Increasing Income Inequality and Policy Options in Korea

Transcription

1 e-labor News No. 129 Issue paper Increasing Income Inequality and Policy Options in Korea Jiyeun Chang and Byung-Hee Lee 1. Introduction Someone once described the life of Koreans as exhausting, unfair and unstable. Such a description is related for the most part to the problem of income inequality. Excessive income inequality, in which differences in income exceed what can be seen as the result of differing amounts of individual effort, does not function as an incentive to work hard, and in itself damages the notion of justice. Every society seeks ways to minimize income inequality; in order to do so, it must first identify the characteristics and causes of the phenomenon. That Korea s income redistribution system is not effective in lowering income inequality has been revealed by previous studies, and it is explained by a lack in the absolute amount of social expenditures (Kim, 2004; Yeo, 2009; Yeo & Song, 2010). Meanwhile, market income calculated on a household basis is shown to be relatively equal compared with other countries, and one recently proposed explanation for this is that it is the result of the lower-income classes maximizing the labor supply of their household members, but more research on this topic is needed (Chang, 2012). With regard to factors that intensify income inequality, attention is focused mostly on the role of wage inequality in the labor market. Questions have been raised about the impact of the increase in the elderly population due to population aging and changes in family structure in which the number of single-parent families increases, or about the correlation between the increase of female participation in the labor market and income inequality, but consensus conclusions have yet to be drawn. These questions remain unanswered, and are what this study seeks to resolve. The solution, as well as policy efforts and programs that require the injection of financial resources, may differ depending on whether the increase in the elderly population or This paper was presented at the international conference New Socio-economic Paradigm, New Social Policies organized by the Korea Labor Institute (KLI) in Seoul on October 19, Jiyeun Chang (Senior research fellow and Director of International Cooperation and Information Office, KLI); Byung-Hee Lee (Senior research fellow, KLI). 1

2 whether wage inequality in the labor market is a key factor of inequality. Even if the increase in single-parent families or the increase in the elderly population is found to intensify income inequality, policies to suppress the phenomena themselves are not viable solutions. Rather, the focus should be on searching for measures to shut off or mitigate their effects. The increased female participation in economic activities has been reported to reduce income inequality (C. Lee, 2007), and this corresponds to the observation that active participation in economic activities on the part of the low-income classes in Korea reduces income inequality. However, we should take caution, as female participation in economic activities does not necessarily lead to such an effect. If women s additional participation in the labor market has a negative impact on income distribution as seen in other developed nations, the burden on policies to redistribute income through income-transfer programs will increase. This study seeks to identify the factors that intensify income inequality in Korea and their mechanisms, and to propose an appropriate social-policy paradigm based on the findings. It is organized in the following manner: Section 2 compares the income-inequality phenomenon in Korea with those in other countries and observes the long-term trends and their characteristics. We examine the cause of the recent increase in income inequality through a factor-decomposition method. In Section 3 we discuss how each key factor of income inequality operates to intensify inequality. In Section 4 we conclude by providing proposals for what kind of distribution/redistribution model Korea s social policies should be based on in order to solve the country s problem of income inequality. 2. Income Inequality and Decomposition of Factors 1) Relative Level and Characteristics of Income Inequality Chang (2012) suggested the conceptual framework shown in Figure 1 to identify the mechanism by which income inequality occurs. The distribution of income that an economically active individual can earn from the labor market is the most basic field in which inequality occurs. Because the economic prosperity of individuals is determined by their household income, the inequality index or poverty rate is calculated based on household income, and in most countries, because labor income accounts for over 3/4 of household income (OECD, 2011), income levels are greatly affected by how many people in each household are working. In short, the level of income inequality is affected by the labor-supply factor of each household. Aside from labor income, income inequality is also affected by the 2

3 distribution of assets. Finally, disposable-income distribution deduces by reflecting by the progressiveness of taxes and the effectiveness of public income transfer programs. 1 (HH) Disposable income inequality (HH) Market income inequality (Labor income + asset income; pre tax) Public income transfer Redistribution of taxes* Asset income factor (HH) Labor income inequality HH labor supply factor (Individual) Labor income inequality (wage + self employed income) (Individual) Wage income inequality Self employed ratio and self employed income inequality factor Labor market wage dispersion Figure 1. Income inequality in various strata and their impact factors. * Includes social security premiums. HH = household. The question of the equality (or inequality) of income distribution in itself is based on the premise of comparison. While it is possible to assert that income distribution is unequal compared with the past or compared with other countries, there is no way to measure the absolute level of inequality without a comparison. Table 1 shows the results of a comparison of Korea s income-inequality levels in various strata with those of other countries as of the mid-2000s. 1 Table 1 reconfirms a fact that is widely known. The distribution of household income in Korea is quite comparable to that of other countries when calculated based on market income, the level of inequality being average or slightly above average. This can be understood as a confirmation of Korea s lack of redistribution policies. 1 The purpose of this figure is to provide an understanding of the income-inequality mechanism. It does not allow for measuring how much each stage intensifies by using Gini coefficients or poverty rates calculated by sequentially adding income for each field. This is because the values differ depending on the order of adding. 3

