Working Moms and Child Care

By Heather Boushey and Joseph Wright[1]

 

May 5, 2004

 

This project was funded by the generous support of the Rockefeller Foundation and the Joyce Foundation. 

David Maduram, Eric Wingerter, and Marya Murray Díaz provided valuable research assistance for this report.

Executive Summary

 

Access to safe and affordable child care is critical for working mothers. Mothers who have stable child care are more likely to stay employed and are able to focus on their jobs, knowing that their children are well-cared for while they are at work.

Most mothers of young children work outside the home and most use child care. In 2001, the latest year with data available, over half of mothers of children under the age of six were employed—three-quarters of employed mothers worked 30 hours per week or more—and nearly all of this group—over 90 percent—reported using some kind of child care.

Child care, especially formal day care, which often provides more educational activities than other kinds of care, is expensive. Even though they spend less on child care on average, mothers in lower-income households spend a much higher share of their total income on child care than do higher-income households. In 2001, mothers who were in the bottom 40 percent in family income, who paid for formal daycare, spent an average of 18.4 percent of their total income on child care, compared to only 6.1 percent among mothers in the highest quintile.

Many families rely on informal child care arrangements. Among working mothers who use child care, about one-third rely on relatives, approximately the same number who use a formal day care setting. However, working mothers who use formal day care tend to be wealthier and better educated than other mothers, indicating that those who rely on informal care may be doing so out of necessity, not out of choice. Mothers in lower-income households use parental care more and are less likely to use formal day care centers, all else equal.

Child care assistance is critical for families struggling with the high cost of child care. Between 1997 and 2001, there was a significant increase in the percentage of working mothers receiving assistance with child care payments from all sources, including government assistance. Working mothers in the bottom 40^th percentile of households received more government child care assistance in 2001, compared to 1997. Even so, research has found that many children eligible for child care subsidies do not receive them. Only about 15 percent of children eligible for federal child care assistance actually receive any funds.

Child care is a problem for all working mothers. However, lower income mothers face the greatest difficulties in securing adequate care. Current legislative proposals are intended to partially address the problem. Senator Olympia Snowe has proposed to add $6 billion over the next five years in additional child care funding to the TANF reauthorization. This compares to a proposal in the House to add $2 billion over the next 5 years to the current $4.8 billion annual appropriation. Expressed as a share of the federal budget, Senator Snowe's proposal is equal to approximately 0.05 percent of projected federal spending over the next five years. The current federal subsidy is equal to approximately $475 for each child under age 15 living in poverty.

 

Kind of Child Care Used by Working Mothers

A generation ago, most mothers of young children spent their days at home with their children: in 1975, only two out of every five mothers with a child younger than six held a paid job. Today, nearly two-thirds of mothers with young children have jobs (Bureau of Labor Statistics 2002) and nearly three-out-of-four employed mothers work more than 30 hours per week. This change has made the search for safe, enriching, dependable and affordable child care a critical issue for millions of American mothers. Working mothers must find child care arrangements that meet their needs for quality, location, flexibility and affordability.

The distribution between different kinds of child care arrangements changed slightly over the period from 1997 to 2001. This time period covers the peak of the economic boom of the late 1990s, through the recession of 2001. Compared to 1997, more working mothers were using formal daycare in 2001 and fewer were using family daycare centers.[2]

Among working mothers, the majority (around 80 percent) report using one of six kinds of child care arrangements for their children under age six[3]:

1)   Parental care—care by the child’s parent, guardian, or stepparent, either at work or at home.

2)   Relative care—care by a relative of the child (including siblings 15 years or older), either in the child’s home or someplace else.

3)   Family daycare—care by a family daycare provider or by someone who is not a relative of the child away from the child’s home.

4)   Nanny or sitter care—care by someone who is not a relative of the child in the child’s own home.

5)   Formal daycare—care in a child or day care center, nursery or preschool, or Head Start program.

6)   Young sibling—child either cares for his or herself or is cared for by a sibling under age 15.

