Income inequality and economic growth in Latvia
Ieva Margevica (10360956)
Bachelor Thesis Economics and Business Specialization: Economics
Faculty Economics and Business
Supervisor: Ieva Rozentale
2 Contents Abstract ... 3 1 Introduction ... 4 2 Literature review ... 5 2.1 Economic growth ... 5
2.1.1 Definition of economic growth ... 5
2.1.2 Measuring economic growth ... 6
2.2 Income Inequality ... 7
2.2.1 Definition of income inequality ... 7
2.2.2 Measuring income inequality ... 7
2.2.3 Causes of income inequality ... 8
2.2.4 Consequences of income inequality... 10
2.2.4.1 Social and political consequences ... 10
2.2.4.2 Economic consequences ... 11
2.3 The case of Latvia ... 12
2.3.1 Economic growth of Latvia (1980-2012) ... 12
2.3.2 Income Inequality in Latvia (1980-2012) ... 13
3.1 Data... 14
3.2 Method ... 15
3.3 Hypothesis ... 16
4 Results ... 16
4.1 Unit root tests ... 17
4.2 Cointegration test ... 19
4.3 Granger causality test ... 20
5 Conclusion ... 20
3
Abstract
This thesis studies income inequality and economic growth nexus in case of Latvia throughout the period from 1980 until 2012. More precisely, it carries out time-series data analysis and investigates Granger causality between Latvia’s real GDP per capita and Gini coefficient. The relationship is examined by applying Johansen cointegration test in vector error correction model (VECM) setting. The empirical results for Latvia indicate that there exists long-term relationship between income inequality and economic growth. Besides, this relationship is found to be bidirectional and negative. Hence, Latvian government is suggested to take action in order to decrease income inequality, which would foster economic growth and encourage even lower inequality levels.
4
1 Introduction
One of the main objectives of any economy is positive economic growth since it indicates an improvement in its performance – increase in development, productivity and living standards. In 2012 Latvia was said to be the fastest growing economy among European Union’s countries. At the same time it was identified as the country with the highest income inequality level in the EU.
High income gap between rich and poor is considered to lead to higher poverty and crime rates as well as unfair political decisions (Bourguignon, 2004; Becker, 1968; Lichbach’s, 1989). The combination of these negative effects is said to slow down economic growth. Accordingly, the relationship between income
inequality and economic growth can be perceived as negative. Therefore, several economies are focused on decreasing income inequality levels within the country rather than boosting economic growth. However, if we look at the case of Latvia the negative relationship is not directly observable, since the economy has experienced simultaneous growth of gross domestic product (GDP) and income inequality
(Domrovskis, 2011; Masso et al., 2012). Hence, it might be the case that the
relationship between income inequality levels and economic growth is positive, which would mean that higher income inequality renders higher economic growth or vice versa. In order to understand whether decreased income inequality rates would prompt Latvia’s economic growth the relationship between the two metrics has to be studied.
Two different opinions about the relationship between income inequality and economic growth exist. The first one admitted by researchers like Gruen and Klasen (2008), Kaldor (1957) and Castello-Cliemnt (2010) posits a positive relationship between economic growth and income inequality. It suggests that income inequality is an essential condition for growth of an economy. It states that the wealthier the rich part of society, the more they save and the more they invest and generate
economic growth. However, the opposite opinion defended by others like Keefer and Kanck (2002) as well as Galor and Moav (2004) is that more unequal societies experience slower economic growth. In their opinion higher Gini coefficient and income inequality leads to lower GDP growth rate and economic growth. They argue that with an increase in inequality level of human capital decreases, socio-political instability increases which discourages investment and decreases economic growth.
Although the relationship between the two measures has been extensively discussed in both older and more recent literature, there still exists an ongoing debate about correlation of income inequality and economic growth. Does it exist? Is it positive or negative? How strong is it? Does it vary with the stage of economy? These questions are still central while discussing the relationship. Therefore, the aim of this paper is to estimate the extent to which income inequality is related to
economic growth.
In order to answer the proposed research question the case of Latvia is studied. The case of Latvia is used since within this country both economic growth and income inequality have been increasing simultaneously. That contradicts the most common belief of the negative relationship between the both terms.
To understand the relationship, first of all, relevant literature on the concept of income inequality and economic growth is studied. Later on, the analysis of Latvia’s income inequality level and economic growth is presented based on statistical data retrieved from the World Bank (2014) and Solt (2014). Finally, the inequality-growth relationship is analyzed by implementing bivariate cointegrated vector error
5
correction model (VECM) approach. Consequently, the work is perorated by
conclusions made from both literature review and empirical analysis. The retrieved results could serve as an aid in finding appropriate policy implications to increase Latvia’s well being.
2 Literature review
In order to gain a better understanding about the concept of the relationship between income inequality and economic growth, first of all, both of these complex
phenomena are studied separately. The study commences with defining economic growth and continues with description of its measurement. Later on, the definition and measurement, causes and consequences of income inequality are presented. Further on, the relationship between the inequality and growth are discussed. Afterwards, the case of Latvia is described and economic growth and income inequality patterns within the country during the period of 1980 until 2012 are introduced. Finally, the main aspects essential for the empirical model are summarized.
2.1 Economic growth
2.1.1 Definition of economic growth
Economic growth along with price stability and full employment is one of the main macroeconomic goals of aggregate economy (Goodwin et al., 2008; Rittenberg & Tregarthen, 2009). It is a long run process that can be obtained by increasing economy’s production of goods and services over time. The positive growth of an economy indicates its level of potential output and suggests that it has a rising ability to produce goods and services. Additionally, economic growth is said to improve nation’s material well-being, and hence is believed to be a good indicator of wealth and living standards of a country. However, in order to translate economic growth into a higher level of living standards the economic growth has to exceed population growth. Growing economy can be described as the state of a country’s economy in which it is getting richer due to higher production leading to higher incomes, which consequently increase demand for goods, increase production even further and generate wealth (Rittenberg & Tregarthen, 2009).
Economic growth has a tendency to occur in stages. The pre-growth stage is described by very basic living conditions within a country. When entering the early stage of growth, country’s labor flows from agriculture into industry and services. Later on, labor flows from agriculture and industry to services. Finally, when country develops further, it starts purchasing modern capital from advanced countries and adopts compatible production techniques, hence, becomes competitive relative to advanced countries. In the later stage country can experience a long-lasting, rapid growth until it reaches levels of advanced countries. Eventually, the phase of unsustainable growth ends and the developing country converges with advanced countries having a moderate growth (Blakemore & Herrendorf, 2009).