4 Table 1. Gini Coefficient for Various Income Definition Strata Country Individual income type HH income type Wage Labor Market Disposable Australia Austria Belgium Canada Denmark Finland France Germany Greece Italy Netherlands Norway South Korea Spain Sweden United Kingdom United States LIS key nation average Note. HH = household; LIS = Luxemburg Income Study. Labor income consists of wage income added to self-employed income. In measuring the Gini coefficient for HH income, a household equivalence scale was applied to HH income, and weighting was deduced by multiplying the HH weighting by the number of HH members and calculating individual rankings. Source: LIS data, mid-2000s. Korea s 2006 Household Income and Expenditure Survey data were submitted to LIS. Meanwhile, Table 1 presents a new puzzle: Why is it that whereas the labor income of individuals in Korea s labor market is extremely unequal, the household market income is relatively equal? According to Figure 1, this phenomenon is the result of the labor-supply pattern of households. If a relatively high number of household members from the low-income brackets that lack human capital participate in the labor supply, income measured on the household level may be found to be distributed equally even if the labor market is unequal. In reality, Chang (2012) found that although the number of labor-income earners in the lower 20% income brackets of developed nations was 0.3 to 0.7 people per household, in Korea, an average of 1.4 people per household participate in labor activities in that same income bracket. Based on this fact, Chang suggested that although Korea has an 4

5 extremely unequal labor market, individual households participate in income activities in a manner that cancels the effect, making the household-income distribution relatively equal. 2) Time-Series Trend of Income Distribution One source of information appropriate to a time-series observation of the income-inequality index is the Korea National Statistical Office s Household Income and Expenditure Survey (HIES), in which total income and expenditures for each household are calculated on a monthly basis through entries in household ledgers. However, because the survey was extended to households throughout the nation only as recently as 2006, a reliable time-series analysis is that conducted on urban households with two or more members. Figure 2. Gini coefficient trends for urban households with two or more members and total households. HH = household. Source: Korea National Statistical Office, Household Income and Expenditure Survey, each year. Figure 2 shows the changes in income distribution since the 1990s as measured by the Gini coefficient. According to this figure, until the mid-1990s, income inequality in Korea was much lower than it is today. It increased sharply with the 1997 Asian Financial Crisis, was followed by some fluctuation, and has continued to increase since the mid- and late 2000s. The level of inequality in all households can only be observed after 2006, but it is possible to assume that even before that, inequality in households with two or more members would have shown a similar pattern, although the overall level would have been higher. Another notable phenomenon depicted in Figure 2 is that the inequality in market income and disposable income has been increasing since the early 2000s. We can interpret this to mean 5

6 that with social security systems having been strengthened since the 1997 crisis, redistribution policies have begun to have stronger effects than previously. Figure 3. Gini coefficient trend for urban households with two or more members. Calculations of Dr. Roh Dae-Myung (Korea Institute for Health and Social Affairs). Source: Korea National Statistical Office, Household Income and Expenditure Survey, each year. In order to determine the Gini coefficients during the period before the early and mid-1990s, when inequality was relatively low, Figure 3 provides data from previous years, although it encompasses only worker households in cities. According to this figure, income was more unequal in the 1980s than in the 1990s. Income inequality is likely to have improved following the activation of labor movements during the labor strikes of In short, income inequality in Korea as observed through time-series data was lowest in the early and mid-1990s, and high in the periods before and after that point. The point at which income inequality begins to decrease corresponds to the point at which wages of laborers began to rise because of labor movements, and the point at which income inequality began to increase again during the Asian Financial Crisis of 1997, after which the Gini coefficient continues an upward trend until the late 2000s. 3) Decomposition of Driving Factors of the Recent Income Inequality What are the driving forces behind increasing inequality since the late 1990s? Four hypothetical factors may affect the recent trend of inequality: population aging, changes in 6

