Ethiopian Economics Association and the Econometric Society. 19th Annual Conference of the African Region Chapter of the Econometric Society. 12th International Conference on the Ethiopian Economy. July 16-19, 2014. Addis Ababa

1. The Future in Mind: Aspirations and Forward-
Looking Behaviour in Rural Ethiopia
Tanguy Bernard (LAREFI/IFPRI)
Stefan Dercon (Oxford/DFID)
Kate Orkin (Cambridge)
Alemayehu Seyoum Taffesse (IFPRI)
July 17, 2014
The 19th Annual Conference of the Africa Region Chapter of the Econometric Society
Ethiopian Economic Association (EEA) Multi-purpose Building, EEA
16/10/2014 1

2.  Do people’s aspirations – an aspect of their mental models concerning
their opportunities and whether and how they can achieve them –
affect whether they invest?
 Randomly assign Ethiopian farmers to receive a ‘vicarious experience’
of the lives of four role models.
 Treatment = one hour of documentaries.
 No other intervention.
 Key findings:
 Improvements in aspirations after screening and after six months.
 Changes in related psychosocial concepts, but not risk aversion or
time preferences.
 Small improvements in savings, credit, demand for credit,
children’s school enrolment and spending on schooling 6 months
after screening.
,
Summary
16/10/2014 2

3. Outline
 Motivation
 Aspirations
 Field experiment – design and findings
16/10/2014 3

4. Motivation – why do poor people underinvest?
 Conceptual – ‘opportunities’
 Focus 1 - ‘external circumstances’ and ‘opportunities’.
 Low returns to investments;
 Unexploited opportunities due to lack of information or
knowledge;
 Social constraints;
 Focus 2 - constraints associated with the manifested attributes of
decision makers
 Identity issues: sense of self;
 Psychological issues: impatience, commitment, and psychological
barriers
Aspirations failure perspective
16/10/2014 4

5. What are Aspirations?
Aspirations:
 are goals or boundary-states sought after with respect to a relevant domain
of choice (future-oriented);
 Aspirations and expectations – preferences vs. beliefs;
 Aspirations are important for analysing and/or addressing
poverty:
 Condition individual behaviour and well-being (motivators )
 Aspirations are context-dependent and changing
 Are distributed unevenly within communities.
Specific Question:
Is it possible to alter poor people’s understanding of the opportunities they face by
actively trying to change their aspirations using an experimental design in a real-world
setting?
16/10/2014 5

6. Specific Question
Is it possible to alter poor individuals’ understanding of the
opportunities they face by actively trying to change their
aspirations using an experimental design in a real-world
setting?
16/10/2014 6

7. Experimental design: individual treatment
 64 villages. Random selection of 6 treatment HH, 6 placebo HH, 6
control HH. Head and spouse treated.
 3 arms:
 Treatment: ticket to view mini-documentaries about similar people who were
successful in small business or agriculture.
o No other intervention.
o 4 x 15 minute documentaries (2 men, 2 women) = 1 hour in Oromiffa
o Examples on CSAE Oxford YouTube channel
 Placebo: local Ethiopian TV show in 15 minute segments.
 Control: surveyed at their home.
 3 rounds of data collection:
 baseline (Sept-Dec 2010),
 aspirations immediately after treatment,
 follow-up six months later (Mar-May 2011).
16/10/2014 7

8. On going experiment
16/10/2014 8

9. Measures of aspirations
 Four dimensions:
 Annual income in cash
 Assets: house, furniture, consumer goods, vehicles
 Social status: do villagers ask advice
 Level of education of oldest child
 Aspirations vs. Expectations:
 What is the level of ___ that you would like to achieve?
 What is the level of ___ that you think you will reach within ten years?
 Overall aspiration index:
퐴푖 =
푘
푘 − 휇푘
휎푘
푘 푎푖
w푖
푘 = individual 푖’s aspiration response to dimension 푘.
푤푖
푎푖
푘 = weight individual 푖 assigned to dimension 푘.
휇푘 , 휎푘 = village sample mean and standard deviation for dimension 푘.
16/10/2014 9

