1.AfS q -oo0 POLICY RESEARCH WORKING PAPER 2400 Child Care and Women s In Romania both the maternal decision to take a job and the Labor Force Participation decision to use out-of-home in Rom ania care are sensitive to the price of child care as well as to the potenrial market wage of the Monica Fong ro(i[her. A decrease in the Michael Lokshin price of child care can increase the number of mothers in the labor force and thus reduce poverty in some households. The World Bank Development Research Group Poverty and Human Resources July 2000 | POLICY RESEARCH WORKING PAPER 2400 Summary findings Fong and Lokshin model the household demand for market care and the decision to engage in paid child care, the mother's participation in the labor force, employment. and her working hours in Romania. Their model * The level of household nonwage income has little estimates the effects of the price of child care, the effect on maternal employment and the demand for child mother's wage, and household income on household care. behavior relating to child care and mothers working In addition to facilitating women's work, outside the home. They find that: kindergartens and creches appear to provide educational * Both the maternal decision to take a job and the and social benefits for children. Close to half the children decision to use out-of-home care are sensitive to the in these facilities have mothers who do not work. price of child care. A decrease in the price of child care Further research is needed to assess the cost and nature can increase the number of mothers who work and thus of these benefits and to determine the appropriate roles reduce poverty in some households. for the private and public sectors in providing, financing, * The potential market wage of the mother has a and regulating such services for working and nonworking significant positive effect on the decision to purchase mothers. This paper-a product of Poverty and Human Resources, Development Research Group-is part of a larger effort in the group to understand the role of gender in the context of the household, institutions, and society. Copies of the paper are available free from the World Bank, 1818 H Street NW, Washington, DC 20433. Please contact Patricia Sader, room MC3- 632, telephone 202-473-3902, fax 202-522-1153, email address psader@worldbank.org. Policy Research Working Papers are also posted on the Web at www.worldbank.org/research/workingpapers. Michael Lokshin may be contacted at mlokshin@worldbank.org. July 2000. (30 pages) The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is toget thefindings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the view of the World Bank, its Executive Directors, or the countries they represent. Produced by the Policy Research Dissemination Center CHILD CARE AND WOMEN'S LABOR FORCE PARTICIPATION IN ROMANIA by Monica Fong and Michael Lokshin' The World Bank Keywords: Romania, Eastern Europe, Labor supply, Child care, Maximum likelihood, Non-parametric estimations Address for correspondence: Michael Lokshin, Development Research Group, World Bank, 1818 H Street NW, Washington DC, 20433, USA. 1. Introduction The research in this paper addresses the problems faced by families with children in a transitional economy and the impact that reforms in the child care system can have on the well-being of children and the employment of women in Romania. Early childhood development programs have been shown to have substantial physical health, nutrition, and family benefits as well as major cognitive and social benefits for children. A large body of research from diverse cultures provides strong evidence that most early childhood development programs of relatively good quality have meaningful short-term effects on cognitive ability, early school achievement, and social adjustment (Reynolds et al. 1997). Ensuring healthy child development is not only an investment in a country's future workforce and economic capacity, but it is also an important mechanism for making use of the human capital vested in women of working age. In addition, the provision of child care services facilitates women's employment and can reduce economic disadvantages for women and their children by increasing their income. A better understanding of child care choices and their effects on women and children in different socioeconomic groups is therefore needed to inform policy discussions on female employment and child care in the context of the economic transition. In the pre-transitional period women in Romania, as in other Communist countries, participated in the labor force on a level higher than in industrialized market economies. Romanian women also worked full-time the whole year round, and there was very little part-time employment. Despite a decline of 12 percent since 1990, the rate of women's labor force participation remains relatively high in Romania, and averaged over 75 percent for women of ages 25-49 in 1997 ( Cace et al.) Such a high level of involvement of women in the economy would not have been possible without the existence of a wide range of government-subsidized child care programs, in the form of creches, kindergartens, and after-school programs. Under central planning government resources were devoted to providing appropriate care for children in center-based programs from the earliest months of life until they entered primary school. 1 The fundamental restructuring ofthe economic system toward a market economy that started in 1989 led to dramatic changes in Romania's socioeconomic environment and strained the existing system of social protection and state-subsidized institutions. A more than 26 percent drop in gross national product (GNP) from 1989 to 1994 (Milanovich 1997) resulted in a widening budget deficit, shrinking government programs, and an erosion of family subsidies (child benefits and child care allowances). Between 1989 and 1997, child benefits were reduced to 65 percent of their pre- transition levels (Zamfir 1998). Most countries in the region have seen a decline in preschool child-care services during the transition. In Romania public child care services declined in quality and parents were asked to carry a larger share of the cost already in the last decade of the socialist regime. Responsibility for creches for children ages 0-2 has been removed from enterprises and transferred to the Ministry of Health, while kindergartens for ages 3-6 remain the responsibility of the Ministry of Education. The number of creches declined from 847 in 1989 to 573 in 1995, and the enrollment of children (2 months to 3 years of age) in creches dropped by half. At the same time, new legislation (law 120/97) provided 65 percent of her previous salary to a mother who wished to care for her child during its first year of life. This was recently extended to age two, and many mothers working in the public sector prefer to stay home on child-care leave when children are young. The number of children (3 -6 years old) in kindergartens has declined by nearly one third since 1989 (Zamfir 1998). According to UNICEF (1997) the proportion of children in kindergartens dropped from 63.3 percent in 1989 to 55.1 percent of relevant population in 1996. Although attendance has declined, the growth of private institutions has afforded greater choice in kindergartens. Kindergarten fees doubled in relation to the average wage in 1990 but have since declined roughly to the pretransition level of 10 percent (Fong 1996). Fees are subsidized, and users currently pay about 51 percent of the total cost of keeping their children in child care facilities. (Zamfir, 1999) For some households, particularly single-parent households, the unemployed, and those subsisting on pensions, the fees for child care may nevertheless present a serious problem. Until recently, there has been little research on the economics of child care outside of the United States. In recent years a certain amount of research on child care has been conducted in the nations of Western Europe, where (as in the United States) growing numbers of women with young children have been entering the workforce (see, for example, Gustafsson and Stafford, 1992; 2 Cleveland, Gunderson and Hyatt 1996; Van Den Brink and Groot i 997). To date only a very limited amount of research has been conducted on child care and women's labor market activity in transitional economies. Lokshin (1999) has investigated the effects of the price of child care, the mother's wage, and