that an increase in tertiary education enrollment rates will not immediately occur as a country transitions into a democracy. Like income inequality, it might take twenty years to see the effects that a transition has on tertiary education enrollment rates (Timmons, 2010, p. 745). The final and arguably most important variable for this study is income inequality. Although there are many ways to measure income inequality, I will use the most well-known and most widely used one: the Gini index. I will use the Gini index as measured by the World Bank as done by many scholars on this topic (such as Reuveny & Li, 2003 and Huber, Nielsen, Pribble & Stephens, 2006). The World Bank has the most data available for my three countries. Unfortunately the Gini coefficient is only measured from 1960 in Costa Rica, from 1966 in Chile and from 1979 in Uruguay, so therefore I can only conduct research from those time periods. Although I will include past democratic experience for these countries as another way to measure their democratic experiences, my main time frame will be from those above mentioned dates. These time constraints should not be incredibly limiting since it is around those time periods that these countries experienced democratic transitions. Uruguay is the only country that might not have enough data but it should not impose extreme constraints on my results.
Furthermore I shall explain why I chose a qualitative research. I believed that it was important to look at the topic more in depth rather than at a more abstract level. On a more in depth level, it is possible to look at different aspects of this theory and to see what part if any part has an effect on income inequality in these three countries. The type of method I used for these studies is a most similar system design. The three countries that I chose are similar in many aspects concerning democracy, history and government which shall be explained in more detail in the next paragraph. These countries do differ in how long they have been a democracy, the amount of democratic stock that they have built and their income inequality levels. Those first two factors should explain the different income inequality levels these three countries have. Thus I chose three cases that have different levels of democratic experience and levels of income inequality, yet are similar in all other aspects in respect to democracy. These three
11
countries are considered to be the most democratic in Latin America today and have had the longest democratic traditions in Latin America and in comparison to some Western European countries (Wiarda & Kline, 2014, p. 127, 237 & 261). In addition, each of these countries has had similar parties in their parliament for the last couple of years. The type of ruling political party (Cusack, Iversen, Soskice, 2007) in the government has also been said to be a factor in explaining different levels of income inequality and by them being quite similar it should rule out this difference. In addition, all of these countries have a proportional representation voting system for parliament and a run-off voting system for the president. Voting systems can also explain differences in income inequality (Cusack, Iversen, Soskice, 2007 and Iversen & Soskice, 2006), so it is useful that this difference can be discarded. The countries also have many similarities in colonial history, population density and the size of their indigenous populations. Although they are not related to democracy, they are all factors that can have an influence on income inequality. There are some important differences between them as well that are not related to democracy or their political systems. These three countries have had different experiences with labor unions in their country and with business related aspects (Wiarda & Kline, 2014). These factors can have an influence on income inequality, but for this research I will disregard these differences, since they are not related to what I am focused on. However these factors might have an influence on the different levels of income inequality in these countries, so they should be taken into account.
Before I continue on to the data, I will explain what exactly I am testing and what results I would consider to support my hypothesis. The main focus on my research is the effect democratic duration has on inequality. According to previous literature, the longer a country has been a democracy, the lower the income inequality levels should be. Thus in order for my results to support this conclusion, all countries should follow this pattern. If one country does not follow this pattern, it already suggests that democracy duration does not influence income inequality. It might have a delay of twenty years as many researchers (Gradstein & Milanovic, 2004 and Sirowy & Inkeles, 1990) have suggested or income inequality levels after a democratic transition might follow an inverse U-shape as Gradstein and Milanovic (2004) and Haggard and Kaufman (2012) have mentioned. If income inequality levels follow an inverse U-shape, it means that they initially increase after a democratic transition before decreasing. With an inverse U-shape there will also be a time lag. Either way income inequality should decrease after around twenty years
12
as a country gains more democratic experience. A second aspect that I will test is if the amount of previous democratic experience that a country has gained since 1838 has an influence on income inequality. Some researchers (Gerring, Bond, Brandt & Moreno, 2005) have suggested that there is a relationship with respect to economic growth but none have looked at income inequality. Thus I will see if the more democratic stock a country has results in a decrease in income inequality levels. In addition I will test if the amount of democratic stock had an influence on how stable income inequality levels remained during the debt crisis in Latin America starting in the 1970s (Wiarda & Kline, 2014). Previous research (Gerring, Bond, Barndt & Moreno, 2005) has suggested that built up democratic stock can add as a buffer during times of crisis.
