The F. A. Hayek Program for Advanced Study in Philosophy, Politics, and Economics welcomed Roland G. Fryer, Jr., the Henry Lee Professor of Economics at Harvard University and faculty director of the Education Innovation Laboratory, for the inaugural Buchanan Speaker Series event on “Education, Inequality, and Incentives.”

2. Education, Inequality, and Incentives
Roland G. Fryer, Jr.
Henry Lee Professor of Economics
Harvard University
HGSE, EdLabs, NBER

3. CONFIDENTIAL
Among cities that participate in NAEP, the magnitude of racial differences in educational
achievement is startling.
Overview
The Achievement Gap
0
10
20
30
40
50
60
70
80
90
100
Percent Proficient, 8th Grade Math, NAEP 2011
white
black
Hispanic
0
10
20
30
40
50
60
70
80
90
100
Percent Proficient, 8th Grade Reading, NAEP 2011
white
black
Hispanic

4. CONFIDENTIAL 3
Why EdLabs Was Founded
Accounting for educational achievement drastically reduces racial and socioeconomic inequality
across a wide range of important life outcomes.
28%
190%
283%
250%
87%
234%
-27%
141%
0.6%
90% 87%
76%
42%
114%
137%
33%
-50%
0%
50%
100%
150%
200%
250%
300%
Wages Unemployment Have Savings Less than 10K in
savings
Negative Net
Worth
Do not own home Any College Public Assistance
Black-White Differences in Economic Outcomes (NLSY79)
Before and After Controlling for 8th Grade Test Scores
Raw B/W Gap After controlling for AFQT

5. CONFIDENTIAL 4
Why EdLabs Was Founded
126%
20%
820%
514%
0%
200%
400%
600%
800%
1000%
0%
20%
40%
60%
80%
100%
120%
140%
Teen Pregnancy Child out of wedlock
Social Outcomes
Raw B/W Gap After controlling for AFQT
0.20 SD
0.059 SD
0.00
0.05
0.10
0.15
0.20
0.25
SF-12 Mental and Physical Health Index
Health Outcomes
Raw B/W Gap After controlling for AFQT
68%
182%
25%
38%
0%
50%
100%
150%
200%
Violent Crime Incarceration
Crime Outcomes
Raw B/W Gap After controlling for AFQT
0.28
-0.27
0.006
1.37
-0.5
0
0.5
1
1.5
Wages Any College
Economic Outcomes
Raw B/W Gap After controlling for AFQT

6. CONFIDENTIAL
22.3 :1
20.4 :1
18.7 :1
17.9 :1
17.2 :1 17.3 :1
16.0 :1 15.6 :1
14:1
16:1
18:1
20:1
22:1
24:1
1970 1975 1980 1985 1990 1995 2000 2005
Student to Teacher Ratio
$5,243
$6,049 $6,268
$7,347
$8,790 $8,949
$10,508
$11,438
$0
$2,000
$4,000
$6,000
$8,000
$10,000
$12,000
1970 1975 1980 1985 1990 1995 2000 2005
Total Expenditure Per Pupil (2008-09 $)) $12,11
6
23.5% 27.5%
49.6% 53.1% 56.8% 61.8%
0%
20%
40%
60%
80%
100%
1961 1971 1981 1991 2001 2006
Percentage of Teachers with a
Master's Degree or Higher
175
200
225
250
275
300
325
1971 1975 1980 1984 1988 1990 1994 1996 1999 2004 2008
Reading and Math Achievement of 9, 13, and 17 year-olds,
1971-2008
9 year-
olds
13 year-
olds
17 year-
olds
Conventional Wisdom Has Failed

7. CONFIDENTIAL
The Problem: Data Sifters
• Social scientists have spent decades trying to tease
out causal relationships from non-experimental data
Major Approach
• Observational Data: examine mounds and mounds
of data, invoking various assumptions, in an effort to
make causal statements
– Legalized abortion lowered crime
– The Endangered Species Act endangers species
How would one go about using this approach in
education?

8. CONFIDENTIAL
A= f(C,H,S,P)
A = achievement
C = the child’s inputs
H = the household’s inputs
S = the school’s inputs
P = the relevant prices
To use economic principles, we need to know the marginal value of
each factor and the marginal cost of provision
The Education Production Function

9. CONFIDENTIAL
• Basic economic principles are not currently being
applied (e.g. MC=MR)
• A new crop of visionary leaders are eager to
implement change (e.g. Arne Duncan, Terry Grier)
• Observational data have not been able to provide
these new leaders with the answers they need to
revolutionize the production of human capital
Field Experiments
Balancing “Unobservables”

10. CONFIDENTIAL
My First Idea:
Maybe we should give kids incentives for
achievement…
Incentives

11. CONFIDENTIAL
-0.5
0
0.5
1
1.5
2
2.5
Dallas: Reading Books D.C.: School-Determined
Inputs
Houston: Math
Assignments
NYC: Test Results Chicago: Report Card
Grades
NYC: Schoolwide Test
Scores
Impact of Incentive Programs on Student Achievement
Incentives
Months of Schooling
Teacher
Incentives
Student Output
Incentives
Student Input Incentives

