The Today and Tomorrow of Kids

Marco Castillo, Georgia Institute of Technology

Paul Ferraro, Georgia State University

Jeff Jordan, University of Georgia

Ragan Petrie, George Mason University

Purpose

Objective:  To understand how school children view the future.

Goal: To use this understanding to reduce the high school drop out rate.

Result: Inform policies and programs on improving educational outcomes.

Time Preference Instrument

  • Decide between $49 in one month or $49+$X in seven months.
  • Measure of time preferences or discount rate.
    • Implied interest rate you would need to be willing to wait for higher amount of money.
    • Also called impatience.
  • 866 8th-graders (13-15 yrs) in Griffin, GA.
    • 3 cohorts, entire school district.
  • 20 separate decisions ($50.83 to $98.02).
  • Paid with Wal-Mart gift cards.
  • Payment private, by school principal.
  • Switch point implies discount rate range.

Basic Result 1

Discount Rates
Male 90.2
Female

Blacks

Whites

White males

White females

Black males

Black females

81.1

91.0

81.4

83.9

78.7

97.4

85.4

Males have higher discount rates than females.

Blacks have higher discount rates than whites.

White males and white females are not significantly different.

Black males are more impatient than black females.

Black boys are most impatient.

Due to risk preferences?

  • May appear to have larger discount rates, when it is really risk aversion.
  • This would imply that black boys are more risk averse than any other group.
  • Collected data on risk preferences in same population to check.
    • Experiment used decisions over 5 lotteries ($25 for sure, increasing in value and risk) – subjects choose one (A, B, C, D or E) – E most risky.
A B C D E
1 $25 1 $40 1 $55 1 $70 1 $85
2 2 2 2 2
3 3 3 3 3
4 4 4 4 4
5 5 5 5 5
6 6 6 6 6
7 7 7 7 7
8 8 8 8 8
9 9 9 9 9
10 10 10 10 10
11 $25 11 $20 11 $15 11 $10 11 $5
12 12 12 12 12
13 13 13 13 13
14 14 14 14 14
15 15 15 15 15
16 16 16 16 16
17 17 17 17 17
18 18 18 18 18
19 19 19 19 19
20 20 20 20 20

Basic Result 2

Black White
Decision Female Male Female Male
1 36 25% 28 19% 34 26% 24 16%
2 24 17% 21 14% 42 32% 36 25%
3 21 15% 30 20% 25 19% 27 18%
4 21 15% 16 11% 12 9% 26 18%
5 39 28% 54 36% 17 13% 33 23%
Average 3.0 3.3 2.5 3.0
Obs 141 149 130 146

Basic Result 3

Black boys have larger discount rates, even when controlling for other variables (including math & reading skills, free & reduced lunch).

This summer we will conduct a phone survey of all families of 8th graders to find out more about what home factors may affect discount rates.

Basic Result 4

  • Discount rates are positively correlated with outcomes (disciplinary referrals) even controlling for other variables that may affect discipline.
  • Disciplinary referrals relate to dropping out of school and lower income later in life.
    • In Spalding school district, among drop-outs, 4.6 referrals.
    • Among our subjects, 1.8 referrals.

Inconsistencies in Risk

(some preliminary evidence)

  • Interested in whether kids exhibit rational preferences over risk.
  • Are they expected utility maximizers?
  • Or, something else?

Risk Preferences Instrument

  • Measures risk preferences and choice consistency.
  • 6 separate decisions over lotteries.
    • One decision over 5 lotteries.
    • Five decisions over $0, $30, $40 (Chew & Waller, 1986, design).
  • Paid with Wal-Mart cards.
  • Payment private, by school principal.
  • 603 8th-graders (13-15 yrs) in Griffin, GA.

Lottery 1

A B
1 $30 1 $30
2 2
3 3
4 4
5 5
6 6
7 7
8 8
9 9
10 10
11 11
12 12
13 13
14 14
15 15
16 16 $0
17 17 $40
18 18
19 19
20 20

Lottery 2

A B
1 $0 1 $0
2 2
3 3
4 4
5 5
6 6
7 7
8 8
9 9
10 10
11 11
12 12
13 13
14 14
15 15
16 $30 16
17 17 $40
18 18
19 19
20 20

Lottery 3

A B
1 $30 1 $0
2 2
3 3
4 4
5 5 $40
6 6
7 7
8 8
9 9
10 10
11 11
12 12
13 13
14 14
15 15
16 16
17 17
18 18
19 19
20 20

Lottery 4

A B
1 $30 1 $0
2 2
3 3
4 4
5 5 $40
6 6
7 7
8 8
9 9
10 10
11 11
12 12
13 13
14 14
15 15
16 $0 16
17 $40 17
18 18
19 19
20 20

Lottery 5

A B
1 $30 1 $0
2 2 $40
3 3
4 4
5 5
6 $40 6
7 7
8 8
9 9
10 10
11 11
12 12
13 13
14 14
15 15
16 16
17 17
18 18
19 19
20 20

Basic Result 5

  • Black and white students are different in how they make decisions.
  • Black boys and girls exhibit “pessimistic weighting.”
    • Weight bad outcomes higher and good outcomes lower.