The next aspect to be discussed is degree of democracy. Here the results of Freedom House will also be looked at, since it is focused on the liberties that citizens have which is an important measure for the degree of democracy. Once again the more free a country is according to the Freedom House scores, the higher its democratic degree is, thus the lower the income inequality should be. The Freedom House results might be slightly lagged in comparison to the Polity IV results since the effects of changes in institutions might take some more time to change, however this lag should have no effect on the results since income inequality also will most likely have a lag. The final thing that I will look at is the amount of people who attained a tertiary education. Like the Freedom House results, they might also be lagged in comparison to the onset of democracy. Tertiary education enrollment rates should be higher when a country becomes more democratic. Thus the higher the tertiary education attainment rates are, the lower the income inequality should be. There is still much debate about this topic in the literature. As has been mentioned earlier certain authors believe that tertiary enrollment should increase with the start of democracy, while others say there is no correlation between regime type and democracy. Therefore I will also make sure that there is indeed a relationship between democracy and tertiary education enrollment in these countries by looking if once a country becomes a democracy, the tertiary education enrollment rates increase. Having explained my research methodology it is time to look at the cases.
13
Case study
For my data collection I will analyze the effects duration of democracy and degree of democracy have on income inequality separately for all of the three countries in order to compare them and determine if there is a pattern in these countries for those two factors. At the end of this section I will go over all of the results combined to give a quick conclusion on the results.
Duration of democracy
In order to see if there is relationship between duration of democracy and income inequality it is useful to include a graph of income inequality levels in each country over the years which can be viewed in figure 1. Income inequality levels started to be measured for each country at different points in time. Costa Rica was the first at 1966 and Uruguay the last at 1978, but these different starting dates should not interfere with the results, since it was at those points that these countries became or started to transition to democracies.
Figure 2 shows the regime type scores for each country according to the data from the Integrated Network for Societal Conflict Research. The data for these countries is taken from 1960 and onwards for the purpose of this research. Countries are considered to be democracies at a score of 6 and full democracies at a score of 10. Costa Rica has been a full democracy since 1960 according to this graph and according to earlier data from the Polity IV index Costa Rica has been a democracy since 1875 and a full democracy since 1890. However this dataset does not include the brief coup d’état Costa Rica had in 1948 (Wiarda & Kline, 2014, p. 367). This factor should be taken in account when comparing figure 1 and figure 2, since according to previous research it could take up to twenty years to see the full effects that democracy has on income inequality.
14 Figure 1
Data was taken and adapted into a graph from all the Gini data created by the World Bank (2014).
If the two graphs are compared and one considers the theory, there are two countries that act according to the theory. Chile is the country that acts most according to the theory. Chile’s score plummeted from 6 to -8 in 1974 which coincides with the drastic increase in income inequality. It transitions back to a democracy at the start of the 1990s. Income inequality only starts to decrease from 2000 onward, however this lag can be explained by the theory. It could take up to twenty years to see the effects that a transition to democracy has on income inequality. The next country that seems to follow the theory is Uruguay. Uruguay transitioned into an authoritarian regime in 1972, however since income inequality was not measured during this time, it is impossible to see if an increase in income inequality followed this transition. However
15
it can be seen that when Uruguay transitioned to a democracy during the 1980s, there was a slight decrease followed by a sharp increase in income inequality. Income inequality started to decrease during the 2000s which has led to Uruguay’s current low scores in income inequality. Once again the time lag and the inverted U-shape income inequality can be explained by the theory. However it should also be noted that at the end of the 1990s there was a Brazilian currency crisis that put Uruguay in a recession for a couple of years (Wiarda & Kline, 2014, p. 241) which could also explain its drastic increase in income inequality levels. Nonetheless Uruguay seems to follow the theory overall.
