Child Care Quality:
Does it Matter and Does It Need to be Improved?

Tables 4 – 13

TABLE 4
Linear Trends Relating Number of Recommended Standards Met to Child Outcomes: NICHD Study of Early Child Care

  Adjusted Mean by No. of Standards Met Linear Trend
  0 1 0 2 3 F P
24 months
No. 15 38 65 29 16    
Mental development, mean (SE) 95.4 (3.0) 93.8 (1.9) 97.2 (1.5) 98.6 (2.2) 99.7 (3.0) 2.19 .14
Behavior problems, mean (SE) 0.34 (0.86) -0.12 (0.54) -0.19 (0.41) -0.40 (0.62) -2.17 (0.84) 4.18 .04
Positive social behavior, mean (SE) 0.31 (0.41) -.02 (0.26) 0.16 (0.20) 0.74 (0.30) 0.86 (0.40) 2.25 .14
36 months
No. 8 31 65 60 86    
School readiness, mean (SE) 36.1 (8.0) 38.9 (4.1) 47.9 (2.8) 51.5 (2.9) 51.8 (2.4) 6.29 .01
Expressive language, mean (SE) 92.3 (4.7) 98.6 (2.4) 95.8 (1.6) 101.4 (1.7) 100.2 (1.4) 3.34 .07
Language comprehension, mean (SE) 95.6 (4.5) 99.1 (2.3) 100.1 (1.6) 105.0 (1.6) 104.7 (1.4) 6.03 .02
Behavior problems, mean (SE) 2.69 (1.25) 0.81 (0.63) 0.31 (0.43) 0.04 (0.45) -1.14 (0.38) 9.59 <.01
Positive social behavior, mean (SE) -0.14 (0.60) 0.06 (0.30) -0.10 (0.21) 0.14 (0.22) 0.27 (0.18) 0.47 .49
Note.  Family income-to-needs ratio and maternal sensitivity were used as co-variates.

TABLE 5
Adjusted Means for Children in Settings That Did or Did Not Meet Specific Child Care Standards: NICHD Study of Early Child Care

  Child-Staff Ratio Group Size Caregiver Education Caregiver Training
  Met Not Met Met Not Met Met Not Met Met Not Met
24 months
No. 43 120 45 118 38 125 57 106
Mental development                
Mean 99.26 95.82 97.87 96.29 97.00 95.84 97.63 95.06
SE 1.78 1.06 1.75 1.08 1.06 1.96 1.14 1.57
Behavior problems                
Mean -1.54 0.07** -0.09 -0.15 -0.30 -0.55 -0.51 -0.07
SE 0.50 0.30 0.50 0.31 0.30 0.56 0.33 0.45
Positive social behavior                
Mean 0.79 0.13* 0.47 0.24 0.33 0.20 0.39 0.15
SE 0.24 0.14 0.24 0.15 0.14 0.27 0.16 0.21
36 months
No. 140 110 157 93 201 49 187 63
School readiness                
Mean 49.77 47.12 49.23 47.56 51.06 38.52*** 51.09 41.22**
SE 1.92 2.17 1.82 2.37 1.57 3.19 1.64 2.82
Expressive language                
Mean 99.96 97.60 99.60 97.77 99.63 96.01 99.40 97.48
SE 1.11 1.25 1.05 1.37 0.92 1.88 096 1.66
Language comprehension                
Mean 103.81 101.02 103.21 101.52 103.50 98.81* 104.03 98.27**
SE 1.07 1.21 1.02 1.33 0.89 1.81 0.92 1.58
Behavior problems                
Mean -0.66 0.57** -0.32 0.23 -0.51 1.49*** -0.41 0.76*
SE 0.30 0.34 0.29 0.37 0.25 0.50 0.26 0.45
Positive social behavior                
Mean 0.31 -0.16* 0.11 0.09 0.13 0.01 0.11 0.09
SE 0.14 0.16 0.13 0.17 0.12 0.24 0.12 0.21
Notes: At 36 months, the ratio standard was 7.1, the group size standard was 14, the caregiver education standard was some college, and the caregiver training standard was some post-high school training in child development, early childhood education, or a related field. Pairs of means shown in italics were significantly different from one another. Child outcomes and child care standards were measured concurrently.

