Impact of child care support on female labor supply, family income and public finance Nicholas-James Clavet and Jean-Yves Duclos CIRPÉE, Université Laval May 2011 Preliminary — please do not quote Abstract We study in this paper the impact of child care support on female labor force participation, family income, and public finance in Quebec (Canada). We first estimate the determinants of female labor supply using a conditional logit. We then simulate the effects of withdrawing existing child care support measures (federal tax deduction, provincial tax credit and child care in-kind support) and of the implementation of recent and alternative child care support measures. Outcomes are heterogenous with family structure and child care support measures. The results show that child care support has a significative effect on female labor supply, family income and public finance. The federal tax deduction, the provincial tax credit and in-kind child care support increase female labor supply and family income but deteriorate public finance. Withdrawing child care support totally would induce a significant fall in labor supply and in family income and an improvement in public finance. Finally, a universal child care allowance would increase labor supply and family income but would worsen public finance — the opposite effect would be observed following a reduction in in-kind child care support. Acknowledgements This research was supported by the Fonds Québécois de Recherche sur la Société 1 et la Culture (FQRSC) as part of a concerted program on the assessment of the impacts of public policy on health and well-being. We are grateful to FQRSC and to its partners, the Ministry of Health and Social Services and the Health Research Fund, for their support. 1 Introduction Many of North America’s children nowadays receive child care outside their home. For example, 80 % of Quebec’s children aged between 3 and 5 receive regular child care (Japel, Tremblay, et Côté 2005) outside their home. This large prevalence of nonparental child care naturally questions the influence of child care on the lives of families and on parental work habits in particular. The issue of child care is also high on the agenda of policy makers. It has been seriously debated in Canada’s last provincial and federal election campaigns. For the last few years, the federal government has been directing transfers to families with preschoolers for the stated purpose of assisting families with child care responsibilities. In Quebec, the different major political parties have tended to favor different modalities for supporting child care, ranging from direct monetary transfers to parents, to development of highly subsidized public “Early Childhood Centers” (ECC), to fiscal relief towards the burden of child care expenditures, and to the encouragement of the provision of private child care services. The Quebec government has also been investing substantial resources in the establishment and maintenance of child care services (see Ministère de la famille et de l’enfance 2002), of which ECC are the primary delivery modality. Although these child care services are offered within a network that is purportedly universal, it is instructive to note that the distribution of these services differs very much across family types and living standards — especially since it suggests that the least-advantaged families are also those that are least likely to receive such forms of child care support (Grenier 2005). The issue of the effects of child care support in Quebec is also connected to the fast increase in female labor participation rate over the past decades. This transformation has had a significant impact on families, partly because they are sometimes headed by low income single females. In recent years, government policies have also been extensively redesigned to encourage the participation of as many people as possible in the labor market, including that of single mothers and of both parents of preschool children. The child care issue is therefore at the heart of multiple effects and objectives, including supporting family incomes and the well-being of children as well as encouraging female labor market participation. This article aims to study the effects of child care support in Quebec on female labor force participation, income families, and public finance. We do this by using a microsimulation model that takes into