10. Results
Small effects on aspiration immediately.
After screening (t=1)
Aspirations Expectations
Treated individual 0.13* 0.13* 0.12* 0.12** 0.12** 0.11**
0.07 0.07 0.06 0.06 0.05 0.05
Placebo individual 0 0 0 0.02 0.03 0.03
0.03 0.03 0.03 0.04 0.04 0.03
Village F.E. Yes Yes Yes Yes Yes Yes
Lagged outcome No Yes Yes No Yes Yes
Controls No No Yes No No Yes
Respondents 1959 1957 1957 1959 1954 1954
16/10/2014 10

11. Results
Small effects on aspiration after 6 months
After six months (t=2)
Aspirations Expectations
Treated individual 0.04* 0.04* 0.03* 0.06*** 0.06*** 0.05**
0.02 0.02 0.02 0.02 0.02 0.02
Placebo individual 0.03 0.02 0.03 0.02 0.02 0.03
0.02 0.02 0.02 0.02 0.02 0.02
Village F.E. Y es Yes Yes Yes Yes Yes
Lagged outcome No Yes Yes No Yes Yes
Controls No No Yes No No Yes
Respondents 2063 2058 2058 2062 2054 2054
16/10/2014 11

12. Elements of the Aspirations Perspective
After screening (t=1)
Aspirations Expectations
Treated individual 0.13* 0.13* 0.12* 0.12** 0.12** 0.11**
-0.07 -0.07 -0.06 -0.06 -0.05 -0.05
Placebo individual 0 0 0 0.02 0.03 0.03
-0.03 -0.03 -0.03 -0.04 -0.04 -0.03
Village F.E. Y es Y es Y es Y es Y es Y es
Lagged outcome No Y es Y es No Y es Y es
Controls No No Y es No No Y es
Respondents 1959 1957 1957 1959 1954 1954
After six months (t=2)
Aspirations Expectations
Treated individual 0.04* 0.04* 0.03* 0.06*** 0.06*** 0.05**
-0.02 -0.02 -0.02 -0.02 -0.02 -0.02
Placebo individual 0.03 0.02 0.03 0.02 0.02 0.03
-0.02 -0.02 -0.02 -0.02 -0.02 -0.02
Village F.E. Y es Y es Y es Y es Y es Y es
Lagged outcome No Y es Y es No Y es Y es
Respondents 2063 2058 2058 2062 2054 2054
16/10/2014 12

13. Observations
• Watching documentaries about role models improves
aspirations compared to a control group and, in some
cases, compared to a placebo group.
– Driven by those with above-median aspirations at baseline.
– No changes in risk aversion or time preferences.
– Also improvements in individuals’ sense that they control their lives.
• Small effects on ’forward-looking behaviour’ - children’s
school enrolment, spending on schooling, hypothetical
desire for credit - that are robust to multiple testing.
– Effects on savings, credit aren’t robust to multiple testing.
• Suggestive evidence that peer effects may reinforce
individual effects.
16/10/2014 13

14. Thank you
16/10/2014 14

15. Conceptual Schema
16/10/2014 15

16. Field Experiment – Design
 16 Screening sites, 4 villages/screening site (2 Intense Treatment, 2 Intense
Treatment village Placebo village
Surveyed : Treatment, 6 households (12 individuals) in every village
Placebo, 6 households (12 individuals) in every village
Control, 6 households (12 individuals) in every village
Non-Surveyed : Treatment, 18 households (36 individuals)/ intense treatment
village
Placebo, 18 households (36 individuals)/ intense placebo village
Placebo),
 36 households/village (18 households surveyed, 18 households not surveyed)
16/10/2014 16