*p<.05; **p<.01; ***p<.001

TABLE 6
Summary of Effect Sizes at 36 Months: NICHD Study of Early Child Care

  Child Care Quality(1) Hours in Care(2) Home Quality(3)
  Whole Sample Extreme Groups(4) Whole Sample Extreme Groups(4) Whole Sample Extreme Groups(4)
  r d r d r d
School Readiness .14*** .39*** .00 .04 .25*** .83***
Expressive Language .14*** .44*** .00 .04 .28*** 1.01***
Receptive Language .09*** .27* .06 .10 .19*** .71***
Behavior Problem (Mother) -.03 -.14 .05 .08 -.06 -.10
Positive Social Behavior (Mother) .03 .13 .03 .15 .18*** .55***
Behavior Problems (Caregiver) -.06 -.25 .04 .12 -.15*** -.56***
Peers Skills -.04 -.05 .06 .06 .17*** -.57***
*p<.05; **p<.01; ***p<.001

Footnotes:

1.  Associations between cumulative positive caregiving and child adjustment at 36 months, controlling for site, mother education, income-to-needs, mother psychological adjustment, home quality, child gender, percent time in centers, and hours in care.

2.  Associations between cumulative child care hours and child adjustment at 36 months, controlling for site, mother education, income-to-needs, mother psychological adjustment, home quality, child gender, percent time in centers, and child care quality.

3.  Associations between cumulative observed home quality and child adjustment at 36 months, controlling for site, mother education, income-to-needs, mother psychological adjustment, child gender, percent time in centers, hours in care, and child care quality.

4.  Extreme group analyses contrasted the bottom and top quartiles.

TABLE 7
Regression Results and Simulated Change in Child Cognition and Language Skills
Simulation: Shift in Quality One Standard Deviation below the Mean to One Standard Deviation above the Mean
  Positive Caregiving Rating† Language Stimulation‡
  Coefficient Change Coefficient Change
15 months (N=595)        
Bayley .39 -- 0.15* 4.79
CDI Vocab. Prod. 1.21* 6.97 0.46* 14.53
CDI Vocab Comprehen. 1.44* 8.30 0.32* 9.96
         
24 months (N=739)        
Bayley 0.78* 4.47 0.06* 3.91
CDI Vocab. Prod 0.64 -- 0.20* 12.03
CDI Vocab Comprehen. 1.21* 6.93 0.15* 9.24
         
36 months (N=856)        
Bracken 0.97* 6.39 0.08 --
Reynell Expressive Lanaguage 0.81* 5.30 0.12* 6.36
Reynell Verbal Comprehension 1.31* 8.62 0.06 --
Source: NICHD Early Child Care Research Network (in press-b).

Notes: Equations also include maternal ratings, income to needs, gender, HOME total score, maternal stimulation, average hours per week in childcare, number times in center care, number times in child care home.

† Reported for model which does not include language stimulation.

‡ Reported for model which includes positive caregiving rating.

* Statistically significant at 5% level

TABLE 8
Studies of Child Care and Parental Employment

CITATION STUDY DESIGN ISSUE ADDRESSED N AGE QUALITY MEASURE POSSIBLE CONTROL FINDINGS
Does Employment Matter?
Blau, D. and P. Robbins, 1988. “C hild Care Costs, and Family Labor Supply.” 1980 Employment Opportunity Pilot Project (EOPP) Mother does not work versus four combination of mother working with purchased care or not, other relatives working or not. Married and single mothers < 7 Average child care expenditures among families in the community who purchased care The study averages child care expenditures across families within a given geographical location. Uses multinomial logit estimation procedure.
-0.38 (average price elasticity of employment over a range of examined child care costs) Significance cannot be determined from available information.