account the complexity of the system of taxes and of various income support measures, and the fact the effects of that system affect a widely heterogeneous population of individuals and households — see e.g. Gupta et 1 Kapur (2000), Harding (1996), and OECD (2004) for a discussion of these models. The model is of an accounting, distributive and behavioral type since it allows to simulate the impact of existing and/or prospective elements of the tax and transfer system on the distribution of living standards as well as on labor supply of families. The two main types of measures to support families — unconditional allowances to parents and (conditional on labor force participation) child care support measures — have theoretical effects that are qualitatively different from one another. Relative to a scenario of no family support, family allowances to parents lead to an income effect (an increase of disposable income) that tends to reduce parental participation in the labor market. Subsidies (conditional on participation in the labor market) to child care introduce both a price effect (an increase in hourly wages net of child care expenses) that encourages participation and an income effect that tends to reduce participation. The value and the impact of child care subsidies also differ significantly by income and the provision mode of these subsidies — they depend for example on whether they are paid directly to families in the form of a refundable or a non-refundable tax credit or as a deduction from taxable income. The effects of these types of policies on labor supply also depend on the interaction between price and income labor supply elasticities as well as on the wage distribution and on activity levels in the population. Estimating precisely these empirical effects requires a detailed modeling of these elements; this is what we do in this article. This also allows us to examine how different forms of family support may differentially affect the net income of families. The microsimulation model used in this article shows that the child care tax credits and deductions coupled with ECC significantly increase family income after taxes, transfers and child care costs. Supporting child care also has a greater impact on the net incomes of single mothers than on those of “married women” (namely, women living in a couple). 2 Model The model used in this article seeks to evaluate first the effects of the current system of child care support on household income, the labor market behavior of mothers with preschool children (between 0 and 4 years), and the finances of provincial and federal governments. The model also allows to quantify the impact of potential changes to child care support. A description of current and alternative child care support modalities system is presented in Table 1. More precise definitions and the costs of the different child care support modalities are presented in Tables 2 and 3. The model is based on an accounting, distributive and behavioral structure. Using 2 socio-economic information on a representative set of households, income profiles before and after taxes and transfers are first computed.1 Since family policies in connection with child care are of particular interest, child care costs are subtracted from income after taxes and transfers (usually called disposable income by Statistics Canada) to obtain a income net of taxes, transfers and child care expenses. Net income varies depending on the type of child care since the costs of these types of child care vary. The probability of making a given choice of labor supply is estimated econometrically through the observation of the choices made by working women in our representative sample. Table 1: Definition of child care support measures Type Definition Provincial Tax Credit Amount deducted from the provincial tax payable (See Appendix A for table and refundable if the credit exceeds of credits) the amount of provincial tax otherwise due. Federal Tax Deduction Amount deducted from federal taxable income. Direct subsidy to ECC These grants take the form of direct grants to ECC; a family contribution of $7 a day is then paid. Note that only full-time (260 days) or part-time (130 days) care is available with ECC. Fam. allowance (alternative measure) $100 per week paid directly to families for each child under 5 years. Direct subsidies to ECC are then removed and the total cost of ECC