Successfully reported this slideshow.
We use your LinkedIn profile and activity data to personalize ads and to show you more relevant ads. You can change your ad preferences anytime.

Inequality in the 21st Century: The Declining Significance of Discrimination

3,511 views

Published on

Roland Fryer (Harvard University), winner of the Calvó-Armengol International Prize, delivered this lecture at the Barcelona Graduate School of Economics on
May 31, 2012 as part of the Prize activities.

About the Calvó Prize: http://www.barcelonagse.eu/Calvo-Armengol-Prize.html

  • Be the first to comment

  • Be the first to like this

Inequality in the 21st Century: The Declining Significance of Discrimination

  1. 1. Inequality in the 21st Century:The Declining Significance of Discrimination Roland G. Fryer, Jr. Robert M. Beren Professor of Economics Harvard University EdLabs NBER
  2. 2. Overview The Achievement GapAmong cities that participate in NAEP, the magnitude of racial differences in educationalachievement is startling. Percent Proficient, 8th Grade Math, NAEP 2011 100 90 80 70 60 50 40 30 20 white 10 0 black Hispanic Percent Proficient, 8th Grade Reading, NAEP 2011 100 90 80 70 60 50 40 30 20 white 10 0 black Hispanic CONFIDENTIAL
  3. 3. Overview Education and Later-Life Outcomes Accounting for educational achievement drastically reduces racial and socioeconomic inequality across a wide range of important life outcomes. Black-White Differences in Economic Outcomes (NLSY79) Before and After Controlling for 8th Grade Test Scores300% 283% 250%250% 234%200% 190%150% 137% 141% 114%100% 90% 87% 87% 76%50% 42% 28% 33% 0.6% 0% -27%-50% Wages Unemployment Have Savings Less than 10K in Negative Net Do not own home Any College Public Assistance savings Worth Raw B/W Gap After controlling for AFQT CONFIDENTIAL 3
  4. 4. Overview Education and Later-Life Outcomes Social Outcomes Health Outcomes140% 126% 1000% 0.25 820% 0.20 SD120% 800% 0.20100%80% 514% 600% 0.1560% 400% 0.1040% 0.059 SD 20% 200%20% 0.05 0% 0% 0.00 Teen Pregnancy Child out of wedlock SF-12 Mental and Physical Health Index Raw B/W Gap After controlling for AFQT Raw B/W Gap After controlling for AFQT Crime Outcomes 200% 182% 150% 100% 68% 50% 38% 25% 0% Violent Crime Incarceration Raw B/W Gap After controlling for AFQT CONFIDENTIAL 4
  5. 5. Overview Economic Impacts What if we could Nearly 30% of the populationhave eliminated theracial achievement gap by 1998? is The racial achievement gap cost 0.8 s.d. behind $310 -$525 billion in lost GDP in 2008 their white peers in math --------------------------------------------------------- and reading And $1.5 – $2.5 trillion cumulatively from 1998-2008 and there is a 1.2%- 2% growth premium for a 1 s.d. increase in academic achievement Based upon Calculations from McKinsey (2009) CONFIDENTIAL 5
  6. 6. 400 450 500 550 600 $2,000 $4,000 $6,000 $8,000 $10,000 $12,000 $14,000 $16,000 South Korea Luxembourg Finland Switzerland Switzerland Source: Education at a Glance 2010: OECD Indicators Japan Norway Canada United States Netherlands Austria New Zealand Denmark Belgium Source: OECD, Program for International Student Assessment (PISA), 2009 Iceland Australia Sweden Germany United Kingdom Estonia Netherlands Iceland Denmark Belgium Slovenia France Norway Canada France Japan Slovak RepublicCONFIDENTIAL Ireland Austria Italy Poland Spain Sweden Australia Czech Republic United Kingdom Germany Hungary Finland Luxembourg Annual Expenditure Per Student, 2007 South Korea United States vs. OECD Countries United States Portugal Ireland New Zealand Mathematics Literacy Among 15-Year-Olds, 2009 Portugal Czech Republic Spain Hungary Italy Poland Greece Israel Slovak Republic Turkey Chile Chile OECD Average OECD Average Mexico Mexico International Results Overview6
  7. 7. Overview Economic ImpactsWhat if we could On average, American have closed the students lag 0.75 international s.d. behind their top rankedachievement gap international peers on the PISA test By 1998? The international achievement gap cost $1.3 -$2.3 trillion in lost GDP in 2008 there is a 1.2%- 2% growth premium for a 1 s.d. increase in academic achievement Based upon Calculations from McKinsey (2009) CONFIDENTIAL 7
  8. 8. Basic Facts• Test Score Gap Does Not Exist at 9 months old• The correlation between 9 month old scores and 12 year old scores is 0.3• Black kids lose ground starting at age 2 CONFIDENTIAL