Blau, D and P. Robins, 1989. “Fertility, Employment, and Child- Care Costs.” Employment Opportunity Pilot Projects (EOPP) An examination of employment, child care costs, births rates and potential behavior effects of child-care subsidies 6170 employed women and 8940 unemployed women     child care costs, wage rates, education, work experience, age Child care costs are estimated to have significant effects on both birth and employment transition. Higher child-care costs are estimated to lower birth rates for unemployed women, but not for employed women. Higher child-care costs also increase the rate of leaving employment and reduce the rate of entering employment.
Brooks-Gunn, et al, 1994. In Ross, C and D. Paulsell, 1998. “Sustaining Employment Among Low-Income Parents: the Role of Quality in Child Care: A Research Review.” Infant Health and Development Program Employment of Mothers Mothers of low-weight pre-mature infants offered home visits   random assignment intervention   The provision of high-quality center-based care to low-income mothers of very young children can increase employment rates and improve the stability of employment over time. Cumulative months of employment for less-educated black mothers were 14.2 for the intervention group and 12.0 for the control group.
Connelly, R. and J. Kimmel, (2000) 1992 and 1993 SIPP Examination of difference in the effect of child care prices on employment and child care mode choice by marital status. Married and unmarried mothers (employed, full & part- time employed)     tax credits, subsidies, price of care, employment status Using a single predicted price of child care married women working full or part-time were no different in their mode of child care choice. Single mothers working full-time were more likely to use center-based care. Estimating separate prices of care for each mode of care, married and single mothers working full-time were both more likely to use center care rather than relative care. Implications include a move towards increased center based care as more women increase full-time work as they reach welfare reform time limits.
Fuller, et al., 1999. “Variation in Poor Children’s Home and Child-Care Settings: Does Maternal Employment Matter?” 2 waves 1997, 1998 Does maternal employment affect choice of child care and child development 948 single motherers entering new welfare programs in CA, CN, FL.   Early Childhood Environment Rating Scale (ECERS), Family Day Care Rating Scale (FDCRS), Arnett scale, Child Observation System maternal education, age, ethnicity, child age, high and low-income groups Children of high and low-employment mothers spend just under 40 hrs/wk in child care. High employment group has established more stable child-care providers averaging 6.8 months in duration compared to 4.9 moths for low employment group (p<.01, n=272). Children of low-employment group were attending centers who scored higher on the ECERS compared to children of high- employment mothers (p<.01, n=133). Women with more work experience select center- based programs more frequently than low- employment mothers.
GAO, 1995 Urban Institute’s 1990 National Child Care Survey and Low-Income Sub-study Mother is employed or not. (see above) (see above) Examined mothers employed or not. Selectivitiy corrected predicted hourly costs from employed who purchase care   Probit estimation. -0.5 for poor mothers, -0.34 for near-poor mothers, and -0.19 for non-poor mothers (calculated at means).
Leibowitz et al, 1992 NLSY Whether mother is employed when child is 3 mo. old and whether mother is employed when child is 24 mo. old first time mothers children at 3 months and 24 months Subsidy available through state and federal income tax credits; assumes the woman works full-time at her predicted wage. Variation among states and over time in state and federal income tax credits for child care Greater tax credits increased early return to work (within 3 mo.) But had little effect on employment of women with older children.
Maume, D. 1991. “Child-care Expenditures and Women’s Employment Turnover.” Longitudinal SIPP, Wave 5, 1985. Looking at employed women from 1985 and the effects of child care payments on employment. 1814 women 15-45 years old at least one child < 13 yrs.   Employment status, child care costs, education & work history, income Child care payments force women to choose between employment and exiting the work force. The level of child care payments has s significant effect on rates of quitting for mothers of pre-schoolers, but not school-age children. The effects of child care expenses on quitting does not vary by mother’s wage. Low and middle-income women will reduce work force attachment as child care expenditures increase.
Meyers,M, 1 Ross, C and D. Paulsell, 1998. “Sustaining Employment Among Low-Income Parents: the Role of Quality in Child Care: A Research Review.” 1993. In Longitudinal CA GAIN program Child care arrangements and mothers employment Mothers participating in the GAIN program       A mother’s assessment of the safety of the child care arrangements and the trustworthiness of the provider were important predictors of whether she was active in job preparation activities or employment 1 yr later.
Parent’s assessment of the learning and social opportunities in child care were not significantly associated with the parent’s job related progress.
When child/caregiver ratios did not meet standards, mothers were twice as likely to drop out of the program.
Peisner-Feinberg, E.S. et al., 1999, “The Children of the Cost, Quality, and Outcomes Study Go To School.” Longitudinal Cost, Quality and Outcomes Study Education of mother, child care quality, and the effect on child development 826 children pre- school through 2nd grade Survey, child/ classroom assessment, teacher-child relationship quality Analysis controlled for effects of maternal education, child gender and ethnicity For children with less highly educated mothers, better quality child care was strongly related to better math skills and fewer problem behaviors through 2nd grade.
Does Income Matter?
Bureau of Census, U.S. Dept. of Labor. “ Work and Family” (1988) Longitudinal (NLSY) child care arrangements and mother’s income and family earnings   < 5 The % of children in relative, non- relative, and center care   The majority of mothers with earnings <$10,000 use relative care (47.3%)