child care must then be paid by the family. ECC rate increase (alternative measure) Fees for ECC increase from $7 to $10 a day, with a consequent reduction of the government direct grant. Universal benefit (recent measure) $100 a month paid directly by federal government to families for each child under 6 years. Figures 1 to 4 present income net of taxes, transfers and costs of child care in two family types by number of hours worked per week. This shows the “budget constraints” 1 This is done using the Canadian Tax and Credit Simulator (CTaCS). A wide variety of government programs is taken into account by the simulator (family allowances, tax credits, etc.) in addition to federal and provincial taxes. This simulator is supplemented with a separate model of welfare transfers built on Quebec’s welfare system in 2004. 3 Table 2: Definition of child care types Type of care Definition Family-based ECC Subsidized daycare in private home. Non-profit ECC Subsidized daycare located in a center not seeking to make profits. For-profit ECC Subsidized daycare located in a center seeking to make profits. Family-based private Non-subsidized daycare located in in private home. Private non-profit center Non-subsidized daycare located in a center not seeking to make profits. Private for-profit center Non-subsidized daycare located in a center seeking to make profits. Child care provided by relatives Child care using the services of relatives. Table 3: Description of the costs of child care types Type of care Cost per day Cost per day for the family for the gvt Family-based ECC 7.0 25.96 Non-profit ECC 7.0 34.61 For-profit ECC 7.0 19.03 Family-based private 32.96 0.0 Private non-profit center 41.61 0.0 Private for-profit center 26.03 0.0 Child care provided by relatives 0.0 0.0 Total daily Cost 32.96 41.61 26.03 32.96 41.61 26.03 0.0 These costs were derived from Lefevbre, Merrigan, et Verstraete (2009) and Lefevbre et Merrigan (2008). 4 Figure 1: Impact of the provincial tax credit and of the federal deduction on income net of taxes, transfers and costs of child care for a single mother using family-based private day care. 29000 27000 25000 23000 21000 without family policy With deduction 19000 With credit and deduction 17000 15000 0 10 20 30 40 50 60 Figure 2: Budget constraints for a single mother by child care type (ECC, private non-profit center and family-based private) 37000 32000 27000 ECC 22000 Private nonprofit day care Private family day care 17000 12000 0 10 20 30 40 5 50 60 of those families and displays the income/work relationship generated by the tax accounting structure of the model. Figures 1 and 2 deal with single mothers, while Figures 3 and 4 focus on married women. Figure 1 presents the budget constraint of a single mother with a preschooler using private family-based child care and earning an average wage (14.37$/hour). One can see that measures to finance child care significantly increase the net income of the single mother and that this effect increases with the number of hours worked. It also shows that the provincial tax credit seems to be slightly more generous than the federal deduction. Figure 2 presents the budget constraints for the same type of women, but with different child care modalities. It shows for example that it may be financially advantageous to use family-based private child care rather than ECC for a single mother when she earns an average wage and works less than 40 hours per week. ECC becomes more interesting than a private non-profit center when a single mother works more than 25 hours per week — after which point the value of the subsidy paid directly to ECC exceeds the value of the sum of the tax credits and deductions received from paying private child care fees. Figure 3 presents the budget constraints of a married mother of a preschooler that uses family-based private child care services, with a hourly wage set to the average wage (16.05$/hour) and a spouse earning an (average) salary of 18.68$/hour and working 40 hours per week. Unlike the graphs for the single mothers that are smoother, those for married women are more broken: this is explained by the fact that such women have a greater household income and that transfers are less important for them. Each bend in the budget constraints represents a change in the personal income tax bracket. In the first segment, the wife’s income is not taxed because it benefits from the