Figure 2
Data was taken and adapted into a graph from the Polity IV annual time-series created by the Integrated Network for Societal Conflict Research (2015).
16
Costa Rica does not act according to the theory. Costa Rica had been a democracy for ten years at 1960. Initially Costa Rica seems to follow the theory. For the first twenty years income inequality scores decreased resulting in an extremely low score for Latin America of 34.6 in 1986. However after that point, income inequality only increased even though Costa Rica remained a full democracy. In the case of Costa Rica, the longer it was a democracy, the higher its income inequality levels became. Income inequality in the case of Costa Rica does not seem to be affected by regime type and democracy. Perhaps the decrease and increase in income inequality levels could be better explained by non-regime related factors such as economic factors or its loss of strength of its labor unions during this time (Wiarda & Kline, 2014).
The second part of democratic duration that needs to be taken in consideration is the amount of democratic experience a country had prior, also known as democratic stock. Democratic stock is measured by adding all the polity scores Polity IV has for each country from 1838 to 2015. Table 1 shows the democratic stock each country has built or lost starting 1960 until 2015 measured at 15-year intervals. According to the theory, countries that have built up more democratic stock should have lower levels of income inequality than countries with less democratic stock.
Therefore according to this theory, Costa Rica should have the lowest levels of income inequality and Uruguay should have the highest levels of income inequality. Currently Costa Rica has the second highest levels of income inequality of all three countries, nearly surpassing Chile’s scores. Uruguay has the lowest levels of income inequality despite having the lowest democratic stock of the three countries. Thus it seems like using democratic stock as a measure of democratic duration contradicts the theory further. Instead it seems to support the conflict perspective which states that authoritarian regimes lower income inequality levels. Nonetheless this perspective is not the answer for explaining the different levels of income inequality for these three countries, since in the case of Chile the start of the authoritarian regime led to an increase in income inequality levels and the return to democracy led to a decrease in income inequality levels.
However there is a part of the theory that it does seem to support. The theory states that income inequality levels will remain relatively stable in times of economic crisis in countries that have gained more democratic stock. There was one important crisis in Latin America during that
17
period that had an effect on these countries: the debt crisis starting in the late 1970s (Wiarda & Kline, 2014). Both Chile and Uruguay were under authoritarian regimes during this time period. Although some data is missing from Uruguay, income inequality levels seemed to have increased in both Uruguay and Chile during that time period. I earlier said that it could be explained according to the theory because of the authoritarian regimes, however the crisis could also be an important factor in explaining their rising income inequality levels. However in Costa Rica income inequality decreased during this time period despite it also having severe foreign debts. Chile and Uruguay, both with relatively little democratic stock when compared to Costa Rica and under an authoritarian regime, did have a sharp increase during the late 1970s and 1980s. Therefore the amount of democratic stock does seem to act as a buffer for the stability of income inequality levels during times of crises. Nonetheless it is difficult to make finite conclusion because of missing data from Uruguay.
Table 1
Year Democratic Stock Chile Democratic Stock Costa Rica Democratic Stock Uruguay 1960 141 592 -63 1975 189 742 -7 1990 125 892 -16 2005 251 1042 134 2015 351 1142 234
Data was taken and adapted to create a table from the Polity IV annual time-series created by the Integrated Network for Societal Conflict Research (2015).
18 Degree of democracy
The next part of this study is to view the effect degree of democracy has on income inequality. The first part of measuring degree of democracy is analyzing how free a country is. In other words the amount of rights and liberties citizens have. This part of degree of democracy is going to be measured by Freedom House scores. Figure 3 shows the amount of civil liberties and political rights the three countries had since 1972. A score of 8 means that a country is not free, while a score of 1 means that a country is at its most free. The Freedom House scores are slightly lagged in comparison to the Polity IV scores like Papaioannou and Siourounis (2008) said, but they do follow along the patterns of the Polity IV scores. Thus they can be used as a measure for degree of democracy.