Most mothers with earnings >$30,000 use non-relative care (46.8%)

Families with total income <$10,000 use relative care 65.0%. Most families with total earnings >$30,000 also use relative care 34.7%.
“Child Care for Low Income Families Workshop” 1995 Longitudinal NICHD Study, Low income families and choice of care 1,364 mothers Data collection at ages 6-,15-, 24-,36-,42- mo. and 7 year birthday ECERS   Higher levels of non-maternal income appear to be necessary for low-income families to puchase higher-quality care.
“Child Care for Low Income Families Workshop” 1995 Longitudinal National Child Care Survey, 1993

Low income children and continuity of care 430 households with income

< $15,000

 pre-school Stability (number of arrangements)   24% of low-income children < 5 and 45% of low-income preschoolers in families headed by an employed single mother were in more than one arrangement on a regular basis.
“Child Care for Low Income Families Workshop” 1995 Longitudinal National Child Care Survey, 1993 Low income families and choice of care 4,400 households children < 13 years Type of child care setting   48% low-income (<$15,000) cared for by parent. Use of non-parental care varied greatly by household type and employment status of mother. Care by a relative for low-income preschoolers with single mother (26%), employed mothers (28%), and mothers in education and training programs (23%). Children under 5 in families headed by employed single mothers rely heavily on family day care (21%) and center-based care (27%).
Hofferth & Wissoker, 1999 1992. In “Why and how Working Women Choose Child Care: A Review with a Focus on Infancy,” 1999 Longitudinal NLSY Income and choice of care   < 6     Higher income, measure both as total household income and as mother’s hourly wage, are associated with a higher probability of choosing center care over sitter, father, or relative care.
Hofferth, S. and D. Wissoker. 1992 "Price, Quality and Income in Child Care Choice.” Longitudinal NLSY Income and choice of care 971 mothers who used some form of non-maternal care as the primary arrangement. < 6 multinomial and universal logit models   Mothers who earn more per hour and families who have higher incomes (other than mother’s earning) are more likely to select center care over other mode. Subsidizing child care expenditure directly through vouchers. Reducing fees (reducing the price by 10% increases the use of center-based programs by 39% (ML) and 27% (UL). Reducing the price by 25% increases the use of center based programs by 114% (ML) and 77% (UL). Increasing tax credits increases the use of center-based programs (all else equal).
Hofferth, S. (1999) Child Care, Maternal Employment, and Public Policy Longitudinal Low-income families and choice of care   < 6
(sub-groups < 2 and 3-5 yrs.)		 Type of care   Higher proportions of children of middle- income use center care (38% of all children). There are significant differences in center care use between working poor (22%) and non- working poor (35%)
Kontos, et al, 1995. In “Why and how Working Women Choose Child Care: A Review with a Focus on Infancy,” 1999 Concurrent Income and choice of care     Child care setting   Mothers with a family income greater than $40,000 were the most likely to select regulated family day care and mothers with a family income of less the $20,000 were the most likely to select relative care.
NICHD. “Poverty & Patterns of Child Care” (1997) Longitudinal

(NICHD )

