Gender Impact of National Pension Reforms in Korea

Transcription

1 Gender Impact of National Pension Reforms in Korea Abstract This paper examines the gender impact of National Pension (NP) reforms in Korea. In 2007, the Korean government introduced fixed-amount of basic pension scheme, child credits and changed entitlement conditions for divorcées and widows and widowers. This paper examines the impact of policy changes for shorter working lives and lower wages, the survived and familial dependency. Findings indicate that the reforms have some positive features. However, the reforms still give better value for higher earnings and uninterrupted employment history, both of which are more characteristic of male workers. The familial dependency is still strongly constructed in the provisions. Protection for the survived remains still weak. Key words Pension, women, simulation model, Korea Introduction Pension policy stratifies the older members of society for differences in life course experience. We expect to see differences in class, gender, health status and ethnicity among the elderly. However, the most significant disparities in pensions occur between men and women. Pension policy is designed to provide high pensions for long and uninterrupted labour market participation; this is more prevalent among men than women. The disadvantaged position of older women has received a fair amount of scholarly attention, especially with regard to the high correlation between this population and poverty (Arber and Ginn, 1991; Groves, 1983, 1991; Walker, 1992, in comparative context Ginn, 1994; Ginn and Arber, 1992; Hutton and Whiteford, 1994; Ginn, 2000). This paper examines the gender impact of Korea s National Pension (NP) reforms. The NP was introduced in 1988 and initially it covered firms with more than ten employees. It is a contributory employment-based system which calculates pensions based on earnings and contribution period. The NP expanded its coverage to smaller firms (five or more employees) in 1992, to those in rural areas in 1995, to the selfemployed in urban areas in 1999 and to employees in workplaces with one or more employees in In 2007, the government introduced a fixed-amount of basic pension scheme. It also introduced child credits and changed entitlement conditions for divorcées and widows. Women tend to have shorter working lives and lower wages due to child bearing and rearing responsibilities. They also tend to live longer than men and are likely to experience severity of income loss in widowhood when they become most economically 1

2 vulnerable. In addition, pension policy constructs and perpetuates a certain type of dependency. This paper examines how the changes in policy rules affect recipients who have shorter working lives and lower wages. It also examines the extent to which the policy changes are effective for recipients who have a longer life expectancy. Finally it examines the extent to which the NP reinforces dependence on family. Researchers (Korean Women s Development Institute KWDI, 2002a; KWDI, 2007) have described the gendered structure of the NP. This paper both contributes to and marks a departure from the discussion. First it compares the gender implications of the new and old pension rules, looking into the extent to which particular elements of the policy protect men and women from pension risks. Second, in order to investigate the impact of pension reform on men and women, this paper creates a simulation of the old and new systems using national representative cross-sectional data. The following section discusses gender differences in labour market participation, women s longevity and the issue of familial dependency in Korea and I bring the features of the NP into this. This is followed by discussion on the changes in the new NP rules in Then, I present a simulation model to examine the extent to which the policy changes can reduce gender disparities in pensions. Finally, I present the results. Gendered pension risks and the NP Differentials in labour market participation and the NP There are significant gender differences in labour market participation in Korea. In 2006, men s labour market participation rate was 74 per cent while women s was 50.3 per cent (National Statistics Office NSO, 2007). Women s participation in the labour market rose from 48.9 in 1996 to 50.3 in 2006 (Table 1). Despite this improvement, Korean women s lifetime labour market participation still takes the form of M, which indicates that women tend to drop out of the labour market in order to bear and raise children. The KWDI (1996) found that in 1994 the proportion of female workers before marriage was 69 per cent; this figure dropped to 37 per cent after marriage and again to 28 per cent after the birth of the first child. The KWDI (2002b) also found that only 22.6 per cent of female workers remained in the labour market after marriage. After the birth of their first child, only 20.6 per cent of women remained in the labour market. 2

3 120 Figure 1 Labour market participation rate by sex and age group in 1996, 2006 % Age group 1996 female 2006 female 1996 male 2006 male Source: Author s calculation based on economically active population survey data Furthermore, much of the increase in women s employment is attributed to the presence of temporary and daily workers. Figure 2 shows the percentage of employed men and women aged by employment status. In 1996, 35.8 per cent of men were regular workers, and 17.4 per cent were irregular workers. In contrast, 13.9 per cent of women were regular and 18.7 per cent, irregular workers. In 2006, the percentage of male employed regular workers decreased by 0.9 percentage points, while the percentage of irregular male workers increased by 5 percentage points. At the same time, the percentage of regular female workers increased by 3.9 percentage points and that of irregular female workers by Seventy-three per cent of the increase in female paid workers between 1996 and 2006 resulted from the increase in the number of temporary and daily workers. 3

