Parents' Child Care Preferences: Patterns among Welfare Mothers


U.S. Department of Health and Human Services

Parents' Child Care Preferences: Patterns among Welfare Mothers

Freya L. Sonenstein and Douglas A. Wolf

The Urban Institute

May 1990

PDF Version: (  PDF pages)

This report was prepared under contract between the U.S. Department of Health and Human Services (HHS), Office of Family, Community and Long-Term Care Policy (now known as the Office of Disability, Aging and Long-Term Care Policy) and The Urban Institute; portions of which were presented at the Annual Meeting of the Population Association, April 21-13, 1988. Additional funding was provided by the Rockefeller Foundation. For additional information about this subject, you can visit the ASPE home page at or contact them at

The opinions and views expressed in this report are those of the authors. They do not necessarily reflect the views of the Department of Health and Human Services, the contractor or any other funding organization.


TABLE 1: Activities Throughout 14 Months by Age of Youngest Child
TABLE 2: Main Child Care Arrangement for Youngest Child by Age
TABLE 3: Main Child Care Arrangement for Youngest Child 0-5 Years by Activity
TABLE 4: Variable Definitions
TABLE 5: Mother's Mean Rating of Care by Type of Arrangements
TABLE 6: Determinants of Mother's Satisfaction with Child Care


The proportion of American mothers with preschool children who are employed has more than tripled since World War II. By 1988, 56 percent of U.S. women with children under the age of 6 were employed. (Zill, 1989). This rise in mothers' rates of employment has led to a watershed in public opinion and policy. Increasingly, the role of the welfare system crafted in the 1930's to assist mothers to stay home with their children has come to be questioned as the majority of mothers have moved into the labor force. Work incentives have been part of the program since the 1960's, but now the system has been reformed by the Family Support Act of 1988 so that the primary goal of the program is to assist AFDC parents, usually female single parents, to become economically self sufficient.

Chief components of the new AFDC program (Aid for Families with Dependent Children) are the required participation of mothers with children over 2 years of age in employment programs and the-guarantee of child care for AFDC mothers who participate in the newly established Job Opportunities and Basic Skills program or other approved education and training activities as well as to mothers who go to work. States have the option to reduce the age cutoff so that mothers with children over 1 year: of age are required to participate in employment programs. A further provision of the Act is that all parents who exit from the AFDC program because of earnings are guaranteed 12 months of transitional child care assistance on a sliding scale fee basis. For many states these new requirements have lead to large scale planning efforts as they have geared up to expand child care opportunities for welfare families.

A fundamental precept of the regulations governing the new child care guarantees is that parents reserve the right to chose their own child care arrangements (Federal Register, 1989). Yet very little is known about the child care choices of AFDC mothers. The last time data were collected about child care of recipients was in 1979 when the AFDC Recipient Survey found that 60 percent of AFDC mothers employed full time did not use available child care subsidies (Hofferth and Sonenstein, 1983). Consequently, States efforts to expand child care options for AFDC parents have proceeded in the absence of any information about what these parents would prefer.

This paper addresses this issue by focusing on the child care arrangements made a by sample of AFDC mothers over a 14 month period starting in September 1, 1983. We have limited our discussion to mothers with children under the age of 6 for two reasons. The child care needs and options for preschool-age children, including those in half day kindergarten programs, typically differ from children who are in school (grades 1-6). Also until recently, AFDC mothers with children under the age of 6 were exempt from registering for the WIN program. A key element of welfare reform is the expectation that mothers with younger children work. In light of this policy shift, information about the current care arrangements made by AFDC mothers with children under the age of 6 becomes critical to projections about the types of child care needed by these mothers.

First we describe the most recent care arrangements used by mothers for their youngest children. We also examine how these child care arrangements vary in terms of objective characteristics (cost and hours of the day) and the mother's subjective assessment of the quality of the arrangement measured along a number of dimensions. Then we examine how these dimensions of quality contribute mother's overall satisfaction with her child care arrangement. Finally the policy implications of these findings for children on welfare are explored.


A major problem regarding child care for welfare mothers is the general lack of information about the types of child care arrangements they use. The Child Care and Self Sufficiency Study was initiated to fill these gaps in knowledge. Its objectives were to describe the child care arrangements made by AFDC mothers who work and to examine whether certain types of arrangements and subsidies enhance the ability of welfare mothers to get off welfare and stay off welfare.