 Type of care with and without income 1,281 < 15 mo quality of care measured by: Home Observation for Measurement of the Environment (HOME) Inventory; Pos. caregiving freq.; pos. caregiving rating. ethnicity, sex, children in family, mother’s education, income -to- needs ratio In some homelike settings, mother’s education predicted some day care quality measure, but income differences accounted for the effects of education. In both in-home care and child care homes, children living in poverty received care that scored significantly lower on most quality measures than children with incomes above poverty. The pattern in centers showed a curvilinear relation between income and quality with children from poor families receiving care that scored higher than that received by children from near-poor families.
NICHD, “Study of Early Child Care,” (1997) Longitudinal Family income and child care usage 1,364 children       Families living in poverty (income needs ration 1.0) were less likely to use center care. Children in near poverty (income to needs ration between 1.00 - 1.99) received lower quality of center care than children in poverty.
Family economics accounted for the amount, the age of entry into care and the type and quality of care infants received.
NICHD, “Study of Early Child Care,” (1999) Longitudinal Income and choice of care     In-home care and family day care   Families using in-home care had the highest income levels for maternal and non-maternal income. The mean non-maternal income was $20,000 higher than the means for families using any other type of care.

Family income was a significant predictor of both the age at which infants entered care as well as the number of hours per week spent in care. 5 groups were compared based on age of entry into nonmaternal care (0-2 mo, 3-5 mo, 6-11 mo, 12-15 mo, not in care at 15 mo.) The families of infants who entered care between 3 & 5 mo. had the highest incomes. Maternal income was greatest in this group. Infants who entered care between 0-2 mo. had famlies with low nonmaternalt incomes - they were more dependent on maternal incomes. Families whose infants were not in care at 15 mo. had relatively high nonmaternal incomes.

TABLE 9.
Ratings of Process Quality by Type of Care and Child Care:
NICHD Study of Early Child Care

Distributions of Positive Caregiving Scores
(percentages)

  Poor Fair Good Excellent
15 Months
Child care homes 7.0 43.5 33.5 16.0
Centers 10.2 61.7 23.4 4.7
All types 7.7 43.2 34.2 14.9
Extrapolated U.S. care (nonparental) 9.0 44.4 34.9 11.7
24 Months
Child care homes 8.5 50.2 27.5 13.7
Centers 11.4 66.3 18.5 3.8
All types 9.7 51.7 27.4 10.2
Extrapolated U.S. care (nonparental) 10.9 52.3 28.3 8.5
36 Months
Child care homes 2.4 63.1 29.2 5.4
Centers 4.4 61.9 30.3 3.4
All types 3.9 59.5 31.5 5.1
Extrapolated U.S. care (nonparental) 4.3 60.6 30.2 5.0
15-36 Months (mean)
All types imputed 7.2 52.7 30.6 9.5
Extrapolated to U.S. 8.1 53.2 29.6 9.0