Working mothers with pre-school children (infants to age five) most commonly chose one of three kinds of care as their primary child care arrangement: care by a relative, formal daycare, or parental care (Figure 1). In 2001, one-third of working mothers who used child care used either relative care (33.3 percent) or formal daycare (28.9 percent) and one-fifth used parental care (20.2 percent). These figures report the proportion of mothers who use these arrangements for each of their children, thus one mother may be counted as many times as she has children under age six in child care. All arrangements that the mother used are included in the analysis.[4]

 

 

 

The kind of child care that a working mother uses is determined by a number of factors, including income and household composition. Table 1 shows how the kind of child care used differs across working mothers in different circumstances. In order to tease out the individual effects of various characteristics, the findings reported in Table 1 control for demographics and other characteristics of the mother and her household. The most important factors determining the kind of child care used are household composition, household income, and hours of work. To a lesser degree, the mother’s educational attainment and her race are associated with the kind of child care she uses for her children. (See Appendix for a complete description of our analysis.)

 

All else equal, single mothers who live with family or friends are often able to benefit from this situation by incorporating other household members into the daycare arrangements for their young children. Nearly half (46.6 percent) of single mothers living with family members have a relative as the primary care arrangement for their child.

Table 1: Differences in childcare arrangements, controlling for demographics, mother's hours of work, and household income
	Familial care							Non-familial care
	Parent		Relative		Sibling			Family		Nanny/Sitter		Formal
Year
1997	20.9%		33.1%		1.5%			16.7%		3.5%		24.4%
1999	21.6		32.7		0.8			16.0		3.0		26.1
2001	20.2		33.3		0.9			13.7	**	3.1		28.9	**
Mother's education level
Less than high school	23.8		36.3		1.0			13.2		5.0		20.8
High school grad	19.6		37.8		1.0			14.3		1.4	***	25.8	*
Some college	13.9		32.2		0.2			17.6		1.2	**	34.9	***
College degree	18.5		22.0		1.2			22.3	*	4.3		31.8	**
Mother's race/ethnicity
White	20.7		32.7		0.5			16.5		2.9		26.7
African American	15.7	***	33.1		1.7	**		12.8	***	1.4		35.3	*
Hispanic	19.6	*	42.6		0.5			15.3	*	2.8	**	19.3	***
Other	24.8		36.6		na			13.2		4.4		21.1
Household composition
Married couple	26.5		30.4		0.4			15.7		2.3		24.7
Cohabitating couple	19.9	*	34.8		2.1	**		12.4		3.5		27.3
Single parent living    alone	9.4	***	33.8	*	0.6			18.3		3.6		34.4	***
Single parent living in family	10.5	***	46.6	***	1.3	**		13.1		2.3		26.2
Single parent living with other adults	18.7	*	30.4		3.2	***		15.3		6.5	***	25.9
Household income
Bottom 20%	26.9		36.7		0.9			13.6		3.0		18.9
20-40%	24.7		37.0		1.1			13.0		3.0		21.2
40-60%	21.5		33.2		0.8			16.1		3.4		25.2	**
60-80%	17.7	***	34.2		0.4			18.3	*	1.6		27.8	***
Top 20%	16.8	***	32.2	*	0.6			14.7		3.1		32.5	***
Mother's hours worked per week
0-20	29.9		36.8		1.1			9.7		3.2		19.3
20-29	25.8	*	38.6		0.5			12.8		3.5		18.8
30-39	20.7	***	31.8		0.8			18.0	***	1.7		27.1	**
40-49	17.1	***	33.8	*	0.6			16.7	***	2.6		29.3	***
50 or more	22.0	**	29.0	**	1.4			14.4		4.5		28.7	**
p> |z|  = 0.10 *; p> |z|  = 0.05 **; p> |z|  = 0.01 ***
Source: CEPR analysis of the Survey of Income and Program Participation, 1996 and 2001 panels
Notes: Data is pooled across topical modules 4 and 10 from the 1996 panel and topical module 4 from the 2001 panel, representing data from the calendar years 1997, 1999, and 2001. Predicted eans are normalized so as to sum to 100 percent; non-normalized sum to greater than 100 percent because some mothers report childcare arrangements for more than one child. Full regression results are available in Appendix along with a discussion of the predicted values simulations. Tests of significance were conducted using the first category of each control variable as the omitted group.