The Freedom House scores for these three countries confirm previous conclusions made. The more political rights and civil liberties a country has according to this index, the lower the income inequality levels should be according to the theory. Chile’s political rights and civil liberties scores were high at the start of the 1970s and slowly decreased both reaching a score of 1 at the start of the 2000s. It was at the start of the 2000s that Chile’s income inequality levels also started to decrease. Uruguay’s scores also increased at the start of the 1970s before decreasing during the 1980s and also reaching a score of 1 for both civil liberties and political rights in the 2000s. Its income inequality scores increased and then decreased a couple of years later, but this lag can be explained by the theory. Both scores for Costa Rica were 1 for most of this time period, although civil liberties did briefly go up to 2 between 1990 and 2005. A score of 2 is considered to be free according to the Freedom House. Therefore Costa Rica was considered to free, yet its income inequality scores increased from 1986 and onwards. Like the results for duration of democracy, Chile and Uruguay act as expected, while Costa Rica does not.
The next part of degree of democracy is education, specifically tertiary education. The more democratic a country becomes, the higher its tertiary education enrollment and attainment rates should be. It is possible that there is a slight lag for that changes can be noticed in enrollment levels. Figure 4 shows tertiary education enrollment levels for all three countries from 1970 until 2016. For all three countries, tertiary education enrollment increased with the onset of democracy. Enrollment also slightly increased during the authoritarian regimes in Chile and Uruguay, but the most drastic increase for these two countries occurred when they transitioned to
19
democracies. Therefore it does seem like there is a relationship between democracy and tertiary education enrollment levels. It might be explained by other factors but for this research I will accept that a relationship exists and tertiary education enrollment can be used as a measure of degree of democracy.
Figure 3
Data was taken and adapted into a graph from the freedom in the world data created by the Freedom House (2016)
As tertiary education enrollment rates increase, income inequality levels should decrease. Even though tertiary education enrollment rates increased during periods of authoritarian rule in Chile and Uruguay, the main and most drastic increase occurred when these countries transitioned into a democracy according to the Polity IV numbers. This increase also occurs alongside a decrease in income inequality levels. Costa Rica’s tertiary enrollment rates
20
gradually increase at the same pace. This gradual and steady increase does coincide with the theory since Costa Rica has been a democracy throughout this whole period. However its enrollment rates are lower in comparison to those of the other countries, suggesting that Costa Rica might have lower rates of degree of democracy with respects to tertiary education. Those low rates might explain the increase in income inequality levels of Costa Rica that occurred after 1986. On the other side these low rates in tertiary education enrollment rates might also suggest that there is not an extremely strong relationship between democracy and tertiary education enrollment. Costa Rica did have the most democratic experience so it should have the highest degree of democracy.
Figure 4
Data was taken and adapted into a graph from UIS Statistics data created by the United Nations Education, Scientific and Cultural Organization (2016).
21
Nonetheless the tertiary education enrollment rates are the only data that might be able to explain the increase in income inequality levels in Costa Rica. In addition, these results might also suggest that democracy might not be the most important factor in explaining income inequality levels. Instead it might be the case that factors associated with democracy such as education might be more useful in explaining income inequality levels. The focus of this research might have been on the wrong aspect of democracy. Still all of these results and conclusions should be taken in with a hint of caution. Costa Rica’s score tertiary education enrollment rates increased as much before 1986 as they did after 1986. Therefore tertiary education enrollment rates cannot explain Costa Rica’s income inequality levels alone but they might be a factor worth considering.