4 % Figure 2. Percentage employed, by employment status, men and women aged male 1996 female 2006 male 2006 female ers Temporary workers Daily workers Employers Self-employed Unpaid family workers Source: Author s calculation based on economically active population survey Gender inequalities can also found in the wages of male and female workers. Amsden, who wrote about wage discrimination in Korea, stated that Not only has Korea set world records with its growth rate in wages, it has also out-competed other countries in its discrimination against women workers (1989: 85). Lee and Lindauer (1991) found 60 per cent of the aggregate gender earnings differential in 1971 and 44 per cent in Official statistics shows that gender differences in wages have remained significant. On average, women s wages were 58 per cent of men s in 1995 and 63.6 per cent in 2005 (Ministry of Labour, 2007). The NP provides benefits based on earnings and contribution periods. Monthly pensions are calculated using the formula, (1.5 (A+B) + (1+0.05N))/12 provided by the National Pension Corporation (NPC). A is the average of the price-indexed average monthly income of all participants for the three years prior to pension payment. B is average amount of the Standard Monthly Income (SMI) of an insured person during his/her insured period. SMI is the amount specified within the range of 45 grades (1st through 45th) on the basis of the insured. N is the number of years exceeding 20 years. The NP provides a full old age pension to those who have contributed for at least 20 years. Reduced old age pensions are provided to those who have been insured for between 10 and 19 years, multiplied by payment rate provided by the NPC. Individuals can be either workplace, individually, voluntarily, or voluntarily and continually insured. All employees and employers from 18 to 59 years of age are automatically insured if they are working at a workplace with at least one employee. 4

5 The self-employed, non-income earners aged 27 or older, and non-income earners under 27 who have paid one or more month contribution can be individually insured. Nonincome earning spouses, non-income earners younger than 27 who have not paid any pension contribution, people protected under the National Basic Living Security Act, and retired government employees can be voluntarily insured. Insured people reaching 60 years of age with an insured period of less than 20 years and those with a special occupation such as miners or fishermen, who are aged 55 or older and entitled to an oldage pension with an insured period from 10 to less than 20 years can be voluntarily and continually insured. Workplace-based insured people and the voluntarily and continuously workplace-based insured persons pay nine per cent of their wages (4.5 per cent by employers and 4.5 per cent by employees). The individually, the voluntarily and continuously individually insured pay nine per cent. Women can be workplace-insured if they are working for a firm with more than one employee, or individually insured if they are unpaid family workers, self-employed, or voluntarily insured as a housewives. It appears, however that there are significant gender differences in NP participations (see Table 1). In 2006, there were 11,293,404 male participants and 6,446,535 female participants. The total number of participants is only 46.3 per cent of population aged 15 and over, 38,300,000 (NPC, 2007). Seventyeight per cent of males and 63 per cent of females aged 15 and over participated in the NP; this is equivalent to 81.3 per cent of the economically active male population and 65.3 per cent of the economically active female population. Among the employed, 84.7 per cent of males and 67.6 per cent of females participated. Among those, 47.5 per cent of women were workplace-insured, 51.8 per cent individually insured, 0.3 per cent voluntarily insured, and 0.3 voluntarily and continually insured. In contrast, 53.3 per cent of male insured were workplace-insured and 46.5 per cent were individually insured. Furthermore, there were significant differences in wages of the insured men and women. Among insured women, 63 per cent earned between $477 and $931 (exchange rate is approximately 1,300 won for a dollar at the time of this writing) while 59 per cent of insured men earned between $1,354 and $2,769 in 2006 (NPC, 2007). Table 1. Men and women s labour market participation and affiliation rates in the NP (2006) Labour market participation rate (%) No. of participants in the NP (A) A/ population aged 15 and over (%) A/economical ly Active population (%) A/ employed (%) 5

6 Male 74 11,293, Female ,446, Total ,739, Source: National pension statistical yearbook (2007) Familial dependency and the NP Welfare policies, assuming that some dependency is both normal and acceptable, may create or reinforce those relationships. Daly (1996) argues that a welfare state creates dependence. The welfare state may do so by granting or denying access to benefits to individuals or to families. When benefits are paid on the basis of family status or need, access to those benefits tends to be collectivized. For wives, divorcées and widows, their right is derived from the marital relationship. In this way, state benefits can mirror or reinforce this relationship. In addition, welfare states tend to stratify families by type, often favouring the male breadwinner family over the dual earner family (Land, 1978; Rake, 1998). An (2005) argues that the government assumes the husband should and/or can look after the spouse and thus does not provide sufficient state support. Based on the assumption regarding the spousal relationship, the NP provides a dependent pension and survivor pension. In addition, a division of pension rights was implemented in 1999 entitling the divorced spouse to half of the former spouse s pension rights if the marriage lasted for at least five years. The division of pension rights is suspended if the divorced dependent spouse remarries. The type of dependency within the NP is not just a matter of between husband and wife. The dependency in the NP is such that the male breadwinner of the family takes responsibility for all other members of the family. For example, dependents who are eligible for dependent pension include spouse, children younger than 18 or those with a first- or second-degree disability, and parents (and parents-in-law) aged 60 or over or with a first- or second-degree disability, supported by the pension beneficiary. An (2005) points out that the amount of pension from the NP could be lower than minimum standard of living not only for single elderly households but also elderly couples. This might imply that those with less support from the state may have to depend upon their family, and especially on their children. Longer life expectancy and the NP In 2005, Korean women s life expectancy was 82 and men s was 75. In addition, the proportion of the female population rises with age. In 2005, 60 per cent of the 65-and- 6

