경제학적 분석 방법에 의한 출산 결정에 대한 정책적 지원의 효과

Abstract

본 연구는 현재 대한민국에서 전방위적으로 시행되고 있는 다양한 출산장려 정책 중에서도 최근 사회적 화두가 되고 있는 영유아 보육/교육비 지원 제도에 초점을 맞추었다. 영유아 보육/교육비 지원 제도는 미취학 아동의 양육비용을 국가가 보조해 줌으로써 자녀를 가질 것인지 말 것인지에 대한 의사 결정을 하고자 하는 잠재적 부모들에게 미래의 양육비용에 대한 기대치를 감소시켜 보다 적극적으로 출산을 하게 하려는 데에 아이디어를 두고 있는데, 이에 대한 분석을 위해 본 연구는, 한국아동패널(Panel Study of Korean Children, PSKC)과 여성가족패널조사(Korea Longitudinal Study of Women and Families, KLoWF)의 결합자료를 이용하여, 개인의 실제 출산 행태와 기대 양육비용 사이의 관계를 보다 직접적으로 분석하고자 시도하였다. 이를 위해 본 연구에서는 커널 매칭 기법(the method of kernel matching)을 이용하여 여성가족패널에 포함된 출산력 정보와 한국아동패널이 보유하고 있는 영유아 보육환경 및 양육비용에 대한 정보를 결합하여 분석에 이용하였다. 패널데이터 매칭 후 로짓회귀 분석을 실시한 결과, 잠재적 모의 경우 기대 양육비용과 실제 출산 결정과의 부적(-) 관련성이 약하게 나타났으나 저소득층과 대졸이하 학력 집단의 경우에만 특히 유의하였다. 출산행태와 양육비용의 관계에 대해서는 많은 이론적 논의들이 있어 왔으나, 이를 실제 대규모 데이터를 이용한 분석을 찾기는 매우 쉽지 않으며, 패널 데이터를 보다 적극적으로 활용하여 기존의 연구들에서 직접적으로 다룰 수 없었던 출산행태와 양육비용의 관계를 보다 명시적으로 살펴보고자 한 데에 본 연구의 의의가 있다고 하겠다.

Ⅰ. Introduction As low fertility rate and late marriage have become serious social issues in Korea, a lot of studies have been conducted to find out the reasons and draw policy steps to aggress them. Existing studies have indicated that the low birth rate in Korea is due to the increase of opportunity cost of childbirth, increased child rearing expenses, and conflicts between work and family since increasing number of women get higher education. A lot of attempts have been made such as financial support and educational support for those who raise young children. This study aims to focus on the effect of Early Childhood Education and Care(ECEC) on the Korean women’s decision-making of childbirth, which have been adopted throughout the country among various childbirth support policies. Financial support for caring and educating children is to support childrearing expenses to reduce the expected amount of money that prospect parents take into account when they thought about having children. Methodologically this study aims to circumvent difficulties with which many existing studies faced by using the data coupled with Panel Study of Korean Children (PSKC) and Korea Longitudinal Study of Women and Families (KLoWF). PSKC started since 2008 has sufficient information on child raising environment and cost. And while KLoWF started around the same time lacks this information, it conducted a study on the history of childbirth among fertile women. This study combined the information on the history of childbirth contained in a study by KLoWF and on child raising cost contained in PSKC based on the method of kernel matching and made the information into a single data. And based on this, the authors want to analyze the relationship between expected child raising cost and childbirth in an individual level. Furthermore, we hope to contribute to current discussions on child birth incentives currently being promoted in Korea by re-evaluating conclusions and implications drawn in existing studies.