The sample consisted of 554 AFDC mothers with children under age 10 randomly drawn from welfare rolls in Boston, Charlotte and Denver. The sample was stratified so that half the mothers had earnings in August 1983 and half did not. Since approximately 5 percent of the AFDC caseload was employed at any single point in time in 1982 (Weder, n.d.), working recipients were heavily oversampled. However sampling weights are applied to make the results descriptive of all mothers with children under 10 on the AFDC caseloads in these 3 cities. Our analyses in this paper are limited to the 382 mothers in the sample with children under the age of six; 296 of these mothers reported using child care at some time in the 14 month period.

Two sets of interviews were conducted with these mothers. The first set took place 8 months after the sample was selected in August 1983, the second was conducted 6 months after that. In the two interviews retrospective information was gathered in a life events format about the respondent's employment, training, school participation, job search and child care arrangements across the 14-month period. In addition, AFDC case record data were abstracted for the 14-month period.


Over the 14 month period 63 percent of the sample reported that they used a child care arrangement for their youngest child at least once. To obtain this information mothers were first asked about any episodes of working, looking for work, or going to school or a training program during this period. If they had participated in any of these activities they were then asked whether anyone looked after their youngest child while they engaged in them. Our definition of child care is therefore tied to arrangements made for the youngest child while the mother worked, went to school or a training program, or looked for work; babysitting for other activities that the mother might engage in is not included.

Over the course of 14 months 43 percent of mothers obtained employment for some period of time. (See Table 1) Another group--about 12 percent--sought employment but not find work. Twenty-four percent participated in training programs or went to school. Only 36 percent of the mothers with children under age 6 did not engage in a job, job readying or job search activities. Since these data were gathered before states began to mount their "welfare to work" demonstrations in earnest, they indicate the amount of effort mothers on welfare in the 3 cities were making without specialized programs.

Work effort did not vary by the age of the youngest child. Mothers with infants were just as likely to work as mothers with older preschool children. Interestingly, mothers of infants were more likely to look for work without success compared to mothers with older children. Twenty-one percent looked for work but found nothing compared to 8 percent of mothers with slightly older children. This may imply that these mothers are more motivated to enter the labor force but may be encountering barriers in terms of arranging care for infants.

Table 2 shows the type of arrangement used for the youngest child. This table is a picture of the most recent main child care arrangement reported by the mother, usually in the first interview. By main child care arrangement we mean the one used for the most number of hours. The predominant form of care is care by relatives. Over half the infants and toddlers and over 40 percent of 4-5 year-olds were cared for by relatives, primarily grandparents. Formal care in centers, early childhood programs and schools is not common for children under age 2, but its use increases with the age of the child. Ten percent of children under 2 years of age are in day care centers, but by ages 4 and 5 one-quarter are in centers as a main arrangement. Head Start is not utilized much as a main care arrangement for these children on welfare even though they would be a prime target group for the program.

Relatives are the predominant caregivers for preschool children no matter what the mother's activity is--employment, looking for work or school/training (see Table 3). For employed mothers, the most common form of care after relatives was family daycare homes, care outside the home by nonrelatives. Only 14 percent used group care, a category combining care in centers with nursery school or Head Start arrangements. For mothers in school and training, group care was just as common as relative care. Almost half of the youngest children of these mothers were in group care. Possibly, training and school programs that mothers are attending provide access to group care; access that is apparently not as available to working AFDC mothers.


The interview was designed to obtain the mother's assessment of her most recent main child care arrangement on a number of objective and subjective measures of its quality, convenience, and cost. In addition she was asked about her satisfaction with her child care arrangement. A full list of variables is available in Table 4. We have compared mother's ratings of their care arrangements to see whether certain types of arrangements are associated with better ratings. Table 5 compares different types of arrangements on these dimensions. Care arrangements have been grouped into five categories: relatives providing care in the child's home; relatives providing care outside the child's home; care in someone else's home who is not a relative-sometimes referred to as family day care; care by a nonrelative in the child's home--sometimes called in-home babysitting; and care in a group setting such as a center, a nursery school, a Head Start program or school. Care provided by the mother while she is employed have been excluded from these analyses. The table shows the mother's assessment of care varied significantly by the type of care that she used.1

Two objective indicators of the quality of care that are often used to measure the quality of child care for licensing purposes are the ratio o children to adults and whether or not the care giver has received specialized training in child development (Ruopp et al, 1979). These two indicators were cornerstones in the federal government's attempts to regulate child care in the seventies (Nelson, 1982). Mothers in our sample reported the number of children cared for together in their youngest child's most recent arrangement as well as the number of adults who were always present. Using these data we calculated a child to adult ratio for each youngest child's main arrangement. Respondents also indicated whether or not their child care provider had received special training in caring for children.