The most well know growth model is based on endogenous growth theory. In line with this model, growth is generated by economy’s internal processes. It is the effect of an increase in both quality and quantity of human and physical capital as well as the innovation that leads to technical progress. This rise can be achieved by
6
investing in human capital, innovation and knowledge. According to the model, these investments raise the productivity of labor and thereby increase the potential
consumption and economic growth. Another important contributor to both investment and economic growth are savings (Hindriks & Myles, 2006; Rittenberg & Tregarthen, 2009). Apart from the endogenous growth model, economic growth can be also looked at from classical or neoclassical growth theory perspectives. The endogenous model is more adequate model and is thereby used in this thesis, since it takes into account technological progress and explains where the technological changes come from.
2.1.2 Measuring economic growth
In order to compare and monitor development of economies across different
countries and times an indicator of economic growth has to be present. A percentage rate of change in gross domestic product (GDP) of a nation serves as such a tool (Hinser, 2010; Blakemore & Herrendorf, 2009). However, when comparing GDP over time the growth might misleadingly appear because of the increase in price levels or due to growth of population. To obtain a meaningful interpretation of GDP growth the impact of inflation and differential population growth across different countries and times has to be eliminated. To avoid the influence of inflation GDP has to be measured in real terms, whereas to prevent the impact of population growth
differences it has to be measured in per capita terms. Such an adjusted indicator - real GDP per capita - provides insight about productivity of the economy and its real living standards (Blakemore & Herrendorf, 2009).
GDP is calculated by taking into account only the amounts of goods and services produced within the country (Rittenberg & Tregarthen, 2009). There are three approaches in calculation of GDP. When implementing output approach GDP is derived by summing up the value of newly generated goods and services less the value of all goods and services used in their creation, plus taxes less subsidies on products. GDP can also be calculated by expenditure approach, where it is defined as sum of private final consumption expenditure, government final consumption expenditure, gross capital formation, exports less imports. Finally, income approach can be used to obtain GDP by summing up compensation of employees, net taxes on production and imports, gross operating surplus and mixed income (Eurostat, 2012). The three methods of GDP calculation should yield approximately the same results. However, the expenditure method is said to derive the most credible data due to the fact that its components are more reliable in comparison with the ones of income or production approach (Pritzker et al., 2014).
Instead of GDP alternative measures of growth such as gross national product (GNP), gross national income (GNI) or net domestic product (NDP) can be used to determine economic growth. While GDP delivers information about the strength of country’s economy GNI and GNP indicate its residents’ performance by accounting also for net income receipts from abroad not only within the country. It is worth noting that GDP does not count for depreciation, unpaid work, black-market activities, it excludes sales of second-hand products as well as transfer payments made by government and it does not reflect country’s natural capital. One of the adjusted forms of GDP is NDP which takes into account depreciation of country’s capital assets. Although economic growth can be measured by adjusted GDPs and national measures, still GDP is the most widely used one. Hence, it facilitates data
7
availability and cross-country as well as over time comparison and is therefore used within this paper.
2.2 Income Inequality
2.2.1 Definition of income inequality
Income is an essential part of any economy. It indicates earnings generated over period of time and the amount that one can spend during the period. Since it stimulates demand income can be considered as an engine of an economy.
Income inequality can be defined as an unequal distribution of income across society. It is stated that income inequality can be captured by comparing the
inequality either between nations based on their average GDP or between
individuals based on each person’s actual income. Income inequality identifies the gap between poor and rich individuals, households or nations. This allows reviewing income inequality at both national and global level (United Nations Development Programme, 2013).
According to the sociology theory, income inequality is one of the forms of social inequality. In general, income can be seen as a monadic attribute of value, which means that one’ s income can be defined without knowing the income of others. However, the process of acquiring income has relational properties since one’s acquisition of income can have an effect on other person’s income. Hence, income inequality can be viewed as ’’a relational process for distributing monadic attribute’’. Even though income is a monadic attribute it has a social content. Therefore, the degree of income inequality can be understood only if income of several individuals is known and can be compared. (Wright, 1979).
2.2.2 Measuring income inequality
Several measures, for instance, Gini coefficient, Theil index, decile dispersion ratio or share of income/consumption of the poorest have been developed for the purpose of evaluating and comparing income inequality. However, it is claimed that Gini coefficient, developed by the Italian sociologist and statistician Corrado Gini in 1912, is the most widely used measure of inequality. It measures the extent to which individuals’ or households’ income distribution deviates from a perfectly equal distribution (The World Bank, 2011).
Gini index can be graphically represented by Lorenz curve and the 45-degree line of equality. In the Figure 1 with cumulative share of total income on the y axis and cumulative share of people from lowest to highest incomes on x axis, Gini coefficient can be obtained by dividing area A with the total area (A+B) under
equality line. If complete income equality was present then Lorenz curve and line of equality would merge leading to Gini coefficient of 0 (The World Bank, 2011).
8
The coefficient takes up values between 0 and 1. The value of 0 stands for a complete equality when everyone receives the same level of income, whereas the maximum value of 1 indicates situation where all income belongs to one person - complete inequality. Hence, lower values of Gini index correspond to more equal income distribution while values close to 1 indicate income inequality (The World Bank, 2011).
Gini index as a relative and simple measure of income and wealth may lead to inaccurate, inconsistent, flawed and deceptive results. Furthermore, it is said to be unable to separate different types of inequalities and appears to be more sensitive to inequalities in the middle part of the income spectrum (Mellor, 1989; Cowell, 1995; Bellu & Liberati, 2006; Krol & Miedema, 2009; Atkinson & Brandolini, 2001).
Despite its limitations and contradictions Gini coefficient is the most popular measure of income inequality. Among the other inequality measures it is said to have the most benefits and is therefore used within this paper. First of all, the common usage of Gini index and its graphical representation of income inequality allows for comparison over time and between countries. Secondly, it is simple to calculate. Besides, it can be calculated both for individual as well as household data. Moreover, the data are readily available and Gini is easy to interpret (Krol & Miedema, 2009).
2.2.3 Causes of income inequality
The most famous hypothesis of the relationship between inequality and economic growth was developed by Simon Kuznets (1955). His work presented an idea that in the early phases of economic growth there is more unequal income distribution followed by a period of stabilization and ending with more equal income distribution during later stages of growth. This conclusion is now called the Kuznets inverted U hypothesis. Kuznets explained the observed trajectory, presented in the Figure 2, by stating that inequality rises in the early stages of growth due to nation’s shifts from rural or agricultural to urban or industrialized production. Later on, when a critical point with a certain level of GDP per capita is reached the inequality starts declining over time due to the overall industrialization.