household income on household behavior and well-being in Russia. The research reported in this paper is prompted by the complexity of the problems faced by families with children in the transition and by the significant impact that reforms in the child care system can have on the political and economic environment, and on the well-being of future generations in Romania. Where child care represents an important mechanism to overcome social inequality experienced by women and children, a better understanding of child care choices and their effects on women and children in different socioeconomic groups can serve to inform policy discussions about the contexts in which economic changes enhance or compromise child development. Based on recent progress in the theory of demand for child care and women's labor supply in the United States, this paper follows the work of Blau and Robins (1988), Ribar (1992, 1995), Connelly (1992), Michalopoulos, Robins, and Garfinkel (1992), Kimmel (1995), Averett, Peters, and Waldman (1997) in jointly modeling households' decisions about child care and mothers' decisions about entering the workforce. The research is based on household survey data from the Romanian Child Care and Employment Survey (RCCES) and data on child care providers from the Romania Child Care Facilities Survey (RCCFS). These two surveys were conducted by the World Bank in the same geographical areas during the same time period and were designed to allow matching data on child care fees and child care quality collected from kindergartens and with the households surveyed in the same communities. The paper is organized as follows: Sections 2 and 3 describe the data and present descriptive statistics on the main factors that influence household behavior. Subsequent sections show the development of the theoretical model, give details of the empirical model, discuss the conceptual issues involved in estimating a consistent model of household child care choice and labor supply. This is followed by a presentation of the results, and an analysis of the estimation results and simulations in Section 6. A discussion of the policy implications and a summary of the findings conclude the paper. 3 2. Data and Variables This research is based on data from the Romania Child Care and Employment Survey (RCCES) conducted by the World Bank in April 1999. The nationally representative sample contains information on 1,505 households with children younger than 12 years of age, and 403 households with children of age 0-6 in kindergartens and/or creches. The initial sample of households for the survey was identified from a stratified three-stage, multistrata sample of residential addresses. In the first stage, the total sample was proportionally divided into several subsamples, using two strata based on the region ( 1 regions) and the size of the locality (rural, small town with population under 50,000, medium-size town with population 50,000-200,000, large town with population over 200,000; and Bucharest). In the second stage, the geographical area units were randomly selected within each subsample. In the third stage, households in the area units were selected using a random route (maximum of 10 households per sampling point). The data set includes information on the individual members of these households, household-specific information, and community-level data. It also contains information on the child care arrangements made for each child in the household, the amount of time spent by each child in formal and informal child care, and the amount of money paid for formal child care during the week ofthe survey. The part ofthe questionnaire administered to each member ofthe household yields data on how much time each household member spent looking after children and was active in the labor market, as well as information on their monthly wages. The second part of the questionnaire, administered to one respondent per household on matters that affected the household collectively, yields information on non-wage household income and on household composition. Parallel to the household survey, a Romanian Survey of Child Care Facilities (RCCFS) collected information about 316 kindergartens and creches that were attended by the children from the household sample interviewed by RCCES. The questionnaire, which was administered to every facility manager, provides detailed information on fees and types of care. With these surveys it is possible to match households with the child care prices they face in the area in which they live and purchase care. 4 3. Descriptive Statistics Child care facilities The system of preschool care in Romania is regulated by Government Decision (Romanian Government 1991). According to this regulation kindergartens and nurseries should provide, free of charge, children's medical assistance, instruction, and education, with the required expenses covered by the budget of the state central administration. Parents contribute up to 75 percent ofthe cost of food for children in child care. Their monthly contribution is set as a percentage of total monthly household income, adjusted by the number of children the household has enrolled in the child care facility. Fees are calculated according to three income brackets. Households with total gross income below 12,000 leis (in 1991 prices) pay 50 percent of the total fee for the first and 40 percent for the second child in the facility. Parents with income between 12,000 and 20,000 leis pay 75 and 55 percent of the fee for the first and the second child. Finally, households with total income above 20,000 leis pay the full fee for the first child and 70 percent of fee for the second. The average fee per-child in the daily program was 173,300 leis per month for the households surveyed. In some facilities parents are also required to pay for the development and academic activities provided. In our sample, 54 percent of child care facilities managers reported that they charge for such activities and that the amount ofthis additional fee is about 20 percent ofthe base fee. Since 1989 there has been a decline in the use of state-provided child care facilities, which has led to underutilization of kindergartens and creches. Table 1 shows the occupancy rates i.e. the ratio ofthe facility capacity to the actual number of children enrolled in the program. These occupancy rates are lower than the numbers from other sources. For example, Evans et al. (1995) reports an occupancy rate of 94 percent at the end of 1993. This discrepancy may indicate that enrollment in child care facilities in Romania is falling faster than the decrease in the number of kindergartens or creches. 5 Household child care arrangements Placing children in child care in Romania is common practice, whether the mother is working or not. About half of the children in the sample were in formal child care, half at home (Table 2). For formal child care, however, more than half of the child care places went to children whose mothers were not working. In other words, formal child care is used as both to allow the mother to work outside the home, and as a benefit in its own right for the child. For women who work, however, child care remains an important factor in allowing them to work, and 62.9 percent of working mothers put their children in formal child care, while 37 percent still keep the child at home. The dual function of child care to allow mothers to work and to benefit pre-school children is also evident if we look at the role of other members of the household (Table 3). The presence of other relatives in the household influences the choice of child care. Of the 81.0 percent of working mothers with other household members who were at work, over 50 percent put the children in child care and the remaining 31 percent kept them at home. When other household members do not work, child care is still used as a benefit for the child, and twice as many of these households put children in child care as keep them at home. This dual function becomes especially clear for households with many children. For mothers who do not work, those with one child are more likely to keep the child at home than put it in child care. Those with two or more children, however, are slightly more likely to put them in formal care than to keep them at home. (Table 4). Mother 's labor force participation The level of mothers' labor force participation declines with household size. For example, in the households of two (single mother family) or three (single mother with two children or two adults with one child), about 42 percent of mothers work. This number declines to 32 percent for the households with five family members and falls even further for larger households (Table 5). A possible explanation for this pattern is that the mother's decision to participate in the labor market is determined by the total household income. Large households have higher total income, and thus the potential contribution from the mother's wage income to total household income is relatively 6 lower than that contribution in smaller households. Large households more often can find it optimal for the mother to stay home with children and thus to provide better-quality care and perhaps other domestic services as well than households with fewer members. The decision for mothers to work also depends on the number of children in the household. Among households with more children, the advantage of the mother's staying home is higher than for households with only one child. Table 6 presents the proportion of households with working mothers by the number of children. Mothers' productivity at home increases with the number of children in the household relative to their productivity in the labor market because of economies of scale that can exist in home-provided care. For households with only one child, more than 40 percent of mothers work. For households with 2 and 3 children, the proportion of mothers in the labor force drops to about 3 0 percent, and this proportion declines to 20 percent for households with four or more children. A special section of the RCCES questionnaire was devoted to questions about mothers' own perceptions of child care and its effects on maternal employment. Among mothers of small children in Romania, 65.5 percent say the price of child care influences their decision to work or not to work. Only a little more than one-third of the mothers consider the lack of kindergartens and creches an obstacle to employment, however. When asked about the major factors that influence the decision to put their children in kindergarten, mothers report that the quality of care provided and the convenience of the facility's operating hours are the most important. At the same time, 20 percent of Romanian households with small children would not use formal child care arrangements even if they were free. 4. Theoretical Model The analysis applies to households with children under 7 years of age. Three forms of child care are available to households in Romania: * informal (home provided and free) care provided by the mother * informal child care provided by other household members * formal (government or market provided) child care For households with children and two parents, the husband is considered a potential provider of free child care. In a household with a single mother who has no relatives living with her, it is assumed that 7 any informal child care is provided by children themselves or relatives who live outside the household. We classify child care arrangements into six categories defined by combinations of the mother's employment status, mode of care (formal and informal), and employment status of other household members (Table 2). The theoretical model used in this paper is based on the assumption that household members make choices about their consumption of child care quality, of market goods, and of leisure. A household's decisions about the quality of child care it wishes to obtain and about the amount of time each member of the household can work are motivated by the desire to achieve the highest level of household welfare. We also assume that households pay a flat fee for child care services. The fees charged by kindergartens are a function of the quality of care provided by the facility and the total level of child care prices within the locality. This important assumption needs further explanation. As we noted, how much Romanian households are charged for child care depends on total household income. Parents of children attending the same kindergarten can pay different fees. Thus, the prices of child care are endogenous to household behavior and in particular to the household members' labor supply decisions. Certain households may decide not to send their mothers to the labor market because this would increase their income and put such households into a higher child care price bracket. The endogeneity of child care prices complicates the model substantially. There are two ways to deal with the problem. We could estimate the structural model of household behavior with a kinked budget constraint under an assumption of endogeneity of prices of child care, or we could ignore the fact that households with different incomes pay different prices for child care and assume that every household whose children attend the same kindergarten pays the same flat fee. Estimating the correct structural model under assumptions of a kinked budget constraint is complicated. Even in simple cases, using the instrumental variables for identification may be criticized on the basis of invalid exclusion restrictions. At the same time, ignoring the differences in fees paid by households would introduce measurement errors in our estimation. Most of the recent literature on child care treats the price of child care as exogenous. The assumption of exogenous child care prices may underestimate the effect of child care prices on poor 8 Romanian households. Yet despite the disadvantages of the simpler method, we still choose it over the more complicated estimation of the structural model, as the ability of such a model to provide any additional insights has yet to be established. In this paper we identify the effect of changes in the price of care through the regional differences in child care fee levels. In the one-period utility maximization problem the household chooses its consumption of a Hicksian composite good G, the average per-hour quality of child care Q, the leisure time of the mother L,W and the leisure time of other household members Lo subject to its budget and time constraints. The household utility function is assumed to be twice-continuously differentiable and quasi-concave: Max U = U(Lm, Lo, G, Q). (1.1) The total quality of child care Q is the weighted sum of the exogenous quality of the child care provided by the mother Qm, the quality of child care purchased on the market Q, and the exogenous quality of child care provided by relatives Qo: QQm=mL+Qp(Hm-To)+QoTo. (1.2) The budget constraint includes total household expenditures on child care as a function ofthe number of children in the household, of the per-unit quality price of child care, of the quality of formal care, and of the time spent by the children in care: G = E + Wm Hm + WoHo-N Pq Qp (Hm-ToO), (13) where E is the exogenous nonwage household income, Hm is the mother's actual work time, Ho is the other household members' actual work time, N is the number of children in the household, Pq is the exogenous price per unit of quality of formal child care, To is the amount of time spent by other household members on child care, Wm is the market wage available to the mother, and Wo indicates the market wage available to the other household members. Finally, the model specifies - under the assumption that children require constant care - the time constraints affecting the mother, the other household members, and the children: Lm Hm =Lo +Ho +To =1 (1.4) Hm-To 20 (1.5) O < To,LHo0,Lm,H < 1 (1.6) 9 The household optimizes the labor supply of its members, chooses the optimal quality of child care for each of the six possible child care/employment mode states, and then chooses the state with the highest utility. 