7 the household labor supply, wage dispersion, and redistribution policies. Each factor can be translated into research questions as follows. First, how much can population aging explain the increasing inequality since the late 1990s? All else being equal, population aging can increase overall income inequality at the current time in Korea, because the proportion of the poor is much higher among elderly Koreans. The proportion of the population 65 and older increased from 5.1% in 1990 to 11% in The relative poverty ratio of the elderly population was 47.1% in Second, how strongly does the pattern of female labor supply affect the household-income inequality of the working-age population? Household-income inequality is largely affected by two highly correlated factors: household labor supply and wage dispersion in the labor market. The increasing number of one-person households and one-parent families tends to increase household-income inequality. Among couples, female labor force participation rates may affect inequality. It is also useful to observe which income classes the female participants belong to. According to Maxwell (1990), who analyzed the CPS data from 1947 to 1985, wives of low-income husbands tended to participate in the labor market at higher rates than their high-income counterparts in the 1950s and 1960s, which means that wives labor supply reduced the household-income inequality during this period in the United States. After labor participation rates for middle-income-earners wives increased in the 1970s, the positive effects of the female labor supply on household-income equality began to diminish. Lee (2008) reported similar stories in the Korean case. He observed expanded inequality of household income in Korea between 1996 and 2000, and he argued that the key driving factor was the increasing dispersion of household heads labor income, while the change in spouses labor supply contributed to lessening the inequality in household income. Third, how much does wage dispersion explain household-income inequality when labor income composes 3/4 of total household income, on average, in OECD countries (OECD, 2011)? Very high rates of low-wage employment and non-regular employment might be the most important factors affecting household-income inequality in Korea, as Lee (2008) observed of the late 1990s. Fourth, how much do taxation and public income transfer affect efforts to reduce inequality in disposable income? Social assistance, unemployment payments, and old-age 7

8 pension systems have been developed during the past several decades in Korea, and we need to ask whether those institutions helped attenuate income inequality. Decomposition by household type: Population aging vs. increasing inequality within the working-age population? In order to investigate whether the increasing income inequality originates in population aging in Korea, we decompose the change of inequality by household type that is identified as working-age (18 to 64 years old) households and elderly households according to the age of their head of household. The measurement of income inequality used in this section is mean log deviation (MLD), defined in the following formula. If the incomes are equally distributed, MLD would be zero., where is the mean income of the population, is the income of household,, and is the number of individuals in the household. Considering the types of households, the total inequality can be decomposed according to the following formula (static decomposition):, where is the component proportion ratio of members of specific household type ( ) among the total population, and is the inequality measure by household type. The first part of the right-hand side of the formula is the within-group component of the inequality, and the second part is the between-group component of the inequality. Table 2 shows the results of the decomposition. MLD in 2011 was 0.245, an increase from in The structure of the population has also changed, and the number of members of elderly households increased. Income inequality increased in both household types, and for elderly households was severe. The within-group component occupies the larger part of the inequality in both years. Although the contribution of elderly households to total inequality increased from 8.7% in 1996 to 25.7% in 2011, the contribution of the working-age-household population was the majority (68.8%) in

9 Table 2. Decomposition of Income Inequality by Household Type (unit: Thousand Korean Won, %) Year and HH type Proportion Mean yearly income MLD Absolute contribution Relative contribution 1996 Elderly HHs , (8.7) Working age HHs , (88.7) Within group inequality (97.5) Between group inequality (2.5) Total (100.0) 2011 Elderly HHs , (25.7) Working age HHs , (68.8) Within group inequality (94.5) Between group inequality (5.5) Total (100.0) Note: HH = household; MLD= mean log deviation. Yearly incomes are deflated by 2010 CPI. Source: Korea National Statistical Office, Household Expenditure Survey, 1996; Korea National Statistical Office, Household Income and Expenditure Survey, The change in income inequality over time can be decomposed as follows (Oxley, Burniaux, Dang, & d Ercole, 1997):. The first part of the right-hand side of the formula represents the change in the within-group inequality when the proportion of each household type is fixed, and the second part represents the change in the between-group inequality when the proportion of each household type is fixed. The remaining three parts represent the change of inequality due to the compositional change of each household type when the within-group and the between-group components are fixed. Table 3 presents the decomposition of the change in inequality between 1996 and The increase in within-group inequality contributed as much as 66.3% to the total increase in inequality, and the change in population composition by household type explains 31.6% of the inequality change between those 2 years. Among the effects of within-group 9