7 over population was female and this figure approached 75 per cent among those aged 85 and over (NSO, 2006). Women s greater longevity means that elderly women are more likely to experience bereavement and consequent reductions in household income later in life. The NP protects widows by providing survivor pensions. If an old-age pensioner, a disability pensioner with a first- or second-degree disability, and an insured person dies, a survival pension is paid to surviving dependents. A survivor pension is paid to surviving dependents with the priority to the spouse, then to children, parents, grandchildren, and grandparents. In addition, to acquire the survivor pension right, a person had to have been supported by the current or former insured person at the time of his/her death. The amount of a survivor pension is calculated by adding the DP and the amount gained by multiplying the old age pension of the dead by the payment rate based on insured period. The payment rate is differentiated by the insured period of the deceased: 40 per cent if the insured period is less than 10 years, 50 per cent for between 10 and 19 years and 60 per cent for more than 20 years insured period. Changes in the NP The pension rule changed in First, under the new rule, the basic old age pension is provided for people over the age of 65. The government pays the basic pension to single people whose monthly income is less than $308 and to couples earning $492. The benefits are fixed at $65 for a single and $103 for a couple. No minimum contributions are required to become eligible for the basic pension. Secondly, the 2007 reform changed the benefit calculation rule that the calculations are no longer based on SMI and they use the insured s income. Third, it provides child credits. From the second child, it gives 12 months of child credits towards contributions periods and from the third child, an additional 18 months. However, child credits should not exceed 50 months. Fourth, for the divorced, division of pension rights continues even after remarriage. The divorced spouse can claim both his/her own pension rights and the divided pension rights. Fifth, the survivor pension term is reduced from five to three years, and is suspended when the survivor reaches 55. If the survivor is eligible for two different types of pension rights, he or she must choose one. In a case where the chosen pension is not the survivor pensions, the NP provides 20 per cent of the survivor pensions. Table 2. Old and new systems of the National Pension Old system (1999) New system (2007) 7

8 Structure Earnings-related pension Basic pension Earnings-related pension Base earnings Standard Monthly Income (SMI) Insured earnings Years of 10 years for the earnings-related 10 years for the earnings-related eligibility pension pension No minimum for the basic pension Child credits none For the second child 12 months credits and 18 months credits for the third and after children The divorced Half of the ex-spouse s pension or his/her own pension, ceased at remarriage Continuously provided at remarriage, co-receipt of the division of pensions and his/her own pension Widows Duration of payment: 5 years Either survivor pension or his/her own pension Duration of payment: 3 years In case of his/her own pension is chosen, 20% of survivor pension is paid Methodology This section discusses a simulation model to examine the gender impact of pension reforms. Several factors make it difficult to analyze how women fare relative to men under the new and old pension systems. First, the new system has not been in effect long enough to be mature. Second, it is impossible to know how someone who is fully covered under the new system will fare in the future. Third, it is difficult to predict precisely what the rate of wage growth and rate of return on investment will be; longitudinal data which tracks actual employment histories of current retirees and workers to estimate their retirement accumulations and entitlements are not available. I solved these problems by constructing synthetic men and women, using cross-sectional data on current behaviour of people of different ages, educational levels and marital status to proxy the lifetime employment, wage and contributions of synthetic cases. I then simulated how these synthetic men and women would fare under the old and new systems. The model has several components. To begin with the model process, I first specify characteristics of the synthetic couples: for example, their level of education, the age at which they marry, the education of their spouses and the ages at which they bear children. These specifications are based on cross-sectional data from NSO. Next, using 8

9 statistical relationship estimated from Economically Active Population (EAP) Survey Data, I predict labour market participation and earnings for these synthetic couples for each year of their lifetime. Then, using the earnings, I model pension contributions and receipts. This methodology assumes that age-specific labour force participation and wage behaviour will remain constant over time. I interpreted these as age effects rather than cohort effects. In reality, cohort effects are undoubtedly involved. Women s labour force participation rates are increasing as women complete more years of formal education. This means that aggregate female labour force participation rates will also rise over time. Changing social norms may bring more women into the paid labour force. Moreover, the work incentives and disincentives in the new pension systems may alter work habits. These potential changes in age-specific female labour force participation rates were not taken into account. This method has been often used in a different vein (Davis and Joshi, 1994; Joshi et al., 1996; Evans, 1996; Evans and Falkingham, 1997; Johnson and Rake, 1999; Rake et al., 2000). An obvious limitation of the method is that it is not possible to generalise from the results, which are highly sensitive to the choice of synthetic cases. The synthetic cases nevertheless are to be chosen to examine the policy impact on specific cases (Evans, 1996; Bradshaw et al., 1993). Rake et al. (2000) argue that the method has the advantage of showing the elements of these policies in detail so that a research can fully explore the link between policies and outcomes. Development of simulation model My synthetic couples are classified into three occupational levels, based on their educational attainment. The model considers middle school and under as low, high school and two years of college as middle, and university level as high educational qualification. I assume that the more highly educated a person is, the older he or she is at the time of marriage. I also assume that the synthetic women marry men who are as educated as they are. The emphasis on educationally matched couples is supported by research (KWDI, 2002c). It shows about half of the wives had the same qualification level as their husbands. Each husband is assumed to be three years older than his wife. Statistics (NSO, 2007) show that average age difference is 3.1 years. The marriages in average synthetic cases are lifelong partnerships. However, I also construct a case in order to examine the impact of divorce. I also assume that the age of the synthetic mothers varies with their educational qualification - the more highly educated the 9