5. Empirical Model The empirical model used in this paper consists of a discrete choice equation for the child care mode and mother's labor supply, an equation for a mother's hours at work, and an equation for children's hours in paid care. For the discrete choice model, the utility that the ith household derives from the choice of the jth discrete alternative can be expressed in linear form as: Q j = Qy + ey = Xjj6j + Z,1yrj + ,i' j = 0,...,5 (2.1) where Q,j is the utility for household i choosing statej, X, is the vector ofthe household characteristics that affect the choice of the ith household and that do not vary by state, Zij is the vector of outcome-specific variables, pi and y are vectors of unknown parameters, and E is a random disturbance that reflects, among other things, unobservable attributes of the alternatives. The probability that household i chooses statej is then: Pr(j) = Pr[Q,2 > Q4i for any j q] = Pr[scj, -Esqi > Xj(I3q i -ji) + Z1j(Yqi - Yji) for any j q] The supply function for mother's hours at work and the demand function for hours spent by children in formal care can be specified in linear form as: ri ki hours mother works Hi = akXi +6pkZ1+ 1k, where k = (2.2) 2 hours children spend in paid care Here, H, is the continuous dependent variable k associated with household i in statej. In the first continuous outcome equation, Hi' is the number of hours that a mother supplies to the labor market (if she works), and Hi2 is the number of hours spent by children in formal child care facilities in those states where formal child care arrangements have been chosen. Xi and Z, are the vectors of the variables defined above, (Pk and ak are vectors of unknown parameters, and (,k is an error term with mean zero. 10 The theoretical model assumes that a household makes simultaneous decisions about the mode of child care it wishes to use, the labor supply of each of its members, the amount of time that each family member spends on child care, and the amount oftime that their children spend in formal care. All ofthese decisions are determined by exogenous characteristics ofthe family and individual family members, both observable and unobservable. Several estimation issues need to be discussed. First, the error terms in the discrete (e) and continuous (4) equations may be correlated across states and among each other. The correlation across states is a correlation among disturbances in the state-specific indirect utility functions (e.g., 2.1). If, for instance, a mother's participation in the labor force is determined by, among the other factors, some unobservable taste for work, this unobserved factor will be part of et for states in which the mother is employed (j = 2,...,5). There may be a correlation between the disturbance in the equation for a household's choice of discrete states and the equation for the amount oftime that women supply on the labor market. Similar correlations can exist for the equations that determine the following: (i) labor supply decisions of the other family members, (ii) time spent on child care by the mother and the other family members, and (iii) household child care arrangements. To account for possible error correlations in a tractable way, we impose a factor structure on the disturbances in equations (2.1) and (2.2): Cy= :-+-pV; 4 ,I = Xi + TVi; 4 2 =i +,;Vi (3.1) where i, is an independent extreme value error, and Xi, and yi are independent normal random variables. V is a factor. This factor is unobservable variables that influence the choices made by households and that is uncorrelated with the explanatory variables. p, , and s are factor loadings that represent the effect of the factor V in each equation. The system of equations (2.1-2.2) with the error structure (3.1) can be estimated by the Semi-Parametric Full Infornation Maximum Likelihood (SPF1ML) method developed by Laird (1978) and Heckman and Singer (1984) and applied to simultaneous equations by Mroz and Guilkey (1992) and Mroz (1999). The above specification assumes that each household has the same choice set of child care arrangements and labor force participation modes. However, a significant proportion of Romanian families do not consider formal care a possible care alternative for their children. The RCCES asks whether households would consider putting their children in formal care if the care were free and of 11 better quality. One-fifth of the households with small children answered that they would not send the children to formal child care facilities. For these groups of families, the conditional contribution of the discrete outcome equation to the likelihood function is calculated based on a restricted set of possible forms of care, i.e., that there are no formal care arrangements in the choice set of these households (equations 5.1 and 5.2): Households that would use all forms of care Households that would not use formal care ePPjXit + PjIl I + Pjj2V2MA eP IX3 + PklV) + PkI2V2h Pr(Y,1 = A VI mV2kt)= Pr(Y,, = Al VI. IV2kt P 3 + E ePkXi+PkivIk+P 2v2mnt 1 + ePkX + PkIVI +Ph2v2kV k=1 k=1 Pr(Y}, = 0VI.m,V2kt)A 1 Pr(MY,, = OIVimV2kt)NA 1 1 + E ePkXi PklVlk + Pt2V2w,,- 1 + e3kXi+Pk1Vl, + Pkt2V2k k=1 k=1 (5.1) (5.2) where Pr(Yi, =jlvI,v2k)A is the probability that household i (which has access to formal child care facilities) chooses state j at time t conditional on factors VI and V2, and Pr(Yj, =jjvlm,v2k)NA is the probability for those households that prefer not to use formal care arrangements. Conditional on the Vs, the likelihood contributions ofthe continuous outcome equations (2.2) are: 1 H' ~x.i a1-zl y I r,vm I i2 _x2a2 _z2y2 V Pr(H,l lvm.) = 1- p( 2 al (6.2) where Hi,' and H,,2 are the dependent variables in the continuous outcome equations (2.2), is the probability density function of standard normal distribution, and a' j, and a2i, are the square roots ofthe variances of the error terms in equations (2.2). Thus, the semi-parametric log-likelihood function for the system of equations (2.1-2.2) with the error structure (3.1-3.3) is: NM 3 = ln( ZJ Pr(1 = j(vm) Pr(Y = jl v.)' Pr(H,'I vm)Pr(Hj2 vm)]), (7) i=l I where N is the number of households in the sample. 12 Choosing a priori numbers of points of supportM, the log-likelihood function 9 is maximized over a's, f's, y's, l's, C's, Ps, and v's. For identification purposes, the two points of support are normalized to equal 0 and 1, respectively. The number of points of support is increased until the difference in the log-likelihoods of consequent maximizations satisfies the convergence criteria. The joint distribution of the error terms (3.1-3.3) is unknown, so the sample statistics of the estimates cannot be derived analytically. It is feasible to estimate the covariance matrix 8 of the coefficients in the model (2.1-2.2) by inverting the Hessian matrix of the second derivatives of the log-likelihood function Z. Dependent variables The dependent variable for the discrete outcome equation is defined according to the possible combinations of amother's employment status, employment status of other household members, and the mode of child care, which are shown in Table 2. These combinations are: (0) - the mother does not work and stays at home with her children; (1)- the mother does not work and children are in kindergarten; (2) - the mother works, the other household members also work, informal child care arrangements are used; (3) - the mother works, the other household members do not work, informal child care arrangements are used; (4) - the mother works, the other household members work, formal child care is used; (5)- the mother works, the other household members do not work, both formal and informal care arrangements are employed. Table 7 shows the distribution ofthe dependent variables for continuous outcome equations, i.e., the time that mothers spent working and the time that children spent in formal care. Both continuous outcomes are observed only among the sample of working mothers or among the sample of children in formal care. 13 Explanatory variables The definitions and descriptive statistics for the explanatory variables in the system of equations (2.1-2.2) are presented in Table 7. Several key variables of interest are discussed in detail below. Price per quality unit ofchildcare fP,).' We estimate the quality-adjusted price of outside-home care using the method suggested by Blau and Hagy (1998). RCCFS collected extensive information about the characteristics of the child care provided and the fees charged by creches and kindergartens. These data are used to estimate a model of fees for formal child care facilities. The quality-adjusted price of an hour of child care is determined by a location-specific hedonic price equation: Pi = ,i + 6 Xi + ei where Pi is the price of the formal care at location i, xi is a vector of variables that represents the characteristics of the facility, , is a vector of coefficients, and s, is an error term. In that specification it can be interpreted as a market-specific quality adjusted hourly price of child care. We use the quality-adjusted price of care to be able to compare the effects of price on household behavior for facilities that offer different quality of care. For example, suppose one facility offers