  9. 9. Basic Facts• Black kids enter kindergarten 0.64 SD (or 8 months) behind their white peers• The gap can be accounted for by 13 simple variables that proxy for Pre-K home environment• The gap grows 0.1 SD per year from Kindergarten through eighth grade• We don’t really know why The Evolution of the Racial Achievement Gap through 8th grade 14 12 10 8 6 4 2 0 9 months Pre-K Kindergarten 1st grade 3rd Grade 5th Grade 8th grade Months behind in math Months behind in reading CONFIDENTIAL
  10. 10. Possible ExplanationsWe tested dozens of hypotheses, including:• Poor parenting• Racist Teachers• Summer Setback• Flawed Standardized Tests• School Quality How parents respond to Lost ground from Fall Kindergarten to Spring 1st grade, before and after accounting for parenting practices, misbehavior, by race ECLSK 0.280% 75%70% 0.15 0.103 0.09760% 0.093 0.09 49% 48% 0.150% 45% 41%40% 0.0530% White 16% 17% Lost Ground20% Black 0 9%10% Reading Reading, Math Math, Hispanic accounting for accounting for 0% parenting parenting Spank child Give child a "time-out" Asian CONFIDENTIAL
  11. 11. Conventional Wisdom Seems Ineffective Percentage of Teachers with a Total Expenditure Per Pupil (2008-09 $)) Masters Degree or Higher $12,11 $11,438 6 $12,000 $10,508100% $10,000 $8,790 $8,949 80% $8,000 $7,347 60% $6,268 $6,049 $6,000 $5,243 40% 53.1% 56.8% 61.8% 49.6% $4,000 20% 23.5% 27.5% $2,000 0% 1961 1971 1981 1991 2001 2006 $0 1970 1975 1980 1985 1990 1995 2000 2005 Student to Teacher Ratio24:1 Reading and Math Achievement of 9, 13, and 17 year- 325 22.3 :1 olds, 1971-200822:1 300 9 year- 20.4 :1 olds 275 13 year-20:1 18.7 :1 250 olds 17.9 :1 17 year-18:1 17.2 :1 17.3 :1 225 olds 16.0 :1 15.6 :116:1 200 17514:1 1971 1975 1980 1984 1988 1990 1994 1996 1999 2004 2008 1970 1975 1980 1985 1990 1995 2000 2005 CONFIDENTIAL
  12. 12. Financial IncentivesDistributed a total of $10 million to kids in 5 cities.A. Input Experiments • Dallas • Houston • Washington DCB. Output Experiments • New York City • Chicago • Teacher Incentives CONFIDENTIAL
  13. 13. Financial IncentivesCONFIDENTIAL
  14. 14. Financial IncentivesCONFIDENTIAL
  15. 15. The Achievement Gap 1 Past Interventions vs. the Racial Achievement GapAnnual Treatment Effect on Student Achievement (in SD 0.8 0.6 0.4 units) 0.2 0 -0.2 CONFIDENTIAL
  16. 16. Results From High-Performing Charters HarlemChildren’s Zone CONFIDENTIAL
  17. 17. Results From Charter Schools 0.6 0.5 B. Studies of High-Performing Schools 0.4 0.3Math A. Broad Surveys 0.2Reading 0.1 0 -0.1 Notes: Solid bars represent experimental estimates. Striped bars represent quasi-experimental estimates. *Oversubscribed Schools CONFIDENTIAL only.
  18. 18. Finding the Vaccine Traditional vs. Non-Traditional School Inputs and School Effectiveness 1.00 0.93 0.79 0.80 0.75 0.75 0.70 0.59Months of Schooling 0.50 0.25 vs. 0.01 0.00 -0.25 -0.31 -0.50 -0.41 -0.40 -0.45 Average Correlation with Reading and Math Effectiveness (in months of schooling) CONFIDENTIAL
  19. 19. An Experiment in Houston: The Five TenetsThe key goal is to translate charter schools’ successful policies into common principles and then transplant them intotraditional public schools. To this end, EdLabs initiated a multi-year study of NYC charters that determined that thefollowing 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 Promise Academy, 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. CONFIDENTIAL
  20. 20. Implementation Increased Time in School The school day was extended in Apollo schools during the 2010-11 school year: 7:45am – 4:15pm Monday through Thursday, and 7:45am – 3:15pm on Fridays. This was an average of an hour longer per school day. The school year was extended by five school days. Apollo students reported for school on August 16, 2010, while the rest of the district began on August 23, 2010. A: Instructional Hours per Year 1800 1600 1400 1200 1000 800 600 400 200 0Bottom line: The difference between instructional time in 2009-10 and 2010-11 amounts toapproximately 30 school days – that’s 6 additional weeks of school for students. CONFIDENTIAL