basic tax credit; in the second segment, the woman is taxed according to the first personal income tax bracket, and in the third segment, according to the second personal income tax bracket. Note that the value of the federal deduction increases with each bracket. Moreover, note that the federal deduction has a monetary impact that is not too different from that of the provincial tax credit in the case of a married mother, unlike the case of a single mother. Figure 4 presents the budget constraints of a married woman with the same characteristics but with different child care services. From 25 hours per week onwards, ECC becomes increasingly more financially attractive than private child care services. The net income advantage over family-based private child care amounts to $473 a year (at 25 hours/week) and to $1,423 for private profit center-based child care. The greater the amount of income and leisure a woman is able to enjoy at a given point on her budget constraint, the more likely it is that she will choose the corre6 Figure 3: Impact of the provincial tax credit and of the federal deduction on income net of taxes, transfers and child care costs of a woman married to a man working full-time and using family-based private child care services 72000 67000 62000 57000 Without family policy 52000 With deduction 47000 With credit and deduction 42000 37000 32000 0 10 20 30 40 7 50 60 Figure 4: Budget constraints of a woman married to a man working full-time with child care of different types (ECC, private non-profit center and family-based private). 77000 72000 67000 62000 ECC 57000 private nonprofit day care 52000 Private family day care 47000 42000 37000 32000 0 10 20 30 40 50 60 sponding level of labor supply. Since a lower level of leisure time is needed to increase income, a trade-off between income and leisure must be made. This is implemented out in the model’s behavioral structure through the estimation of a statistical model (a conditional multinomial logit) that estimates the labor supply preferences (based on observed and unobserved characteristics) of women in Quebec with children aged 0 to 4 years. The sample is representative of the Quebec population of such women. The technical description of the model can be found in Appendix B. 3 Data The main source of data used by the model comes from the 2004 Social Policy Simulation Database (SPSD) produced by Statistics Canada. The primary component of this database is the Survey of Labour and Income Dynamics (SLID). For the purposes of this paper, we also need to impute additional variables, such as gross hourly wages, the net value of the residence, the value of financial assets and the net worth of vehicles owned. The additional databases used by the model to impute these variables date mostly from 2004. 8 The selected sample excludes all women under age 18 and over 65 as well as full-time students and disabled persons. Individuals reporting income from self-employment as well as married women with no children of preschool age (between 0 and 4 years) are also removed from the sample. Individuals working more than 70 hours per week are omitted to avoid outliers. The work behavior of the spouses is not modeled: the labor supply of men is supposed to be fixed. 4 Simulation The model explains the labor supply behavior of women with children 0-4 years in the presence of the current child care support system, which includes the provincial tax credit, the federal tax deduction and the direct subsidies to ECC. The model can also help anticipate the reactions of female labor supply to possible changes to that system. Two alternative measures of supporting child care are considered in addition to the federal measure introduced in July 2006 — the Universal Child Care Benefit. The first alternative measure abolishes the direct ECC subsidies and provides a yearly allowance of $5,200 to women for each child aged between 0 and 4 years. The second alternative measure increases the price paid by families for ECC services from $7 to $10 a day. These two measures have recently been proposed and debated in Quebec. Finally, the universal child care benefit provides a monthly transfer of $100 per child under 6 years, while not modifying other support measures. These measures are summarized in Table 1. The results of the simulation are shown in Appendix C. References Cogan, J. (1981): “Fixed Cost and Labour Supply,” Econometrica, 49, 945–964. Gong, X. et A. van Soest (2002): “Family structure and female labour supply in Mexico City,” Journal of Human Resources, 37, 163–191. Gouriéroux, C. et A. Monfort (1996): Simulation-Based Econometric Methods, Core Lectures, Oxford University Press. Grenier, M. (2005): “Un enjeu oublié de la politique des services de garde à 5$: les effets distributifs des subventions en nature,” Mémoire de maı̂trise, UQAM, Montréal. Gupta, A. et V. Kapur (2000): Microsimulation in Government Policy and Forecasting, Amsterdam: North-Holland Elsevier Science. 