Trends in inequality worldwide (Gini coefficients)

Section 2 Impact of trade on income inequality As described above, it has been theoretically and empirically proved that the progress of globalization as represented by trade brings benefits in the form of economic growth through productivity improvement and other means. However, there are voices of criticism saying that the progress of globalization is widening inequality by causing negative effects on employment in some regions and industries, for example. Therefore, this section provides an overview of changes in global income inequality and analyzes in more detail changes in income inequality in advanced countries and the factors contributing to the changes. 1. Current status of global income inequality The Gini coefficient (after tax and redistribution) is widely used as an indicator of income inequality. We will look at the current status of global income inequality using the Gini coefficient. Concerning global income inequality, the Gini coefficient declined from 0.703 in 1990 to 0.623 in 2010 as shown in Figure Ⅱ-1-2-1-1. In particular, since 2000, inequality has been narrowing rapidly. 203 While the main reason for that is China s economic growth, there are also other background factors such as the fast catchup being played by India and other Asian emerging countries. The significance of the impact of China s growth is apparent in the results of an analysis showing that if China is excluded, global inequality would have widened because of the low growth of sub-saharan Africa and Latin America over the same period. 204 Figure II-1-2-1-1 Trends in inequality worldwide (Gini coefficients) 203 Concerning the global Gini coefficient, there are some statistical problems, such as the difference in the data sources between advanced and emerging countries, differences in the definitions of foreign exchange, purchasing power parity (PPP) and income, and the need for population-weighted average adjustment. According to Branko Milanovic (2006), for example, the global Gini coefficient stayed stable or slightly increased between 1988 and 2002. 204 Bourguignon. F. (2015), p. 36. 414

Notes: The number of target countries is 106 consisting of 34 advanced countries and 72 emerging countries, covering over 90% of the world population. Source: The Globalization of Inequality, François Bourguignon (2015) P.42 Table 1. However, although the Gini coefficient is declining as a trend, it should be kept in mind that the coefficient of 0.623 for the whole world in 2010, the most recent year for which data is available, is still at a very high level compared with the coefficient of less than 0.4 for most advanced countries. An indicator that expresses income inequality as the ratio of the income level of the top 10% of income brackets to that of the bottom 90% as opposed to the Gini coefficient, which represents inequality across a whole country also shows a downtrend in the inequality between the top and the bottom (Figure II-1-2-1-2). Even so, the income level of the top 10% worldwide in 2010 was 63.5 times as high as the level of the bottom 90%). Figure II-1-2-1-2 Trends in inequality worldwide (comparison of income between the top 10% and the bottom 90%) Notes: The number of target countries is 106 consisting of 34 advanced countries and 72 emerging countries, covering over 90% of the world population. Source: The Globalization of Inequality, François Bourguignon (2015) P.42 Table 1. The rise of emerging countries, including China and other Asian countries, has been cited as a reason for the narrowing of international inequality worldwide as shown in the trend in the ratio of the income level of the top 10% to the income level of the bottom 90%. Figure II-1-2-1-3 shows changes in per-capita GDP in BRIC countries (Brazil, Russia, India and China) and OECD member countries. While per-capita global GDP grew by a factor of 2.37 between 1990 and 2015, per-capita GDP in the BRIC countries grew at a higher rate, by a factor of 3.82, from 1,812 dollars to 6,917 dollars, over the same period (Figure II-1-2-1-4). The rate of growth in 415

per-capita GDP for the BRIC countries is also higher than the average growth for OECD member countries, a 2.07-fold increase from 17,473 dollars to 36,095 dollars. In particular, China recorded a 25.4-fold increase over the same period, outstanding growth compared with other countries. In consideration of this, coupled with China s huge population, it is presumed that the country s trend has had a significant impact on the narrowing of global inequality. Figure II-1-2-1-3 Changes in per-capita GDP in emerging countries/advanced countries and worldwide Notes: The figures are those equivalent to US dollars as calculation results of changes in per-capita GDP in BRIC countries (Brazil, Russia, India and China) and OECD member countries, and worldwide. Source: World Development Indicators, World Bank. 416