several developmental programs and has a low teacher-child ratio. In other words, that facility provides a high quality of child care but charges a relatively high price for its services. The other institution does not offer such high quality of care, and the price it charges is low. Directly comparing the prices of these two facilities is not possible because these prices are charged for different services. The methodology suggested above allows us to adjust prices for difference in the quality of provided care and thus makes these quality-adjusted prices comparable. Mother's offered wage W The wage rates available to each mother have been imputed using Mincer's and Polachek (1974) type earning function regression with a control for selectivity (standard Heckman correction)2 run on a subsample of working women for whom hourly wage data were available. The hourly wage has been calculated as a ratio of the women's monthly earnings and the total number of hours they worked during the month the survey was administered. In the absence of 2 Regression coefficients for the wage equations are shown in Appendix 2. For identification in the selection equation, we use the standard set of household characteristics that can influence the mother's labor force participation decision but are uncorrelated with the potential wage rate. 14 data on the total amount oftime a mother had worked during the preceding month, the imputations were made based on the number of hours worked during the week of the survey. In the wage regression, the following explanatory variables have been used to predict mothers' experience, her marital status, and the amount of time she had been in her current mainjob. Imputations are made based on the women's predicted hourly wages, with the job tenure of nonworking mothers being equal to zero. Here the offered wage is assumed to be a wage that a mother could earn if she were to start a new job. Offered wages of other household members (JffLThe wage rates available to other household members are calculated in a similar way to the wage rates available to mothers. Different regressions were run to predict wages for household members of different ages and genders. After the imputations, two methods were used to obtain the wage W0. Under the first specification the offered wage of other household members is equal to the lowest wage earned by any household member except the mother. The second specification uses the average wage of all working household members as an explanatory variable in the model. Nonwage household income (E): Nonwage income is measured as household monthly income from all sources other than wage income. This may include social security transfers, private transfers, in-kind income, and income from home production. The structure of household income changed over the rounds ofthe survey, and certain adjustments were made to ensure compatibility ofthe income data across all of the survey rounds. Other explanatory variables include some individual characteristics ofthe mother such as her age and level of education, household demographics and size, the number of children in the household and their ages, the number of pensioners in the household, and the household's geographical characteristics. 6. Results The results ofthe estimation ofthe system of simultaneous equations (2.1-2.2) are shown in Table 8(a) for the discrete outcome equation and Table 8(b) for the continuous outcome equations. 15 Estimated coefficients The estimated coefficients ofthe household child care/labor supply equation confirm the predictions of the theoretical model. An increase in the price of the unit quality of child care decreases the probability for the household to chose paid care (modes 1,4, 5) and increases the probability of choosing the states where children stay home. The change in the child care prices has the strongest and statistically most significant effect on the state where the mother does not work and children are in formal care (mode 1). An increase in potential market wages of the mother increases the likelihood that the mother will participate in the labor market. The coefficient on the log of the mother's wage is positive for all states where the mother works, and it is statistically significant for modes 3, 4, and 5. Younger mothers are more likely to work and children of younger mothers are more likely to be in formal care. Relative to mothers with only primary education, mothers with higher levels of education are less likely to work and to have their children in kindergartens or creches. Households may prefer that the mother stays home with children as higher-educated mothers can provide a higher quality of care for their children. Households in the rural areas of Romania are less likely to use formal child care, and the probability of using kindergartens and creches declines when the distance from the facility is greater than a 20-minute walk. The results ofthe estimation ofthe continuous outcome equations ofthe mother's hours at work and the hours children spent in the child care facilities are presented in Table 8(b). The coefficients of the hours of work equation show apositive, although insignificant, effect ofthe mother's wage rate on the number of hours the mother spent working. An increase in the wage rate of other household members decreases the mothers' hours at work, but mothers from the larger households work more. Mothers with higher work tenure also tend to work longer hours. Significant differences in the hours of work are observed among the different occupations, where women employed in industry and in education work longer hours than other women. The estimation ofthe model ofthe hours that children spent in kindergarten is consistent with the theoretical prediction results although the explanatory power of the model is weak: higher wages ofthe mother increases the number of hours children spend in care and an increase in the price of care 16 has a negative effect on the hours. As the assumption of exogenous prices for child care tends to underestimate this effect, the latter may be regarded as a minimum estimate, however. Simulations To examine the effects of policy instruments on household behavior, we simulate how households would respond to changes in the specific parameters used in the model. In a given simulation, a certain value of the variable of interest is assigned to all the households in the sample. The simulated probabilities for the discrete model outcomes and simulations for the continuous models are generated for each household at every point in time by integrating over the estimated heterogeneity distribution and averaging the probabilities across the sample. Next, the value of the variable of interest is changed, and this changed value is assigned to the whole sample of the households. Then the new set of simulated probabilities is generated. The effect of the changes in the particular parameter is calculated as a difference in these simulated probabilities. The simulated distributions of the probabilities for the discrete outcomes are shown in Table 9. A 10 percent increase in the mother's potential wage increases the rate of maternal employment. For both states where the mother does not work (states 0 and 1, i.e. children at home or in child care), we observe a decline The decrease is stronger in the state where the mother does not work and stays home with children (state 0). The number of households in this category drops by 7.2 percent as a result of an increase in mothers' wage. Under this new policy households would also be more likely to choose states that employ formal care arrangements. The proportion of households with working mothers that would use formal care increases by 8 percent. The total effect of the change in the wage rate is a 10.9 percent increase in the rate of women's labor force participation together with a 4.3 percent increase in the use of formal child care facilities. Changes in the price of child care would have a smaller effect on the level of maternal employment and on the use of formal care. The simulated increase in the price of care by 10 percent would result in a 1.2 percent decline in the number of working mothers and a 2 percent decrease in the number of households that use creches or kindergartens. Thus, a policy that fully subsidized formally provided child care could increase the rate of women's labor force participation by as much as 12-15 percent. 