Discipline

Discipline referrals incurred by a child have been found to be a good predictor of the child’s decision to drop out of school and of lower average lifetime earnings.

Hypotheses

  • The odds of a student being referred for disciplinary action increases if the student is male, black, in special education classes, or is poor.
  • The gender and race of the teachers who refer students for disciplinary action have a significant impact on the first hypothesis.

Basic Result 6

  • As hypothesized:  the odds of a student being referred for disciplinary action increases if the student is male, black, chronically absent, or is poor.
  • Gender of teacher was not significant but race was although not in the way hypothesized.
  • Black female teachers were more likely to discipline students.
  • Special teacher/student dynamic between black males.
  • Further research will link this to student impatience and other educational outcomes such as dropping out.

Implications

  • We see differences in how children view the future as well as how they view likely outcomes.
  • We see connections between impatient decision making and discipline.
  • Does impatience cause discipline problems for black boys or does the disproportionate application of discipline over the course of a school career, even if administered by a black or male teacher, affect how African American children see the future?
  • One implication of this research is the need for more immediate rewards than those long-term rewards suggested by high school graduation.
  • Differences in discount rates may mean changing the level of those rates is less important than making rewards more immediate and sequential.
  • If preferences are stable from a young age, uniform incentive and rewards programs may have the perverse effect of making the gaps between children larger.
  • Children who would already excel and graduate will take further advantage of incentives and reinforce gaps and pessimistic outcomes.
  • This implies the need for targeted incentives.
  • All of this argues for finding out when children form their preferences and risk perceptions and what is behind that formation.
  • This will then require the changing of these discount rates rather than just designing new incentive programs that increase gaps.

Dropping Out: Does Location Matter?

Jeff Jordan, University of Georgia

Purpose

The purpose of this paper is to:

  • Explore the determinants of dropping out.
  • Examine whether differences in graduation rates and basic determinants have changed since the 1980’s.

Our Approach

We use recent and past national representative data sets to provide an in depth analysis of high school dropout rates in the U.S.

Specifically we:

  1. Use recent, geo-coded nationally representative data from the National Longitudinal Survey of Youth in 1997 (NLSY97)
  2. Use geo-coded household level data from a similar cohort of youth as the NLSY97 but who attended high school in the late 70’s and early 80’s (NLSY79) to examine whether the graduation rates and determinants have changed over the last 30 years.

Data

NLSY79: 12,686 young men and women who were 14-22 years old when first surveyed in 1979, and surveyed annually through 1994 (biannually since).

NLSY97: 9,000 youths, 12-16 years old by 1997.

Data Grouping

Based on the Beal Codes we grouped individuals in four categories: those living in:

cities (zone 1) suburban/metro area (Zone 2) adjacent non-metro (Zone 3) remote non-metro (Zone 4).

High School Graduation Rates (%)

NSLY79

All      Zone 1   Zone 2   Zone 3   Zone 4

80.3    80.1       81.1        80.1        78.6

NSLY97

77.1    76.9       77.3        77.1        77.1

Determinants of Graduation

Testing for:

  1. Individual and family characteristics (race, gender, income, “hard times”, net worth, household size and composition, relation of youth to adults in house, education of parents, poverty ration, attending public school).
  2. Peer characteristics:
    • Percent of peers that are part of a gang
    • Do drugs
    • Skip class
    • Want to attend college
  3. Geographic context:
    • Total population
    • Crime rates
    • Median family income
    • Racial composition
    • Unemployment
    • Employment by sector

Results

  1. High School dropout rates are similar throughout the urban-rural continuum.
  2. Overall, dropout rates are three percentage points higher in the 2000s than in the 1980s across all locations.
  3. The determinants of dropping out are also similar across the urban-rural continuum.
  4. Family level characteristics are far more predictive of dropping out than geographic attributes and appear to operate in similar ways across locations.

These determinants are gender, race (in some cases), family assets, the presence of both biological parents, maternal attributes, and peer characteristics.

  1. The biggest exception, after adding all control variables, is that Black and Hispanic males are at an advantage in the most rural areas in terms of graduating.
  2. Females are more likely to graduate than males, holding all other variables constant.
  3. Differences in graduation rates among races mostly disappear when peer and some location attributes are included.
  4. Graduation rates for Black students have declined in the 2000s compared to the 1980s.
  5. The difference in graduation rates between public and private schools in the 1980s have disappeared in the 2000s.
  6. Dropout prevention policies need to focus on the effect of different family arrangements, poverty, and peer surroundings, regardless of location.