10 woman, the older she is when her first child is born. These synthetic women have their first child two years after marriage and the rest at two-year intervals. The analysis focuses on mothers with one child. However, by simulating scenarios of two children, I give a more graphic illustration of the different experiences in the family and the labour market. Time out of the labour market reduces the total amount of lifetime wages. For women, this time out of the labour market reflects time spent on unpaid domestic work (primarily childcare) and in this analysis, the mean length of time out of employment for women is set up as five years. My synthetic couples marital, fertility and education specifics are as follows. I chose a typical job for the synthetic people of low and middle and high education drawn from national statistics. Mr. and Mrs. Mid have a median level of education and hold middle-skills occupations (the wife works in a clerical position and the husband is a machine operator). Mrs. Mid starts working at age 22, marries at age 26, and has her first child at 28. Mr. Mid starts working at 25, marries at 29, and has his first child at 31. Mr. and Mrs. High have high educational attainments and high-skill jobs (the wife is a teacher and the husband is a professor). Mrs. High starts working at 24, marries at the age of 28 and has her first child at 30. Mr. High starts working at age 27 after university graduation and military service, marries at 31 and has the first child at the age of 33. Mr. and Mrs. Low have low educational qualifications and low-skill jobs (both husband and wife are manual labourers). Mrs. Low marries start working 20, marries at 21, and gives birth at 23. Mr. Low starts working at 22, marries at 24, and has the first child at 26. The synthetic cases are modeled to be dual earner (DE) households with one child. For comparisons, I created male breadwinner (MB) households with two children. In the male breadwinner households, the wives leave their jobs after the birth of the first child. The synthesized cases are not an average of those in the real world, nor are their lives as complicated. The projected lifetime employment status, wages and pension generated by the model are not forecasts. The synthetic cases are just distillations of reality, upon which they cast some light. 10

11 Predictions of labour market behavior National statistics in Korea divides the employed into six categories: regular workers, temporary workers, daily workers, the self-employed, unpaid family workers, and employers. In the model, both for men and women, I predict the outcomes not employed, or employed regular or employed irregular which includes temporary and daily employment. Therefore, a multinomial logistic model is used to model the synthetic cases labour market participation. Variables such as sex, marital status, educational qualifications, and job types are considered to affect the probability of the cases' participation in the labour market. Table 3 shows the logistic regression outputs of the labour market participations. The reference category is not employed. The outputs indicate the effect of sex, marital status, education, and job type on the labour market participation are statistically significant. Table 3. Logistic regression outputs on labour market participation Employment status Coef. Std. Error Exp(B) Intercept 1.425**.000 Male.271** Female 0*.. Single.835** Married -.298** Divorced -.244** Widowed 0*.. Low 1.060** Middle -.320** High 0*.. Senior officers ** Professionals.160** Technicians 0.820** Clerks.983** Service workers ** Sales workers ** Skilled, agricultural, forestry and fishery workers ** Craft and related trades workers **

12 Plant, machine operators and assemblers -.596** Elementary occupation 0*.. Irregular work Intercept 3.005**.000. Male -.643** Female 0*.. Single.797** Married 0.942** Divorced -.660** Widowed 0*.. Low.176** Middle.366** High 0*.. Senior officers ** Professionals ** Technicians ** Clerks ** Service workers ** Sales workers ** Skilled, agricultural, forestry and fishery workers ** Craft and related trades workers ** Plant, machine operators and assemblers ** Elementary occupation 0*.. Log-likelihood *** ***this reduced model is equivalent to the final model because omitting the effect does not increase the degree of freedom. ** p <.000 *this parameter is set up zero because it is redundant. Anyone with a probability of employment over 0.5 is counted as employed, even though there might be an up to 49 per cent chance of his or her being out of paid work. Among those who are presumably employed, if the probability of regular employment is over 0.5, the assumption is that the worker is a regular worker; otherwise I assume that the worker is employed on irregular basis. I find that regardless of education, men are predicted to be either regularly or irregularly employed between age 15 and 59. The average probability of men with low education being regularly employed is 0.19, of 12

13 being irregularly employed, 0.39 and of not being employed, The comparable probabilities of those with middle education are 0.37, 0.34 and Those with high education have a 0.61 probability of being regularly employed, a 0.13 probability of being irregularly employed and a 0.24 probability of not being employed. I also find that women also are likely to be either regularly or irregularly employed between 15 and 59. Women with low education have a 0.07 probability of being regularly employed, a 0.50 of probability of being irregularly employed, and a 0.42 probability of not being employed. Women with middle education have a 0.26 probability of being regularly employed, a 0.49 probability of being irregularly employed and a 0.25 probability of not being employed. Those with high education have probability of being reemployed on regular basis as 0.55, for being irregularly employed as 0.26 and 0.18 for not being employed. I conduct multiple linear regressions to estimate the monthly wages of the synthetic cases. The predicted wages are based on age, sex, employment status, marital status, education, job types and employment status. Table 4 displays the outcomes. For sex, male is the reference category, for marital status it is single, for education, low education, for job type it is senior officers and for employment status, it is regular work. It shows that the explanatory factors significantly affect the wages, with not employed with the strongest influence. On average, between the ages of 15 and 59, highly educated men are predicted to earn $1,523 per month while those with middle education, $846 and those with education $586. Less educated women are predicted to earn $252 per month, middle-educated women earn $505 and highly educated women earn $1,038. Gender differences in these predicted wages are apparent: women with high education are predicted to earn a monthly income that is 68 per cent of that of their male counterparts. For the middle and low educated women, the sex ratios are predicted to be 60 per cent and 44 per cent, respectively. Table 4. Linear regressions outputs on wages Coef. Std. Error t Constant ** Female ** Married ** Divorced ** Widowed ** Middle 5.965** High **