Figure 1: Cumulative share
of total income by income groups
9
Several researchers have worked on testing Kuznets theory and have found evidence that there is no trend for countries to follow Kuznets inverted U hypothesis (Fields, 1989; Deininger & Squire, 1998; Rodrick, 1999; and Krueger, 2002). For instance, Kuznets hypothesis has been invalidated by the combination of economic growth and relative equality in low-income countries of East Asia as described in works of Fei et al. (1980), Wellisz and Saw (1993) and Birdsall et al. (1995). In fact, recent findings of Bourguignon (2004) showed that the relationship between
economic growth and inequality is country specific and such a generalization as U hypothesis cannot be made.
The recent rise in income differences between rich and poor are argued to be mostly driven by increased inequality in wages (Angel, 2011) and free market
capitalism (Piketty, 2014) . Apart from that there are several other determinants of income inequality and increased income dispersion.
One of the factors linked to free market economies and having an impact on inequality level is the openness to international trade. It is argued that reduced import tariffs, increased amount of imports and exports drive down the income gap. It is claimed that when countries open up for international trade the most exports come from agriculture sector. Generated export opportunities increase demand for
agriculture goods and generate more profits to agriculture sector. Since agriculture sector is usually run by poorest part of the society openness to trade generates more profits to poor and reduces the gap (Jaumotte et al., 2008). Trade restrictions would be expected to lead to adverse, negative effects. In addition to trade, movements towards capitalism and financial globalization affect the inequality level. Both
capitalism and financial globalization stimulate foreign direct investment. Investing in foreign country is associated with rising inequality since it is usually oriented at high skilled sectors of host country (Cragg & Epelbaum, 1996). As FDI’s increase
employment opportunities and income benefits for those who already have relatively higher income foreign direct investment contributes to income inequality (Jaumotte et al., 2008).
Apart from components of free market economies that create differences in wages, there are also other determinants of wage and thus income gap. Firstly, it is immigration into a country that usually increases low-skilled labor force and drives down compensation for lower skilled workers. Lower skilled workers usually are the poorest part of the society. Hence, lower wages for lower-skilled workers increase income inequality (Lerman, 1999). Secondly, it is technological progress contributing to differences in wages and later on to income distribution. Advanced technology increases demand for more skilled labor as well as replaces lower skilled with
10
machine labor. The increased demand for more competent workers creates new jobs and opportunities for highly skilled ones and decreases it for lower skilled people. This tendency thus increases highly skilled and diminishes lower skilled worker wages and income. Hence, the advancement of technology can be considered disadvantageous for lower skilled, poorer part of society and, therefore, is said to be positively related with income inequality (Jaumotte, Lall, & Papageorgiou, 2008).
Among other factors that impact income distribution one can also mention changes in family structure (Percheski, 2008), access to education (Becker &
Murphy, 2007), amount and progressivity of taxes (Alberto & Rodrick, 1994) and rent seeking (Stiglitz, 2012). Beyond that better democracy can positively influence the situation of income inequality, while corruption and having no direct access to sea or ocean can worsen it (Aradhyula et al., 2007).
2.2.4 Consequences of income inequality 2.2.4.1 Social and political consequences
There are several reasons why income inequality is an important concept for empirical investigation. Income inequality as discussed in the book of Grottschalk and Justino (2006) has negative effects on stock of human capital. British
researchers Wilkinson and Pickett (2009) have noted that countries with higher inequality level have higher risk of facing social problems like teenage births, child conflicts, drug usage, lower life expectancy, poorer educational performance, distrust among strangers, parenting problems and issues of social mobility. This relationship is said to be caused by division into the classes and the competition in between them. Another social impact of income inequality mentioned by Becker (1968) and Sala-i-Martin (1996) is increased crime rate and violence. This is supported by authors like Daly et al. (2001) and Lederman et al. (2002) who have identified high and significant correlation between inequality and violence, crime and homicide rates.
Income inequality is said to cause not only social but also political conflicts. Lichbach’s (1989) evaluation of different studies on the topic of relationship between economic inequality and political conflicts has acknowledged positive correlation between inequality and political problems. As noted by Stiglitz (2012) one of the main issues related to income inequality is fairness - many consider income of the richest not fairly obtained. The author stated that the feeling of unfairness decreases society’s trust in government, country, and democracy, political and jurisdictional system. On top of that, according to Gottschalk and Justino (2006) illiterate and poor - people with the lowest income level - are often excluded from active participation in political life. As believed by authors, these individuals either cannot afford to vote or their votes are corrupted by political candidates. Following the idea presented by Weyland (1998) this leads to fiscal goals that do not take into consideration the needs of poor. That hereafter creates political arena that benefits the rich. Hence, higher inequality results in unfair political decisions.
Findings of Wilkinson and Pickett (2009) have also showed an increase of health problems like obesity, psychological stress, depression and other stress related illnesses due to the increased inequality level. In accordance with authors the previously mentioned social and health problems are not restricted to the lower class but rather extended to the whole society.
11
2.2.4.2 Economic consequences
The existing literature mentions four main channels through which inequality can affect economic growth independent of the state of economy. Gruen and Klasen’s (2008) experiments have proved that high relative income plays a major role for perceiving higher welfare. Authors have explained it by the fact that higher income distribution can encourage people to work more efficiently since the rewards of additional effort are higher in comparison to the situation where everyone is
rewarded equally. In their work Rooth and Stenberg (2012) have acknowledged this argument by showing that increased income inequality in Swedish regions boosted economic growth. This theory has also been approved in Mahy et al.’s (2011) paper where higher wage differences are associated with increased productivity and hence growth.
In accordance with both Kaldor (1957) and Castello-Climent (2010) income inequality encourages investments which in return stimulates economic growth. Both authors suggested that the higher income level the higher marginal propensity to save. Therefore, as proposed in both papers, if savings and investment rates are positively related, societies with more unequal income distribution will have a higher growth rate.
Finally, Galor and Tsiddon (1997) in their paper reported that capital
concentration is essential to construct new activities with high initial costs. Hence, according to them a country with more unequal income distribution would have a higher output growth in comparison with more equal one.
In contrast to the previously mentioned positive effects of income inequality on economic growth, there are also widespread reasons why income inequality reduces economic output. In their paper Keefer and Knack (2001) have acknowledged that unequal societies are less socio-politically stable. Hence, violent protests, ethnic tensions and social polarization are more common in such environment. As noted by authors surroundings of reduced security of property and contract rights discourage investment, which ultimately reduces economic output.