 

 

Similarly, 6.5 percent of single mothers who live in a household with other adults (who could be friends, partners, or housemates) have someone caring for their children in the child’s home. Married mothers are more likely than other household types to have one of the two parents providing the primary care for the child (26.4 percent) and single mothers who live alone are most likely to use formal care (34.4 percent).

Income is also a predictor of what kind of child care a working mother uses, once we control for other characteristics of the mother. Working mothers living in high-income households are most likely to use formal daycare for their children, probably because they have sufficient income to afford this kind of care. After adjusting for other factors, working mothers in the top 20 percent of households are 13.6 percentage points more likely to use formal care than those in the bottom 20 percent. Household income is also important in who uses parental care: working mothers in higher income families are less likely to use parental care than are lower income mothers.

 

Mothers with short workweeks most often use parental care, all else equal. Mothers who work less than 30 hours per week—one-quarter of working mothers—are more likely than mothers with longer workweeks to have a parent as the primary caretaker. Longer workweeks are also associated with a greater likelihood of using formal daycare. Formal daycare may provide more reliable hours of care for mothers who need to be at work for long hours each day.

 

Educational attainment and race are also factors important in determining what kind of child care a mother will use, although these appear to be less important than other characteristics. Working mothers with more education are more likely to use formal care; they are also less likely to use relative care and parental care, but these differences are not statistically significant. African American and Hispanic mothers are less likely than white mothers to have a parent caring for their child; they are also less likely to use a family daycare center. Hispanic mothers are least likely to use formal daycare centers.

 

Cost of Child Care

The kind of child care that working mothers use is based on the preferences of the mother, as well as her ability to pay for various kinds of care. Child care is expensive: on average, in 2001, a working mother using formal daycare paid $92.30 per week per child, which adds up to an annual cost of $4,615 in 2002 dollars (this calculation assumes two weeks off for vacation—although many low-income mothers do not get vacations ) (Figure 2). Nearly all mothers using formal or family daycare paid for it and, in 2001, on average, this payment took up 9.0 percent of family income for formal daycare and 7.4 percent for family daycare. Working mothers are less likely to pay for relative care, but when they do, it can be a substantial burden: in 2001, on average, costs were $66.20 per week, or $3,310 for a 50-week year. From 1997 to 2001, there was no significant change in how much mothers paid for child care in inflation-adjusted dollars.

 

The expense of formal daycare explains why mothers in higher income households are more likely to use this kind of care (Table 2). On average, higher-income mothers pay more for child care:  over the period from 1997 to 2001, on average, higher-income mothers paid $110.57

per week per child  for formal child care, compared to lower income mothers who paid $63.32. For family daycare, higher income mothers pay $75.19 per week, while low-income mothers pay only $42.82. 

 

Child care expenses comprise a larger share of household income for lower-income mothers, compared to higher-income mothers. Among working mothers in the bottom 40 percent of households, those who used formal care paid 18.4 percent of their total household income towards their child care. Working mothers in the top 20 percent of households, however, spent only 6.1 percent of their total household income on child care (Figure 2). The high share of total household income spent by low-income mothers on child care is consistent with analysis of the costs of child care relative to other goods and services in a basic family budget. These budgets detail family budgets for low-income families, using low-cost goods and services. They find that, on average, parents (both single parent and married-couple parents) with one child must spend about one-fifth of their family budget on child care, about the same as the share actually spent by low-income households on child care (18.4 percent) (Boushey et al. 2001) .