Group care had a significantly higher mean child to adult ratio, 5.28, than all the other forms of care. Care provided by relatives either in or outside the child's home had the lowest ratios at 1.58 and 1.65 respectively. Care by non-relatives either inside or outside the child's home had child to adult ratios that were slightly higher than relative care situations, at 2.56 and 2.62. By this criterion, group care had the poorest rating. However on the second criterion, child care training, care givers in group care were significantly more likely to have had special training than those in other settings.

In addition to reporting objective characteristics of their child care arrangements, the mothers were asked to assess other dimensions of its quality: her provider's experience in caring for children, her child's opportunities to learn new things, and the level of supervision, discipline and safety provided in her arrangement There were significant differences between the types of care on three of these dimensions. Mothers reported that their children had more frequent opportunities to learn new things in group care than in the other types of care. Care provided by nonrelatives either in or outside the child's home had the poorest rating on this dimension; care provided by relative fell in between. Related caregivers and caregivers in group settings were reported to have approximately the same levels of experience caring for children and to have significantly more experience than nonrelated caregivers. A similar pattern emerged for mothers' reports of the safety precautions taken by the caregiver to prevent accidents. Related caregivers and caregivers in group settings were rated as significantly more care u t an nonrelated caregivers. These data indicate that mothers in the sample rated the quality of care provided by relatives and group care as significantly better than care by nonrelatives on several dimensions.

Another dimension of the mother's assessment of her youngest child's care situation is her perception of the child's feelings. Mothers in the sample were asked to rate their youngest child's happiness with the care arrangement and their child's feelings about the person caring for them. On both these aspects of care, care provided by nonrelatives got significantly poorer scores than other types of arrangements. Children in family day care or babysitting arrangements were reported to be significantly less happy with the care situation and less loving towards the caretaking person than children in group care or cared for by relatives.

Care arrangements also varied significantly in terms of the mothers' perceptions of their convenience of hours and location. Care by relatives outside the child's home was consistently rated less convenient in terms of hours and location than other forms of care. Care by nonrelatives out of the home came in second from the bottom on convenience of hours and location. Related to this assessment of convenience ate the actual number of hours and the time of day that care is required. Youngest children spent significantly more hours per week in group programs and in family day care arrangements than they spent in relative care or in non-relative in-home care. On average children were in group care or family day care for 30-31 hours per week in contrast to 24-25 hours per week for the other arrangements.

Children cared for by relatives spent significantly more hours in care after 6:00 pm and before 7:00 am on weekdays than children in other arrangements. The average hours per week was 8.7 for relative out-of-home care, 6.4 for in-home relative care, 5.2 for out-of-home non-relative care, 3.6 for non-relative in-home care and 2.2 for group care. Indeed a substantial proportion of children in all forms of care except group arrangements required care outside of business hours. While only 21 percent of children whose main arrangement was a center or school needed evening or early morning care, the range was 40-56 percent of children in the other forms of child care. There were no significant differences between the types of care on weekend care requirements.

The dependability of the child care arrangement can also be very important to the AFDC mother who is trying to maintain a job or participate regularly in an education or training program. The mothers were asked two questions about the number of days lost from work, school or training in the last 8 months because (1) the youngest child was sick or (2) the regular child care arrangement was not available. Mothers using group care reported a significantly higher number of days lost because of the child's sickness than mothers using other forms of care. On average mothers with children in group care lost 5.6 days over the 8 month period compared to between 1.3 and 2.6 days lost when children were in other forms of care. Mothers with children in nonrelative in-home care lost the fewest days because of a child's illness. There were no significant differences between the different kinds of arrangements on days lost because the arrangement was not available. All forms of care appeared equally reliable.