Another channel of negative effects of inequality discussed by Galor and Moav (2004) is related to decrease of human capital in unequal societies. Authors in their work stressed credit market imperfections that prevent people with lower
incomes to fully realize their potential in knowledge building and education thereby inhibiting society’s economic growth. They also noted that this negative effect of income inequality through human capital channel has become more crucial since in current knowledge-based economies the economic importance of schooling has increased.
Regarding the previous studies on the topic of income inequality and economic growth relationship as well as income inequality’s causes and
consequences there has not been established one particular direction and pattern of positive or negative influences that the two variables follow. However, at all times, when studying the relationship one has to take into account the effect income inequality and economic growth has on human and physical capital, technological progress, immigration rates, social and political equality, investment and saving rates, the amount of foreign direct investments and international trade patterns. Economic growth’s and inequality’s effect on these parameters vary across different countries dependent on their past and present economical as well as the social situation, hence, no overall judgment about the relationship can be applied.
12
2.3 The case of Latvia
2.3.1 Economic growth of Latvia (1980-2012)
As a post Soviet state Latvia has had a difficult history that has influenced its economic growth. Ever since the end of the Second World War up until 1990 both Latvia’s political and economical life were entirely sovietized and country’s economic growth was closely related to the economy of the Soviet Union.
Latvia’s GDP path starting from 1990 until 2012 is characterized in the Figure 2. It shows that after regaining independence in 1990 Latvia’s economic model was changed from the Soviet centralized economic planning to a free market economy. Nevertheless, during the first years of regaining independence Latvia’s GDP was falling. It has been noted that from 1990 to 1993 the GDP declined for at least 49 per cent (United Nations Economic Commission for Europe, 2000). The decline was related to the Soviet budget deficit, changes of economical model, increasing prices and a lack of foreign currency reserves that could have stimulated country’s trading opportunities (Dombrovskis, 2011; Kajaks, 2013). Only in 1994 Latvia’s economy steadied. This was a result of a signed EU Free Trade Agreement, recovery in light industry, stabilization of the local currency and development of commercial and financial markets. In 1995 Latvia’s economy was hit by a banking crisis related to the bankruptcy of Latvia’s largest bank Banka Baltija. Due to the crisis in 1995 Latvia’s GDP contracted by 2.1 per cent (Dombrovskis, 2011; Bitans & Purvins, 2012). When Latvia’s economy recovered and reached a GDP growth of 8.3 per cent in 1997 it again got interrupted by Russian financial crisis in 1998. Since Latvia was no longer dependent on Russian market its growth rate fell only by 3.3 per cent in 1999 (OECD, 2000). Thanks to the Latvian entrepreneurship, rise in manufacturing output, growth in transport and communication sectors and growth of investments related to foreign capital inflows the economy regained its stability very rapidly. After 2000 for 4 consecutive years Latvia’s growth rate was on average 7.6 per cent, which was significantly above EU average (Ministry of Economics of the Republic of Latvia, 2005). Since joining EU in 2004 until the middle of 2008 Latvia was considered as the fastest growing economy in EU. In 2006 its GDP reached 12.2 per cent. In 2008 Latvia was hit by the financial crisis and the burst of credit bubble. Hence, in 2009 the country experienced the most dramatic economic downfall in EU - its GDP fell by 18 per cent. Economic situation improved in 2010 and by 2011 Latvia was among the fastest growing economies in the EU again. From 2010 to 2011 Latvia’s GDP increased by 5.5 per cent while other EU economies grew by 1.5 per cent. The rapid growth this time was explained by change in the economic structure that moved the
focus from private consumption to exports (''Svarīgāko rādītāju prognozes'', 2011). In
2012 Latvian economy grew even more than predicted and reached the growth rate
of 5.6 per cent (Bicevska, ''IKP kāpums Latvijā'', 2013).
Figure 2: Latvia’s GDP path 1990-2012 (IMF, World Economic Outlook database,
13
2.3.2 Income Inequality in Latvia (1980-2012)
During the period of 1980-2012 Latvia experienced a lot of changes in its macroeconomic environment. Being a part of Soviet Union, regaining the
independence and establishing its own economic model has also caused changes in the level of income inequality within the country (Masso et al., 2012).
In the Soviet times (1980-1990), as mentioned in the research of Reutersward (2003), income inequality in the Baltic States was low and similar to other Soviet countries. Referring to the author, the Soviet system was stimulating income equality within Baltic countries since it was oriented in achieving equality in wages and non-labor incomes, full employment and state ownership of means of production, increased social transfers and negative attitude towards private entrepreneurship. Later on, after regaining independence, the Baltic States began an active
participation in entrepreneurship and capitalist market relations, which resulted in an increase in real income, but also led to increased differences between social
classes. Hence, at the beginning of 1990s income inequality dramatically increased in all three Baltic countries and was the highest among other Central and Eastern European countries. This increase was caused by the growing difference in wages and other individual qualifications (Reutersward, 2003). In the second half of the 1990s inequality stabilized in Estonia and Lithuania, but Latvia continued to
experience an increase in inequality throughout the whole period after regaining the independence. According to Fofack and Monga (2004) this continuous increase can be explained by increased risk-rewards in the post-socialism period, increased importance of education and limited opportunities for the poor both education and asset wise.
From the three countries Estonia always used to have the highest inequality level. However, in 2003 Latvia reached Estonian inequality level and since then continued to show the highest inequality scores. In 2005 inequality in Latvia steadied and the inequality rates stopped increasing. This is explained by government actions undertaken due to the parliamentary elections in the upcoming year. In order to gain more votes the government increased support for farmers and social transfers, which drew down income inequality. The improvements were not long lasting and in the next year differences in incomes increased again. During this time it was the real
14
estate bubble that caused sharp increase in wages and income from property, while other income, including pensions and social benefits grew more slowly (Bicevska, 2012). Inequality decreased again during the crisis starting from 2008. This tendency is explained by changes in the income of retired people relative to the dynamics of workforce wages. Besides, during this period of time Latvia experienced decrease in income differences between cities and rural areas due to EU’s direct agricultural aid. However, in 2010 additional taxes were introduced on some support measures in Latvia, which decreased income for rural areas and again increased income inequality (Masso et al., 2012).
In 2012 income inequality in Latvia improved slightly and Gini coefficient declined from 35.7% in 2011 to 35.2% in 2012. Although, the equality in Latvia started to recover, in 2012 the country still had the highest inequality rates in EU. Similar indicators were recorded in Spain, Portugal and Greece. In Latvia’s neighboring countries Lithuania and Estonia Gini coefficients were lower and amounted to 32% and 32.5% (Central Statistical Bureau of Latvia, 2014).