14 Professionals ** Technicians ** Clerks ** Service workers ** Sales workers ** Skilled, agricultural, forestry and fishery workers ** Craft and related trades workers ** Plant, machine operators and assemblers ** Elementary occupation ** Irregular work ** Not employed ** R square* ** p <.000 * this model explains 60.4 per cent of variance in wages. Result and discussion Labour market participation and wages Table 5 shows the simulation outcomes of the cases probability on labour market participation and monthly wages over the working life. Mr. Low (DE) is predicted to be irregularly employed between ages 22 and 59. Before getting married he is predicted to earn $316 per month and after marriage he earns $669 per month. Mrs. Low (DE) is simulated to be out of the labour market for five years, until age 27. Before giving birth, she is simulated to be working as irregular worker with average monthly earnings of $123 and after a five-year break for child rearing, she returns to the labour market as an irregular worker with average monthly wages of $339. Mr. Mid (DE) is simulated to be employed as a regular worker without interruption between 25 and 59. Before marriage he earns $1,222 and after $1,575. Mrs. Mid (DE) is predicted to be regularly employed between 22 and 26 with average monthly wages of $1,013. After marriage, her monthly earnings are predicted to increase to $1,366. Giving birth affects Mrs. Mid s labour market participation because she takes five years away from her job to raise her child. At 33, she returns to the labour market as an irregular worker with monthly wages of $648. Mr. High (DE) is predicted to be regularly employed between 27 and 59, with earnings of $1,827 before marriage and $2,180 afterwards. Mrs. High (DE) is regularly 14

15 employed between 24 and 28, earning $1,496. After marriage her monthly earnings increase to $1,849. She leaves the labour market at 30, after the birth of her first child, and returns at age 34. She then returns to labour market as an irregular worker earning $1,131. Table 5. Simulation outcomes of the synthetic cases of dual earner (DE) couples on labour market participation and monthly wages Age Employment status Monthly wages ($) Mr. Low Irregular work Irregular work Mrs. Low Mr. Middle Mrs. Middle Mr. High Mrs. High Irregular work Irregular work Not employed Irregular work Not employed Irregular work Not employed Irregular work ,222 1,575 1,013 1, ,827 2,180 1,496 1, ,131 Table 6 shows simulated results of the synthetic cases of male breadwinner couples. The simulated predictions on the labour market participations and monthly wages are the same as those of the husbands in the dual earner couples. I assumed that the wives in the male breadwinner families will be unemployed after the birth of the first child. On the one hand, Mrs. Low (MB), Mrs. Mid (MB) and Mrs. High (MB) are simulated to have significantly shorter working lives. Mrs. Low (DE) has 35 working years, Mrs. Mid (DE), 33 years and Mrs. High, 31 years. On the other hand, Mrs. Low (MB) works for three years, Mrs. Mid works for six, and so does Mrs. High. 15

16 Table 6. Simulation outcomes of the synthetic cases of male breadwinner (MB) couples on labour market participation and monthly wages Age Employment status Monthly wages ($) Mr. Low Irregular work Irregular work Mrs. Low Mr. Middle Mrs. Middle Mr. High Mrs. High Irregular work Irregular work Not employed Not employed Not employed ,222 1,575 1,013 1, ,827 2,180 1,496 1,849 0 Simulated contributions and pensions A few caveats about the simulating contributions and pensions are in order. First, according to the benefit calculation rule, the insured period included is the period during which contributions are paid. Any period with payment of postponed contributions is taken as an insured period. Periods of unpaid contributions and of contribution exemption are excluded from the insured period. In short, the NPS regards years with contributions as insured periods whether or not the insured is in the labour market. I cannot be sure if those employed on an irregular basis pay contributions to the NP. But, in the interest of simplicity, the assumption is that each case makes contributions while employed and employers pay the contribution for the employees. Second, in order to calculate the pensions, I need to know the average monthly income of all participants in the programme three years before their pensions start, which is A in the calculation formula. In the absence of data, I use A value provided by NPC in 2006 which was $1,205. Third, a husband s pension income includes the DP amount, which allows for a spousal supplement but not a child supplement, as we assume that the children do not live with their retired parents. The annual dependent pension amount provided by NPC 16