The cost of care can be another component of the mother's assessment of her youngest child's care. However, the proportion paying for care varied widely by the type of care. Only 28 percent of the care provided by relatives in the child's home was paid for compared to 78 percent of the care provided outside the home by a nonrelative. In between, 46 percent of group care, 60 percent of nonrelative care inside the home and 44 percent of relative care outside the home involved an out-of-pocket cost. These data indicate that child care by relatives is not necessarily free care, especially if this care is provided outside the child's home. Neither does care by relatives necessarily carry a lower unit cost when it is paid for. Although in-home care by relatives is one of the cheapest form of care at $.74 an hour, out-of-home care by relatives costs more per hour on average ($.90) than out-of home care by non-relatives ($.75) or group care ($.69). For those paying for care the average weekly cost could be as high as $40.83 for nonrelative in-home care or $25.49 for nonrelative out-of home care and as low as $13.23 for group care. The average weekly costs of care by relatives fell in between these two poles at $19.03 for in-home care and $21.82 for out-of-home care. Mothers who paid for care were asked whether they considered the cost of their most child care arrangement very reasonable, reasonable, somewhat reasonable or too expensive. The care arrangements that cost the most per week, care by non-relatives in or out of the home, were rated significantly less reasonable than the other arrangement types.

The mothers were also asked two questions which provide measures of their satisfaction with care. One simply asked the mothers to rate their, satisfaction with their child care arrangement for their youngest child; the second asked mothers to indicate what type of child care arrangement would be their first choice for their youngest child if they could chooser any at all. Answers to this question asked early in the interview could be compared to the actual child care arrangement of the child to indicate whether the respondent was actually using an arrangement which was her first choice. The results of comparisons across care types on these two measures show parallel patterns. In terms of the satisfaction measure, mothers using group care were significantly more satisfied than mothers using other forms of care. Conversely nonrelative care either in or out of the home received the lowest satisfaction ratings. On the second measure, the data indicate that with the exception of mothers using group care, very few of the respondents had a child care arrangement that was their first choice. While over half (53%) of mothers using group care had an arrangement that was their first choice, only one third of mothers using in-home relative care, and one quarter of mothers using out-of-home relative care had their first choices. The proportions were even smaller for the users of non-relative care; it was 6 percent for out-of-home users and 10 percent for in-home users.


Mothers' ratings of types of care varied substantially on various dimensions of child care quality, but it was not clear which of these dimensions were most important to mothers. To examine this question, we analyzed which ratings contributed the most to mother's reported satisfaction with her child care arrangement. The original list of variables had used the professional literature on child care quality to identify salient dimensions of quality; the subsequent analyses focus on what appears to be important from the AFDC mother's perspective.

We used an ordered logistic multiple regression specification (Walker and Duncan, 1967) to fit separate models of mother's reported satisfaction with child care when her youngest child is 0-2 years old and when her child is 3-5 years old. In these analyses satisfaction levels were coded 0 for "not very or not at all satisfied," 1 for "somewhat satisfied," 2 for "mostly satisfied," and 3 for "completely satisfied." Initially full models were specified containing the dimensions of child care "quality" described in Table 4 in addition to other variables anticipated to affect mother's satisfaction with child care. These other variables included demographic descriptors of the mother (race, education level, the number of children under 9 years old that she had, and her city of residence), descriptors of the child care arrangement (type, number of arrangements she had simultaneously, hours and time of day in main child care arrangement) and descriptors of her main activity when she used this care arrangement ( working versus looking for work or going to school). Subsequently the models were reduced to incorporate the predictors contributing most to variation in the dependent variable.

The reduced models of mothers' satisfaction with child care are displayed in Table 6. They show that many of the significant independent predictors of mothers' satisfaction with care for 0-2 year-olds are quite different from the predictors of satisfaction with care for 3-5 year-olds. A Likelihood Ratio Test confirmed that the models of satisfaction for the two age groupings should not be pooled.

When a mother's youngest child is under 3, her satisfaction with her child care arrangement appears to be most strongly related to the child's age in the negative direction and positively associated with and her perceptions that the location and hours of care are convenient and that the amount of adult supervision is adequate. There are also significant site differences in satisfaction levels with mothers from Boston reporting higher levels of satisfaction than those in Charlotte or Denver. In addition the more children there are in the arrangement relative to adults, the less likely it is that she will be satisfied. Other factors positively associated with satisfaction at the .10 significance level are mother's education level and her employment status (working versus being in school or looking for work). Factors negatively associated with satisfaction at the .10 level are the weekly cost of care, the need for care on Saturdays or Sundays, and care provided by relatives in the child's home (versus care provided by non-relatives).

