We have shown that acts are not purely random .... Used a major credit cardâVisa, Mastercard, etc.? â« Used a major .
Using Behavioral Economics and Field Experiments To Make The World A Better Place
John A. List University of Chicago, NBER, and RFF
R.I.P.
Rey Dorantes, 14
Antonio Fenner, 16
Oscar Marquez, 17
A Problem for the Other Sex
Teen Births in Developed Countries
US: Similar to Rwanda and Sudan 70
Preg per 1,000 women (15-19)
60
50
40
30
20
10
0 France
United Kingdom
India
Thailand
Rwanda
USA
Sudan
Source: UNFPA, State of World Population (2003)
Cambodia
What links these 2 Issues? n
We know very little about their causes, and less about useful remedies
n
Using economics and field experiments we can provide solutions
For Teen Shootings n
We have shown that acts are not purely random but predictable with an economic model
n
Now that we can identify those most at risk, we can target dollars to those boys
n
Field experiments have taught us that: ¨ Our
pecuniary and non-pecuniary incentive schemes have not worked ¨ These boys need a ‘complete’ intervention where a father figure enters their lives and stays for the formidable years (age 15-19)
For Teen Births n
Teen births are also predictable with an economic model
n
Field experiments have taught us that: ¨A
mixture of pecuniary and non-pecuniary incentives can help the female optimally time child birth ¨ When we provide reasons to have optimism for the future (e.g., a college tuition fund) the first pregnancy is pushed back considerably
A Big Problem for Both Sexes Public Education
How Do Most Economists Think About Empiricism in Education?
Economists as Data Sifters Economists have a semi-automatic approach of writing down a model, downloading mounds and mounds of secondary data, invoking various assumptions, and making causal statements
Conventional Wisdom n
We need teachers with higher degrees, we need lower student/teacher ratios, and we need to spend more to make public education work
Teachers are receiving more years of education
Percentage of Teachers with a Master's Degree or Higher 100% 90% 80% 70% 60% 50% 40% 30% 20% 10%
23,5%
27,5%
1961
1971
49,6%
53,1%
56,8%
61,8%
1981
1991
2001
2006
0%
13
Student to Teacher Ratios Have Dropped Student to Teacher Ratio 24:1
22,3 :1
22:1
20,4 :1
20:1
18,7 :1 17,9 :1
18:1
17,2 :1
17,3 :1
16:1 14:1 1970
14
1975
1980
1985
1990
1995
16,0 :1
15,6 :1
2000
2005
Spending Has Risen $12.000
$10.508
$10.000 $8.000 $6.000
$12,116
Total Expenditure Per Pupil (2008-09 $)
$6.268 $5.243
$8.790
$8.949
1990
1995
$11.438
$7.347
$6.049
$4.000 $2.000 $0 1970 15
1975
1980
1985
2000
2005
Results Have Not Materialized 100%
75%
HS Graduates as a ratio of 17 year-olds 76,9%
73,6%
71,4%
72,4%
1975
1980
1985
73,4%
69,3%
69,8%
75,4%
50%
25%
0% 1970
16
1990
1995
2000
2005
74,7%
Results Have Not Materialized Reading and Math Achievement of 9, 13, and 17 year-olds 325 300
9 year-olds
275
13 year-olds
250 17 year-olds
225 200 175 1971 1975 1980 1984 1988 1990 1994 1996 1999 2004 2008
17
How to Solve n
We must understand that schools are not just about teaching children. They are about teaching ourselves what works and why
Empirical Alternative? n
Field Experiments
Using the world as a lab…rather than stand by as a passive observer, generate new data Who are the subjects in my field experiments?
Have you n
Flown United Airlines in the past few years?
n
Purchased from Amazon.com in the past 10 years?
n
Used a major credit card—Visa, Mastercard, etc.?
n
Used a major search engine for retail purposes?
We are ALL lab rats now
My Education Lab: Chicago Heights
Chicago
Chicago Heights
Why Chicago Heights? n
A major reason why public education is failing--urban schools
§
Less than 50% High School graduation rate Modal ninth graders reading at a 4th grade level Observables similar to other urban settings across U.S.
§ §
Our Work n
We have conducted field experiments to explore the education production function from pre-k through high school.
n
We have looked at teachers, mentors, families, communities, friend networks, administrators, etc.
24
An Example Within Schools
Teacher performance pay…has become a big hot button issue around the world.