17 A 10 percent increase in the potential market wage of household members other than the mother would not have a significant effect on the household choice of child care mode or the mother's labor supply. As already discussed, these should be regarded as minimal estimates due to the exogeneity assumption. We also fail to discover any significant effect of the changes in household nonwage income on household behavior. This result seems to be consistent with the findings in other transitional economies (Lokshin 1999). The Romanian results confirm those found in other countries. Table 10 presents estimated elasticities of female labor force participation and use of formal care for studies conducted in the United States, Brazil, Russia, and Romania. Among all these studies, Romanian results show the lowest response of women's labor force participation to a change in the price of child care. However, the elasticity of the use of formal care is found to be highest in Romania, indicating that child care prices play a strong role in households' decisions about the use of formal care. This comparatively high elasticity ofthe household demand for child care with respect to child care costs may be attributed to the fact that a significant proportion of the households where mother is not working also use child care facilities. For households with working mothers child care costs may be seen as a fixed cost of maternal employment, i.e., for every hour the mother works she has to pay for the cost of care. In these cases household decide to use child care facilities as a substitute for the maternal care when the mother is at work. Any changes in the modes of care (for example a switch from kindergarten to home care) in response an increase in child care prices may lead to substantial changes in the labor supply ofthe household members and even to mother's exit from the labor market. In households where the mother does not work but the child is in formal child care, on the other hand, the decision not to use child care facilities would not have such a dramatic effect. Switching from paid care to home care may affect developmental outcomes for the children and mother's leisure time, but the impact on the household income is likely to be insignificant. The low elasticity of female labor supply with respect to child care costs may be explained by the fact the child care is subsidized for the poorest household in Romania. Child care costs can be thought of as a fixed cost of maternal employment. For every hour the mother spends in the labor market she has to pay a cost of children care, which decreases her effective wage. In the better-off households the wages earned by the mother may be significantly higher than her reservation wage, when she would leave the labor market. Thus change in the cost of care would not decrease the 18 mother's effective wage below her reservation wage in better-off households, but in poorer households, and households where the mother's education and labor market experience is low, the situation is different. The potential market wages of the mother can earn on the market are not much higher than her reservation wages. An increase in the cost of care could therefore lead the mother to leave the labor market and stay home with the child. At present, however, the Romanian government subsidizes child care for and such policies buffer the impact of the changes in care prices on the poorest households. If the subsidy were removed maternal employment may be expected to decline, especially for young women and women with low education, little labor market experience. Distributional impact of changes in price of child care and changes in mothers' wage rate In the previous sections we showed the effects of various policy instruments on household behavior. An important issue in implementing a particular policy measure is the distributional impact of the policy. Changes in the price of child care or the mothers' wage rate may improve the well-being of households with small children, but these policies can affect households with different levels of income differently. Figure 1 presents a non-parametric estimation3 of the effect of a 10 percent decrease in the price of child care on the mother's probability of being employed by per capita household income. The figure indicates that the policies that decrease the costs of care are most effective for households with above-average incomes. The elasticity ofthe change inthe wage rate is about 0.11 for the poorest Romanian households and reaches 0.19 for households with per capita income around 900,000 lei per month (a 60 percent difference). The weak responsiveness of low-income households to the change in child care price could be explained in part by the above-mentioned differences in fees for households with different incomes. Poor households pay only 50 percent of child care facility fees, and the effect of changes in fees would accordingly be smaller for them. A somewhat different picture is shown in Figure 2, which presents a non-parametric estimation of the distributional impact of a 10 percent increase in the maternal wage on the mother's probability of being employed. This policy is more effective for families with lower incomes. The effect of an 3 For the graphs we used the program for locally-weighted smoothed scatter plots. 19 increase in mothers' wages is almost twice as high for the poorest households than for households from the right side of income distribution. As a policy instrument that seeks to increase the maternal labor force participation of low-income families, an increase in wage rates appears preferable to a decrease in the price of child care. 7. Conclusions This study estimates the effects of child care costs on the labor decisions of women in Romania. The estimation ofthejoint model ofhouseholds' child care choices, mothers' labor supply decisions, and household demand for formal child care confirms the predictions of the theoretical model developed in this paper. The estimation indicates that economic incentives have a powerful effect on the work behavior of women with children in Romania. The level of wages available to them and the costs of child care can both be expected to affect women's labor force participation and labor supply decisions. Child care costs affect which child care arrangement households choose. When the costs of formal care are high, this discourages households from using formal child care and increases the number of households that rely on informal care. These findings match the subjective perceptions of Romanian women that high child care costs and low availability are barriers to women's employment. Government subsidies for child care may increase the number of mothers who work, thus increasing the incomes of poor households and lifting some families out of poverty. The simulations in this paper show that measures such as subsidies aimed at reducing the costs of market child care are effective in increasing the number of mothers who work and the number of hours that they work. However, the effect of such policies on the poorest households in Romania is less significant. In Romania formal child care plays the important additional role of providing pre-school education for children, regardless of the mother's labor market participation. While non-working mothers are more likely to keep their children in home care than working mothers, close to half ofthe children attending formal care have mothers who do not work. Further analysis is needed to determine whether this is a factor of the age of the child, with very young children being kept at home, and older children sent to formal care for the social and educational benefits provided by formal child care. The nature and extent of these benefits will need to be assessed in order to determine how kindergartens and creches can best prepare and equip Romanian children for their formal education. Further research 20 is also needed to determine the roles of the public and the private sector in the provision, finance and regulation of such services for both working and non-working mothers. 