17 in 2006 was $154. It is important to note that the DP is paid to the insured only if the dependent spouse does not have his or her own pension entitlements. Fourthly, the basic pension is provided to those with monthly income of less than $308 for a single household and $492 for a couple at age 65. The income includes earnings and other income courses adjusted by annual interest rates. However, I do not simulate incomes sources other than the pensions from the NP. Instead, consider the pension amount as the income at age 65 to be applied to the basic pension. Fourth, in cases where there are fewer than ten contribution years, lump-sum benefits are provided. These lump-sum benefits are calculated by multiplying the contribution amount by interest rates. In 2006, the interest rates applied was 3.8 per cent. Fifth, prices are constant and I do not allow for either inflation or indexation. Table 7 shows the simulated monthly contributions and pensions of dual earner couples under the old rule of the NP. Mr. Low (DE) is predicted to contribute $29 per month for 38 years and receive a pension of $230 for 20 years until he dies at 79. Mrs. Low (DE) is predicted to pay $15 for 35 years and receive $192 per month for 23 years. Mr. Mid (DE) pays $67 for 35 years and receives $345 for 20 years. Mrs. Mid (DE) is predicted to pay $33 per month for 33 years and receive $243 for 23 years. Mr. High (DE) pays $95 per month for 33 years and receives $415 for 20 years. Mrs. High (DE) pays $55 for 31 years and receives $304. Total pensions are $422 for Mr. and Mrs. Low, $587 for Mr. and Mrs. Mid and $720 for Mr. and Mrs. High. Table 7. Simulated monthly contributions and pensions of dual earner couples under the old rule ($) Mr. Mrs. Total Contributions Pensions Contributions Pensions Contributions Pensions Low Middle High Table 8 shows the simulated monthly contributions and pension of male breadwinner couples under the old rule. Mr. Low (MB), Mr. Mid (MB) and Mr. High (MB) receive DP of $13 per month. Mrs. Low (MB) pays $13 for three years while she is working. Mrs. Mid (MB) pays $50 for six years and Mrs. High (MB) pays $73 for six years. Since they do not have ten years of contributions, they are entitled to lump-sum benefits when they reach 60. Mrs. Low is entitled to $474, Mrs. Mid to $3,742 and Mrs. High to $5,484. The figures in tables are monthly amount for 23 years. Monthly couple total 17

18 pensions are $245, $371 and $448. Table 8. Simulated monthly contributions and pension of male breadwinner couples under the old rule ($) Mr. Mrs. Couple total Contributions Pensions Contributions Pensions Contributions Pensions Low Middle High Table 9 shows simulated monthly contributions and pension of dual earner couples under the new rule. Mr. Low (DE) is predicted to contribute $29 per month for 38 years and receive a pension of $232 for 20 years. Mrs. Low (DE) is predicted to pay $15 for 35 years and receive $192 per month for 23 years. Mr. Mid (DE) pays $69 for 35 years and receives $342 for 20 years. Mrs. Mid (DE) is predicted to pay $33 per month for 33 years and receive $243 for 23 years. Mr. High (DE) pays $102 per month for 33 years and receives $435 for 20 years. Mrs. High (DE) pays $50 for 31 years and receives $304 for 23 years. At age 65, Mr. and Mrs. Low s total income is estimated to be $424; since this is less than $492, they receive $103 per month for 15 years. Thus, until Mr. Low s death, they are modeled to receive $527 per month. Table 9. Simulated monthly contributions and pension of dual earner couples under the new rule ($) Mr. Mrs. Couple total Contributions Pensions Contributions Pensions Contributions Pensions Low Middle High Table 10 shows simulated monthly contributions and pension of male breadwinner couples under the new rule. Male cases show the same trajectories of contributions. With DP of $13, Mr. Low (MB) is estimated to receive $245, Mr. Mid (MB), $355 and Mr. High (MB), $448. Mrs. Low (MB) pays $12 for three years while she is working. Mrs. Mid (MB) pays $50 for six years and Mrs. High (MB) pays $73 for six years. Since they do not make contributions for ten years, they are eligible for lump-sum benefits when they reach 60. Mrs. Low (MB) is entitled to $448 and Mrs. Mid (MB) 18

19 $3,803 and Mrs. High (MB) $5,427. In addition, Mrs. Low (MB), Mrs. Mid (MB) and Mrs. High (MB) are entitled to child credits for their second child. It gives them $2 for Mrs. Low (MB), $16 for Mrs. Mid (MB) and $23 for Mrs. High (MB) on a monthly basis. At age 65, Mr. and Mrs. Low (MB) s incomes are $247, Mr. and Mrs. Mid (MB) s income is $371 and Mr. and Mrs. High (MB) s incomes are $471, which are less than $492. Thus, they are entitled to the basic pension of $103 per month for 15 years. Mr. and Mrs. Low (MB) incomes are $350, Mr. and Mrs. Mid (MB), $474 and $574 for Mr. and Mrs. High. Table 10. Simulated monthly contributions and pension of male breadwinner couples under the new rule ($) Mr. Mrs. Couple total Contributions Pensions Contributions Pensions Contributions Pensions Low Middle High First, the results shows that the new pension rule still gives better value for higher earnings and uninterrupted employment history, both of which are more characteristic of male workers. The synthetic female cases in dual earner couples have shorter working lives and lower wages than do their male partners. This results in gender differences in pensions that under the old rule, Mrs. Low s (DE) pension is lower than her husband s by 16 per cent, Mrs. Mid s (DE) is 30 per cent lower and Mrs. High (DE) is lower by 27 per cent than their husbands. The new system does not do much in this regard, as the pension amount of Mrs. Low (DE) is lower by 17 per cent, for Mrs. Mid (DE) by 29 per cent and for Mrs. High by 30 per cent. Women are likely to drop out of labour market to bear and raise children. For Mrs. Low (MB), Mrs. Mid (MB) and Mrs. High (MB), under the old rule, being unemployed after the first child means a 99 per cent, a 96 per cent and a 95 per cent reduction in pension compared to their husbands. Compared to Mrs. Low (DE), Mrs. Mid (DE) and Mrs. High (DE), they have lower pensions by 99, 94, and 93 per cent. Under the new rule, the disadvantages remained similar. They have pensions lower by 99 per cent, a 96 per cent and a 96 percent compared to their husbands. Compared to Mrs. Low (DE), Mrs. Mid (DE) and Mrs. High (DE), they have lower pensions by 99, 94, and 94 per cent. For the first time in history, Korea s government is compensating women for their 19