Overall, growing economy is characterized by higher incomes, increased consumption, increased demand for goods and more production; hence it is one of the most important goals of any nation including Latvia. The growth could be achieved by increased quality and quantity of human and physical capital;
innovation; technical progress – investment in technology, innovation, knowledge; savings. As discussed in the earlier mentioned literature income inequality might both directly and indirectly affect these growth determinants and lead to changes in economy’s development and vice versa. With regard to the case of Lavia, the lowest inequality levels are associated with pre-growth stage when country was a part of Soviet Union and faced basic living conditions; sharply increasing inequality appeared in early growth stage and continued to increase in middle and pre-advanced growth stages which is explained by unequal access to education,
openness to international trade and foreign direct investment; inequality in Latvia has stabilized couple of times but permanent steadiness has been present only since the final stage of growth. Hence, one might conclude that in case of Latvia economic growth goes together with higher income inequality levels. However, there are other factors influencing the relationship and the causes of both income inequality and economic growth might not be directly related. Since there is no consensus on the direction of the income inequality effect, it is essential to study the relationship in order to be able to address the high income inequality rates in the most appropriate manner.
3 Methodology
This paper aims to investigate the relationship between income inequality and economic growth in Latvia. Even though this relationship can be influenced by such factors as poverty levels, inflation and foreign direct investment, within this paper the simple specification for the relation GINI-GDP is discussed. It is estimated by
bivariate cointegrated VECM approach regarding only 2 data sets as suggested by Risso et al. (2013) who have studied the relationship in Mexico.
3.1 Data
To study the long run relationship between income inequality and economic growth in Latvia real GDP per capita and an estimate of Gini index are analyzed over the
15
period from 1980 until 2012. Due to the limited data availability, first of all, the time period from 1986-2012 is analyzed. Secondly, following Grun and Klasen’s (2000) suggestion, the period from 1990 to 2008 is studied separately. The authors believe that patterns of life satisfaction in transition countries have to be analyzed some time after the initial transition shock, hence, it allows to focus on the impact of income and its distribution and eliminates the effect of the transition. Since, Latvia’s economy has experienced shock related to change of economical model in 1990 and a financial crisis in 2008, the ‘neutral’ period from 1990 to 2008 is studied. Even though this leads to a sample of only 18 observations, the data from year 1990 are more reliable since the prior data were generated in Soviet Union times and might not be correct.
Data about Latvia’s real GDP per capita are gained from the World Bank database. The World Bank is calculating annual GDP by production approach. Hence, GDP is defined as output of goods and services less intermediate consumption plus taxes and less subsidies on products. The calculation is done without deduction for depreciation of fabricated assets or for depletion and degradation of natural resources. To retrieve GDP in per capita terms gross
domestic product is divided by midyear population. To gain data that are expressed in real terms constant-dollar value is used.
To obtain data about the income inequality Standard World Income Inequality Database created by Solt (2014) is used. This database is formed from standardizing observations that have been collected from such statistical offices as Eurostat,
OECD Income Distribution Database and the World Bank. It provides all available income inequality data for the broadest sample of years and countries. Within this paper variable Gini-net is used as an estimate of Gini index. It is an estimate of inequality equalized by household disposable income using Luxemburg Income Study data as the standard.
As suggested by Risso et al. (2013) both real GDP per capita and Gini-net are expressed in logarithmic form. This enables the model to interpret percentage
change in the explanatory variable in terms of percentage changes in the dependent variable.
3.2 Method
Since both real GDP per capita and Gini-net data are collected over period from 1986 to 2012 time series analysis is performed. To answer the research question about existence, volume and direction of correlation between income inequality and economic growth Granger causality has to be tested. The regression model and method of testing Granger causality depends on the stationarity of each variable separately as well as on their cointegration. The aim of this analysis is to study the following equations:
ln(GDP)(t) = a1 + b1ln(Gini)(t) + u1(t)
ln(Gini)(t) = a2 + b2ln(GDP)(t) + u2(t)
To begin with, it has to be estimated whether time series of Gini-net and real GDP per capita are stationary or have a stochastic trend – long-term movement of a variable that varies over time. In order to test for the stationary augmented Dickey and Fuller (ADF), Phillips and Perron (PP) as well as Kwaitkowski-Phillips-Schmidt-Shin (KPSS) tests are performed. If both variables are non-stationary, their order of integration is estimated by repeating the same tests on their differences until
16
If Gini-net and real GDP per capita are both I(1) vector autoregression (VAR) method is used to describe the Granger causality. When variables are integrated of the same order other than 1 then vector error correction model (VECM) is used. Otherwise, an alternative procedure introduced by Toda and Yamamoto (1995) can be employed to study the Granger causality.
Finally, the Wald test is performed to test the causality direction between real GDP per capita and Gini index.
3.3 Hypothesis
Since there has not been conducted a prior research about the relationship of
income inequality and economic growth in case of Latvia, the hypothesis is based on the previous study regarding the relationship in other post Soviet countries that share the same history of transition from planned to market-oriented economies. In Nina Torm’s (2003) discussion paper focusing on poverty, inequality and economic growth nexus in Armenia, Kazakhstan, Kyrgyzstan, Moldova, Tajikistan and Uzbekistan suggests that even though these countries did not share the same human
development level, composition of economies or transition strategy, the transition process has eventually lead them in the same economic and social paths. Based on this paper that has been composed regarding several researches on the topic of poverty, income distribution and its social and economic effects within the previously mentioned countries, the following hypotheses are developed:
1. There exists a long term relationship between economic growth and income inequality;
2. Income inequality Granger causes economic growth and this relationship is negative;
3. Economic growth Granger causes income inequality. 4 Results
Plots of the time series of both Latvia’s Gini-net and real GDP per capita are displayed in the Figure 3. It clearly shows that Latvia’s economy and GDP has experienced 2 breaks in 1990 and 2008. The first break is explained by change of the economical model whereas the second break has occurred due to the financial crisis. During the period of 1990 until 2008 we can see that Latvia’s GDP and Gini-net both exhibit distinct trends and are growing. To overcome the influence of breaks the period of 1990 to 2008 is studied separately. The whole period from 1986 to 2012 is studied as well in order to have a larger sample.
17
4.1 Unit root tests
To study the stationarity of series unit root test has to be performed. In order to generate more reliable and robust results three tests, firstly, Augmanted Dickey-Fuller (ADF) and Philips-Perron tests and secondly Kwaitkowski-Phillips-Schmidt-Shin (KPSS) test is applied.