21 References Averett., S., Peters, H., Waldman, D., (1997) "Tax Credits, Labor Supply, and Child Care." The Review of Economics and Statistics, 79, No. 1. Blau, D., Robins, L., (1988) "Child care cost and family labor supply." Review of Economics and Statistics 70(3): 374-81. . (1989) "Fertility, employment and child care cost." Demo2raphv 26(2): 287-99 . (1991) "Child care demand and labor supply of young mothers over time." Demo2raphv 28(3): 333- 351 Blau, D. (1989) "The labor market effects of child care subsidies." Unpublished manuscript. University of North Carolina at Chapel Hill, April. . (1990) "The economics of child care" New York: Russel Sage Foundation . (1993) "The supply of child care labor." Journal of Labor Economics, 1993, vol 11, no.2 Blau, D., Hagy, A., (1998) "The demand for quality in child care" Journal of Political Economy. 106(1): 104- 139. Cleveland, G., Gunderson, M., Hyatt, D., (1996) "Child care costs and the employment decision of women: Canadian evidence." Canadian Journal of Economics, XXIX, No. 1 Connelly, R., (1992) "The effect of Child care costs on married women's labor force participation." Review of economics and statistics 74(1), 83-90 Connelly, R., DeGraff, D., Levinson, D., (1996) "Women's employment and child care in Brasil" Economic development and cultural change 44: 619-56 Evans, J., with Karwowska-Struczyk, M., Korintus, M., Herseni, I., Komazheva, B., (1995). "Who is caring for the children? An exploratory survey conducted in Hungary, Poland, Bulgaria and Romania. " Report, The Consultative Group on Early Childhood Care and Development. Fong, M., (1996), "Romania: gender in the transition", Report, Poverty and Social Policy Department, The World Bank, Washington DC Gustafsson, S., Stafford, F., (1992) "Child care subsidies and labor supply in Sweden" Journal of Human Resources 27(1): 204-229. Heckman, J., (1974) "Effects of child care programs on women's work effort." In Economic of the family, ed. T. Schultz, 491-518. Chicago: University of Chicago Press. ,(1979) "Sample selection bias as a specification error." Econometrica 47: 153-61 ,Singer, E., (1984) "A method of minimizing the impact of distributional assumptions in econometric models for duration data." Econometrica. 52: 271-320 Kimmel, J.,(1995). "The effectiveness of child care subsidies in encouraging the welfare-to-work transition of low-income single mothers." American Economic Review. Papers and Proceedings May. Laird. N., (1978). "Nonparametric maximum likelihood estimation of a mixing distribution," Journal of the American Statistical Association, 73(1978), 805-811 Lokshin, M., (1999) "Household child care choices and mother's labor supply in Russia." Dissertation, University of North Carolina at Chapel Hill, Chapel Hill Michalopoulos, C., Robins, P., Garfinkel, I., (1992) "A structural model of labor supply and child care demand." Journal of Human Resources 27(1) 166-203 Milanovich, B., (1997), "Income, inequality and poverty during the transition from planned to market economy." The World Bank, Washington DC Mincer, J., Ofek, H., (1982) "Interrupted work careers." Journal of Human Resources 17:3-24. Mincer, J., Polachek, S., (1974) "Family investment in human capital: earnings of women," Journal of Political Economy (Supplement), 82:S76-S108 Mroz, T., Guilkey, D., (1992) "Discrete factor approximation for use in simultaneous equation models with both continuous and discrete endogenous variables" Working paper series. Carolina Population Center. University of North Carolina at Chapel Hill. 22 _______ ,(1996) "Applications of discrete factor models in labor economics" Preliminary draft. Department of Economics and Carolina Population Center, University of North Carolina at Chapel Hill. _____, (,I 999) "Discrete factor approximations in simultaneous equation models: estimating the impact of a dummy endogenous variable on a continuous outcome." forthcoming in Journal of Econometrics National Committee for Statistics, (1998) "Survey of Labor Force in households" (AMIIGO), Bucharest Reynolds, A., Mann, E., Miedel, W., Smokowski, P., (1997) "The state of early childhood development intervention, effectiveness, myths and realities, new directions." Focus, University of Wisconsin- Madison: Institute for Research and Poverty. Vol. 19, No. 1. Ribar, D., (1995) "A structural model of child care and the labor supply of married women." Journal of Labor Economics 13(3): 558-97 ______. (1992) "Child care and the labor supply of married women: reduced form evidence." Journal of Human Resources 27(1): 134-65 Romanian Government (1991) "The organization of nurseries and kindergartens and the establishment of subsistence contributions to be paid by children's parents to these units." Government Decision #360/1991, Bucharest, Romania Van Den Brink, M., and Groot, W., (1997) "A household production model of paid labor, household work and child care" De Economist, v145 n3, pp. 325-43. Wong, R., Levine, R., (1992) "The effect of household structure on women's economic activity and fertility: evidence from recent mothers in urban Mexico." Economic Development and Cultural Change, Vol. 41 No. 4: 89-102. Zamfir, C., (1998) "Women's participation in the labor market and consequent implications in child rasing", Preliminary report, Institute of life quality research, Bucharest UNICEF (1997) "Education for All?" The MONEE project Regional Monitoring Report, No. 5., Florence: UNICEF International Child Development Center. 23 Table 1: Utilization of state-provided child care facilities Full-week Full-day Part-day program program program Occupancy rate (%) 82.7 80.6 86.4 Percent of facilities providing service 14.2 55.1 63.6 Source: RCCFS 1999. Table 2: Household child care arrangements Working mother Non-working mother Both Child at home 13.9 (110) 35.6 (283) 49.5 (393) Child in child care 23.4 (186) 27.3 (217) 50.7 (403) All 37.3 (296) 62.9 (500) 100 (796) Table 3: Child care arrangements of working mothers Others work Others don't work Both Child at home 31.0 (92) 6.1 (18) 37.1 (110) Child in child care 50.9 (151) 12.0 (35) 62.9 (186) All 81.9 (243) 18.1 (52) 100 (296) Table 4: Household child care arrangements by number of children # of Mother working Mother at home All children 0-6 years Formal Home Formal Home Formal Home Total old care care care care care care 1 21.3 (170) 12.8 (102) 19.1 (152) 28.5 (227) 40.5 (322) 41.3 (329) 81.7 (651) 2 2.01 (16) 1.0 (8) 7.8 (62) 6.28 (50) 9.8 (78) 7.3 (58) 17.9 (136) 3 0.0 (0) 0.0 (0) 0.5 (4) 0.6 (5) 0.5 (4) 0.6(5) 1.1 (9) 4+ 0.0 (0) 0.0 (0) 0.0 (0) 0.0 (0) 0.0 (0) 0.0 (0) 0.0 (0) All 23.3 (186) 13.8 (110) 27.3 (218) 35.4 (282) 50.8 (404) 49.2 (392) 100.0 (796) 24 Table 5: Mothers' labor force participation by household size Household size Percent of households Percent of households with working mothers 2 1.6 41.7 3 35.9 43.4 4 33.7 37.4 5 26.7 32.1 6 8.3 28.8 7 6.9 21.8 Total 100.0 Source: RCCES Table 6: Mothers' labor force participation by number of children in the household Number of children Percent of households Percent of households with working mothers 1 50.5 41.8 2 37.3 34.7 3 7.2 29.8 4+ 5.0 20.0 Total 100.0 Source: RCCES 25 Table 7. Summary statistics for the explanatory variables, the means and standard deviation. Mean Std. error Log of mother's wage 6.62 0.17 Log of wage of other household members 6.30 1.76 Mother's age 29.49 5.80 University degree 11.05 Technical/vocational education 20.10 Primary education 68.84 Household characteristics Total household nonwage income 382.74 563.46 Number of children 1.72 3.77 Number of pensioners 0.15 0.45 Household size 4.16 1.36 Rural regions 50.25 Town 10,000- 50,000 14.94 Town 50,000-200,000 11.30 Distance to facility: 10-20 minutes walk 34.43 47.54 Distance to facility: More than 20 minutes walk 17.21 0.37 Geographic area Transilvania 30.78 Muntenia 31.41 Moldova 19.72 Bucharest 18.09 Mother's occupation Agriculture 3.39 Industry 25.24 Construction 2.91 Trade 19.90 Health 8.25 Education 15.04 Banking 2.43 Other 22.81 Mother's tenure in months 84.39 64.31 26 Table 8(a): Semi-parametric estimation of the discrete outcome model of household child care choice and women 's labor supply Mother works Children at home Children in formal care Mother does not work, children in formal care Other household Other household Other household Other household members do not work members work members do not work members work Coefficient St. error Coefficient St.error Coefficient St. error Coefficient St. error Coefficient St. error Logaveragepriceofchildcare') -11.38** 3.98 6.37 5.41 13.26 14.36 0.58 0.51 -12.61 9.41 Logofmother'swage 9.30 6.83 37.32 35.29 76.26* 54.11 59.56** 31.02 95.26** 40.42 Log of other household members -0.88 0.84 5.62 4.88 -2.17 1.36 1.01 1.18 -4.01*** 1.09 Mother's age 3.73* 2.13 3.56 2.79 2.72 4.45 10.48*** 3.14 9.90 4.67 Mother's age squared -4.88* 2.93 -4.82 3.84 -2.45 5.90 -14.68*** 4.51 -14.76** 3.52 University degree -1.29** 0.63 -0.81* 0.46 -2.36** 0.86 -1.54** 0.60 -2.23** 0.88 Technical/vocational education -0.27 0.25 -0.07 0.34 1.44* 0.71 -0.85** 0.40 1.13* 0.54 Primary education Reference Household characteristics Total household non-wage income -4.56* 2.85 -0.94 2.30 -5.05 3.44 -4.49* 2.72 2.09 3.06 Number of children -1.28 1.56 -8.74*** 2.35 -1.76 7.42 -8.28** 2.40 1.37 3.05 Number of pensioners 3.66 2.40 -3.23 3.59 20.46** 7.78 0.35 3.19 6.70 2.75 Household size 1.94 1.24 4.15** 1.49 -9.82 6,22 1.84 1.69 -2.08 2.49 Rural regions Reference Town 10,000-50,000 -0.68** 0.34 0.37 0.37 0.16 0.96 0.45 0.35 -0.31 0.75 Town 50,000-200,000 -0.95** 0.41 0.30 0.38 1.23 0.79 -0.49 0.42 -0.59 0.79 10-20 minutes walk -0.48* 0.23 -0.25 0.33 0.72 0.73 0.12 0.29 1.10* 0.52 More than 20 minutes walk -0.52 0.30 0.30 0.35 1.36* 0.77 -0.02 0.37 -1.04 1.14 Transilvania -0.02 0.33 0.63 0.38 0.15 0.81 0.70 0.40 2.64** 1.25 Muntenia -0.04 0.33 0.03 0.40 -0.18 0.84 1.04** 0.40 3.52** 1.25 Moldova 0.55* 0.34 0.02 0.44 -0.37 1.02 0.67 0.42 3.61** 1.26 Bucharest Reference Constant -6.70 5.27 -3.94*** 0.42 -6.08*** 0.54 -5.87*** 0.80 -7.56*** 0.81 i) All variables are normalized to be in [0,1] range Note: The mode where mother does not work and children stay home is used as a reference; * is significant at 10% level; ** at 5% level; *** at 1%. 27 Table 8(b): Semi-parametric estimation of the continuous outcome equations of the mother 's work hours and the hours children spend at child care facility. Hours of work Hours in formal care Coefficient Std. Error Coefficient Std. error Log average price of child care') 6.87 9.04 -53.88 36.71 Log of mother's wage 11.25 19.62 19.72 76.63 Log of other household members -3.90** 1.67 4.34 6.83 Mother's age 1.77 3.03 1.61 12.38 Mother's age squared -0.03 0.04 -0.03 0.18 University degree Technical/vocational education -6.34 5.90 7.43 24.18 Primary education Reference Household characteristics Total household nonwage income -0.01 0.00 -0.01 0.02 Number of children -5.43 3.81 43.23** 15.50 Number of pensioners -2.48 5.55 27.55 22.71 Household size 4.90* 2.77 4.56 11.39 Rural regions Reference Town 10,000-50,000 7.23 6.98 20.03 28.17 Town 50,000-200,000 1.30 7.82 29.83 31.79 10-20 minutes walk 1.82 5.27 2.93 21.53 More than 20 minutes walk 0.74 6.63 -4.08 26.97 Transilvania 5.57 7.62 5.26 30.55 Muntenia 3.01 7.59 18.78 30.84 Moldova -6.07 7.85 37.79 31.96 Bucharest Reference Mother's occupations Agriculture 15.57 18.22 X Industry 46.37*** 8.47 X Construction 35.60** 18.93 X Trade 30.78*** 7.60 X Health 23.73** 12.25 X Education 47.54*** 10.80 X Banking 21.22 17.75 X Other Reference Mother's tenure in months 0.44*** 0.05 X Constant -105.55 124.16 80.26 476.43 ') All variables are normalized to be in [0,1] range Note: * is significant at 10% level; ** is significant at 5% level; *** is significant at 1% level. 28 Table 9: Simulation of the effects of various policies on the household choices of child care mode and mother's labor supply (proportion of households in each state) Mother does not work Mother works Children home Children in Children home Children in formal care formal care Other Other Other Other household household household household members members members members (State 0) (State 1) do not work work do not work work (State 2) (State 3) (State 4) (State 5) Increase in mothers 'wage rate by 10% Sample average 0.378 0.267 0.131 0.022 0.158 0.044 Sample average+10% 0.306 0.230 0.148 0.033 0.207 0.075 Change (%) -7.193 -3.696 1.744 1.115 4.961 3.068 Increase in child care price by 10% Sample average 0.375 0.261 0.128 0.020 0.174 0.042 Sample average+10% 0.395 0.258 0.117 0.018 0.177 0.036 Change (%) 1.998 -0.301 -1.102 -0.250 0.300 -0.646 Increase in other household members wage by 10% Sample average 0.382 0.272 0.114 0.020 0.178 0.033 Sample average+10% 0.381 0.270 0.120 0.020 0179 0.032 Change (%/e) -0.138 -0.275 0.527 -0.051 0.059 -0.121 Increase in household non-wage income by 10% Sample average 0.374 0.264 0.128 0.022 0.172 0.041 Sample average+10% 0.376 0.262 0.129 0.021 0.170 0.041 Change (%) 0.269 -0.232 0.059 -0.023 -0.131 0.059 Table IO: Comparison of the elasticities of the effect of child care cost on labor supply Study (year) Data Estimated elasticities (increase in child care cost) Labor force participation Use of formal care This paper Romania Child Care and Employment Survey -0.17 -0.41 Lokshin Russia Longitudinal Monitoring Survey -0.19 -0.11 (1999) Cleveland, Morley, Canadian National Child Care Survey -0.39 -0.2 Hyatt, (1996) Connelly Panel of Survey of Income and Program -0.20 Not available (1992) Participation 1984, USA Blau and Robins 1980 Baseline Household Survey of -0.38 Not available (1988) Employment Opportunity Pilot ProLect, USA 29 2.0% Mean per capita income 1.5%- 1.0% 0.5% 0.0% 0 900 1800 Per capita income Figure 1: Non-parametric estimation ofthe distributional impact of a I10 percent decrease in the price of child care on the mother's probability of being employed, by per capita income 1 5.0% Mean per capita income ,, D 10.0% .c . -C 050% - 0.0% 0 900 1800 Per capita income Figure 2: Non-parametric estimation of the distributional impact of a 10 percent increase in mothers' wage rates on the mother's probability of being employed, by per capita income 30 Policy Research Working Paper Series Contact Title Author Date for paper WPS2378 Disintegration and Trade Flows: Simeon Djankov June 2000 R. Vo Evidence from the Former Soviet Caroline Freund 33722 Union WPS2379 India and the Multilateral Trading Aad tya Mattoo June 2000 L. Tabada System after Seattle: Toward a Arvind Subramanian 36896 Proactive Role WPS2380 Trade Policies for Electronic Aaditya Mattoo June 2000 L. Tabada Commerce Ludger Schuknecht 36896 WPS2381 Savings and the Terms of Trade Pierre-Richard Ag6nor June 2000 T. Loftus under Borrowing Constraints Joshua Aizenman 36317 WPS2382 Impediments to the Development and Thorsten Beck June 2000 E. Mekhova Efficiency of Financial Intermediation 85984 in Brazil WPS2383 New Firm Formation and Industry Thorsten Beck June 2000 E. Mekhova Growth: Does Having a Market- or Ross Levine 85984 Bank-Based System Matter? WPS2384 Are Cost Models Useful for Telecoms Daniel A. Benitez July 2000 G. Chenet-Smith Regulators in Developing Countries? Antonio Estache 36370 D. Mark Kennet Christian A. Ruzzier WPS2385 The Rise, the Fall, and ... the Antonio Estache July 2000 G. Chenet-Smith Emerging Recovery of Project John Strong 36370 Finance in Transport WPS2386 Regulators and the Poor: Lessons Richard Green July 2000 G. Chenet-Smith from the United Kingdom 36370 WPS2387 The Long and Winding Path to Private Antonio Estache July 2000 G. Chenet-Smith Financing and Regulation of Toll Manuel Romero 36370 Roads John Strong WPS2388 The Role of Special and Differential Constantine Michalopoulos July 2000 L. Tabada Treatment for Developing Countries in 36896 GATT and the World Trade Organization WPS2389 Vietnam: On the Road to Labor- Patrick Belser July 2000 H. Sutrisna Intensive Growth? 88032 WPS 2390 The Social Rate of Return on David Canning July 2000 H. Sladovich Infrastructure Investments Esra Bennathan 37698 Policy Research Working Paper Series Contact Title Author Date for paper WPS2391 Are the Poor Protected from Budget Martin Ravallion July 2000 P. Sader Cuts? Theory and Evidence for 33902 Argentina WPS2392 What Factors Appear to Drive Private Dipak Dasgupta July 2000 S. Crow Capital Flows to Developing Countries? Dilip Ratha 30763 And How Does Official Lending Respond? WPS2393 Will the Euro Trigger More Monetary Patrick Honohan July 2000 A. Yaptenco Unions in Africa? Philip R. Lane 31823 WPS2394 Tax Evasion, Corruption, and the Waly Wane July 2000 H. Sladovich Remuneration of Heterogeneous 37658 Inspectors WPS2395 Decentralizing the Provision of Health William Jack July 2000 H. Sladovich Services: An Incomplete Contracts 37698 Approach WPS2396 Aid Dependence and the Quality of Stephen Knack July 2000 P. Sintim-Aboagye Governance: A Cross-Country 38526 Empirical Analysis WPS2397 Verifying Exchange Rate Regimes Jeffrey Frankel July 2000 E. Khine Eduardo Fajnzylber 37471 Sergio Schmukler Luis Serven WPS2398 Determinants of Currenl Account Cesar Calder6n July 2000 H. Vargas Deficits in Developing Countries Alberto Chong 38546 Norman Loayza WPS2399 Managers, Investors, and Crises: Graciela Kaminsky July 2000 E. Khine Mutual Fund Strategies in Emerging Richard Lyons 37471 Markets Sergio Schmukler