10 inequality, the increase in income inequality within working-age households explains the majority of the inequality (53.7%). In sum, the increase in income inequality within the working-age population was the most important factor, and the proportional change in elderly households was the second most important factor, in explaining the increasing inequality during the last 15 years of the period. Table 3. Decomposition of Change of Income Inequality by Household Type ( ) Contribution Total Net effect of within group inequality Elderly HH Working age HH Subtotal Net effect of between group inequality Composition effect Absolute Relative Note. HH = household. Source: Korea National Statistical Office, Household Expenditure Survey, 1996; Korea National Statistical Office, Household Income and Expenditure Survey, Effects of household labor supply and wage dispersion on inequality. We found the growing inequality of employment income to be a critical factor in overall income inequality. Then what makes the inequality of employment income greater? Next we try to identify the effects of the labor supply of household heads and spouses, and the effects of family structure as well as the effects of wages and salaries of working individuals. Two frequently asked research questions are closely related to this analysis. First is whether a spouse s, particularly a female spouse s, additional labor supply positively or negatively affects inequality. Second, how does the increasing number of one-person households and single-parent families affect income inequality? Here again we focus on the working-age population, excluding households headed by elderly people. Inequality is measured as the mean income difference between the first and tenth decile of disposable income. Following Lee (2005), we identify mean household income in a specific income decile as follows:, where subscript indicates the household head, is labor income (employment and business income), P is the employment rate, and is the proportion of households whose head and spouse live together; stands for other household members income, capital 10

11 income, private and public income transfer, and tax and social security contribution combined. The inequality measure,, is defined as the log difference between the mean incomes of the highest 10% and lowest 10% of households:, where is the highest 10% and is the lowest 10%. The inequality measure,, is approximately expressed as follows:, where is the proportion of each income source. For example, is the proportion of the spouse labor income in the total household income. The change of inequality between two time points can be decomposed as follows:. The first part of the right-hand side of the formula represents the contribution of the change in the household head s labor income. The second part reflects the contribution due to change in the household heads employment rate, and the third part reflects the relative contribution of change in proportion of household heads labor income to total household income. If the distribution of household heads labor income is more unequal than that of household total income, then the growing proportion of household heads labor income increases household-income inequality. With this method, we compare the households in the same decile of income between two time points. In other words, if the income gap between the highest and the lowest income groups increased in 2011 compared with 1996, then we observe the change in labor income, labor supply, family structure, and so on in those two years. The deciles distribution ratio increased from 5.9 in 1996 to 7.1 in 2011 (Table 4). The dispersion of household heads labor income increased, and that of spouses labor income also rose. Spouses employment rates increased in both low-income and high-income groups. The proportion of couple families decreased in both income groups, with a larger drop in the low-income group. Table 5 presents the decomposition of change in household-income difference between the highest and the lowest income groups in 1996 and The dispersion of labor 11

12 income seems to be more responsible than the change in labor supply for the increasing inequality in the recent 15 years. The difference in labor income is greater than the difference in total household income (122%). In particular, the household heads labor-income dispersion explains 111% of total household income, which means that labor market inequality has been a key driving force of household-income inequality. The change in labor supply also explains some of the household-income inequality. On the other hand, redistribution policies such as public transfers and tax contributions seem to contribute to a reduction of the income gap between the high- and low-income groups. Table 4. Composition of Household Income by Income Groups (Working-Age Households) Item 1st 10th 1st 10th Average yearly income (thousand won) HH disposable income 4,741 28,179 7,005 49,558 Labor income by heads 4,313 20,251 6,066 37,103 Labor income by spouse 1,266 9,035 1,906 16,109 Labor income by other families 193 2, ,404 Capital income 234 2, Private transfer income ,233 Public transfer income ,017 Tax and social security contribution 399 1, ,756 Rate/ratio Employment rate of heads Employment rate of spouses Proportion of couple family Proportion Labor income by heads Labor income by spouse Labor income by other families Capital income Private transfer income Public transfer income Tax and social security contribution Note. HH = household. Source: Korea National Statistical Office, Household Expenditure Survey, 1996; Korea National Statistical Office, Household Income and Expenditure Survey,

13 Table 5. Decomposition of Change in Income Inequality: Gap Between the Highest and the Lowest Income Groups (Working-Age Households) Item Estimate Contribution (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) (17) (18) (19) Labor income (2+5) Labor supply (3+6) Family structure (7) Labor income by other family members (9) Capital income (11) Private transfer income (13) Public transfer income (15) Tax and social security contribution (17) Composition of income ( ) Note. HH = household; o = other family member s labor income; c = capital income; PRT = private transfer income; PUT = public transfer income; TAX = tax and social security contribution. Source: Korea National Statistical Office, Household Expenditure Survey, 1996; Korea National Statistical Office, Household Income and Expenditure Survey,