Table 2: Average payment for child care  for children under age six and share of household income
(2002 dollars)
	Relative Care			Family Daycare			Formal Daycare
	Average payment per week per child	Payment share of household income		Average payment per week per child	Payment share of household income		Average payment per week per child	Payment share of household income
Year
1997	$53.70	7.6%		$58.86	7.7%		$88.70	10.4%
1999	55.08	6.7		62.39	6.8		82.60	9.2
2001	66.18	7.1		56.12	7.4		92.33	9.0
Mother's education level
Less than high school	39.80	6.4		49.91	7.6		57.20	13.6
High school grad	48.09	7.6		47.66	7.7		74.85	10.8
Some college	62.66	7.2		56.95	6.6		81.04	9.4
College degree	72.04	6.7		74.64	7.5		107.27	8.2
Mother's race/ethnicity
White	55.53	6.4		59.05	7.1		88.60	9.1
African American	52.59	8.2		54.59	7.5		70.29	9.5
Hispanic	65.89	7.9		59.90	7.8		97.95	13.3
Other	87.91	7.1		86.61	9.6		120.19	8.6
Household composition
Married couple	61.81	6.5		62.66	6.5		95.15	8.4
Cohabitating couple	54.61	8.8		33.91	5.0		79.53	9.0
Single parent living alone	61.58	11.4		49.09	13.9		72.33	16.1
Single parent living in family	52.41	5.5		50.97	5.6		71.38	7.1
Single parent living with          other adults	37.33	3.9		69.18	7.6		91.91	12.4
Household income
Bottom 40%	43.07	11.8		42.82	12.7		63.32	18.4
40-60%	51.30	7.5		51.07	7.2		76.35	11.3
60-80%	59.89	5.8		64.78	6.4		80.73	7.7
Top 20%	78.75	4.5		75.19	4.4		110.57	6.1
Source: CEPR analysis of the Survey of Income and Program Participation, 1996 and 2001 panels
Notes: Data is pooled across topical modules 4 and 10 from the 1996 panel and topical module 4 from the 2001 panel, representing data from the calendar years 1997, 1999, and 2001. Data for nanny care were dropped due to insufficient observations.

 

The high cost of child care is prohibitive for many families. The Children’s Defense Fund has found that in 14 states, the average annual cost of child care is more than double tuition at a state university (Ewen and Hart 2003) . High costs come at a time when families can least afford it—when the parents are young, starting out in their careers, and when they are also saving up for major life purchases such as a home and paying off student debt. Further, infant care is more expensive than toddler care (Schulman 2000) , so that the numbers reported in Table 2 underestimate costs for mothers of infants and toddlers.

 

Child care assistance

 

More mothers received more help paying for child care in 2001, compared to earlier years. Overall, the share of mothers who paid for child care and received any help with those payments rose from 11.3 percent in 1997 to 16.0 percent in 2001, a statistically significant change. Some of this increased assistance was from non-governmental sources—friends, fathers, relatives—and some was from the government or employers (Table 3).

 

There was a substantial increase in child care assistance through Temporary Assistance to Needy Families (TANF) and the Child Care Development Fund (CCDF) after welfare reform was implemented in 1996. The rules regarding who gets child care assistance and what kinds of care quality are set at the state level, but usually child care assistance goes to formal daycare settings, rather than to pay relatives to care for children. Figure 3 shows that overall, the change in the share receiving government assistance was not statistically significant. However, the changes in the share of working mothers with children in formal daycare receiving government assistance did increase significantly, from 2.9 percent in 1997 to 5.7 percent in 1999, and 6.5 percent in 2001, indicating that new federal child care monies spent after welfare reform in 1996 were reaching more children.

 

More mothers would likely use formal care if they could afford it or if they were provided assistance to pay the high costs of this kind of care. Studies consistently show that formal care is the most reliable and provides the highest quality care, and that mothers report wanting to use it. However, the high costs of care put it out of reach for many low-income mothers. Subsidies provided by the government do not do enough to help most low-income mothers who need help paying for child care. A report by the Department of Health and Human Services found that only 15 percent of children eligible for federal funds for child care assistance received any aid in 1999 (U.S. Department of Health and Human Services 1999) . Thus, most mothers had to find alternative ways of acquiring child care for their children while they are at work, be it looking to relatives, other informal arrangements, or finding more affordable formal daycare, if possible.