Brief summary of results
All of these cases have different results regarding the relationship between duration and degree of democracy and income inequality. Chile is the country that most clearly follows the theory that democracies lead to lower levels of income inequality. It is perhaps the reason why Chile is so often cited and used as an example in papers studying these sorts of relationships (such as Gerring, Bond, Barndt & Moreno, 2005, p. 352). Uruguay has more ambiguous results due to some missing data and due to its inexplicable low levels of income inequality. I assumed that it did follow the pattern, but I do not want to make conclusive statements. Costa Rica, on the other hand, was the clear outlier. It did not follow the theory. Out of all the three countries, it has the most democratic experience yet its income inequality levels have not decreased since the 1980s. They initially decreased before they steadily increased after about 40 years since the last coup d’état. Although tertiary education enrollment did increase with the onset democracy, it did not increase as sharply as in the other countries. Once again there is no relationship with the duration and degree of democracy and income inequality in Costa Rica.
These results are surprising if one considers the theory. Costa Rica, the country with the longest democratic traditions, has rising income inequality levels. Therefore even though the other two countries seem to follow the pattern, Costa Rica demonstrates that this theory cannot explain income inequality on its own. Perhaps other factors come to play when considering Costa Rica. Therefore it is also important to consider more economic explanations such as trade
22
unions, overall privatization and economic policies in these three countries especially in Costa Rica. However it might also be important to look at factors often times related to democracy such as education and health that could explain income inequality levels. In addition it is surprising that Uruguay has such low income inequality levels with arguably the least amount of democratic history and experience. Although Uruguay did somewhat follow the pattern and therefore it somewhat confirms the theory, it is quite confusing why its income inequality levels are so low compared to the rest. Democracy alone cannot explain its low levels of income inequality. In fact it almost seems like the less democratic experience a country, the lower the income inequality is. Thus in the case of democratic stock my research seems to support the theory that authoritarian regimes create lower levels of income inequality more. Still broad conclusions cannot be made from this comparison alone. The main thing that should be taken away from this comparison is that regime type alone cannot explain income inequality levels.
Conclusion
There are different parts of the theory that these case studies agree with and parts of the theory that my case studies do not. These cases do not agree with the notion that the country with the most democratic experience and stock built up has the lowest levels of income inequality, which is the main part of the theory. It does confirm other aspects of the theory. The amount of democratic stock built up did seem to matter on how much countries’ income inequality levels were affected during times of economic crisis. The countries with the least amount of built up democratic stock had a significant increase during times of crisis while the country with the most democratic stock had a decrease in income inequality levels. In addition education as a measure of degree of democracy had an influence on income inequality levels. The country with the lowest tertiary education enrollment rates, thus a lower degree of democracy, also had increasing income inequality levels. Although it can be questioned whether education can be used as a measure for degree of democracy, education in these countries do have the expected effect on income inequality levels according to previous literature.
The most important part of the theory was not confirmed by my research.. The reason why this relationship was not confirmed by my research is most likely because previous literature (Przeworski & Limongi, 1993) did not consider many non-Western countries or
23
countries that were not previously part of the Soviet Union. Most research is focused on Western countries and Western democracies are different in their histories and democratic traditions. In addition, my research focused on three countries including Costa Rica. Costa Rica is often discarded or viewed as an outlier in previous studies (Huber, Nielsen, Pribble and Stephens, 2006) that were large n-studies and did include non-Western countries. Since I focused on three countries, it was not as easy to ignore Costa Rica as a clear outlier. Perhaps if I had included another Latin American country such as Argentina in place of Costa Rica, I might have had different results. Previous studies were not wrong in their conclusions. Their focus was simply broader and on different types of countries than the ones I studied.
The implications of this study are for these three countries during the time periods that I studied, but these conclusions could be considered when explaining income inequality levels in other countries in Latin America with similar backgrounds and histories. Researchers should also consider to look at other explanations for the high income inequality levels in Latin America, because its authoritarian past alone cannot explain the reason why high income inequality levels exist as Costa Rica clearly demonstrates. There must be some other reasons such as perhaps their colonial pasts, economic factors and the strength of trade unions. More research should be done on the topic but the main explanation for high income inequality levels in Latin America cannot solely be explained by only focusing on its authoritarian past. In the future for scholars in Latin America or for other regions where scholars like to apply this theory to explain high levels of income inequality, it is important to consider other factors than solely their authoritarian past.