12. CONFIDENTIAL
My Second Idea:
Maybe we should give teachers incentives
for achievement…

13. CONFIDENTIAL
-0.7
-0.6
-0.5
-0.4
-0.3
-0.2
-0.1
0
NY State ELA NY State
Math
Regents ELA Regents
Math
Months of Schooling
Treatment Effects of Teacher Incentive Program : NYC
Elementary
Middle School
High School
Incentives

14. CONFIDENTIAL
0
0.05
0.1
0.15
0.2
0.25
Loss Treatment Gain Treatment Any Treatment
Math Scores
Treatment Effects of Teacher Incentive Program : Chicago
Incentives

15. CONFIDENTIAL
My Third Idea:
Maybe we should give parents incentives for student
achievement…

16. CONFIDENTIAL
0
0.05
0.1
0.15
0.2
0.25
Standardized Cognitive
Score
Standardized Non-
Cognitive Score
LATE Treatment Effect (in standard
deviations)
Treatment Effects in Parent Academy
"Cash" Treatment
"College" Treatment
Pooled Treatment

17. CONFIDENTIAL
My 379th Idea:
Copy someone else…
Results From High-Performing Charters

18. CONFIDENTIAL
Results From High-Performing Charters
Harlem
Children’s
Zone

19. CONFIDENTIAL
Charter Results
-0.1
-0.05
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
Math
Reading
A. Broad Surveys B. High Performers

20. CONFIDENTIAL
Finding the Vaccine
-0.41
0.01
-0.40
-0.45
-0.31
0.79 0.75
0.93
0.80
0.70
0.59
-0.50
-0.25
0.00
0.25
0.50
0.75
1.00
Months of Schooling
Traditional vs. Non-Traditional School Inputs
and School Effectiveness
Average Correlation with Reading and Math Effectiveness (in months of schooling)
vs.

21. CONFIDENTIAL
Meta-Analysis: Method for Finding Field Experiments
• What Works Clearinghouse - single study and quick reviews that
met standards without reservations
• Education Literature Reviews (Almond & Curie; Fryer; Heckman &
Kautz; Nye, Turner, & Schwartz; Yeager & Walton)
• Searching known databases (e.g. ERIC, JSTOR, EconLit)
• Narrowed our focus to studies that satisfied the following
criteria:
1. Randomized Controlled Trials
2. Treatments affect students between ages 0 - 18 (i.e.
excluding students in colleges)
3. Student achievement in math and/or reading tests as
outcome variables
4. State tests or norm referenced tests
5. Highly Developed Countries (top quartile of the HDI)

22. CONFIDENTIAL
Meta-Analysis: Papers Found
Paper Accounting
Number of Papers
Hits From Broad Search ⇡ 8,000
Selected for Further Review 838
College Sample/Outcomes -42
Design Issues -87
Countries w/o Very High HDI -57
Insufficient Info -22
Paper Not Located -10
No Standardized Reading or Math -337
Repeat Paper -71
Sample Issues -18
Total Included 194

23. CONFIDENTIAL
Meta-Analysis Results

24. CONFIDENTIAL
The key goal is to translate charter schools’ successful policies into common principles and then transplant them into
traditional public schools. To this end, EdLabs initiated a multi-year study of NYC charters that determined that the
following five policies and practices have the greatest correlation with student achievement:
More Time in School
• Extended day, week, and school years are all integral components of successful school
models. In the case of Harlem Children’s Zone’s PromiseAcademy, students have
nearly doubled the amount of time on task compared to students in NYC public schools.
Small Group Tutoring
• In top performing schools, classroom instruction is supplemented by individualized
tutoring, both after school and during the regular school day.
Human Capital Management
• Successful charters reward teachers for performance and hold them accountable if they
are not adding value.
Data Driven Instruction and Student Performance Management
• In the top charter schools, students are assessed frequently, and then, in small groups,
re-taught the skills they have not yet mastered.
Culture and Expectations
• In successful schools, students buy into the school’s mission and into the importance of
their education in improving their lives.
Edlabs’ Research Findings: Finding the Vaccine

25. CONFIDENTIAL
-0.1
-0.05
0
0.05
0.1
0.15
0.2
0.25
0.3
Average
Charter
National
Average NYC
Charter
AUSL
Turnaround
Harlem
Children's
Zone (MS)
Average KIPP
(MS)
Apollo MS/HS Apollo ES Denver
Summit
Schools
Treatment Effects in Context
Math Reading
Turnarounds In Context

26. CONFIDENTIAL
Long-Term HCZ Effects
0
5
10
15
20
25
Enrolled in College Enrolled in 4-year College
College-Going
0
0.05
0.1
0.15
0.2
0.25
0.3
Math Reading
Woodcock Johnson
-0.06
-0.04
-0.02
0
0.02
0.04
0.06
0.08
0.1
0.12
Mental Health Physical Health Index Nutrition Index
Health Outcomes
0
1
2
3
4
5
6
Pregnancy (Girls) Incarceration (Boys)
Social Outcomes (Odds Ratios)