25
U.S. Press: Merit Pay Doesn’t Work n
“Teacher Performance Pay Alone Does Not Raise Student Test Scores” – Vanderbilt News, September 2010
n
“The long, failed history of merit pay and how the Ed Department ignores it” – Washington Post, September 2010
n
“Why Teacher Performance Pay Won't Work” – Huffington Post, November 2012
Press: Merit Pay Doesn’t Work n “$75M
teacher pay initiative did not improve achievement” – Chalkbeat.org, March 2012
n “An
education reform without merit: Performance pay doesn't work, so let's stop wasting money” New York Daily News, January 2014
n “John,
teachers will never respond to merit pay incentives—they value the kids not the money!” CPS Teacher’s Union, dozens of times
…But, teachers do like money!!
n
“Teacher's Union Endorses Bonus Plan Not Based on Performance“ – ABC News, July 2014
Trying Something New with Teachers § Teacher incentives: ¨ If
their students improved test performance teachers could earn as much as $8,000.
¨ The
expected value of the reward was $4,000—an 8% increase in their annual salary.
©John List, University 29 of Chicago
Field Experiment § Framing of rewards ¨Teachers who were randomized into receiving the reward were either given the reward as a gain or a loss.
¨Gains: The reward was distributed at the end of the school year as a bonus.
¨Loss: Teachers received $4,000 at the start of the year and had to write a check to us if they earned less than $4,000. ©John List, University of Chicago
Field Experiment All Teachers Randomization Group 1: Gain, Traditional bonus
Group 3: Control, paid nothing
Group 2: Loss, paid $4000 upfront
Experimental Results: Gain Works A Little Standard Deviations 0,35
0,3
0,25
0,2
0,15
0,1
0,05
0 Gain
Loss
Racial Achievement Gap
©John List, University 32 of Chicago
Experimental Results: Clawback Works Well Standard Deviations 0,35
0,3
0,25
0,2
0,15
0,1
0,05
0 Gain
Loss
Racial Achievement Gap
©John List, University 33 of Chicago
But….One Clear Message n It
is difficult to introduce ‘broad game changers’ late in the student experience
nI
suspect that this much human potential per capita has not been wasted since the dark ages
34
So What Did We Do?
Started Our Own Pre-K!
Chicago Heights Early Childhood Center • Recently finished our 4th year serving in need children aged 3-5
Overarching Experimental Design
3,000+ Children Randomization All Day Pre-K; n=334
Control; n = 1844
Parent Academy; n=372
All Day Pre-K & Parent Academy; n=450
Some Takeaways n
Through field experimentation we have developed a Pre-K curriculum that provides significant expansion of cognitive and executive function (non-cognitive) skills
n
After 4 months in our program the majority of our inner-city kids surpass the average American child in cog. and non-cog test scores
Some Takeaways n
We have developed a parent academy that shows the optimal approach for a number of kids is through their parent (especially Hispanics)
n
We have resources in place to follow every child over the life-cycle, allowing us to track reversion and life-time outcomes
We Have Only Touched the Tip of the Iceberg
Why do women earn less than men? ¨
Early in the job search there are key differences:
v
women shy away from merit pay incentive schemes
v
in ambiguous bargaining situations women will exercise caution whereas men (especially the low skilled types) will ask for higher wages
v
data thus far are consistent with nurture playing an important role
Why Do People Give To Charity? n
Many people give primarily because of selfish motives rather than altruism
n
We can enhance the pool of donors and donations if we understand why people give
Great Big Externalities n
One of the biggest problems humanity faces is climate change
n
How can we induce households to adopt green technologies? ¨ Field
experiments have taught us that social norms work for the very first adoption, prices for deeper adoptions
In Closing: 3 Major Challenges for Policymakers 1.
Evaluating the effectiveness of an implemented policy Did the 2012 Australian Tobacco Plain Packaging / graphic health warning legislation curb smoking?
The Policy
Evidence: Prevalence
47
3 Major Challenges of Policymakers 2.
Projecting the likely effectiveness of a policy in environments different from the one in which it is experienced Would Tobacco Plain Packaging legislation work in Europe?
3.
Forecasting the effects of a new policy, never before seen Would behavioral economic interventions, such as leveraging loss aversion, curb smoking?
Knightian Wisdom
n
Knight (1921): ‘The existence of a problem in knowledge depends on the future being different from the past, while the possibility of a solution of the problem depends on the future being like the past.’
n
A Revision (today): ‘The existence of a problem in knowledge depends on the future being different from the past, while the possibility of a solution of the problem depends on the future being like an experiment.’
How Far Has Economics Advanced in the Last 20+ Years? My Very First Referee Report (1994) n
This author clearly does not understand experimental economics, which is best done in the laboratory…….I strongly advise rejecting this “field experimental” paper.
My most recent referee report
n
This author performs a field experiment, which clearly is the preferred approach to answer questions within economics……. ……but I still advise rejecting the paper