Figure II-1-2-1-4 Comparison of growth rates of per-capita GDP (1990-2015) Notes: This figure shows growth rates of per-capita GDP in BRIC countries (Brazil, Russia, India and China) and OECD member countries, and worldwide between 1990 to 2015. Source: World Development Indicators, World Bank. When looking at global income inequality, the Theil index may be used as an indicator of inequality as divided into domestic inequality and international inequality 205 (Figure Ⅱ-1-2-1-5). Concerning international inequality, the Theil index declined by 0.255 from 0.734 in 1990 to 0.479 in 2010, and the rate of decline was high in the period after 2000 in particular. On the other hand, as for domestic inequality, the index rose gradually but steadily, from 0.215 in 1990 to 0.244 in 2010. The catchup by emerging countries, including China and other Asian countries, is a factor behind the narrowing of international inequality, as was mentioned in relation to the Gini coefficient trend. Regarding the widening domestic inequality, François Bourguignon, a former economist at the World Bank, observed in his book 206 that domestic inequality is growing only moderately on the whole because the significant widening of inequality in some countries (particularly advanced countries) is partially offset by narrowing in other countries, mainly emerging economies. 205 Like the Gini coefficient, the Theil index is an indicator of inequality. While the Gini coefficient only shows inequality across a whole group, the Theil index is notable in that it can decompose a group into mutually independent factors (e.g. gender and income brackets). Here, the Theil index is decomposed into international inequality and domestic inequality. The total value of these two components represents the overall value of the Theil index. The index is represented by a value between 0 and 1. It takes the value 0 in the case of complete equality and approaches 1 in proportion to the degree of inequality. It was devised by Henri Theil (1924-2000), a Dutch econometrician. 206 Bourguignon, F. (2015) 417

Figure II-1-2-1-5 Trends in international inequality and domestic inequality in the world Notes: The number of target countries is 106 consisting of 34 advanced countries and 72 emerging countries, covering over 90% of the world population. Source: The Globalization of Inequality, François Bourguignon (2015) P.42 Table 1. The World Bank also uses the poverty rate as an indicator of inequality. The poverty rate represents the share of people with an income level of 1.90 dollars per day or less in the entire population. The global poverty rate dropped from 37.1% in 1990 to 12.7% in 2012, while the number of people below the poverty line decreased from 1,958 million people in 1990 to 896 million people in 2012 207. As far as the major indicators of inequality are concerned, international inequality worldwide has apparently been on a downtrend since 1990. 2.Current status of domestic inequality in advanced countries Figure II-1-2-2-1 shows the results of international comparison concerning the Gini coefficient (after tax and income transfer) representing domestic inequality in advanced countries. The Gini coefficient for the United States rose from 0.35 in 1990 to 0.39 in 2014. The Gini coefficient for the United Kingdom, which stood at 0.36 in 1990, declined temporarily after the global economic crisis but rebounded and came to 0.36 in 2013. For Japan, although data are not available for all years during the period, the Gini coefficient in 2012 was 0.33, almost flat from 0.32 in 1995, despite some fluctuations in between. The Gini coefficient for Germany rose moderately, from 0.26 in 1990 to 0.29 in 2013. For France, the Gini coefficient rose more moderately than in Germany, from 0.28 in 1996 to 0.29 in 2013. For Sweden, one of the Nordic countries where inequality is said to be relatively small, the Gini coefficient in 2013 was 0.28, the lowest level among the countries and a rise from 0.21 in 1991, but an uptrend has become notable since 2004. On the whole, inequality is widening as a trend among advanced countries except for Japan. 207 http://www.worldbank.org/ja/news/feature/2014/01/08/open-data-poverty 418