14 4) Summary In order to determine the factors driving inequality, we compared income inequalities between Korea and other countries through different viewpoints and time-series trend changes and analyzed the changes that occurred in the last 15 years, during which the inequality has escalated. The results are summarized as follows. First, comparisons with other countries indicate that Korea s market-income dispersion is well balanced. This suggests that active labor market participation by the low-income class has offset the inequality in individual labor incomes. While household market income is relatively balanced, the dispersion of disposable income drops to the average level of compared countries. This is because Korea s redistribution system does not function as well as those in developed countries. Second, a time-series approach to income inequality shows that inequality declined significantly in the late 1980s, before escalating again in the mid-1990s. The changes coincide with the activation of labor strikes in 1987 and the financial crisis in We have noted an increasing gap between market income inequality and disposable-income inequality. Third, among the four hypotheses considered to explain the increasing income inequality over the last 15 years, wage inequality seems to have had the greatest impact. The aging of Korean society is another factor. During the past 15 years, the inflow of female labor had no effect on easing income inequality, whereas taxes and income transfer helped reduce it. 3. Mechanisms of Income-Inequality Intensification In this section, we explore in greater depth the mechanisms that are known to intensify or ease income inequality: (1) wage inequality in the labor market, (2) spouses participation in economic activity, and (3) the effects of redistribution policies. 2 The gap in the labor income of household heads was revealed to be the main factor intensifying the inequality of household income over the last 15 years. Some research suggested that the labor-income inequality of household heads was offset when low-income females participated in the labor market (Lee Chul-Hee, 2007), but recently the effect seems to be fading. Once the active labor-market participation of those in the low-income classes reaches 2 Aging of the population is a key factor, but it is not discussed, because it has simple and clear implications from a policy-option viewpoint. 14

15 its limit, the wage inequality in the labor market will lead directly to inequality in household income. Hence, redistribution policies must play a greater role in easing income inequality. It is clear that the public income transfer must be gradually increased, but there are limits to reducing inequality through income transfer without tending to the wage gap. 1) Wage Inequality in the Labor Market Figure 4 depicts the results of observing the wage inequality in the Korean labor market since the early 1980s. 3 The decrease of wage inequality in the early and mid-1980s contradicts the trend of income inequality in working households, but the plunge in wage inequality in 1987 and the sharp increase of inequality in the late 1990s are in line with the trends in household-income inequality. This confirms that household-income inequality trends form the basis of wage inequality in the labor market (Figure 7). Figure 4. Changes in wage inequality: Gini coefficient ( ). Source: Ministry of Employment & Labor, annual Basic Wage Structural Survey; Cheon Byeong-Yu et al. (2012), paper presented at GINI Also of note in the figure is that the Gini coefficient for hourly wages is always higher than that for monthly wages. This is unprecedented in other OECD countries (OECD, 2011). In developed countries, the monthly wages of part-time workers are low. Hence, wage differences can largely be explained by the size of the labor supply, whereas in Korea, people working long hours earn a low monthly income. In Korea, the difference in hourly wages is large, and the gap is narrowed through long working hours. What can explain the increase in wage inequality beginning in the late 1990s? A comparative study by the OECD reviewed the subject with three hypotheses: globalization, 3 The basic wage structural survey data used here do not provide information for businesses with five workers or fewer. Hence, the overall inequality is likely to be underestimated. The data are used for comparison with HH income inequality. 15

16 technological changes, and weakening employment protections. Evidence of globalization increasing inequality was not discovered, whereas the other two factors were found to have a definite influence (OECD, 2011). Yet, some evidence suggests that globalization may have had an impact on increasing inequality in Korea. The national income s dependency on trade increased beginning in the mid-1990s (Figure 5). We believe that this increase is attributable to the increase in trade with underdeveloped countries, such as China, which had a large impact on low-income workers. Figure 5. Trade dependency of national income. Source: Korea National Statistical Office; Cheon Byeong-Yu et al. (2012), paper presented at GINI conference (not published). This is connected to decrease in or stagnation of the labor share (the share of labor in the national income) since the late 1990s (Figure 6). The labor share has been affected by the proportion of the self-employed. It is noteworthy that the labor share decreased or remained stagnant, although the shares of the self-employed decreased and of paid workers increased. The deterioration of the labor share seems to result from the increased percentage of non-regular and low-wage jobs. Like other OECD countries during the period, Korea weakened its employment protections. The EPL (employment protection legislation) index weakened, 4 while the labor unionization rate dropped from 20% in the late 1980s to the current 10%. An increase in minimum wages is hoped to help ease the wage inequality, but in reality, its effects are not definite. The influence rate of the minimum wage increased in the 2000s, but this implies that the cumulative number of below-minimum-wage workers has increased. 4 4 The EPL index, which is announced by the OECD after compiling data, shows that the employment protection index for regular workers remained the same, at 2.37, from 2003 to 2008, while the employment 16