Bibliography:
Acemoglu, D., Johnson, S., Robinson, J. A., & Yared, P. (2009). Reevaluating the modernization hypothesis. Journal of Monetary Economics, 56(8), 1043-1058.
Boix, C. (2010). Origins and persistence of economic inequality. Annual Review of Political Science, 13, 489-516.
Cusack, T. R., Iversen, T., & Soskice, D. (2007). Economic interests and the origins of electoral systems. American Political Science Review, 101(03), 373-391.
De Ferranti, D., Perry, G. E., Ferreira, F. H., & Walton, M. (2004). Inequality in Latin America: Breaking with history?. Washington, DC: World Bank Publications.
24
Freedom House (2016). Freedom in the world 2016 [Data file]. Retrieved from: https:// freedomhouse.org/report/freedom-world/freedom-world-2016
Gerring, J., Bond, P., Barndt, W. T., & Moreno, C. (2005). Democracy and economic growth: A historical perspective. World Politics, 57(03), 323-364.
Gradstein, M., & Milanovic, B. (2004). Does liberté=égalité? A survey of the empirical links between democracy and inequality with some evidence on the transition economies. Journal of Economic Surveys, 18(4), 515-537.
Greig, A., Hulme, D., & Turner, M. (2007). Challenging global inequality: Development theory and practice in the 21st century. New York: Palgrave Macmillan.
Haggard, S., & Kaufman, R. R. (2012). Inequality and regime change: Democratic transitions and the stability of democratic rule. American Political Science Review, 106(03), 495-516.
Hoffman, K., & Centeno, M. A. (2003). The lopsided continent: inequality in Latin America. Annual Review of Sociology, 29(1), 363-390.
Huber, E., Nielsen, F., Pribble, J., & Stephens, J. D. (2006). Politics and inequality in Latin America and the Caribbean. American Sociological Review, 71(6), 943-963.
Integrated Network for Societal Conflict Research (2015). Polity IV annual time-series, 1800-2015 [Data file]. Retrieved from: http://www.systemicpeace.org/inscrdata.html Iversen, T., & Soskice, D. (2006). Electoral institutions and the politics of coalitions: Why some
democracies redistribute more than others. American Political Science Review, 100(02), 165-181.
Murtin, F., & Wacziarg, R. (2014). The democratic transition. Journal of Economic Growth, 19(2), 141-181.
Papaioannou, E., & Siourounis, G. (2008). Democratisation and growth. The Economic Journal, 118(532), 1520-1551.
Przeworski, A., & Limongi, F. (1993). Political regimes and economic growth. The journal of economic perspectives, 7(3), 51-69.
Reuveny, R., & Li, Q. (2003). Economic openness, democracy, and income inequality: an empirical analysis. Comparative Political Studies, 36(5), 575-601.
Sirowy, L., & Inkeles, A. (1990). The effects of democracy on economic growth and inequality: A review. Studies in Comparative International Development (SCID), 25(1), 126-157.
25
Stasavage, D. (2005). Democracy and education spending in Africa. American journal of political science, 49(2), 343-358.
Timmons, J. F. (2010). Does democracy reduce economic inequality?. British Journal of Political Science, 40(04), 741-757.
Tsounta, E., & Osueke, A. (2014). What is behind Latin America’s declining income inequality? (No. 14-124). Washingston, DC: International Monetary Fund.
United Nations Educational, Scientific and Cultural Organization (2016). UIS Statistics [Data file]. Retrieved from: http://data.uis.unesco.org/
Wesson, R. G. (1982). Democracy in Latin America: Promises and problems. New York: Praeger.
Wiarda, H. J., & Kline, H. F. (Eds.). (2014). Latin American politics and development. Boulder: Westview Press.
World Bank (2014). All the Ginis [Data file]. Retrieved from: http://www.worldbank.org/en/ research/brief/all-the-ginis