Figure II-1-2-2-1 Gini coefficients in terms of disposable income (after income transfer) Source: OECD Stat. According to data compiled by the World Bank, the Gini coefficient for China fell slightly, from 0.43 in 2008 to 0.42 in 2012, but the level was still higher than the coefficients for the United States and the United Kingdom, which were the highest among advanced countries, indicating the large extent of domestic inequality in China, although this may not necessarily be a precise comparison due to the difference in the data sources. Figures II-1-2-2-2 and II-1-2-2-3 show the shares of the top 1% and top 10% income earners in overall income. In particular, the share of the top 1% is higher than 20% in the United States the only country with a share higher than 20% and the share of the top 10% has jumped up to just below 50% there, and these shares have been on an uptrend. In France, the share of the top 1% has remained below 10% since 1990, and the share of the top 10% has also stayed around 30%. It may be said that inequality in France has remained stable at a relatively low level among major countries. In Germany, inequality between the top 1% and 10% and the rest has also remained small although not as small as in France. In Japan, after falling below 10% in 1992, the share of the top 1% showed an uptrend until 2007, but thereafter, it stayed almost flat until 2010. The share of the top 10% followed a trend similar to the trend for the top 1%. 419

Figure II-1-2-2-2 Ratios of the income level of the top 1% (before income transfer) Source: OECD Stat. Figure II-1-2-2-3 Ratios of the income level of the top 10% (before income transfer) Source: OECD Stat. 3.Analysis of the factors of income inequality by the IMF (2007) One example of theory concerning the relationship between trade and wages is the Stolper-Samuelson theorem, 208 which shows that in advanced countries, wage inequality between 208 Stolper, W.F. and P.A. Samuelson (1941), 58-73. 420

highly skilled workers with a high level of academic achievement and workers with low-level skills widens as a result of a rise in wages for the first group of workers and a decline in wages for the second group due to competition with emerging countries. As there are studies 209 whose results quantitatively showed that manufacturing jobs in some regions have decreased in the United States since 2000 because of a rapid increase in imports from China, concerns over the widening of inequality due to the impact of trade are probably growing. Meanwhile, various items other than trade globalization have been pointed out as determinant factors of wage inequality, including technological innovation, labor policy and education level. Although it is extremely difficult to identify the extent of effects generated by individual factors, the conventional wisdom among economists is that technological innovation is the main factor of the skilled labor wage premium (the difference between the high wages for high-skill workers and the low wages for low-skill workers). 210 There are not many studies that have examined inequality based on comparison of multiple factors. Below, the research results 211 shown in the IMF s World Economic Outlook 2007 report will be cited as a representative example. The IMF conducted a comparative analysis of the Gini coefficient levels using indicators of globalization (e.g. the ratio of exports to GDP and the ratio of inward foreign direct investment to GDP), an indicator of technological innovation (the ratio of ICT investment to gross fixed capital formation) and other indicators (e.g. the education level and the share of workers by industry) based on panel data covering 51 countries, including 20 advanced countries and 31 emerging countries, for the period from 1980 to 2006. The IMF varies the explanatory variables from region to region. The major estimation results are as shown in Figure II-1-2-3-1. The model used is a fixed-effect model. 209 David H. Autor, David Dorn, and Gordon H. Hanson (2013), p. 2121-2168, etc. 210 Douglas A. Irwin (2015), p. 142. 211 For the details of individual indicators, see IMF (2007), Chapter 4, Appendix 4.1.Data Sources and Methods. 421

Figure II-1-2-3-1 IMF 2007 analysis results of factors to expand Gini coefficients (advanced countries and emerging countries) Gini coefficient (explained variable) 0.45 Explanatory variable Level of contribution Technology (ICT/total capital accumulation) 0.43 Inward foreign direct investment/gdp 0.36 Others -0.05 Tariff reduction -0.12 Export/GDP -0.17 Notes: The number of target countries is 51 consisting of 20 advanced countries and 31 developing countries. Source: IMF (2007), Chapter 4 Globalization and Inequality, Table 4.1, Figures 4.9 and 4.10.As a result of the decomposition of changes in the Gini coefficient into factors, the IMF concludes that technological innovation has made the biggest contribution to rising income inequality. Among advanced countries alone, globalization (outward foreign direct investment) has a greater impact than technological innovation (Figure II-1-2-3-2). Of the globalization factors, outward foreign direct investment has the greatest impact, followed by inward foreign direct investment. On the other hand, imports and exports from emerging countries contribute to narrowing inequality. Among emerging countries alone, the impact of technological innovation is greater than worldwide, and another difference is that globalization as a whole (total effects of inward foreign direct investment, tariff reduction and exports) contributes to narrowing inequality. The decomposition of globalization into inward foreign direct investment, tariff reduction and exports shows that inward foreign direct investment contributes to widening inequality while tariff reduction and exports contribute to narrowing inequality (Figure II-1-2-3-3). These differences are explained by the fact that the impact of the expansion of globalization in the financial field appear earlier in advanced countries while in emerging countries, the impact of globalization due to trade appears earlier. 422