 

Table 3: Share of working mothers receiving help to pay for child care for children under age six
	Received help paying for child care
	Relative Care			Family Daycare			Formal Daycare
	Any help	Government help		Any help	Government help		Any help	Government help
Year
1997	6.3%	0.0%		6.1%	3.6%		11.3%	2.9%
1999	2.5	1.0		5.1	2.4		10.5	5.7
2001	5.4	0.8		10.0	3.7		16.0	6.5
Mother's education level
Less than high school	9.9	0.0		7.8	6.0		9.9	7.6
High school grad	3.6	0.0		10.8	6.2		12.4	6.5
Some college	4.1	0.9		7.5	2.4		17.4	7.7
College degree	4.6	1.1		3.8	1.6		8.1	1.3
Mother's race/ethnicity
White	4.5	0.6		6.5	3.5		11.3	3.4
African American	6.0	1.5		7.2	2.9		21.0	15.8
Hispanic	4.8	0.0		3.5	0.0		12.3	3.6
Other	0.0	0.0		19.4	5.7		8.9	0.0
Household composition
Married couple	2.5	0.0		3.9	3.9		6.9	1.4
Cohabitating couple	0.0	0.0		26.3	26.3		15.9	6.0
Single parent living alone	7.6	2.2		13.1	13.1		25.1	11.4
Single parent living in   family	9.0	1.7		9.2	9.2		25.1	16.5
Single parent living with other adults	11.3	0.0		15.9	15.9		21.0	7.1
Household income
Bottom 40%	4.3	3.1		16.6	11.6		25.5	15.4
40-60%	3.4	0.0		7.3	2.0		7.9	2.2
60-80%	4.5	0.0		4.8	0.9		12.7	5.5
Top 20%	6.6	0.0		3.6	2.0		9.8	2.1
Source: CEPR analysis of the Survey of Income and Program Participation, 1996 and 2001 panels
Notes: Data is pooled across topical modules 4 and 10 from the 1996 panel and topical module 4 from the 2001 panel, representing data from the calendar years 1997, 1999, and 2001. Means for nanny care were dropped due to insufficient observations.

 

Conclusions

The “typical” American family with young children is one in which the mom is at work and the children are in child care. For most mothers, this means added expenses as well as the daily hassles of transporting children to and from their child care locations. Most mothers use some type of “familial” care (parental care or relative care) as the primary source of care for their children. These arrangements are almost as reliable as formal daycare and are less expensive to use. Low-income mothers are less likely than higher-income mothers to use formal daycare, indicating that the cost barriers may be prohibitive to using this kind of care even though it might be preferable for other reasons, such as the kind of educational activities some formal daycare centers offer.

Slightly less than one-third of mothers put their young children under age six in formal care. While formal care is generally of higher quality and is one of the most reliable forms of care, it is also the most expensive. Moderate and lower income families generally cannot afford such care without assistance. Mothers who cannot rely on family support or afford formal care must rely on informal child care arrangements (family daycare or nanny/sitters[5]). These child care arrangements are often of questionable quality and are the least reliable, often forcing mothers to miss work (Fuller et al. 2001) .

The movement to formal care occurs as mothers’ hours and commitment to staying employed increases, and as income increases. As women become more entrenched in the labor market, they often move from informal to center-based care (Fuller et al. 2001) . Some of this is because their incomes rise, but some of it is because formal settings are more reliable in the long-run. The use of formal child care is associated with increased employment durations for mothers, as is receiving assistance in paying for child care (Boushey 2002) . Therefore, in the future it is likely that there will be a continuing shift from informal and familial care settings to formal care.

One-third of mothers rely on relatives to provide care for their children. The high use of relatives to provide child care among poorer households indicates that this kind of care may be more of a necessity than a choice. This may be in part a result of welfare reform, which pushed low-income mothers into paid employment. High reliance on relative child care may also be partially attributable to the fact that child care costs have risen faster than wages. However, relative care may not be a stable child care arrangement over time. Relatives often have to find regular employment themselves. It cannot be assumed that relative care will meet the child care needs of all households.