17 Figure 6. Trend of paid worker percentage and labor share ( ). Source: Korea National Statistical Office; Cheon Byeong-Yu et al. (2012), paper presented at GINI conference (not published). Year to year increase Influence rate Figure 7. Increase in minimum wage and influence rate. The minimum wage was applied to businesses with 10 or more employees until 2000, when it was applied to all businesses. Influence rate = (number of beneficiaries / number of subject workers) 100. Source: Korea National Statistical Office. 2) Labor Supply of Spouses We previously noted that the number of household members participating in the labor market in the low-income classes was higher in Korea than in other countries. This allows us to predict that the difference will be the same if the term is changed to female economic activity. protection index for non-regular workers dropped from 1.67 to When compared with other countries, the index for regular workers is slightly above average, while the figure for non-regular workers is slightly below average. Protection against layoffs stands at 1.88, which is much lower than compared countries (OECD, Employment and Labour Statistics DB). 17

18 Changes in female labor-supply patterns affect household-income distribution. Household-income inequality is affected by two highly related factors. One is the labor supply of households, and the other is the wage distribution of the labor market. The increase in female heads of single-parent households and their economic participation is a key factor, but it is important to note the economic participation ratio of female spouses in couple households. In other words, the level of economic participation of female spouses in each income class is the most crucial factor. According to Maxwell (1990), who analyzed U.S. CPS data for the period 1847 to 1985, the wives of low-income male household heads were shown to have a higher rate of economic participation until the 1960s. In the 1970s, however, the labor market participation of wives of middle-income households was significant. After this point, the positive influence of female economic activity on household-income distribution began to disappear. Lee (2005) also revealed that the labor supply is becoming a smaller factor in driving household-income inequality in the United States, while the wage distribution of the labor market is having an increasing effect. This finding supports Maxwell s research. In a comparison of the pre-crisis (1996) and post-crisis (2000) eras of Korea, Lee (2007) discovered that during this period household-income inequality increased, which was believed to be caused by the increasing wage inequality for household heads. During this period, female economic participation reduced household-income inequality. In other words, inequality increased during this time despite the increase in the female labor supply from low-income households. Based on this evidence, Lee (2007) suggested that creating more jobs to promote female employment would reduce inequality. Because the female labor participation rate of low-income households is high enough, however, the female labor participation rate of middle-income households will be increased in the future. If certain social characteristics hinder the spread, the female economic participation rate or employment rate will remain stagnant for a long time. Korea is in this situation, as the female economic participation rate has increased by a mere 3 percentage points in 25 years. The current study analyzes the change between 1996 and 2011 to update the research by Lee (2007) and found that an increase in spouses labor supply had no effect on easing inequality. Considering that market-income inequality increased consistently during this period, in Korea the increase in the household labor supply offset household-income inequality resulting from wage inequality, but the effects are clearly weakening. 18

19 Figure 8 compares the employment rate of spouses for each income class. The individual income of household heads was divided into five classes a total of six groups, including those with no income. Aside from Korea, the United States, and Germany, the employment rate for wives increased as the income level of the household head increased. In Korea, the female employment rate is low in general, but the gap peaked for spouses of high-income household heads. In Korea, the labor participation of spouses was clearly higher for household heads with no income, whereas the participation rate of spouses was low for high-income household heads. This seemed to reduce the inequality of household income. It seems unlikely that spouse labor-supply patterns by income level of household heads would converge into the same shape in a short period of time, but assuming that Korea will narrow the gap with developed countries, it is difficult to expect female earnings to ease household-income inequality. Korea US UK France Germany Sweden Denmark Figure 8. Spouse employment rate by income class of household head. Notes: 1) Analysis conducted on couple households, whose heads are under 65 years of age. 2) Household heads with no income set at 0, heads with income divided into five classes. Among couple households, the percentage of household heads with no income stood at 14.3% in Korea, 30.4% in the US, 28.7% in the UK, 31.9% in France, 36.7% in Germany, 19.3% in Sweden and 22.3% in Denmark. Source: LIS (mid- and late 2000s). Like other developed countries, Korea will experience the fading effects of the female labor supply s contribution to equalizing income distribution. We can expect female labor participation to increase, but the increase in the labor supply will be faster in highly educated, middle-income females, as the labor supply from low-income females is already 19

20 saturated. In conclusion, the increasing earnings inequality of household heads will no longer be offset by the low-income labor supply of female spouses. Figure 9. Relative poverty rates: Derived from market income and disposable income. Relative poverty rate refers to the households within income levels below 50% of middle income. Urban = urban households with two or more members, including the paid worker; whole = households around the country with one or more members. Source: Korea National Statistical Office, Household Income and Expenditure Survey data, Table 6. Poverty-Rate Reduction Effect by Income-Transfer Program Country Demogrant + Pension Market income + Unemployment benefits + Other social insurance benefits + Public assistance + Overall public income transfer Korea Germany Sweden UK Note. Relative poverty rate based on households with income below 40% of the middle income level. Households included in analysis limited to heads between the ages of 15 and 64. Source: Yeo Yu-jin & Song Chi-ho (2010). Original source: Korea s data from Welfare Panel Study 2008, data for other countries based on LIS data for mid-2000s. 3) Tax and Income Transfer The redistribution effect of tax and income transfer in Korea continues to fall short of that in developed countries. The ineffective redistribution effect of tax and income transfer is primarily revealed by the fact that the Gini coefficient derived from market income does not show a large difference with the Gini coefficient derived from disposable income (Table 1). As seen in the aforementioned Gini coefficient trend, the effect of 20

