Economics of UPK
Benefits and Costs of Quality Early
Childhood Education for All
Presentation at a Conference on
The Law and Policy of Universal Preschool
Chicago, IL
October 13, 2006
W. Steven Barnett, Ph.D.
National Institute for Early Education Research
www.nieer.org
1
Benefits of Early Education
Many studies in the US and elsewhere


Increased test scores and school success
Better social behavior
Three benefit cost analyses
 Increased earnings and tax revenue
 Lower government costs
 Lower crime costs
2
Three Benefit-Cost Analyses
with Disadvantaged Children
Abecedarian Chicago
High/Scope
Year began
1972
1985
1962
Chapel Hill, NC Chicago, IL Ypsilanti, MI
Location
Sample size
111
1,539
123
Matched
Design
RCT
RCT
neighborhood
Ages
Program
schedule
6 wks-age 5
Ages 3-4
Full-day, year Half-day,
round
school year
Ages 3-4
Half-day,
school year
3
High/Scope Perry Preschool:
Educational Effects
Program group
No-program group
15%
Special Education
(Cog.)
34%
49%
Age 14 achievement
at 10th %ile +
15%
66%
Graduated from high
school on time
0%
45%
10%
20%
30%
40%
50%
60%
70%
4
High/Scope Perry Preschool:
Economic Effects at Age 27
Program group
No-program group
Earn $2,000 +
monthly
29%
7%
36%
Own home
13%
41%
Never on welfare as
adult
0%
20%
10%
20%
30%
40%
50%
5
High/Scope Perry Preschool:
Economic Effects at 40
Source: Schweinhart et al., 2005
Program group
No-program group
60%
Earned > $20K
40%
76%
Employed
62%
76%
Had Savings
Account
0%
50%
10%
20%
30%
40%
50%
60%
70%
80%
6
High/Scope Perry Preschool:
Arrests per person by age 27
Felony
Program
0.7
No program
1.2
Misdemeanor
0.5
1.5
0.0
Juvenile
2.3 arrests
2.5
1.0
2.0
0.6
3.0
4.0
4.6 arrests
5.0
6.0
7
Perry Preschool: Crime Effects at 40
Source: Schweinhart et al. 2005
Program group
No-program group
36%
Arrested > 5X
55%
33%
Violent Crime
48%
14%
Drug Crime
34%
0%
10%
20%
30%
40%
50%
60%
8
Abecedarian : Academic Benefits
Program group
No-program group
25%
Special Education
48%
31%
Grade Repeater
55%
67%
HS Graduation
51%
36%
4 Yr College
0%
13%
10%
20%
30%
40%
50%
60% 70%
80%
9
Abecedarian Reading Ach. Over Time
READING SCORE
105
100
95
90
85
TREATMENT
CONTROL
80
8
10
12
14
16
18
20
22
AGE (Years)
10
Abecedarian Math Achievement Over Time
MATH SCORES
105
TREATMENT
CONTROL
100
95
90
85
80
8
10
12
14
16
18
20
22
AGE (Years)
11
Chicago CPC: Academic and
Social Benefits at School Exit
Program group
No-program group
50%
HS Graduation
39%
14%
Special Education
25%
23%
Grade Repeater
38%
17%
Juvenile Arrest
0%
25%
10%
20%
30%
40%
50%
60%
12
Economic Returns to Pre-K
for Disadvantaged Children
Cost
Benefits
B/C

Perry Pre-K
$16,264
$277,631
17.07

Abecedarian
Chicago
$36,929
$ 7,417
$139,571
$ 52,936
3.78
7.14

13
Other Relevant RCT’s
with Long-Term Follow-UP





Milwaukee Project
CARE
IHDP (age 18 follow-up, not disadv.)
Houston PCDC
Mauritius Preschool Study
14
15
Could Universal Pre-K Produce
Similar Benefits for the Middle Class?
Middle class children have fairly high rates of the
problems that preschool reduces for low-income
children.
Reducing these problems could generate large benefits.
Income
Lowest 20%
20-80%
Highest 20%
Retention
17%
12%
8%
Dropout
23%
11%
3%
Source:US Department of Education, NCES (1997). Dropout rates in the United
States: 1995. Figures are multi-year averages.
16
Cognitive Readiness Gap—Half as Big at
Median as for the Poor (bottom 20%)
Abilities
Scores
Abilities of Entering Kindergarteners by Family Income-National Data, Fall 1998 (reported by NIEER from ECLS-K)
60.0
55.0
Reading
School Readiness Gap
Math
50.0
General
Knowledge
45.0
40.0
Low est 20%
2nd Low est
20%
Middle 20%
2nd Highest
20%
Highest 20%
Fam ily
Incom e
17
Social Readiness Gap
Social
Scores
Social Skills of Entering Kindergarteners by
Family Income
9.80
9.60
9.40
School Readiness Gap
9.20
9.00
Social
Skills
8.80
8.60
8.40
Family
Income
8.20
8.00
Lowest 20%
2nd Lowest 20%
Middle 20%
2nd Highest 20%
Top 20%
18
Effects of Today’s Programs






Two new rigorous state pre-K studies
National Impact Study of Head Start
Michigan School Readiness Program
Cost Quality and Outcomes
NICHD Child care Study
EPPE in England
19
Oklahoma’s Universal Pre-K




3,028 children in Tulsa public schools
Rigorous RD design
Gains for all SES & ethnic groups
Literacy and Math gains



Smaller than Perry and Abecedarian
Similar to CPC
Larger gains for minority and poor children
Source: Gormley et al. (2004). CROCUS/Georgetown University
20
NIEER Evaluation of
5 State Pre-K Programs






5,071 children in 5 States
OK and WV are universal
MI, NJ, & SC targeted
Gains from Pre-K in all 5 states
Gains in language, literacy & math
All children gain, low-income gain more
Source: Barnett et al. (2005). NIEER/Rutgers University
21
Is Targeting More Cost-Effective?
Targeting is costly and imperfect



Poverty is a moving target
Need is not defined by targeting alone
Targeting is not perfect
Benefits do not stop at the poverty line


Middle class has similar problems
Benefits decrease gradually with income
22
Economic Comparison of
Targeted and Universal Pre-K
Targeted Programs Have Lower Total Cost
Universal Programs Have Higher Benefits
--they can reach more of the target children
--greater diversity in the classroom increases gains for
disadvantaged children
--some benefits gained for all or most children
Under Plausible Assumptions Universal is Better
Investment
Source: Barnett (2004). Maximizing returns from pre-kindergarten
education. Federal Reserve Bank of Cleveland Research Conference.
23
High Quality Preschool Programs
Needed to Produce Benefits
 Well-educated preschool teachers
 Adequate teacher compensation
Small classes and reasonable teacher:child ratios
Strong supervision
High standards and accountability
24
Conclusions

Pre-K for all can be a sound investment

Too few children attend now

Quality and intensity is too low

Universal can be more cost-effective

We can target within universal
25