Since ADF is said to have low power for highly persistent series it is complemented by Philips-Perron test. The null hypothesis for both tests is non-stationarity identifying that the variable had one unit root I(1). If the null hypothesis is rejected it means that the variable is stationary. Both of these tests are left-tailed, hence, the null hypothesis of non-stationarity is rejected when test statistic is lower than critical value. First of all, the original data are examined. If non-stationarity is found in the original data the first differences are being tested. Later on, the second differences are tested. The testing continues until stationarity is found and the null hypothesis is rejected. In each level the data are examined by looking at the equations with intercept, intercept and trend and none. When intercept is used a constant is included in the equation; when intercept and trend is used both constant and trend are included; when none is used neither constant nor trend are included in the equation. The null hypothesis is rejected only in case if variable is stationary in all three situations. The difference at which variable is stationary indicates the maximum integration order.
The previously mentioned tests for stationarity, are said to have a low power of rejecting null hypothesis. Therefore, to avoid this KPSS test is taken to
complement ADF and PP tests. KPSS test has a null hypothesis of stationarity, hence, if the null hypothesis is rejected it means that the variable follows random walk and is non-stationary. KPSS test is right-tailed, meaning that null hypothesis is rejected when test statistic is higher than critical values. The same as in ADF and PP tests, firstly the original data are tested and then testing continues for differences until the null hypothesis does not fail to be rejected. In order to be able to reject the null hypothesis test statistic has to fall into rejection region in both cases - with constant and trend and solely a constant.
The results of unit root tests for Latvian real GDP per capita and Gini-net in levels and differences are presented in Table 1 for time period of 1986 to 2012 and in Table 2 for time period of 1990 to 2008. Numbers in the table are the t-statistics results analyzed by the tests and subscripts a, b and c indicate significance at 10%, 5% and 1% respectively. According to the tests, Latvian real GDP per capita as well as Gini-net taken in logs are integrated processes of second order I(2) independent of which time period is regarded. Since, data from both periods have revealed similar
20 00 40 00 60 00 80 00 1 00 00 G DP 1980 1990 2000 2010 year 20 25 30 35 40 G IN I_ N E T 1980 1990 2000 2010 year
18
results the following analysis is conducted regarding solely the period of the largest sample (1986-2012).
Table 1: Unit root test results (1986-2012)
ADF Phillips-Perron KPSS
lnGDP lnGini lnGDP lnGini lnGDP lnGini
Levels Intercept -0.220 -1.886 -0.658 -1.471 1.57c 2.59c Intercept and trend -1.481 0.988 -1.646 0.057 0.497c 0.551c None 0.832 4.371 0.610 2.828 First differences Intercept -2.749a -1.914 -2.785a -1.854 0.409a 0.863c Intercept and trend -2.897 -2.634 -2.916 -2.534 0.156b 0.362c None -2.747b -1.390 -2.779c -1.258 Second differences Intercept -5.425c -7.981c -5.515c -7.503c 0.0391 0.142 Intercept and trend -5.311c -8.568c -5.390c -8.244c 0.0397 0.0448 None -5.547c -8.158c -5.650c -7.653c a
Null hypothesis rejected at 10%
b
Null hypothesis rejected at 5%
c
Null hypothesis rejected at 1%
Table 2: Unit root test results (1990-2008)
ADF Phillips-Perron KPSS
lnGDP lnGini lnGDP lnGini lnGDP lnGini
Levels Intercept 0.306 -6.594c -0.178 -5.504c 1.46c 1.81c Intercept and trend -4.420c -0.118 -6.520c -0.254 0.335c 0.447c None 0.964 6.293 0.712 4.149 First differences Intercept -2.369 -1.124 -2.199 -1.090 0.742c 1.41c Intercept and trend -2.391 -2.689 -2.144 -2.760 0.195b .0821 None -2.261b -1.764a -2.126b -1.812a Second differences
19 Intercept -6.312c -5.879c -5.856c -5.356c 0.111 0.0482 Intercept and trend -8.597c -6.195c -7.871c -5.487c 0.0337 0.0462 None -6.157c -5.732c -5.713c -5.212c a
Null hypothesis rejected at 10%
b
Null hypothesis rejected at 5%
c
Null hypothesis rejected at 1% 4.2 Cointegration test
To test for stationary and empirically meaningful relation among non-stationary variables their cointegration has to be tested. Since both logarithm of Gini-net and
real GDP per capita are integrated of the same, 2nd order, the cointegration is tested
via Johansen method in VECM framework.
Firstly, the optimal VECM model is derived by selecting optimal number of lags based on sequential modified LR test statistic (LR), final prediction error (FPE), Akaike’s Information Criterion (AIC), Schwarz’s Bayesian information criterion (SBIC) and the Hannan and Quinn information criterion (HQIC). As presented in Table 3, according to LP, FPE, AIC and HQIC a lag length of 3 is suggested.
Table 3: The maximum lag length test results
Lag LogL LR FPE AIC SBIC HQIC
0 9.93055 NA 0.00172 -0.689613 -0.590875 -0.664781
1 97.3381 174.82 1.2e-06 -7.94244 -7.64623* -7.86794
2 103.133 11.59 1.1e-0.6 -8.09851 -7.60482 -7.97435
3 109.062 11.858* 9.2e-07* -8.26627* -7.5751 -8.09244*
4 112.199 6.2746 1.0e-06 -8.19125 -7.30261 -7.96776
*lag order selected by criterion
Secondly, taking into account the predetermined lag length the number of cointegrated vectors is tested via Johansen test. With Johansen’s approach the null hypothesis of no cointegrating vectors is tested against alternative of one or more vectors. For possible cointegration at least one cointegrating vector should be present. To deliver more robust results trace instead of eigenvalue statistics are regarded. The testing procedure begins with the test for zero cointegrating equations and a maximum rank of zero. The testing ends with the first null hypothesis that is not rejected. The rank number of the hypothesis that is not rejected identifies the number of cointegrating vectors. Table 4 presents test statistics and their critical values of the null hypotheses of no cointegration (line 1) and one or fewer
cointegrating equations (line 2). As displayed in Table 4, the test detects existence of one cointegrating vector approving an empirical long-run relationship between economic growth and income inequality.