  21. 21. Implementation Human Capital In addition to finding nine new principals, teacher turnover spiked to 53% in Apollo schools over the summer of 2010. Value-added data shows that teachers who returned as Apollo teachers had a much stronger history of increasing student achievement in every subject, relative to those who left. Teacher Departure Rates Teacher Value Added 0.260% 0.150% 040% Language Math Reading Science Social Apollo -0.1 Studies Schools30% -0.2 Left -0.3 Stayed20% Comparison -0.410% Schools -0.50% -0.6 2005-06 2006-07 2007-08 2008-09 2009-10 -0.7 CONFIDENTIAL
  22. 22. Implementation High Dosage Differentiation: Tutoring and Double-Dosing•All sixth and ninth grade students received daily 2:1 tutoring in math•Seventh, eighth, tenth, and eleventh graders received an extra reading or mathcourse if they had tested behind grade level in the previous year• All told, middle school students received approximately 215 hours oftutoring/double-dosing, and high school students received 189. CONFIDENTIAL
  23. 23. Implementation Data-Driven Instruction•In addition to required HISD assessments, Apollo schools administered twoadditional comprehensive benchmark assessments in four core subjects:math, reading, science, and social studies.•After each assessment, teachers received student-level data and used this tohave one-on-one goal-setting conversations with students. CONFIDENTIAL
  24. 24. Implementation Culture and ExpectationsAt the end of the 2009-10 school year, The New Teacher Projectinterviewed all teachers in what would become Apollo schools. Those whoreturned for the 2010-11 year showed a demonstrably strongercommitment to the Apollo 20 philosophy. Interview Responses 3.5 3 2.5 2 1.5 Left Stayed 1 0.5 0 No Excuses Alignment Student Commitment Student with Mission Achievement to Students Motivation CONFIDENTIAL
  25. 25. Implementation Culture and ExpectationsReports from student focus groups provide a lens into the culture shift.1. Pre-Treatment: There were lots of fights and “wilding out” all the time. Teachers didn’t give homework. People just showed up and basically went through the motions. Observers noted rowdy hallways, messing around, not taking school very seriously.2. Treatment Fall: The extended school day was a big shift. Constant complaints of exhaustion. Everyone’s tired. The students are tired. The teachers are tired.3. Treatment Spring: “The food in the cafeteria sucks.” “I had a hamburger that wasn’t any good.” • Student: “The apples taste like soap.” • Project Manager: “Next time I visit I’ll figure out why the apples tasted like soap.” From a teacher in Fall 2011: “The sixth graders from last year who are seventh graders now have started to shift the whole school culture. The climate is really changing – it’s calmer everywhere, and there are no more fights.” CONFIDENTIAL
  26. 26. First-Year Results In Math, we see positive and statistically significant results in both middle and high school. The gains in grades that received high-dosage tutoring were dramatic. The reading results are mixed. While high schools performed extremely well, there is little evidence of success in middle school – indeed some estimates are negative. Apollo Treatment Effects0.8 0.7260.70.6 0.4840.50.4 0.363 0.277 Math Reading0.3 0.237 0.215 0.165 0.1880.2 0.125 0.113 0.1160.1 0.062 0 6th 7th and 8th -0.008 All Middle 9th 10th and 11th All High Pooled-0.1 -0.067-0.2 CONFIDENTIAL
  27. 27. Results In ContextPooling all grades together, the results are strikingly similar to thoseachieved by the Harlem Children’s Zone Promise Academy Middle Skilland KIPP – two of the country’s most recognized charter operators. Treatment Effects in Context 0.3 0.277 0.26 0.25 0.229 0.2 0.15 Math 0.09 0.09 Reading 0.1 0.062 0.04 0.047 0.05 0 Average NYC Charter Harlem Childrens Zone Average KIPP (MS) Apollo Year 1 (MS) CONFIDENTIAL
  28. 28. Cost-Benefit AnalysisUsing an estimate of the correlation between test scores and futureearnings, we can calculate a rough rate of return for the first year of theApollo experiment and compare it to other popular educationinterventions. Initiative Cost/Student IRR Apollo 20 $1,837 21.66 % “No Excuses” Charter School $2,496 18.50 % Early Childhood Education $8,879 7.6 % Reduced Class Size $3,501 6.20 % CONFIDENTIAL
  29. 29. The Path ForwardThis year, we’re expanding the program to:1. Eleven randomly selected elementary schools in Houston2. Seven schools – in their own feeder pattern – in Denver.The model is far from perfect, but it’s enough to get started.We can save 10 million kids by 2020. CONFIDENTIAL

×