II. Theoretical Background and Literature Review Existing research on the determinants of birth can be categorized as follows; demographic sociological approach, economic approach, and cultural approach. Demographic approaches focus on demographic factors such as the age of women when they get married for the first time, marriage partners, and educational level while economic approaches explain childbirth based on utility and cost. Cultural approaches emphasize the influence of cultural factors such as opinions on childbirth and a notion of preferring a son to a daughter. Of course, these approaches cannot be separated completely and thus can be overlapped to some extent. In socio-demographic analysis on the determinants of childbirth points out that the more the educational level of women, the higher the employment rate and longer education period, which make increase the women's age at the time of marriage. As higher education leads to a higher salary, this becomes a high opportunity cost if a woman stops working for childbirth. That is why higher education of women is a negative impact on childbirth(e.g., Ryu Gi-cheol & Park Young-hwa, 2009; Lee Sam-sik et al., 2005; Choi Jun-wook & Song Hyeon-jae, 2010) Economic approaches to childbirth focus on the usefulness of childbirth and the accompanying costs to analyze the determinants of childbirth. These explain how individual and social factors such as age and education level have an impact on childbirth in the process of married women's decision on providing labor. An important feature of economic approachers is that it assumes children as goods influencing utility function. Aforementioned Becker (1960, 1973, and 1976) suggested Quantity Quality Tradeoff moel explicitely considering the number of children (quantity) and quality and this model has been used to many empirical studies using the data from each country. Cho Youn-young (2006) and Choi Kyeong-su (2008) adopted life-cycle income hypotehsis of women and theoretically analyzed how consumption, labor, and childbirth are related and tried to verify this using domestic and international statistics. Cho (2006) conducted mock tests and concluded that aids for supporting children's education and deduction of income proportional to one's income are more effective than common supports such as childcare benefit. Choi (2008) calculated potential cost of childbirth and concluded that to enhance the fertility rate, there need ways to mitigate the impact of decreased labor, especially the ones that could facilitate returing to a labor market after childbirht. In a study of childcare support for children aged 0 to 5, Hong Seok-chel and Kim Jeong-ho (2012) assumed that childcare benefit has an impact on increasing economic participation of mothers of young children. But they assumed that this does not have a significant impact on the childbirth of additional children and asserted child support benefit and labor market policy should be implemented complementarily. As pointed out by Lee Sam-sik et al. (2005), there have been a lot of studies on the determinants of childbirth. Ultimately, childbirth is determined by the supply and demand for children and the prospect of realizing individual (or couple)'s childbirth motive is influenced by projected cost. This cost not only includes financial cost, but includes time, social norms, health, and psychological costs as well as economic status and social standing. Since it is very hard to examine the determinants and causes for changes, this study focuses on the effect of all expenses entailed by childbirth, which is the focus of eonomic approaches. The author especially tries to analyze expected child rearing cost –the ECEC expenses(the child care center and kindergarten fees) - on childbirht. While costs in economics cover not only direct costs, but also indirect costs, this study only handles care and education costs, directly paid by parents. This is timly in that childbirth promotion policies in Korea focus on reducing costs. While there have been a lot of studies on the effectiveness of policies that provide financial incentive, it does not seem like there is a clear consensus among researchers. For example, Laroque and Salanie (2014) who analyzed the impact of child rearing support in France using utility maximization model of standard life-cycle concluded that € 150 a month has a significant impact on increasing birth rate and has a particularly significant effect on children who are born third or more. In the meantime, Choi Jun-wook and Song Hyeon-jae (2010) Hong Seok-cheol and Kim Jeong-ho (2012) did not find any significant impact on women in terms of childbirth. However, these studies have a limitation in that they used an intention to give birth as a dependent variable, rather than the actual childbirth. As mentioned in the introduction, for prospects parents, child rearing costs is an expected value that has not been realized yet. But since it is not easy to get information that can analyze the actual childbirth and unrealized future child rearing cost, this has not been fully examined in existing studies. The following sections discuss the way to circumvent this difficulty using statistical materials and suggest brief empirical results.