Concerning inward foreign direct investment s contribution to widening inequality, it is presumed that wages for high-skill workers with a high level of academic achievement rise in advanced countries because foreign companies enter high-skill industries there. On the other hand, there is a risk that businesses and industries offering high income may disappear from Japan unless the country wins global competition to attract such businesses and industries. Conversely, if Japan succeeds in attracting inward foreign direct investment, it may gain such benefits as productivity improvement for individual companies and industry and promotion of open innovation. Meanwhile, it is possible to deal with inequality through domestic policy measures (e.g. labor and education policies measures) other than trade and investment policy measures, so Japan needs to actively accept inward foreign direct investment. Figure II-1-2-3-2 IMF 2007 analysis results of factors to expand Gini coefficients (advanced countries alone) Gini coefficient (explained variable) 0.57 Explanatory variable Level of contribution Outward foreign direct investment /GDP 0.81 Technology (ICT/total capital accumulation) 0.39 Inward foreign direct investment/gdp 0.31 Tariff reduction -0.03 Export/GDP -0.12 Share of imports from developing countries -0.21 Others -0.24 Internal debt -0.35 Notes: The number of target countries is 20 advanced countries. Source: IMF (2007), Chapter 4 Globalization and Inequality, Table 4.1, Figures 4.9 and 4.10. 423

Figure II-1-2-3-3 IMF 2007 analysis results of factors to expand Gini coefficients (Emerging countries alone) Gini coefficient (explained variable) 0.32 Explanatory variable Level of contribution Technology (ICT/total capital accumulation) 0.53 Inward foreign direct investment/gdp 0.29 Others -0.14 Tariff reduction -0.16 Export/GDP -0.19 Notes: The number of target countries is 31 developing countries. Source: IMF (2007), Chapter 4 Globalization and Inequality, Table 4.1, Figures 4.9 and 4.10. 4.Analysis of factors contributing to widening inequality in recent years The IMF s study described in 3. above covered the period from 1980 to 2006, so it is possible that the study did not fully cover the period during which the impact of the expansion of exports from China following the country s accession to the WTO in 2001 appeared in earnest. Therefore, we conducted a new study covering the period from 2001 to 2014. 212 The horizontal axis represents the rate of change in the Gini coefficient in the case of a 1% change in the value of each index. The IMF s analysis covered a sample group including many emerging countries. However, as the possibility cannot be ruled out that the factors contributing to inequality may be different between advanced and emerging countries, our analysis was conducted on the basis of panel data covering only advanced countries. 213 In addition, while we excluded tariffs and inward foreign direct investment, 212 For some countries covered by the research, there may be some data deficiencies during the research period. In our research, data deficiencies were complemented by estimating the values of the missing figures based on the compound annual growth rate (CAGR) when the distance in time between the deficient data and the available data most close in time to it is six years or shorter. 213 The research covered the following 22 countries: Australia, Austria, Belgium, Canada, the Czech Republic, Denmark, Estonia, Finland, France, Germany, Ireland, Italy, Japan, Luxembourg, the Netherlands, Portugal, Slovakia, Spain, Sweden, Switzerland, the United Kingdom and the United States. 424