20 shorter working lives by offering child credits. As the child credits apply for the second and subsequent children, this change does not help our dual earner couples. The new rule increases the pension amount by 26 per cent for the Mrs. Low (MB), 15 per cent for the Mrs. Mid (MB) and 15 per cent for the Mrs. High (MB). Despite the positive effects, however, it is doubtful how many married women would enjoy this as compensation of their loss in labour market participation and wages, considering the fact that Korea has recorded the lowest fertility rate in the world, 1.18 in Second, women outlive men. In order to examine the extent to which the NP protects the survivorships, I assume that the synthetic male cases die at age 79 which means our female cases can be entitled to the survivor benefits from age 76. As the male cases are simulated to contribute to the NP for more than 20 years, the survivors are entitled to 60 per cent of the deceased s old age pensions. Under the old rule this comes to $138 for Mrs. Low, $205 for Mrs. Mid and $249 for Mrs. High. The wives in the dual earner couple find that the survivor pensions are lower than pensions derived from their own contributions. As a result, they choose to receive their own pensions. In contrast, Mrs. Low (MB), Mrs. Mid (MB) and Mrs. High (MB) receive the survivor pensions. As a result, the difference between Mrs. Low (DE) and Mrs. Low (MB) is $54, the difference between Mrs. Mid (DE) and Mrs. Mid (MB) is $36. The difference between Mrs. High (DE) and Mrs. High (MB) is $55 per month. It implies that 35 working year for the Mrs. Low (DE) and 33 working years for the Mrs. Mid (DE) and 31 working years for Mrs. High (DE) may become meaningless as Mrs. Low (MB), Mrs. Mid (MB) and Mrs. High (MB) who spend their lifetime as a housewife have similar incomes. Under the new rule, 60 per cent of the deceased old age pensions are $139 for Mrs. Low, $205 for the Mrs. Mid and $261 for Mrs. High. Mrs. Low (DE), Mrs. Mid (DE) and Mrs. High (DE) find the survivor pensions lower than the pensions derived from their own contribution and thus chose their own pensions. This gives them an additional 20 per cent of the survivor pensions which are $46 for Mrs. Low (DE), $68 for Mrs. Mid (DE) and $87 for Mrs. High (DE). Total pensions for Mrs. Low (DE) is $186, for Mrs. Mid (DE) is $274 and for Mrs. High (DE) is $348. This increases the pensions of Mrs. Low (DE) by 24 per cent and of Mrs. Mid (DE) by 28 per cent and Mrs. High (DE) by 29 per cent compared to the old rule. It indicates that having been a worker can help the surviving spouses. Despite these positive effects, however, the extent to which the new rules for the survivorship would protect vulnerable elderly widows is doubtful. In 2006, the Minimum Cost Living (MCL) for a single household was $322. Under the new rule, pensions of Mrs. High (DE) are higher than the MCL, but not for the Mrs. Low (DE), or Mrs. Mid (DE). Neither is this the case for Mrs. Low (MB), Mrs. Mid 20

21 (MB) and Mrs. High (MB). It implies that without sufficient help from other sources, particularly from family, the surviving elderly women would suffer most. Third, the Korean state provisions for financial security in old age appear to construct a familial dependency. The basic pension under the new rule would increase the pension amounts by 20 per cent for Mr. and Mrs. Low (DE), 41 per cent for Mr. and Mrs. Low (MB), 27 per cent for Mr. and Mrs. Mid (MB) and 22 per cent for Mr. and Mrs. High (MB). However, the state provisions are insufficient for minimum cost of living. In 2006, the MCL for a couple was $539 per month (MHW, 2007). Mr. and Mrs. Low (DE), Mr. and Mrs. Low (MB), and Mr. and Mrs. Mid (MB) have pensions less than the MCL (Table 9 and Table 10). It implies that unless they supplement their income from other sources, from their children for example, they are likely to fall into poverty. Familial dependency can be also found in the protection of the divorced. Let us construct a situation where our synthetic couples divorce when the wife is 37 and the husband is 40. The divorced wife remarries five years later. After the remarriage she continues to do paid work. Mr. and Mrs. Low are married for 17 years, Mr. and Mrs. Mid for 12, and Mr. and Mrs. High for 10. The old age pension amounts for the marriage periods are $208 for Mr. and Mrs. Low and $220 for Mr. and Mrs. Mid and $221 for Mr. and Mrs. High. The simulated monthly divided pensions are $104 for Mrs. Low, $110 for Mrs. Mid, and $110 for Mrs. High. Under the old rule, in the case of remarriage, the division of pension rights is not given to the divorced wife with the understanding that her financial well-being in later life is the responsibility of her husband, not of the state. Under the new rule, Mrs. Low (DE), Mrs. Mid (DE) and Mrs. High (DE) receive the divided pensions in addition to their own pensions which give them $296, $352, and $414, respectively. Mrs. Low (MB), Mrs. Mid (MB) and Mrs. High (MB) receive divided pensions of $104, $109 and $110, respectively. In contrast to the situations under the old rule, the new rule is more generous to the divorced. Nonetheless, the 17 years of marriage do not provide security for Mrs. Low (DE). Having been married for more than 10 years does not help Mrs. Low (MB), Mrs. Mid (MB) and Mrs. High (MB) if they become divorced. Although they would be supported by their second husbands in old age, technically the amounts are lower than the MCL for a single household. Thus I note that it is a good change that remarriage does not preclude pension entitlements. However, it appears that the divorced women will have to find financial security by remarrying or depending on their children. 21