When a mother's youngest child is 3 to 5 years old, the factors emerging as strong predictors of mother's satisfaction with child care are the age of her child (in the negative direction), living in Charlotte (versus Denver), her employment status (working versus being in school or looking for work), the number of non-business hours care is required, and her perceptions of how happy the child is with the arrangement, how frequently the child gets to learn new things, how adequate is the level of adult supervision and how convenient are the hours of arrangement. Two other strong negative predictors of mother's satisfaction are the total number of hours care is required and her perception of how experienced her provider is in caring for children.

Common to both models of mother's satisfaction are the negative effects of the child's age, and the positive effects of the mother's perception that the hours of care are convenient and that the level of supervision is adequate. In addition the significant negative effects of weekend care found among mothers with children under 3, is mirrored among mothers with children 3-5 years old but just fails to reach significance at the .10 level. These commonalities suggest that for children in both age groupings, the convenience of the hours of care and the adequacy of adult supervision are considered to be important child care characteristics by their mothers.

Beyond these common concerns about child care among AFDC mothers with preschool children of all ages, several dimensions of child care quality appear salient only at particular stages of childhood. Most notably mothers with children under the age of three are significantly more satisfied with their childcare if its location is judged to be convenient and if the child to adult ratio is lower. Convenience of location is not a significant predictor of mother's satisfaction when children are older. The child to adult ratio is also not a significant predictor of satisfaction for the older age group. Moreover, the sign for the coefficient is no longer negative for mothers with older children, suggesting a positive relationship between the child to adult ratio and satisfaction. The importance of location for mothers with very young children appears intuitively correct given their need to carry very young children and their various paraphernalia--bottles, diapers, and changes of clothes. Similarly, the importance of a low child to adult ratio for mothers using care for very young children appears to be consistent with the greater dependency of these children. As children become older, we can speculate that parents become concerned that their children have playmates.

Mothers with children between the ages of 3 and 5 are significantly more satisfied with their child care when they think that their child feels happy about the arrangement and has an opportunity to learn new things. They are also more satisfied with their care when they think the caregiver has less experience caring for children, These dimensions of care are not significant independent predictors of mothers' satisfaction with care for children under 3. While the first two findings appear consistent with the developmental growth of children, the finding about the negative effect of a caregiver's experience seems counterintuitive. Perhaps caregiver's experience is masking the effect of an unmeasured variable like enthusiasm, but we can only speculate why this unusual pattern emerges.

Once measures of mothers' objective and subjective assessments of child, care characteristics are entered into the satisfaction models, only in-home relative care as a child care type (contrasted to non-relative care remains a significant independent predictor of dissatisfaction for children under age 3. A similar pattern occurs for children between the ages of 3 and 5, but the significance level was .16. This finding suggests that even after controls are entered for mother's characteristics, employment hours, and the objective and subjective characteristics of the care arrangement, some unmeasured dimensions of in-home care by relatives continue to exert an independent influence on mother's reduced satisfaction with this arrangement.2


We have presented results for a sample of AFDC mothers drawn in 1983 from the caseloads of three cities. As such the results can only be extrapolated to these three caseloads for the time frame of August 1983-October 1984. Yet, given the general absence of information about childcare in relationship to welfare, we are drawn to speculate about the possible policy implications of the results for a wider population.

We would like to comment first on the amount of productive activity among sample mothers over the 14 month period. At a time predating the full scale implementation of welfare to work initiatives, 43 percent of the mothers with children under age 6 had a job during the period; 24 percent went to school or a job training program; 12 percent looked for work without sucess; and only 36 percent did not participate in any of these activities. Moreover mothers with a youngest child under age 2 were just as likely to work as mothers of 2-5 year-olds, and they we re more likely to have looked for work without success. Our sample represents cross-sections of the AFDC caseloads of the 3 cities; long term welfare recipients are therefore over-represented because of length biased sampling (Cox and Isham, 1980). Cumulatively sample mothers had been on welfare for over 5 years, on average, since their first applications. Even so, the data show that almost two-thirds of them looked for jobs, got jobs, or participated in job readying activities over a 14 month period. This high level of activity belies stereotypes and program assumptions about welfare recipients. It is clear that the majority of these recipients voluntarily got jobs, or tried get work or get ready for work. These high levels of activity without special programs suggest that the employment, training and child care provisions of the Family Support Act will find a receptive audience.