TABLE 10
Requirements for Child: Staff Ratio and Group Size By State

States Age of Children
  9 months 18 months 3 years 4 years
Alabama 6:1

6

 8:1

8

 12:1

12

 20:1

20
Alaska 5:1

NR

 6:1

NR

 10:1

NR

 10:1

NR
Arizona 5:1/11:2

NR

 6:1/13.2

NR

 13:1

NR

 15:1

NR
Arkansas 6:1

NR

 9:1

NR

 12:1

NR

 15:1

NR
California 4:1

NR

 6:1

12

 12:1

NR

 12:1

NR
Colorado 5:1

10

 5:1

10

 10:1

20

 12:1

24
Connecticut 4:1

8

 4:1

8

 10:1

20

 10:1

20
Delaware 4:1

NR

 7:1

NR

 12:1

NR

 15:1

NR
District of Columbia 4:1

8

 4:1

8

 8:1

16

 10:1

20
Florida 4:1

NR

 6:1

NR

 15:1

NR

 20:1

NR
Georgia 6:1

12

 8:1

16

 15:1

30

 18:1

36
Hawaii 4:1*

8

 6:1*

12

 12:1

NR

 16:1

NR
Idaho 6:1

NR

 6:1

NR

 12:1

NR

 12:1

NR
Illinois 4:1

12

 5:1

15

 10:1

20

 10:1

20
Indiana 4:1

8

 5:1

10

 10:1

NR

 12:1

NR
Iowa 4:1

NR

 4:1

NR

 8:1

NR

 12:1

NR
Kansas 3:1

9

 5:1

10

 12:1*

24

 12:1*

24
Kentucky 5:1

10

 6:1

12

 12:1

24

 14:1

28
Louisiana 6:1

NR

 8:1

NR

 14:1

NR

 16:1

NR
Maine 4:1

12

 5:1

15

 10:1

30

 10:1

30
Maryland 3:1*

6

 3:1*

9

 10:1

20

 10:1

20
Massachusetts 3:1/7:2

7

 4:1/9:2

9

 10:1*

20

 10:1*

30
Michigan 4:1

NR

 4:1

NR

 10:1

NR

 12:1

NR
Minnesota 4:1

8

 7:1

14

 10:1

20

 10:1

20
Mississippi 5:1

10

 9:1

10

 14:1

14

 16:1

20
Missouri 4:1

8

 4:1

8

 10:1

NR

 10:1

NR
Montana 4:1

NR

 4:1

NR

 8:1

NR

 10:1

NR
Nebraska 4:1

NR

 6:1

NR

 10:1

NR

 12:1

NR
Nevada 6:1

NR

 8:1

NR

 13:1*

NR

 13:1*

NR
New Hampshire 4:1

12

 5:1

15

 8:1

24

 12:1

24
New Jersey 4:1

20

 7:1

20

 10:1

20

 15:1

20
New Mexico 6:1

NR

 6:1

NR

 12:1

NR

 12:1

NR
New York 4:1

8

 5:1*

10

 7:1*

14

 8:1*

16
New York City 4:1

8

 5:1

10

 7.5:1

15

 12:1

20
North Carolina 5:1

10

 6:1

12

 15:1

25

 20:1

25
North Dakota 4:1

NR

 4:1

NR

 7:1

NR

 10:1

NR
Ohio 5:1

12

 7:1

14

 12:1

24

 12:1

28
Oklahoma 4:1

8

 6:1

12

 12:1

24

 15:1

30
Oregon 4:1

8

 4:1

8

 10:1

20

 10:1

20
Pennsylvania 4:1

8

 5:1

10

 6:1

20

 6:1

20
Rhode Island 4:1

8

 6:1

12

 9:1

18

 10:1

20
South Carolina 6:1

NR

 6:1

NR

 13:1

NR

 18.1

NR
South Dakota 5:1

20

 5:1

20

 10:1

20

 10:1

20
Tennessee 5:1

10

 7:1

14

 10:1

20

 15:1

20
Texas 4:1/10:2

10

 9:1/18.2

18

 17:1/34:2

34

 20:1/35:2

35
Utah 4:1

8

 4:1

8

 12:1

24

 15:1

30
Vermont 4:1

8

 4:1

8

 10:1

20

 10:1

20
Virginia 4:1

NR

 5:1

NR

 10:1

NR

 12:1

NR
Washington 4:1

8

 7:1

14

 10:1

20

 10:1

20
West Virginia 4:1

NR

 4:1

NR

 10:1

NR

 12:1

NR
Wisconsin 4:1

8

 4:1

8

 10:1

20

 13:1

24
Wyoming 5:1

NR

 5:1

NR

 10:1

NR

 15:1

NR
Notes:  NR means Not Rated.

Information was obtained from a 1999 report prepared by The Center for Career Development in Early Care and Education at Wheelock College.