As the need for formal child care rises, and as costs rise, little is being done in Washington at present to help mothers find and pay for quality child care. Many states have had to cut back on child care programs over the past few years as they have dealt with serious budget constraints. Most of the programs the federal government has established are not funded at sufficient levels and this problem has been further exacerbated by the states’ fiscal crisis, which has led to cutbacks in child care assistance in many states (Johnson, Lav, and Ribeiro 2003) . Recent analysis has found that at least 13 states decreased their state investment in child care assistance in 2002 (Ewen and Hart 2003) . As a result, working mothers are likely to continue to face serious obstacles to obtaining high quality, reliable, and affordable child care.

Appendix I: Data and Method

 

Data

 

The data for this analysis come from the 1996 and 2001 panels of the Survey of Income and Program Participation. The information on child care usage comes from topical modules four and ten in the 1996 panel and topical module four in the 2001 panel. The topical modules cover March to June of 1997, March to June, 1999, and for 2001, October to December, 2001 and January 2002. Family income and work participation data come from the longitudinal waves that correspond to these topical modules. Child care is evaluated for all children in the mothers’ subfamily, including her own children (adopted or biological) as well as any children for whom she is the guardian (foster children, step-children, or other children). In general, child care questions were asked of the mother if there was a child in the family; however, about three percent of respondents are men. These observations were not included in this analysis.

 

Child care type covers the first five children in the household under age six. Overall, less than one-third of children are in more than one kind of child care arrangement each week. For this report, we include all kinds of child care that the mother uses for her children. Because any one mother can have more than one child, and therefore use more than one kind of child care, the data in Tables 1 and 2 were normalized to 100 percent.

 

The cost of child care is the amount paid per week for the primary child care arrangement across all the mother’s children under age six who have that kind of care. The values are all in 2002 dollars and averages exclude mothers who did not pay for care.

 

Not all working mothers were asked the child care questions and this share changed over the three years. Table A1 shows, however, that this does not appear to have affected the distribution of the sample. In 1997, 91.1 percent of working mothers were asked child care questions, but only 88.6 percent were in 2001, a statistically significant difference. However, the distribution of mothers across educational attainment, race, and household income does not change significantly, leading us to conclude that this smaller sample will not bias our findings.

 