When they engaged in these productive activities, AFDC mothers in our samples relied heavily on relatives, and on grandparents in particular to provide child care. This pattern of reliance on relatives for care in our AFDC sample is similar to national data for non-married employed mothers gathered in 1984-1985 in the SIPP survey (see Table 2). A nagging issue in the public debate about child care is the lack of attention to care provided by relatives as a piece of the puzzle. In the mid-seventies Woolsey (1976) drew attention to this gap, suggesting that many parents prefer that relatives take care of their children. Our analyses suggest that although in- home care by relatives gets high ratings from welfare mothers on some of the dimensions that appear related to satisfaction with care--convenience of location and hours, low child to adult ratio, and child's happiness, when these factors are controlled, use of this form of care results in reduced satisfaction (compared to non-relative care). In the same analyses, care by relatives outside the child's home does not have the same negative effect. Perhaps in-home care by relatives carries hidden interpersonal costs or implies lack of real choice, problems that may not emerge when out-of-home relative care is used.

What are the attributes of care that contribute to higher levels of mothers' satisfaction with the arrangement? The AFDC mothers using child care in our samples appeared to be concerned with the adequacy of adult supervision and the convenience of the hours of care. Mothers with youngest children under the age of 3 are also concerned about the convenience of the location of the arrangement and lower child to adult ratios. Mothers with older preschool children are concerned about whether their children are happy in the arrangement and whether, they get the opportunity to learn new things. These characteristics of care appear more powerful than child care type (group care, non-relative care, relative care) in explaining mother's satisfaction with the arrangement. Indeed when the same models of satisfaction were run without the child care type variables the results were essentially the same except that the training of the child care provider emerged as a predictor of mother's satisfaction for children under the age of 3 (p=.07).

These findings suggest that certain characteristics of care--convenient hours and location, adult supervision, child to adult ratios, learning opportunities, and child's happiness--are more important to parents than the care type itself in explaining their satisfaction with care. Policies aiming to maximize parental choice and satisfaction should therefore foster approaches that will enhance all types of child care on these important dimensions. Some of these care characteristics readily lend themselves to policy interpretation. For example, a large proportion of the mothers in the sample worked hours that are not part of the regular business day, before 7:00 am and after 6:00 pm. Other mothers worked on weekends. Since convenient hours are an important factor in parents satisfaction, targeting efforts to expanding the availability of child care in the "off hours," early morning and evening would provide care that might be more responsive to parents' needs. Similarly efforts could be targeted to improve the geographic accessibility of infant care, reduce child to adult ratios for infant care, and improve the learning opportunities in care for older preschool children.

Our findings support the expansion of a diverse range of child care options for AFDC mothers as the child care guarantees of the Family Support Act are implemented. Just focusing on expanding the traditional service system of centers and family day care will not be enough. Some parents use and prefer care by relatives and friends. Concentrating on increasing the supply of child care during regular business hours will only be a partial solution. Many parents work other hours. If AFDC parents are to be empowered to choose their own child care arrangements, new approaches are needed to enhance the existing care system so it is more responsive to the preferences of these parents.


Allison, P.D. (1982) Discrete Time Methods for the Analysis of Event Histories. S. Leinhardt (ed.) Sociological Methodology 1982. Washington, D.C. Jossey-Bass Publishers.

Cox, D.R. and V. Isham (1980) Point Processes London: Chapman and Hall.

Federal Register (1989) "Aid to Families with Dependent Children; Job Opportunities and Basic Skills Training Program and Child Care and Supportive Services: Final Rule." Friday, October 13. 42146-42267.

Hofferth, S. (1987). Statement on Child Care in the United States Made Before the Select Committee on Children, Youth, and Families, July 1.

Hofferth, S. and Sonenstein, F. (1983) An Examination of the Effects of Alternative Approaches to Financing Day Care for AFDC Children. Washington, D.C.: The Urban Institute.

Nelson, J.R. (1982). The Federal Interagency Day Care Requirements. C.D. Hayes (ed) Making Policies for Children. Washington, D.C., National Academy Press, 151-205.

Ruopp, R., J. Travers, F. Glantz, and C. Coelen (1979). Children at the Center: Final Report of the National Day Care Study, Cambridge, Massachusetts, Abt Associates.