9 Harding, A. (1996): Microsimulation and public policy, Amsterdam: NorthHolland Elsevier. Japel, C., R. Tremblay, et S. Côté (2005): “La qualité des services à la petite enfance: résultats de l’Étude longitudinale sur le Développement des Enfants du Québec (ÉLDEQ),” Éducation et Francophonie, 23, 7–27. Kamionka, T. (1998): “Simulated Maximum Likelihood Estimation in Transition Models,” Econometrics Journal, 1, C129–C153. Keane, M. et R. Moffitt (1998): “A Structural Model of Multiple Welfare Program Participation and Labor Supply,” International Economic Review, 39, 553–589. Laroque, G. et B. Salanié (1993): “Simulation-Based Estimation of Models with Lagged Latent Variables,” Journal of Applied Econometrics, 8, S119–S133. Lefevbre, P. et P. Merrigan (2008): “Child-Care Policy and the Labor Supply of Mothers with Young Children: A Natural Experiment from Canada,” Journal of Labor Economics, 26, 519–548. Lefevbre, P., P. Merrigan, et M. Verstraete (2009): “Dynamic labour supply effects of childcare subsidies: Evidence from a Canadian natural experiment on low-fee universal child care,” Labour Economics, 16, 490–502. Ministère de la famille et de l’enfance (2002): “Cadre de référence de l’évaluation de la qualité des services de garde,” Rapport technique, Gouvernement du Québec. OECD (2004): “Indicators of Unemployment and Low-Wage Traps (Marginal Effective Tax Rates on Employment Incomes),” Document de travail 18, OECD Social, Employment and Migration, Paris. Soest, A. V. et M. Das (2001): “Family labor supply and proposed tax reforms in the Netherlands,” De Economist, 149, 191–218. A Table of Child Care Tax Credit Table 4 presents the parameters of the child care tax credit in Quebec. 10 11 Household income ($) sup. without to exceeding 0 28 705 28 705 29 765 29 765 30 830 30 830 31 890 31 890 32 950 32 950 34 015 34 015 35 080 35 080 36 145 36 145 37 205 37 205 38 265 38 265 39 330 39 330 40 390 40 390 41 460 Rate of tax credit % 75 74 73 72 71 70 69 68 67 66 65 64 63 Household income ($) sup. without to exceeding 41 460 42 520 42 520 43 580 43 580 44 645 44 645 45 705 45 705 46 765 46 765 47 835 47 835 48 895 48 895 49 965 49 965 51 025 51 025 52 085 52 085 53 150 53 150 54 215 54 215 55 280 Rate of tax credit % 62 61 60 59 58 57 56 55 54 53 52 51 50 Household income ($) sup. without to exceeding 55 280 56 340 56 340 57 400 57 400 58 465 58 465 59 525 59 525 60 595 60 595 61 655 61 655 62 715 62 715 63 780 63 780 64 840 64 840 65 905 65 905 66 970 66 970 68 030 68 030 69 095 Rate of tax credit % 49 48 47 46 45 44 43 42 41 40 39 38 37 Household income ($) sup. without to exceeding 69 095 70 155 70 155 71 220 71 220 72 280 72 280 73 345 73 345 74 410 74 410 75 470 75 470 76 535 76 535 77 595 77 595 78 655 78 655 79 725 79 725 and more Table 4: Parameters of Quebec’s child care tax credit, by yearly household income Rate of tax credit % 36 35 34 33 32 31 30 29 28 27 26 B Econometric Model In order to conduct appropriate policy simulations, the labor supply model must explain individual behavior in a theoretically consistent manner. In particular, the Slutsky restrictions must be satisfied. Unfortunately, the highly non-linear (and often nonconvex) budget sets make this task particularly demanding if we treat hours of work as a continuous variable. To ease the task, it is now customary to follow Soest et Das (2001) and focus on choices of a discrete number of weekly hours of work. Thus, let the choice set facing an individual be {HT 1 , HT 2 ,. . . , HT p }, where p is the number of possible choices of labor supply. Individuals are assumed to maximize a well-behaved utility function defined over leisure, l, and net income, y, taking into account time and income constraints: max U i (li , y i ) s.t. y i ≤ y i (li , w) and li ≤ DT. (1) Hours of leisure, (li = DT − HT i ), are given by the time endowment, DT , from which we subtract hours of work, HT i . We arbitrarily fix DT = 80 hours per week.2 Net income equals earnings, wHT i , plus exogenous non-labor income, N , plus government transfers, B, minus income taxes, T , and minus child care costs, CC i (Keane et Moffitt 1998): y i (HT i ) = wHT i + N + B(wHT i , N, Xi ) − T (wHT i , N, Xi ) − CC i , (2) where X is a vector of household variables. We use a translog utility function because of its well-known properties and flexibility. It is defined as: U i (li , y i ) = β1 log(li ) + β2 log(li )2 + β3 log(y i ) + β4 log(y i )2 . (3) This specification of the utility function is locally flexible to the second order and does not impose the quasi-concavity of preferences.3 As is customary, preference heterogeneity is introduced in the leisure term: β1 = α0 + α1 log(Age) + α2 log(Age)2 + α3 NB018 + α4 (P reschool > 0) + υ. 2 (4) According to Gong et van Soest (2002), the parameter estimates are relatively insensitive to this particular normalization. 3 The marginal utility of income must be positive for the model to be theoretically consistent — see Soest et Das 2001. 12 The dummy variable (P reschool) is equal to one when a preschooler is present in the household and the variable N B018 is the number of children between 0 and 18 years. Preference for leisure may also vary with unobserved characteristics. A random component term υ is therefore added to equation (4). This term is assumed to be independently and identically distributed as a normal random variate with mean zero and variance σ 2 . To allow for optimization errors, we also assume that the utility function is random: ui (li , y i ) = U i (li , y i ) + ξ i . (5) This assumption is made to account for the possibility that individuals may not know their budget sets perfectly, or for the fact that their optimal choice may not correspond to the discrete choices assumed in the model. For the purpose of identification, ξ i is assumed to be independently and identically distributed as a Type-I extreme value random variate (Gumble distribution). According to equation (1), an individual will choose HT i if U i is greater than the utility associated with the other alternatives. Given the stochastic specification of the model, the probability that this will happen, conditional on a given modality for child care, CCM , is given by: exp (U i (li , y i ) |CCM i ) . Pr U i ≥ U j ∀ j = Pp j j j j j=1 exp (U (l , y ) |CCM ) (6) The discrete labor supply literature has generally found that such models tend to underpredict the number of individuals with HT = 0. This will occur if the “fixed costs” associated with work are omitted from the analysis. Indeed, Cogan (1981) long ago insisted that both the monetary costs (commuting, daycare, etc.) and non-monetary cost (psychic costs, stress, etc.) be accounted for explicitly in labor supply models. Obviously, some of these costs are difficult to measure but may be proxied by demographic variables. Fixed costs must then be subtracted from income if HT > 0. The problem with this is that income minus fixed costs may be negative, a possibility that cannot be dealt with due to the translog utility function. Gong et van Soest (2002) recently introduced the notion of fixed income for not working. Instead of subtracting a fixed cost to work, a fixed income can be added to the income at zeros hours of work, making inactivity a relatively more attractive alternative. Both approaches have the potential of capturing the bunching at zero hours of work. For practical reasons, the model for single mothers is thus based upon the “fixed costs” approach, while the models for single females is based upon the “fixed income” approach. 13 Fixed incomes and fixed costs are incorporated into the model by replacing U (y i , li ) by U (y 0 + F I, l0 ) and U (y i − F C, li ) ∀i > 0, respectively. More precisely, F I = γ0 + γ1 ln (Age) F C = δ0 + δ1 (P reschool > 0). (7) (8) Equation (7) supposes that fixed income depends on age. It may be conjectured that older individuals face a higher benefit to staying at home. In the model, this is akin to having a larger fixed income at zero hours of work. Likewise, equation (8) states that the fixed costs of work are associated with the presence of preschoolers. The two functions could in principle be made to depend on a richer set of covariates. Various specifications have indeed been estimated; to save on the