Appendix Table 1: Samples
		Total				Sample 1
	Employed adult mothers in topical modules				Share of employed adult mothers answering child care questions
	1997	1999	2001		1997	1999	2001
Observations in sample	2,465	2,665	2,510		2,245	2,388	2,224
Sample as percent of total	100.0%	100.0%	100.0%		91.1%	89.6%*	88.6%
Distributions within sample
Mother's education level
Less than high school	9.9	10.7	10.4		8.9	10.4	9.4
High school grad	31.4	30.1	28.6		30.6	28.9	28.1
Some college	34.9	32.5	35.2		35.4	32.9	35.7
College degree	23.9	26.7	25.8		25.2	27.9	26.8
Mother's race/ethnicity
White	64.9	65.9	63.2		66.6	67.3	64.1
African American	17.2	15.6	16.2		16.8	14.8	15.7
Hispanic	13.5	14.2	15.7		12.5	13.6	15.2
Other	4.4	4.4	5.0		4.2	4.3	5.0
Household income
Bottom 20%	10.4	7.2	9.8		10.6	7.1	9.9
20-40%	16.0	14.9	13.9		15.5	14.9	14.1
40-60%	23.0	19.6	20.7		22.7	19.4	20.9
60-80%	26.3	27.3	27.9		26.7	28.0	27.3
Top 20%	24.5	31.0	27.7		24.6	30.7	27.9
		Sample 2			Sample 3
	Share of employed adult mothers using any kind of child care				Share of employed adult mothers answering childcare questions and using any child care
Observations in sample	2,171	2,447	2,102		2,094	2,287	2,033
Sample as percent of total	88.1%	91.8%***	83.7%***		84.9%	85.8%	81.0%***
Distributions within sample
Mother's education level
Less than high school	8.9	10.3	9.6		8.6	10.1	9.0
High school grad	30.7	29.6	27.7		30.4	29.0	27.6
Some college	35.3	32.6	36.3		35.4	32.7	36.4
College degree	25.1	27.5	26.5		25.6	28.2	27.1
Mother's race/ethnicity
White	65.5	66.4	63.7		66.6	67.5	64.2
African American	17.0	15.6	16.3		16.6	15.0	15.9
Hispanic	12.9	13.7	14.9		12.4	13.3	14.8
Other	4.7	4.3	5.1		4.4	4.2	5.1
Household income
Bottom 20%	9.8	6.8	9.5		10.0	6.9	9.7
20-40%	15.7	14.7	13.7		15.6	14.7	13.9
40-60%	22.7	19.4	20.8		22.5	19.4	20.9
60-80%	26.4	27.8	27.8		26.7	28.1	27.3
Top 20%	25.4	31.2	28.3		25.3	30.9	28.2
p> |z|  = 0.10  *; p> |z|  = 0.001  ***
Source: CEPR analysis of the Survey of Income and Program Participation, 1996 and 2001 panels
Note: Sample 3 is used in the report. Information for the calendar year 1997 is from topical module 4 of the 1996 SIPP panel; 1999 is from topical module 10 (1996 panel); and 2001 is from topical module 4 (2001 panel). Fathers make up between 3 to 4 percent of employed adults with children under six answering child care questions and using some type of child care arrangement; less than 4 percent of employed men with children less than six years of age answered the childcare questions (2.76 percent in 1997, 3.50 percent in 1999, and 3.31 percent in 2001). Fathers are excluded from our analysis, however.

 

Method

 The values reported in Table 1 are from logit regressions, where the kind of child care used is the dependent variable. The model includes demographic characteristics and income and employment-related characteristics of the mother. The variables included are (with omitted variables in parentheses): age (18-25 year olds), educational attainment (less than high school grad), race (white), hours worked (less than 20 per week), household status (married), household income (bottom 20 percent) and year (1997). Table A2 shows the regression results for all six logit models.

 

After we estimated the logit regression models, we estimate the predicted probability of having a particular type of childcare arrangement shown in Table 1 by conducting simulations that calculate a distribution of expected values of the predicted probabilities. For example, we calculate 1,000 expected values of the probability of having formal care as first differences (different from the omitted variable). We set all other explanatory variables at their mean value. This simulation provides us with a substantively meaningful assessment, using controls, of the effect of certain individual characteristics on the probability of having a particular type of childcare. In the model for sibling care, we drop Hispanic because it perfectly predicts a zero outcome.

 

 