22 Conclusion This paper has examined the gender impact of pension reforms in Korea. While the reforms have some positive features, such as enhanced access through provision of child credit and flat rate basic pension, better protection for divorce and survivorship, this analysis reveals how much still remains to be done. The new pension rule still gives better value for higher earnings and uninterrupted employment history, both of which are more characteristic of male workers. The familial dependency is strongly constructed in the provisions. Protection for elderly widows remains inadequate. The methodology employed, despite its limitations, has been proven to make the policy more sensitive to gender differences in pensions. The lives of Korean women today are more complicated than they were in the past. The assumption that married women do not need social security in their own right is at odds with reality. The divorce rate has been rising; three couples in ten eventually divorce. This increase in divorce is unlikely to be reversed. The decline of marriage as a lifelong contract makes a wife s reliance on her husband for income in later life an even more risky strategy. The growing participation of women in the labour market means that more women will expect and need their own pension income. The Korean government has started to recognise women s pension needs, but has not yet fully met those needs. The needs of women must be fully recognised, understood, and met. Women will welcome more access to pension, but the factors affecting the final level of pensions need to be truly women-friendly. Furthermore, women should have more opportunities for good jobs, which can be important for their economic independence in later life. Without these, pension policy will fail to acknowledge women s right to a quality of life based on economic security. Bibliography Amsden, A Asia s next giant: South Korea and late industrialization. Oxford, Oxford University Press. An, M. Y Ageing and gender inequality: gender impact of pension reforms. D.Phil. thesis, Oxford University Arber, S. and Ginn, J Gender and later life: a sociological analysis of resources and constraints. London, Sage Publication. Bradshaw, J., Ditch, J., Holmes, H., and Whiteford, P Support for children: a comparison of arrangements in 15 countries. York, Social Policy Research Unit. Daly, M Social security, gender and equality in the European Union. Brussels, Commission of the European Communities. 22

23 Davies, H. and Joshi, H Sex, sharing and the distribution of income, in Journal of Social Policy, Vol. 23, No. 3. Evans, M. 199). Families on the dole in Britain, France and Germany. Welfare State Programme Discussion Paper no. 118, London, STIRCED, LSE. Evans, M. and Falkingham, J Minimum pensions and safety net in old age: a comparative analysis. Welfare State Programme Discussion Paper no. 131, London, STICERD, LSE. Ginn, J. and Arber, S Towards women s independence: pension systems in three contrasting European welfare states, in Journal of European Social Policy, Vol. 2, No.4. Ginn, J Gender and pensions in Europe: current trends in women pension acquisition, in P. Brown and R. Crompton (eds.) Economic restructuring and social exclusion. London, UCL Press. Ginn, J Gender, pensions and life course: how pensions need to adapt changing family form, Bristol, The Policy Press. Groves, D Members and survivors; women and retirement-pensions legislation, in J. Lewis (eds.) Women s welfare, women s rights. London, Croom Helm. Groves, D Financial provisions for women in retirement, in M. Maclean and D. Groves (eds.) Women s issues in social policy. London, Routledge. Hutton, S. and Whiteford, P Gender and retirement incomes; a comparative Analysis, in S. Baldwin and J. Falkingham (eds.) Social security and social change: new challenges to the Beveridge model. London, Harvester Wheatsheaf. Johnson, P. A. and Rake, K Comparative social policy research in Europe. Social Policy Review 10. Joshi, H., Davies, H., and Land, H The tale of Mrs typical. London, Family Policy Studies Centre. Korean Women s Development Institute Korean women s labour market participation and employment structure. Seoul, KWDI. (In Korean) Korean Women s Development Institute. 2002a. Research report on women recipients of social insurance benefits. Seoul, KWDI. (In Korean) Korean Women s Development Institute. 2000b. The fourth survey on the women s labour market participation. Seoul, KWDI. (In Korean) Korean Women s Development Institute. 2002c. Research report on livelihood and welfare need of different types of families, Seoul: KWDI. (In Korean) Korean Women s Development Institute Research report how to enhance women s pension rights in Korea. Seoul, KWDI. (In Korean) 23