degrees of freedom, the most parsimonious specification that nevertheless fits the data well was retained. One last modification is made to the standard model to accommodate the bunching of weekly hours of work around 40. We write: F C = δ0 + δ1 (P reschool > 0) + δ2 (HT = 40). (9) The parameter δ2 proxies a fixed effect that captures the utility/convenience possibly associated with working the normal forty hours per week. Finally, note that equation (6) is written conditionally on a given realization of the child care mode, CCM . The unconditional probability is obtained by integrating (6) over the probablity distribution of CCM : Z i exp (U i (li , y i ) |CCM i ) 2 j Pp f CCM ; µ, σ dCCM . (10) Pr U ≥ U ∀ j = j j j j j=1 exp (U (l , y ) |CCM ) Equation (10) does not in general have a closed-form solution. We thus simulate the integration by drawing R = 50 draws from the distribution of CCM for each observation and compute the average probability: R X i exp U i (li , y i ) |CCMqi 1 j c U ≥U ∀j = . Pr P R q=1 pj=1 exp U j (lj , y j ) |CCMqj (11) The maximization of the √ simulated likelihood function yields consistent and efficient parameter estimates if N /R → 0 when R → +∞ and N → +∞ (see Gouriéroux et Monfort 1996). The literature suggests that R = 20 appears adequate — see Laroque et Salanié 1993, Kamionka 1998. 14 C Tables of Results Table 5: Effects of the withdrawal of current child care support measures on hours worked by two subgroups of women with young children (% of hours worked Initially) Prov. credit Fed. deduction $7/day fees without meas. Percentage variations intensive margin Married women -0.22 -0.17 -2.18 -2.70 Single mothers -1.65 -0.87 -2.12 -15.88 Percentage variations extensive margin Married women -0.79 -0.90 -1.91 -11.93 Single mothers -1.17 -1.25 -3.25 -16.63 Total percentage variations Married women -1.01 -1.07 -4.09 -14.63 Single mothers -2.82 -2.12 -5.37 -32.51 15 Table 6: Effects of alternative child care worked by two subgroups of women with worked initially) Family allowance Percentage variations intensive margin Married women -1.64 Single mothers 0.41 Percentage variations extensive margin Married women 1.43 Single mothers 12.70 Total percentage variations Married women -0.21 Single mothers 13.11 support measures on hours young children (% of hours Univ. benefit $10/day fees -0.41 0.69 -0.08 -0.35 0.23 3.33 -0.73 -3.07 -0.18 4.02 -0.81 -3.42 Table 7: Effects of withdrawal of current child care support measures on hours of work of two subgroups of women with preschool children Married women Single mothers Initial total hours worked yearly 101,017,000 13,824,000 ∆ Tot. hours worked yearly after withdrawal Provincial tax Credit -1,017,000 -390,000 Federal tax deduction -1,085,000 -293,000 ECC subsidies ($7-a-day fees) -4,135,000 -560,000 All measures -14,783,000 -4,495,000 Initial average hours worked yearly 933 1,280 ∆ Av. hours worked yearly after withdrawal Provincial tax credit -9.4 -36.1 Federal tax deduction -10.0 -27.1 Seats grant for $7 a day fees -38.2 -68.8 All measures -136.6 -416.1 16 Table 8: Effects of alternative child care support measures on hours of work of two subgroups of women with preschool children Married women Single mothers Initial total hours worked yearly 101,017,000 13,824,000 ∆ Total hours worked in the pop. after withdrawal Family allowance -21,938 1,813,000 Universal benefit -186,508 555,000 $10-a-day fees for ECC -820,434 -473,000 Initial average hours worked yearly 933 1,280 ∆ Average hours worked after withdrawal Family allowance -1,9 213.8 Universal benefit -1,7 51,4 $10-a-day fees for ECC -7.6 -43.8 Table 9: Effects of withdrawal of current child care support measures on income after taxes, transfers and child care expenses in two subgroups of women with preschool children Married women Single mothers Initial total income 5,030,000,000 280,000,000 ∆ Tot. Inc. After Withdra. Provincial tax credit -48,271,000 -4,986,000 Federal tax deduction -45,850,000 -4,453,000 ECC subsidies ($7-a-day fees) -160,400,000 -863,000 All measures -490,300,000 -49,912,000 Initial average income 46,000 26,000 ∆ Av. inc. after withdra. Provincial tax credit -446 -462 Federal tax deduction -424 -412 ECC subsidies ($7-a-day fees) -1,482 -106 All measures -4,529 -4,621 ∆ Av. inc. after withdra. (%) Provincial tax credit -1.1 -1.9 Federal tax deduction -1.0 -1.3 ECC subsidies ($7-a-day fees) -3.0 -1.1 All measures -10.0 -17.0 17 Table 10: Effects of alternative child care support measures on income after taxes, transfers and child care expenses in two subgroups of women with preschool children Married women Single mothers Initial total income 5,030,000,000 280,000,000 ∆ total income Family allowance 46,071,000 30,629,000 Universal benefit 99,029,000 6,829,000 $10 per day ECC fees -40,318,000 -4,407,000 Initial average income 47 000 26,000 ∆ average income Family allowance 3,986 2,836 Universal benefit 915 632 $10 per day ECC fees -372 -408 ∆ percentage income Family allowance 9,0 11.4 Universal benefit 2,1 2,6 $10 per day ECC fees -0,8 -1.6 Table 11: Change in government expenditures (net of effects on taxes and transfers) of withdrawal of child care support measures for married women Variables Federal tax Provincial tax Federal transfers Provincial transfers Social assistance Retirement pen. cont. Employment insurance ECC subsidies Net provincial cost Per capita prov. cost Net federal cost Per capita fed. cost Total cost Per capita total cost ∆ after tax cred. withdrawal -787,000 -1,752,000 977,000 -44,801,000 610 000 -581,605 -278,000 0 -41,857,000 -387 2,041,000 19 -39,815,000 -368 ∆ after tax ded. withdrawal 25,328,000 -2,068,000 -11,544,000 -600,000 525 000 -666,000 -314,000 -5,101,000 -2,450,000 -23 -36,558,000 -338 -39,009,000 -360 18 ∆ after ECC grant withdrawal -48,622,000 -6,809,000 29,053,000 201 400 000 95,000 -3,205,000 -1,396,000 -524,800,000 -313,300,000 -2,894 79,071,000 730 -234,300,000 -2,164 ∆ withdrawal of all measures 1,445,000 -30,922,000 87,70,000 -38,002,000 6,037,000 -9,339,000 -4,426,000 -524,800,000 -516,500,000 -4,772 11,750,000 109 -504,800,000 -4,663 Table 12: Change in government expenditures (net of effects on taxes and transfers) of withdrawal of child care support measures for single mothers Variables Federal tax Provincial tax Federal transfers Provincial transfers Social assistance Retirement pen. cont. Employment insurance ECC subsidies Net prov. cost Per capita prov. cost Net federal cost Per capita fed. cost Total cost Per capita total cost ∆ after tax cred. withdrawal -43,000 -177,000 23,000 -3,924,000 1,374,000 -207,000 -92,000 0 -2,166,000 -201 158,000 15 -2,008,000 -186 ∆ after tax ded. withdrawal 3,155,000 -127,000 140,000 190,000 992,000 -190,000 -84,000 -2,409,000 -910,000 -84 -2,931,000 -271 -3,841,000 -355 ∆ after ECC grant withdrawal -1,837,000 -417,000 1,155,000 27,256,000 2,121,000 -314,000 -47 000 -52,869,000 -22,760,000 -2,795 3,134,000 385 -19,626,000 2,410 ∆ withdrawal of all measures 225,000 -2,881,000 2,231,000 664,000 16,930,000 -2,554,000 -1,147,000 -52,869,000 -29,841,000 -2,763 3,152,000 291 -26,689,000 -2,471 Table 13: Change in government expenditures (net of effects on taxes and transfers) of alternative child care support measures for married women Variables ∆ family allowance ∆ univ. benefit ∆ $10 per day fees Federal tax 4,622,000 18,587,000 -7,016,000 Provincial tax 11,773,000 22,497,000 -1,717,000 Federal transfers -3,813,000 -12,491,000 3,809,000 Provincial transfers 12,206,000 -6,318,000 363,000 Social-assistance -1,415,000 -5,514,000 344,000 Retirement pen. cont. -35,000 -279,000 -512,000 Employment Insurance -6,000 -98,000 -244,000 Subsidies 79,306,000 714,000 -58,592,000 Universal Benefit 169,400,000 Net Provincial Cost 78,358,000 -33,336,000 -55,656,000 Per Capita Prov. Cost 6,780 -308 -514 Net Federal Cost -8,429,000 138,381,000 11,069,000 Per Capita Fed. Cost -729 1,278 102 Total Cost 69,929,000 105,045,000 -44,586,000 Per Capita Total Cost 6,051 970 -412 19 Table 14: Change in government expenditures (net of effects on taxes and transfers) of alternative child care support measures for single mothers Variables ∆ family allowance ∆ univ. benefit ∆ $10 per day fees Federal tax 77,000 1,281,000 -616,000 Provincial tax 6,774,000 1,441,000 -79,000 Federal transfers -586,000 -931,000 386,000 Provincial transfers 36,824,000 -1,047,000 -292,000 Social-assistance -24,670,000 -7,400,000 2,193,000 Retirement pen. cont. 747,000 249,000 -219,000 Employment insurance 377,000 118,000 -105,000 Subsidies 8,808,575 1,070,000 -7,441,000 Universal benefit 14,233,000 Net provincial cost 13,441,000 -9,068,000 -5,242,000 Per capita prov. cost 1,244 -840 -485 Net federal cost -1,040,000 11,902,000 1,108,000 Per capita fed. cost -96 1,102 103 Total cost 12,401,000 2,834,000 -4,134,000 Per capita total cost 1,148 262 -383 20