21 wealth-redistribution policies is increasing (Figure 2 and Figure 3). Tax and public income transfer should lower the poverty rate. Their effects show an ascending trend, although absolute figures remain low (Figure 9). The poverty-rate reduction effect of each income-transfer program is provided in Table 6 (Yeo & Song, 2010). 4. Conclusion: Selection of Labor Policies and Income-Security Policies The intensification of income inequality was studied based on the wage inequality of the labor market, aging of the population, changes in the household labor supply, and the effectiveness of wealth-redistribution policies. From a times-series approach, income inequality dropped significantly in the late 1980s and began to intensify after the mid-1990s. The changes coincide with the activation of labor movements in 1987 and the financial crisis in During the last 15 years, the most noteworthy factor driving income equality is the increase in wage inequality in the labor market. The rise and drop in wage inequality overlap with changes in household income and the share of trade in national income. It has been revealed that increased trade with developing countries (China) and trade-driven economic growth increased the percentage of irregular and low-income jobs, delivering a heavier blow to the low-income working population. During this time, employment protection was greatly weakened, as the negotiating power of workers was undermined. The rise and fall of wage inequality coincides with the changes and trends in household income. Even so, wage inequality did not fully translate to household-income inequality thanks to the active labor-force participation of the low-income classes. However, it is unlikely that such effects will continue. The labor market participation of low-income females offset the wage inequality of household heads, but analyses of recent years show that such effects have disappeared. Considering that the labor market participation of spouses of high-income household heads in Korea shows a large difference with females in developed countries, it is unlikely that household labor-supply patterns will continue to mitigate inequality. Once the labor market participation of low-income classes is saturated, the wage inequality in the labor market will directly translate to household-income inequality. Hence, redistribution policies will have to play a larger role in easing income inequality. Public income transfer must be gradually increased, but there are limits to reducing inequality through income transfer without tending to the wage gap. 21

22 With labor-market wage inequality leading to household-income inequality, how should employment and income-security policies be crafted? Solving poverty though public assistance and increasing social investment oriented general services seem to be inevitable tasks. In determining the combination of employment policies and income-security policies for tackling given tasks, there will be two options based on approach: problem recognition and reality-based judgments. They are not completely mutually exclusive, but they are clearly different directions. Wage and income inequality have increased under the policy direction of creating many low-income jobs and increasing flexibility. For a long time, the prevailing logic was that stricter regulation of layoffs would increase irregular jobs, and that an increase in working standards, such as minimum wages, would boost unemployment by causing small businesses to go bankrupt. The negative result is what we see today. A possible option is to create low-income jobs and guarantee income with tax assistance. However, it is time to change strategy and take the high road by combining stable jobs that offer social insurance and social investment oriented general services (Figure 10). Guarantee of Minimum Income In-work Security Social Service Public A ssistance Employ ment Pr otection + Incr eased Social Insur ance Education, Medical, Childcar e Creation of Low Income Jobs + EIT C Figure 10. Policy options for easing wage and household-income inequality. References Chang, Jiyeun. (2012). Level of inequality according to various strata of income definition and constituting factors. Trends and outlooks (Vol. 85). Seoul: Society for Social Sciences Studies/Park Youngryul Publishing Company. 22