III. Empirical Analysis To analyze the impact of prospect cost on childbirth, this study used two statistics. The first statistics is Panel Study of Korean Children (PSKC) developed by the Ministry of Health and Welfare and Child Care Policy Research Institute. This is a longitudinal study of 2,150 household that gave birth to a child in a standard medical clinic throughout the country from April 2008 to July 2008. Annual survey was conducted and six times of panel data until 2012 has been accumulated. For a survey targeting parents, household features, child rearing feature, phychological features of parents, and the development of children are included. For a study targeting children, major developmental tests have been conducted. For a study targeting the teachers of childcare institutions, the environment of institutions, behavioral characteristics of children, and the features of teachers are measurd. This material is very useful in that it is a nation-wide material on the growth and developmental environments of children in Korea. But, since this is a study for already-born children, there is a limitation in using this to analyze birth rate. That is why this study used Korea Longitudinal Study of Women and Families (KLoWF)(Korea Longitudinal Study of Women and Families, KLoWF). These materials are longitudinal data of women who live in Korea aged 19 to 64 started sincce 2007 (the first study) to 2012. The size of the original same was 9,007 and up to the fourth wave, the holding rate of sample was 76%. Unlike Panel Study of Korean Children (PSKC), this study analyzed the history of childbirth of sample women which suggests that there were 198 childbirth in the first survey, 121 in the second survey, 116 in the third survey and 71 in the fourth survey. As the subjects of this study were women regardless of childbirth experience, it can be used to analze birth rate. However, this material lacks information on nurturing environment and costs, which is complemented by the data coupled with Panel Study of Korean Children (PSKC) and Korea Longitudinal Study of Women and Families (KLoWF). To combine the materials by two different panels, this study used the method of Kernel matching. To measure the non-similarity of observed values included in two data, this technique defines the distance between observed values and gives a particular-type distribution to the distance. And then it gives a weighted value in accordance with the distance between observed values; if the distance is far, an observed value will be low and if the distance is close, an observed value will be high. Korea Longitudinal Study of Women and Families (KLoWF) is the material used in this study to analyze childbirth intention, while PSKC included very detailed information on nurturing environment and costs. Weighted average of nurturing cost variable by PSKC was imputed to KLoWF. This study defined the women aged 19 to 39 are of childrearing age and used the women of these age among observed values in KLoWF were used for analysis. And to measure the distance between observed values included in two data, the age of women, educational level, marriage status, employment, work hour, number of household, familty structure (nulear family and extended family), household income, home-owning status, household assets, and household liabilities from the first wave of two data (Panel Study of Korean Children (PSKC) 2008 and Korea Longitudinal Study of Women and Families (KLoWF) 2007) were extracted as the candidate variables. To avoid the distortion in results by arbitrarily selecting matching variables, based on Akaike Information Criteria (AIC) and Bayesian Information Criteria (BIC), the author excluded the number of household members and work hours from the matching variables in the actual analysis. A dependent variable primarily used in this study is the childbirth behavior of the women in childbearing age. Women subject to this age were 6228 among samples in the first year and there were 506 cases of childbirth from 2007 (first year) to 2012 (fourth year). For analysis, logit regression model is introduced as shown below. As prospect cost of nurturing cost variable 'C' does not exist in the material of KLoWF, it was extracted from PSKC and was coupled with KLoWF based on matching method. To do that, variables indicating the frequency of using child support institutions and montly average of the use were used. One needs to be careful in interpreting results since expenses were the monthly costs for children aged 3 while intention to give birth is influenced by the entire potential expenses throughout the nurturing period. First, this cost variable does not capture the cost of raising children before the age of 3. Second, this variable captures the expenses directly paid to child support services even after 3. Third, this cost variable does not capture nurturing cost after entering into elementary school, even if we assume that the same amount of money is used to raise children before they enter into elementary school. That is why analysis presented in this study are limited to the effectiveness of expenses directly paid to child care services from 3 years old to before entering into elementary school. It is neccessary to inform that there are fundamental limitations in this study in that the impact of other costs (for example, the opportunity cost of mother) and educational cost after elementary school on childbirth is not the scope of this study.

Prospect costs have a negative impact on childbirth in a limited level, which was distinguished in women who had low education level whereas it is not distinguished in women who had higher education. Similar aspect was seen in that increased cost of nurturing children has a negative impact on childbirth, but this is only confirmed among low income class. This means that changes in nurturing cost rarely have any impact on childbirth of higher education and high income class who have relatively small amount of burden in terms of child raising cost. While there was a certain level of impact for low education and low income class, statistical significance was not strong and the scale of marginal effect (dp/dx) was very limited. This suggests that the decision on whether to give birth will not change even with 100,000 KRW a month support. There have been similar conclusions in existing studies, for example, Hong Seok-cheol and Kim Jeong-ho (2012) and You Hye-mi (2011), who used self-response data. There can be a lot of interpretations about this. One thing that needs to be mentioned first is that the financial support used in the analysis is far less than the prospect costs that potential parents expect and thus does not have a significant impact on the actual childbirth. The costs needed to raise a child considered by potential parents are not just expenses needed when their child is infant or baby, but the entire expenses needed util they grow up to be adults. And the costs needed in this period should not just include direct cost paid to child care services, but inlcude opportunity cost in labor market, psychological and potential costs. In this regard, the proportion of costs paid to child care services is rapidly reduced compared to the entire cost of raising children and this is why governmental support for the former expenses has a little impact on promoting birth rate.

IV. Summary and Policy suggestions This study used materials coupled with Panel Study of Korean Children (PSKC) and Korea Longitudinal Study of Women and Families (KLoWF) to analyze the relationship between the actural childbirth and prospect nurturing costs rather directly. The method of kernel matching was used to analyze the information of childbirth history coupled with nurturing environment and cost included in Panel Study of Korean Children (PSKC). The results in this study suggested that there is a very weak relationship between prospect cost and the actural childbirth, which was more distinct in low-education and low-income class. But in this class, the relationship is still weak. This can be a simultaneous impact from various factors on childbirth intention as well as technical limitations adopted in this study. Furthermore, as nurturing cost used in this study only captures direct costs paid to child support institutions when children were 3 to 7 years old (before they enter elementary school), it does not capture the entire costs expected throughout the whole life-cycle. In other words, while the results of this study suggest that the supports for child rearing costs and education cost do not have a significant impact on childbirth, this does not mean that this study denies the effectiveness of various policies to promote childbirth given that child rearing and education costs account for a very little. And the focus of this study lies in the fact that this study tried an exploratory attempt to combine Panel Study of Korean Children (PSKC) and Korea Longitudinal Study of Women and Families (KLoWF), using the method of kernel matching. The author hopes that there will be more adundant stdies using the same or similar statistical methods.