Table 4: Johansen cointegration test results
Maximum Rank Trace Statistic 5% critical value
0 21.4394 15.41
1 1.7200* 3.76
2
20
Finally, the vector is expressed in an equation using Vector Error Correction Model (VECM). From the estimates of cointegrating regressions, presented in Table 5, the following relationship is estimated:
ln(GDP) = 15.35 – 6.91ln(Gini) ln(Gini) = – 2 .22 – 0.14 ln(GDP)
Table 5: Vector Error Correction Model
ln(GDP)(t) = a1 + b1ln(Gini)(t) + u1(t) ln(Gini)(t) = a2 + b2ln(GDP)(t) + u2(t)
beta Coef. Std.err beta Coef. Std.err
ln(Gini) -6.907113 0.9628118 ln(GDP) -0.1447783 0.0260465
constant 15.34981 NA constant -2.222319 NA
Both estimated equations refer to a negative relationship between income inequality and economic growth. The first estimation in terms of elasticity indicates that an increase of 1% in Gini-net results in a decrease of 6.91% in real GDP per capita. According to the second estimation an increase of 1% in real GDP per capita leads to a decline of 0.14% in Gini-net.
4.3 Granger causality test
To understand the direction of relationship between Gini-net and real GDP per capita Granger causality has to be tested. As suggested by Risso et al. (2013) the standard Wald test is utilised to determine the causal relationship. Wald test follows an
asymptotic Chi-square distribution and its null hypothesis is no-causality. The results of the test are reported in Table 6.
Table 6: Granger Causality/ Wald test results
Chi-sq df Prob > Chi-sq
Dependent variable: ln(Gini) ln(GDP)
22.975 2 0.000
Dependent variable: ln(GDP) ln(Gini)
8.7236 2 0.013
According to these results there exists a bidirectional causality in between income inequality and economic growth. In the case of Latvia the null hypothesis of no-causality from logarithm of real GDP per capita to logarithm of Gini-net can be
rejected at 1% significance, whereas the hypothesis of no-causality from logarithm of Gini-net to logarithm of real GDP per capita can be rejected at 5% significance level.
5 Conclusion
This thesis analyzes income inequality and economic growth nexus in case of Latvia during the period of 1986 to 2012. As a part of post Soviet Union countries Latvia has experienced a movement from planned to market-oriented economy and that has significantly changed country’s economic growth and income inequality patterns. Ever since gaining independence country’s economic growth has been mostly
21
improving while at the same time the gap between rich and poor in Latvia has widened. Since in the current literature there has not been established a final conclusion about the growth-inequality relationship, the findings of this paper have generated results that might help to improve Latvia’s economic growth while adjusting for income inequality.
The main findings of this paper have been generated based on results of Johansen cointegration test, VECM estimates and Granger causality test. Solely regarding the information gathered from literature one might believe that in case of Latvia income inequality and economic growth follow the path of Kuznets’s curve where income inequality rises due to economy moving from pre growth to final stage of growth. However, the results of Johansen cointegration test disapprove the
instability of relationship as it goes through phases of growth. More importantly, the results go in line with the first hypothesis and indicate that in Latvia there exists a long-term relationship between income inequality and economic growth.
Furthermore, VECM estimated equations refer to a negative relationship between the two variables meaning that higher income inequality rates prevent economic growth while an improvement of economic growth stimulates equal income
distribution. In addition, results of Granger causality test support the second and the third hypotheses pinpointing that the causality is indeed bidirectional. However, it has to be noted that income inequality disrupts economic growth more drastically than economic growth improves equality. This implies that Latvian government could improve country’s economic growth by implementing policies that decrease income inequality.
In order to improve the results generated within this paper, further research could incorporate in its analysis other sources of growth such as levels of education and R&D. Besides, it would be of a great value to perform the research based on a larger sample, regarding wider time frame. Additionally, the work could be
complemented by analysis that use alternative measures for income inequality. Finally, as the EU’s direct agriculture aid has affected income distribution between Latvian cities and rural areas, further research could be based on comparison of the relationship in between those areas.
22
6 Bibliography
Angel, G. (2011). Press Release for Divided We Stand: Why Inequality Keeps
Rising. Paris: OECD.
Aradhyula, S., Rahman, T., & Seenivasan, K. (2007). Impact of International Trade
on Income and Income Inequality. Portland: American Agricultural Economics
Association .
Atkinson, A. B., & Brandolini, A. (2001). Promise and Pitfalls in the Use of 'Secondary' Data-Sets: Income Inequality in OECD Countries as a Case Study. Journal of Economic Literature , 39 (1), pp. 771-799.
Becker, G. (1968). Crime and Punishment: An Economic Approach. Journal of
Political Economy , 73 (1), pp. 169-217.
Bellu, L., & Liberati, P. (2006). Inequality Analysis - The Gini Index. Food and Agricultur Organization, United Nations.
Bicevska, A. (2013, March 11). ''IKP kāpums Latvijā''.
Bicevska, A. (2012). ''Kā krīze ietekmēja dažādu iedzīvotāju grupu pirktspēju''.
Birdsall, N., Ross, D., & Sabot, R. (1995). Inequality and Growth Reconsidered: Lessons from East Asia. World Bank Economic Review , 9 (3), pp. 477-508. Bitans, M., & Purvins, V. (2012). The Development of Latvia's economy (1990-2004).
In M. e. Bitans, The Bank of Latvia XC (p. 319). Riga: Bank of Latvia. Blakemore, A., & Herrendorf, B. (2009). Economic Growth: The importance of
education and technological development. Arizona: W. P. Carey School of
Business.
Bourguignon, F. (2004). The Poverty - growth - inequality triangle. New Delhi: Indian Council for Research on International Economic Relations. Castello-Cliemnt, A. (2010). Inequality and growth in advanced economies: an
empirical investigation. The Journal of Economic Inequality , 8 (3), 293-321. Central Statistical Bureau of Latvia. (2014). ''During the year, household disposable
income have grown by 5.3%''. Riga: Central Statistical Bureau of Latvia.
Cowell, F. (1995). Measuring Inequality (2nd Edition ed.). London: Prentice Hall. Cragg, M., & 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), pp. 99-116.
Daly, M., Wilson, M., & Vasdev, S. (2001). Income Inequality and Homicide Rates in Canada and the United States. Canadian Journal of Criminology , 43 (2), pp. 219-236.
Deininger, K., & Squire, L. (1998). New Ways of Looking at Old Issues: Inequality and Growth. Journal of Development Economics , 57 (2), 259-287.
Dombrovskis, V. (2011). Latvia's Post-Soviet Transition. In A. Aslund, & V. Dombrovskis, How Latvia Came through the Financial Crisis, pp. 5-16. Washington, DC: Peterson Institute for International Economics.
Eurostat. (2012). National accounts - main GDP aggregates and related indicators. Luxembourg: European Comission.
Fei, J. C., Rainis, G., & Kuo, S. W. (1980). Growth with Equity: The Taiwan Case . New York: Oxford University Press.
Fields, G. (1989). Changes in Poverty and Inequality in Developing Countries. World
Bank Research Observer , 4 (1), 167-186.