Appendix Table 2: Logit regression results
	Familial care				Non-familial care
Dependent variable: Kind of child care	Parent	Relative	Sibling		Family	Nanny/Sitter	Formal
Mother's age (18-25 omitted)
25-34 years old	0.049	-0.231	1.444		-0.105	-0.36	0.189
	(0.39)	(2.17)*	(2.60)**		(0.78)	(1.27)	(1.64)
35-44 years old	0.093	-0.485	2.119		-0.059	-0.039	0.38
	(0.60)	(3.51)**	(3.60)**		(0.35)	(0.12)	(2.66)**
45-54 years old	-0.184	-0.411	1.698		-0.453	-0.301	1.115
	(0.49)	(1.27)	(1.49)		(1.06)	(0.39)	(3.47)**
Mother's educational level (less than high-school omitted)
High school grad	-0.218	0.119	0.14		0.138	-1.26	0.34
	(1.10)	(0.70)	(0.21)		(0.60)	(3.20)**	(1.69)
Some college	-0.252	0.023	-1.099		0.285	-0.705	0.524
	(1.29)	(0.14)	(1.46)		(1.26)	(1.99)*	(2.63)**
College degree	-0.05	-0.269	0.022		0.462	-0.03	0.459
	(0.24)	(1.42)	(0.03)		(1.89)	(0.07)	(2.13)*
Mother's race/ethnicity (white omitted)
African American	-0.527	-0.198	1.002		-0.481	-0.916	0.243
	(3.19)**	(1.57)	(2.38)*		(2.90)**	(2.21)*	(1.92)
Hispanic	-0.271	0.213	-0.551		-0.271	-0.233	-0.692
	(1.75)	(1.62)	(0.71)		(1.63)	(0.68)	(4.50)**
Other	0.297	0.258	NA		-0.268	0.426	-0.323
	(1.50)	(1.37)			(1.15)	(1.14)	(1.63)
Mother's hours worked per week (0-20 omitted)
20-30 hours	-0.328	-0.023	-0.856		0.263	0.043	-0.106
	(1.72)	(0.12)	(1.02)		(1.03)	(0.11)	(0.50)
30-40 hours	-0.574	-0.277	-0.332		0.761	-0.696	0.513
	(3.09)**	(1.56)	(0.46)		(3.21)**	(1.57)	(2.62)**
40-50 hours	-0.9	-0.268	-0.686		0.601	-0.263	0.537
	(5.52)**	(1.70)	(1.04)		(2.75)**	(0.75)	(3.05)**
50 or more hours	-0.541	-0.509	0.141		0.415	0.288	0.527
	(2.29)*	(2.20)*	(0.18)		(1.42	(0.64)	(2.23)*
Household arrangement (married couple omitted)
Cohabitating couple	-0.434	0.204	1.568		-0.321	0.362	0.103
	(1.88)	(1.01)	(2.27)*		(1.18)	(0.73)	(0.47)
Single parent living alone	-1.306	0.254	0.312		0.239	0.479	0.598
	(6.84)**	(1.72)	(0.50)		(1.33)	(1.43)	(3.79)**
Single parent living in family	-1.221	0.843	1.049		-0.232	-0.042	0.081
	(7.01)**	(6.50)**	(2.03)*		(1.39)	(0.11)	(0.59)
Single parent living with other adults	-0.456	0.073	2.151		0.029	1.139	0.119
	(1.83)	(0.33)	(3.44)**		(0.11)	(2.85)**	(0.51)
Household income (0-20% omitted)
20-40%	-0.266	-0.143	0.256		-0.147	-0.051	0.047
	(1.28)	(0.81)	(0.39)		(0.63)	(0.12)	(0.24)
40-60%	-0.32	-0.173	0.074		0.231	0.183	0.425
	(1.58)	(0.97)	(-0.11)		(1.03)	(0.45)	(2.14)*
60-80%	-0.577	-0.108	-0.507		0.411	-0.614	0.581
	(2.79)**	(0.6)	(0.69)		(1.80)	(1.34)	(2.89)**
Top 20%	-0.729	-0.322	-0.132		0.051	0.056	0.738
	(3.40)**	(1.71)	(0.18)		(0.21)	(0.13)	(3.56)**
Year (1997 omitted)
1999	0.068	-0.03	-0.851		-0.098	-0.241	0.034
	(0.60)	(0.29)	(1.97)*		(0.82)	(0.99)	(0.33)
2001	-0.058	-0.056	-0.623		-0.294	-0.232	0.196
	(0.50)	(0.55)	(1.46)		(2.35)*	(0.95)	(1.84)
Constant	0.724	0.242	-5.011		-1.995	-2.031	-2.227
	(2.60)**	-0.97	(4.74)**		(5.90)**	(3.84)**	(7.59)**
Observations	2,525	2,525	2,525		2,525	2,525	2,525
p> |z| = 0.10 *; p> |z| = 0.001 ***; z-statistic in parenthesis.
Source: CEPR analysis of the Survey of Income and Program Participation, 1996 and 2001 panels
Notes: Data is pooled across topical modules 4 and 10 from the 1996 panel and topical module 4 from the 2001 panel, representing data from the calendar years 1997, 1999, and 2001.