23 Cheon, B. et. al. (2012). Growing inequality and its impact in Korea. GINI conference Country Report (unpublished). Choi, Sung-Eun. (2011). Study on the social welfare effect analysis of transfer expenditure. Public Finance Studies, 4(4), Cowell, Frank A., & Fioro, Carlo V. (2010). Inequality decompositions (GINI Discussion Paper No. 4). Amsterdam: Amsterdam Institute for Advanced Labour Studies, University of Amsterdam (AIAS). Förster, M. (2000). Trends and driving factors in income inequality and poverty in the OECD area (OECD Labour Market and Social Policy Occasional Paper No. 42). Paris: OECD. Hong, Kyung-Jun. (2011). Analysis of the distribution effects of public income transfer: with focus on the working poor. Social Welfare Policy, 38(2), Jasso, Guillermina, & Kotz, Samuel. (2007). Two types of inequality: Inequality between persons and inequality between subgroup (IZA Discussion Paper No. 2749). Bonn: Institute for the Study of Labor, University of Bonn.. Jenkins, S. P. (1995). Accounting for inequality trends: Decomposition analyses for the UK, Economica, 62(245), Jeong, Jin-Ho, & Choi, Gang-Sik. (2001). Decomposition of factors of working household income inequality. Economic Studies, 49(3), Kang, Shin-Wook et al. (2011). Social index of Korea seen through 2011 Korea Welfare Panel Data. Seoul: Korea Institute for Health and Social Affairs. Kang, Sung-Ho, & Choi, Ok-Geum. (2011). Comparative analysis on the effects of transfer income on poverty and income inequality mitigation: With focus on general households and agricultural and forestry households. Rural Economy, 34(1), Kang, Sung-Ho, & Lim, Byung-In. (2009). Inequality and polarization of old-age income and improvement effects of public pension. Social Security Studies, 25(2), Kang, Sung-Ho. (2011). Estimations of the size of the blind spot in the national pension scheme and the effects of pension system maturization on mitigating old-age poverty. Public Finance Studies, 4(2), Kim, Jae-Ho, & Jung, Ju-Yeon. (2012). Analysis of changes in income inequality and polarization following the introduction of basic old-age pension. Korean Association for Policy Studies Journal, 21(1). 23

24 Kim, Jin-Wook. (2004). Study of the effects of Korea s income transfer system on the reduction of income inequality and poverty. Social Welfare Policy, 20, Kim, Tae-Wan, Kim, Moon-Gil, Jeon, Ji-Hyun, & Han, Min-A. (2010). National Basic Living Security System: Livelihood guarantee assessment and policy directions. Seoul: Korea Institute for Health and Social Affairs. Koo, In-Hoi, Lee, Se-Hee, & Moon, Hye-Jin. (2010). Effects of the National Basic Living Security System on labor, income, and poverty: Estimation through double difference analysis. Korean Social Science, 44(1), Lee, Byung-Hee, & Kang, Shin-Wook. (2007). Trend analysis of recent income distribution, public transfer, and tax redistribution effects, Committee on Polarization and Livelihood Issues. Lee, Byung-Hee. (2007). Effects of labor market instability on income inequality. Economic Development Studies, 13(2), Lee, Chul-Hee. (2005). Rising family income inequality in the United States, : Impacts of changing labor supply, wages, and family structure (NBER Working Paper No ). Lee, Chul-Hee. (2007). Decomposition of factors of household income inequality increase: Effects of changes in wage, employment, work hours, household structure Joint Seminar on Economic Studies Report. Lee, Seung-Ho, & Koo, In-Hoi. (2010). Assessment of the appropriateness of the National Basic Living Security System s criteria for support obligator selection. Health and Social Affairs Studies, 30(1), Lerman, Robert I., & Itzhaki, Shlomo. (1985). Income inequality effects by income sources: A new approach and applications to the United States. Review of Economics and Statistics, 67, Maxwell, Nan (1990), Changing Female Labor Force participation: Influences on Income Inequality and Distribution, Social Forces, Vol. 68, No. 4, pp OECD. (2008). Growing unequal? Income distribution and poverty in OECD countries. Paris: Author. OECD. (2011). Divided we stand: Why inequality keeps. Paris: Author. Oxley, H., J. Burniaux, T. Dang, & d Ercole, M. M. (1997). Income distribution and poverty in 13 OECD countries (OECD Economic Studies No. 29). Paris: OECD. Seok, Jae-Eun. (2010). Analysis of the effects of the public pension system through the double difference model. Social Security Studies, 26(3),

25 Seok, Sang-Hoon. (2010). Analysis of basic old age pension s effect on reduction of old-age poverty. Elderly Welfare Studies, 50, Shorrocks, A. F. (1982). Inequality decomposition by factor components. Econometrica, 50(1), Shorrocks, A. F. (1984). Inequality decomposition by population subgroups. Econometrica, 52(6), Yeo, Yu-Jin. (2004). National Basic Living Security System support obligator selection criteria and blind spots for poverty: With focus on the relationship between public and private support. Health and Social Affairs Studies, 24(1). Yeo, Yu-Jin. (2009). Income redistribution effects of public transfer and taxation. Social Security Studies. Yeo, Yu-Jin et al. (2005). Poverty and inequality: Trends and factor decomposition. Seoul: Korea Institute for Health and Social Affairs. Yeo, Yu-Jin, & Song, Chi-Ho. (2010). Redistribution effects of public transfer programs: Comparison of Korea, Germany, Sweden, UK. Social Security Studies, 26(4),