Fofack, H., & Monga, C. (2004). Dynamics of Income Inequality and Welfare in
23
Galor, O., & Moav, O. (2004). From Physical to Human Capital Accumulation: Inequality and the Process of Development. Review of Economic Studies , 71 (4), 1001-1026.
Galor, O., & Tsiddon, D. (1997). The Distribution of Human Capital and Economic Growth. Journal of Economic Growth , 2 (1), 93-124.
Goodwin, N., Nelson, J., Harris, J., Torras, M., & Roach, B. (2008). Macroeconomics
in context. Armonk, N.Y.: M.E. Sharpe.
Gottschalk, R., & Justino, P. (2006). Overcoming Inequality in Latin America: Issue
s and challenges for the twenty-first century. New York: Routledge.
Gruen, C., & Klasen, S. (2008). Growth, inequality, and welfare: comparisons across space and time. Oxford Economic Papers , 60 (2), 212-236.
Grun, C., & Klasen, S. (2000). Growth, Income Distribution, And well-Being In
Transition Countries. Munich: University of Munich.
Hindriks, J., & Myles, G. (2006). Endogenous Growth. In J. Hindriks, & G. Myles,
Intermediate Public Economics (2nd ed., p. 720). Cambridge: MIT Press.
Hinser, K. (2010). Economic growth: indicators not targets. Nature , 468 (1), 897. Jaumotte, F., Lall, S., & Papageorgiou, C. (2008). Rising Income Inequality:
Technology, or Trade and Financial Globalization? International Monetary
Fund.
Kajaks, J. (2013). Petijums par ekonomisko un socialo situaciju Baltijas valstis:
Latvija. Bruxelles: Eiropas Ekonomikas un sociālo lietu komiteja.
Kaldor, N. (1957). A Model of Economic Growth. The Economic Journal , 67 (268), 591-624.
Keefer, P., & Kanck, S. (2002). Polarization, Politics and Property Rights: Links Between Inequality and Growth. Public Choice , 111 (1-2), 127-154. Krol, A., & Miedema, J. (2009). Measuring Income Inequality: An Exploratory
Review. Region of Waterloo Public Health.
Krueger, A. B. (2002, April 4). When It Comes to Income Inequality, More than just Market Forces are at Work. The New York Times , p. c2.
Lederman, D., Fajnzylber, P. R., & Norman, L. (2002). Inequality and Violent Crime.
Journal of Law and Economics , 45 (1), 1-40.
Lerman, R. I. (1999). U.S. Wage Inequality Trends and Recent Immigration. The
American Economic Review , 89 (2), 23-28.
Lichbach, M. I. (1989). An Evaluation of "Does Economic Inequality Breed Political Conflict?" Studies. World Politics , 41 (4), 431-470.
Mahy, B., Rycx, F., & Volral, M. (2011). Does wage dispersion make all firms productive. Scottish Journal of Political Economy , 58 (4), 455-489. Masso, J., Espenberg, K., Masso, A., Mierina, I., & Philips, K. (2012). Growing
Inequalities and its Impacts in the Baltics. Country Reports for the Baltic States - Estonia, Latvia, Lithuania.
Mellor, J. (1989). John W. Mellor (June 2,Dramatic Poverty Reduction in the Third World: Prospects and Needed Action. International Food Policy Researc
Institute , 18-20.
Ministry of Economics of the Republic of Latvia. (2005). Macroeconomic
development of Latvia. Riga: Ministry of Foreign Affairs of Latvia.
Organization for Economic Cooperation and Development. (2000). OECD Economic
Surveys: The Baltic States. Paris: OECD.
24
Pritzker, P., Doms, M., & Moyer, B. (2014). Measuring the Economy: A Primer on
GDP and the National Income and Product Accounts. Washington, DC:
Bureau of Economic Analysis, U.S. Department of Commerce.
Reutersward, A. (2003). Labour market and social policies in the Baltic countries. Paris: OECD.
Risso, W. A., Funzo, L. F., & Carrera, E. J. (2013). Economic growth and income distribution in Mexico: A cointegration exercise. Economic Modelling , 35 (1), 708-714.
Rittenberg, L., & Tregarthen, T. (2009). Economic Growth. In L. Rittenberg, & T. Tregarthen, Principles of Economics (p. 895). New York: Flat World Knowledge.
Rodrick, D. (1999). Where Did All the Growth Go? External Shocks, Social Conflict, and Growth Collapses. Journal of Economic Growth , 4 (4), 385-412.
Rooth, D., & Stenberg, A. (2012). The Shape of the Income Distribution and Economic Growth: Evidence from Swedish Labor Market Regions. Scottish
Journal of Political Economy , 59 (2), 196-223.
Sala-i-Martin, X. (1996). Transfers, Social Safety Nets and Economic Growth. Washington, DC: International Monetary Fund.
Solt, F. (2009). Standardized World Income Inequality Database. Social Science
Quearterly , 90 (2), 231-242.
Solt, F. (2014). The Standardized World Income Inequality Database. SWIID Version 5.0.
Stiglitz, J. (2012). The Price of Inequality: How Today's Divided Society Endangers
Our Future. New York: WW Norton & Co.
''Svarīgāko rādītāju prognozes''. (2011). Latvijas Ekonomika.
The World Bank. (2011). Measuring Inequality. Washington, DC: The World Bank Group.
The World Bank. (2014). World Development Indicators. Washington, DC: Global Economic Prospects.
Torm, N. (2003). The Nexus of Economic Growth, Employment and Poverty during
Economic Transition:An Analysis of Armenia, Kazakhstan,Kyrgyzstan, Moldova, Tajikistan and Uzbekistan. Geneva: Recovery and Reconstruction
Department, International Labour Office.
United Nations Development Programme. (2013). Humanity Divided: Confronting
Inequality in Developing Countries. New York: United Nations, Bureau of
Development Policy.
United Nations Economic Commission for Europe. (2000). Economic Survey for
Europe. New York: United Nations.
Wellisz, S., & Saw, P. (1993). Mauritius. In R. Findlay, & S. Wellisz, The Political
Economy of Poverty, Equity, and Growth: Five Open Economies. New York:
Oxford University Press.
Weyland, K. (1998). Democracy without Equity: Failures of Reform in Brazil. The
Journal of Developing Areas , 33 (1), 158-160.
Wilkinson, R. G., & Pickett, K. E. (2009). Income Inequality and Social Dysfunction.
Annual Review of Sociology , 35 (1), 493-511.
Wright, E. (1979). Class Structure and Income Determination. New York: Academic Press.