U.S. Bureau of the Census. (1987) Current Population Reports, Series P-70, No 9, Who's Minding the Kids? Child Care Arrangements: Winter 1984-1985, U.S. Government Printing Office, 1987.

Walker, S.H. and Duncan, D.B. (1967) "Estimation of the Probability of an Event as a Function of Several Independent Variables," Biometrika, 54, 167-179.

Weder, W.A. (no date) Current AFDC Recipient Characteristics and Analysis of Selected Caseload Changes Between May 1981 and May 1982. Washington, D.C.: Office of Family Assistance, Division of Research, Evaluation and. Statistics, U.S. Department of Health and Human Services.

Woolsey, S.H. (1976) Pied Piper Politics and the Child Care Debate. Washington, D.C.: The Urban Institute.

Zill, N. (1989) Basic Facts about the Use of Child Care and Preschool Services by Families in the U.S. Washington, D.C.: Child Trends


  1. Analysis of Variance was used to test for statistically significant differences between the care types.

  2. In models run without the characteristics of care variables, inhome care by relatives was a negative predictor of mother's satisfaction for children ages 3-5 using the same controls for mothers, characteristics that are used in the models that have been reported. Out-of-home care was also a significant independent predictor of reduced satisfaction.

TABLE 1. Activities Throughout 14 Months by Age of Youngest Child(Weighted Percentages)
  0-1 2-3 4-5 Total
Worked   41.3     42.1     44.8     42.7  
Looked/No Work   21.1 7.7 8.1 11.6
School/Training 23.8 22.6 26.3 24.0
No Activity 30.6 40.1 34.5 35.7
Unweighted N 106 161 115 382
TABLE 2. Main Child Care Arrangement for Youngest Child by Age(Weighted Percentages)
  Age of Youngest Child
  0-1     2-3     4-5     Total     SIPP*  
Care in Child's Home
   By Brother/Sister 3 0 5 3 --
   By Grandparent 24 16 19 20 16
   By Other Relative 10 12 5 9 11
   By Non-relative 3 4 0 2 3
Care Outside Home
   By Grandparent 12 12 4 10 8
   By Other Relative 5 12 8 9 6
   By Non-relative 18 7 16 13 24
Day Care Center 10 20 25 18 20
Nursery School/Head Start   3 4 5 4 7
School 0 0 12 4 1
Mother Cares for Child 12 12 0 8 4
Unweighted N 87 116 93 296  
* Employed Non-married Mothers, Youngest Child Under 5, Survey of Income and Program Participation Winter 1984-1985, U.S. Bureau of the Census (1987), Table 3.
TABLE 3. Main Child Care Arrangement for Youngest Child 0-5 Years by Activity(Weighted Percentages)
    Employed     Looking for Work     School/Training  
Relative in Home 35.6 42.3 30.0
Relative out of Home 23.8 20.6 17.8
Nonrelative out of Home   21.8 10.5 3.4
Nonrelative in Home 4.4 2.1 0.0
Group Care 14.4 24.5 48.8
Unweighted N 172 69 49
TABLE 4. Variable Definitions
KIDSADUL Ration: number of children cared for together/number of adults always present in arrangement
TRN Provider has received special training in caring for children
XPER Experience of provider in caring for children
LERNEW How often child is given chance to learn new things
DISC When child misbehaves person caring for child provides just enough discipline (vs. too much/too little)
SUPER Level of adult supervision provided child is just right (vs. too much/too little)
CARF Safety precautions taken to prevent accidents by provider (Extremely careful, Somewhat careful, Careful, of In need of improvement)
HAP Child's feelings about the arrangement (Very happy, Happy, Indifferent or Not too happy)
FEEL Child's feelings about the person who cares for him/her (Loving, Friendly, Indifferent or Dislikes)