which were treated as components of globalization in the IMF study, from among the explanatory variables, we adopted labor policy and the proportion of workers with university degrees as additional explanatory variables for our analysis. As the explained variable, we adopted the Gini coefficient before redistribution as the IMF (2007) did. We conducted an F test and a Hausman test, and we adopted a random effect model for our analysis based on the test results. Consequently, it was found that the main factor contributing to the widening inequality in advanced countries is technological innovation (ICT investment) and that trade, along with other factors, including education policy, work to narrow inequality (Figure II-1-2-4-1). Figure II-1-2-4-1 Analysis results of factors to expand recent inequality Explanatory variable Rate of change t value ICT investment/gdp 0.075 4.48*** Finance (household financial assets/total financial assets) 0.014 0.59 Labor policy cost/gdp -0.020-2.25** Proportion of workers with university degrees -0.036-2.16** Trade value/gdp -0.038-2.04** Notes: * significant level of 10%; ** significant level of 5%; *** significant level of 1% Number of samples: 294 Adjusted R-Squared: 0.25 Notes: 1: Based on IMF 2007, METI expanded the target period of analysis from 1980-2006 to 2000-2014 and modified the number of target countries to 23 OECD member countries. 2. The horizontal axis shows the ratio (%) of changes in the Gini coefficient in the case of a 1% For the details of the indicators used in the research, see Supplementary Analysis 3. 425

change in the respective indices. Source: World Bank, IMF, OECD and other organizations. However, as promoting ICT investment is essential for Japan to raise its economic growth potential, as will be discussed in detail in Chapter 4, Japan s international competitiveness may decline and only low-wage jobs may remain in Japan if sufficient ICT investment is not made in the country. Therefore, Japan must continue to actively promote ICT investment while implementing domestic policies (e.g. labor and education policies) other than trade and investment policies in order to deal with inequality. In some advanced countries, financial assets held by households in higher income brackets are sometimes cited as a factor contributing to widening inequality, so we adopted financial assets as an additional explanatory variable. However, our analysis did not show any statistically significant results with respect to financial assets. 214 5.Analysis of factors contributing to wage inequality by the OECD (2011) In 2011, the OECD conducted a regression analysis using the ratio of the labor wages (weekly wages) of the top 10% wage earners to the wages of the bottom 90% as the explained variable unlike the IMF, which adopted the Gini coefficient as the explained variable and trade, technology, financial deregulation (FDI Regulatory Restrictiveness Index), the labor unionization rate, product market regulation, job protection regulation and tax as explanatory variables. 215 The major results are as shown in Figure II-1-2-5-1. The results indicated that trade and financial deregulation do not have a significant impact on wage inequality. On the other hand, like the IMF s analysis, the OECD s analysis showed that technology (an increase in expenditures on science and technology activities) has a significant impact. The results suggest that technological innovation has a greater impact on wage inequality than trade and financial deregulation (FDI Regulatory Restrictiveness Index), while education contributes to narrowing inequality. 214 As for the impact of financial assets on income inequality, see Supplementary Analysis 1. 215 OECD (2011) 426

Figure II-1-2-5-1 OECD analysis results of factors concerning inequality of wages (2011) Wage inequality (rates of the annual changes in labor wages between the top 10% and the bottom 90%) 0.472 Explanatory variable Level of contribution to the changes above Trade Financial deregulation (FDI Regulatory Restrictiveness Index) Organization/policy 0.424 Technology (private R&D expenditure/gdp) 0.320 Others 0.229 Education (ratios of workers with higher education) -0.501 Notes: The term Trade refers to the weighted average of the import penetration level and the export density. The term Organization/policy refers to labor union participation level, product market regulations, employee protection policies, taxation, and other elements. The term Others refers to female employment rates and other elements. Source: OECD (2011), p.122. 6.Summary For Japan, free trade is the prerequisite for economic development, and its benefits are understood widely among the Japanese people. On the other hand, in light of the recent international situation, the benefits of free trade are not necessarily obvious, and there are voices of concern over and discontent with the risk that free trade may be expanding inequality. As shown above, free trade not only expands the economic pie but also brings other benefits on both macroeconomic and microeconomic fronts, including a rise in the purchasing power for consumers and productivity improvement for producers. Concerning the argument that trade may be expanding inequality, we conducted analysis in reference to the analyses conducted by the IMF and the OECD and presented analysis results showing that the widening inequality is due in large part to technological innovation and that trade, as well as labor and education policies, contributes to 427

narrowing inequality. On the other hand, although free trade is not a factor contributing to inequality, it is necessary to develop a system whereby many people can enjoy the benefits of trade by appropriately combining labor, education and other policies that are considered to contribute to narrowing inequality if we are to obtain understanding from people who are anxious or dissatisfied. 428