TABLE 11
Regression Results for Period and Cumulative Effects at 24 and 36 Months:

  Mean Standard Deviation Coefficient Same Period Only Coefficients Cumulative Impact-
1SD to +1 SD		 Impact-
1SD to +1 SD
CDI VOCABULARY PRODUCED †            
24 mo.: language stimulation 58.62 30.92 0.178 0.338 11.027 20.91
15 mo.: language stimulation 26.51 15.79   0.30   9.36
24 mo.: positive caregiver rating 3.96 2.87        
6–15 mo.: positive caregiver rating 14.63 2.89        
Total         11.027 30.27
Reynell: Vocabulary Comp.            
24 mo.: language stimulation 50.56 27.09   -0.115   -6.25
15 mo.: language stimulation 58.62 30.92   0.195   12.08
24 mo.: positive caregiver rating 19.5 3.29 0.133 0.787 0.87 5.18
6–15 mo.: positive caregiver rating 13.96 2.87        
Total         0.87 11.00
Source: NICHD, 1999b.

Notes: Only variables that are statistically significant at 5% level are reported.

† Mean = 28.82

‡ Mean = 28.73

TABLE 12
Descriptive Statistics for Child Care Standards at Four Ages: NICHD Study of Early Child Care

Feature No. Mean SD Minimum Maximum Recommended Level Classes Meeting Recommended Level, %
6 months
Child-staff ratio 97 4.26 2.31 0.81 15.06 3 36
Observed group size 97 7.86 4.05 1.63 30.13 6 35
Caregiver training 97 1.70 1.55 0.00 4.00 2 56
Caregiver education 97 2.97 1.07 1.00 6.00 3 65
15 months
Child-staff ratio 118 4.14 1.30 1.50 7.88 3 20
Observed group size 118 8.53 3.28 2.38 23.38 6 25
Caregiver training 118 1.64 1.38 0.00 4.00 2 60
Caregiver education 118 2.89 0.89 1.00 5.00 3 69
24 months
Child-staff ratio 163 5.22 1.70 1.68 10.95 4 26
Observed group size 163 10.66 4.78 3.38 37.75 8 28
Caregiver training 163 1.84 1.37 0.00 4.00 2 65
Caregiver education 163 3.06 0.86 1.00 5.00 3 77
36 months
Child-staff ratio 250 6.98 2.32 1.92 14.90 7 56
Observed group size 250 13.20 4.63 3.50 32.63 14 63
Caregiver training 250 2.10 1.33 0.00 4.00 2 75
Caregiver education 250 3.24 0.91 1.00 6.00 3 80
Source:  NICHD Early Child Care Research Network, 1999, “Child Outcomes When Child Care Center Classes Meet Recommended Standards of Quality,” American Journal of Public Health 89: 1072–1077, Table 1.

TABLE 13
Marginal Effects of the Probability of Choosing among Modes
of Care for the Youngest Child for Single Mothers
(Multinominal Logit Estimation)
Variable Relative Care Home-based Care Center-Based Care
Predicted Probability of full-time employment -3.096* -2.951** 6.046***
(-1.91) (-2.27) (3.52)
Predicted wage 0.946* 1.000** -1.947***
(1.84) (2.42) (-3.60)
Education -0.050* -0.042* 0.093***
(-1.72) (-1.87) (3.05)
State’s average monthly AFDC payment per family -6.E-4* -2.E-4 9.E-4*
(-1.84) (-1.01) (2.51)
Predicted Probability of full-time employment -3.141*** -0.258 3.398***
(-3.17) (-0.37) (3.34)
Predicted wage 1.095*** 0.204 -1.299***
(2.66) (0.68) (-3.13)
Education -0.108*** 0.043 0.151***
(-2.81) (-1.55) (3.78)
State’s average monthly AFDC payment per family -6.E-4 3.E-4 3.E-4
(-1.57) (0.97) (0.86)
Significance level: *10%, ** 5%, ***1%

Additional variables include: predicted price of care, predicted price of relative care, predicted price of home-care, predicted price of center-based care, age, nonwhite, non-labor income, youngest child is an infant, presence of other preschoolers, presence of children ages 6-12, presence of children ages 13-17, presence of other adults, urban residence, and state’s average Medicaid expenditure per enrollee.

Source: Connelly and Kimmel (2000), Tables 5 and 6.

Where to?

Top of page
Table of Contents
Executive Summary
Full Report

Home Pages:
Human Services Policy (HSP)
Assistant Secretary for Planning and Evaluation (ASPE)
U.S. Department of Health and Human Services (HHS)

Last updated 11/28/01