CVHR Convenience of hours of the person caring for child
MAINHOUR   Number of hours in main arrangement per week
OFFWEEK Number of hours per week care required on weekends or before 7:00am and after 6:00pm
MONFRI Care required before 7:00am or after 6:00pm Monday-Friday
SATSUN Care required Saturday or Sunday
CVLOC Convenience of the location of arrangement
HOMESIK Number of Days R stayed home from work/school/training in past 8 months because regular child care arrangement was not available  
HOMEUNA Number of Days R stayed home from work/school/training in past 8 months because child was sick
PAYMAIN Pays for Care
WKLYCST Weekly out-of-pocket cost for main care arrangement ($)
HRLYCST Hourly out-of-pocket cost for main care arrangement ($)
AFFORD The cost of the arrangement is very reasonable, reasonable, somewhat reasonable, or too expensive
SATIS Mother's satisfaction with care arrangement
HASCHOIC Child is in arrangement that is mother's first choice
TABLE 5. Mother's Mean Rating of Care by Type of Arrangement
Variable Type of Arrangement
  Relative  In Home   Relative  Out   Non-Rel.  Out   Non-Rel.  In Home GroupCare
KIDSADUL 1.583 1.651 2.618 2.556   5.282***  
TRN 0.111 0.177 0.240 0.111 0.931***
XPER 2.375 2.613 2.098 2.000 2.466***
LERNEW 2.227 2.355 1.980 1.778 2.603***
SUPER 0.889 0.952 0.940 0.875 0.911
DISC 0.841 0.726 0.863 0.750 0.891
CARF 2.578 2.548 2.294 2.000 2.362**
HAP 2.472 2.387 2.176 1.778 2.322**
FEEL 2.854 2.839 2.471 2.333 2.500***
CVHR 2.753 2.475 2.529 2.667 2.712**
MAINHOUR   23.93 24.81 31.14 24.94 30.40**
OFFWEEK 6.402 8.715 5.241 3.539 2.150**
MONFRI 0.433 0.446 0.404 0.556 0.208**
SATSUN 0.105 0.092 0.149 0.111 0.222
CVLOC 2.865 2.274 2.549 2.778 2.576***
HOMESIK 1.733 2.167 2.588 1.333 5.644***
HOMEUNA 1.652 2.015 0.961 2.111 1.254
PAYMAIN 1.720 1.559 1.216 1.400 1.541***
WKLYCST 19.03 21.82 25.49 40.83 13.23***
HRLYCST 0.74 0.90 0.75 1.70 0.69*
AFFORD 3.378 3.000 2.529 2.667 3.190***
SATIS 2.081 2.118 1.745 1.800 2.459***
HASCHOIC 0.340 0.250 0.059 0.100 0.525***
*** p<0.01** p<0.05* p<0.10
TABLE 6. Determinants of Mother's Satisfaction with Child Care (Ordered Logistic Regression)
Variable Youngest Child Age 0-2 Youngest Child Age 3-5
  Coefficient     Standard Error     Coefficient     Standard Error  
ALPHA1 -13.555 5.843 -8.785 7.739
ALPHA2 -15.797 5.943 -11.229 7.801
ALPHA3 -17.469 6.003 -14.528 7.915
INDXAGE -1.587 0.570*** -2.643 0.948***
BLACK 0.866 1.164 0.423 1.743
HISPANIC 0.162 1.671 -1.910 1.711
BOSTON 2.519 1.215** 1.215 1.311
CHARLOTTE   -0.124 0.795 3.480 1.585**
HIGHGRAD 0.502 0.272* 0.371 0.320
WKINGD -1.932 1.124* 4.445 1.439***
MAINHOUR 0.004 0.003 -0.016 0.005***
NUMARR -0.831 1.076 2.575 2.737
OFFWEEK 0.046 0.043 0.383 0.106***
SATSUN -3.151 1.678* -4.876 3.121
REL_IN -1.969 1.077* -1.971 1.394
REL_OUT 0.459 1.394 2.032 1.926
GROUP 1.778 1.219 1.420 2.150
TRN 0.505 0.814 -1.862 2.450
KIDSADUL -0.579 0.258** 0.421 0.313
WKLYCST -0.001 0.000* -0.001 0.001
XPER 0.285 0.449 -1.762 0.779**
FEEL 0.139 0.723 -0.853 0.841
LERNEW 0.262 0.414 2.143 0.888**
HAP 0.869 0.931 2.916 0.973***
CVHR 1.144 0.511** 2.613 0.932***
CVLOC 2.401 0.535*** 0.052 0.471
SUPER 3.077 1.139*** 5.260 1.714***
  N = 90-2 Log Likelihood 158.82LLR p<0.0001 N = 99-2 Log Likelihood 89.05LLR p<0.0001
*** p<0.01** p<0.05* p<0.10

"ccpref.pdf" (pdf, 442Kb)

Note: Documents in PDF format require the Adobe Acrobat Reader®. If you experience problems with PDF documents, please download the latest version of the Reader®