17. Field Experiment – Baseline Correlates of
Aspirations
Income
aspiration
Wealth
aspiration
Education
aspiration
Social status
aspiration
Aspiration
index
Age -0.000 0.001 0.002 0.006 0.003
(0.000) (0.002) (0.003) (0.003)** (0.001)**
Age² 0.000 -0.000 -0.000* -0.000* -0.000
(0.000) (0.000) (0.000) (0.000) (0.000)***
Gender (Male=1) 0.008 0.062 0.258 0.096 0.104
(0.002)*** (0.036)* (0.051)*** (0.049)** (0.021)***
Education (Read/write=1) -0.000 0.068 0.333 0.312 0.152
(0.002) (0.070) (0.051)*** (0.073)*** (0.027)***
R2 0.03 0.01 0.08 0.04 0.07
N 1,964 1,967 1,932 1,957 1,865
* p<0.1; ** p<0.05; *** p<0.01; Screening site fixed effects included but not reported; Robust standard errors in
parentheses
16/10/2014 17

18. Balance
All
Treatment
(T)
% compliance by treatment status 95 93.8 96.2 100
Age
(complete
d years)
Baseline Standarized ---- Aspiration
Income Wealth
0.02 0.32 0 0.84 0.15 0.86 0.1 0.14 0.03 0.35 0.05 0.43 0.09 0.08* 0.04 0.12
0.02 0.32 0 0.93 0.05 0.94 0 0.89 0.05 0.15 0.01 0.83 0.04 0.55 0.01 0.5
* p<0.1; ** p<0.05; *** p<0.01
Placebo
(P)
Education
(Read/write
=1)
Gender
(% male)
Children's
Education
Social
Status
Difference:
T-C, p-value
Difference:
P-C, p-value
Sample balanced on gender, literacy, age and most outcomes
Control
(C)
Aggregate
16/10/2014 18

19. Field Experiment - Compliance and Potency of Treatment
Assessment of Documentaries and Placebo
Treatment
(standard
error)
Placebo
(standard
error)
Difference
(p-value)
Liked a lot what I saw?
0.95 0.73 0.22
(0.02) (0.01) (0.00)***
Discussed it a lot with my neighbours
0.87 0.71 0.15
(0.01) (0.02) (0.00)***
Discussed it at least once with neighbours
over the past two weeks
0.32 0.21 0.11
(0.02) (0.02) (0.00)***
Content generated a lot of discussion
within community
0.92 0.72 0.20
(0.01) (0.02) (0.00)***
* p<0.1; ** p<0.05; *** p<0.01
16/10/2014 19

20. Field Experiment - Compliance and Potency of Treatment
Table 5 – Relevance of documentaries
How does his/her present condition
fares compared to yours today?
He/she is
worse off
We’re about
the same
He/she is
better off
How did his/her
initial condition
fared compared
to yours five years
ago?
He/she was worse off 9.35 1.40 40.19
We were about the same 4.83 2.49 12.15
He/she was better off 6.70 1.71 21.18
Cell proportions are reported. The totals of all cells add up to 100. N=642
16/10/2014 20

21. Impact on Aspirations - Estimation
strategy
16/10/2014 21

22. Treatment and Placebo Effects on Aspirations
(1) (2) (3) (4)
Treatment
0.027 0.026
(0.018) (0.018)
Placebo
0.016 0.015
(0.018) (0.018)
# peers w/treatment
0.026 0.021
(0.010)*** (0.009)**
# peers w/placebo
0.001 -0.022
(0.010) (0.012)*
Baseline aspiration
0.132 0.157 0.132 0.157
(0.062)** (0.050)*** (0.062)** (0.050)***
Constant
0.053 0.038 0.095 0.018
(0.035) (0.036) (0.037)*** (0.037)
R2 0.06 0.07 0.06 0.08
N 1,210 1,258 1,210 1,258
* p<0.1; ** p<0.05; *** p<0.01; Screening site fixed effects not reported; Robust standard errors in parenthesis
16/10/2014 22

23. Treatment and Placebo Effects on Expectations
(1) (2) (3) (4)
Treatment
0.051 0.049
(0.023)** (0.023)*
Placebo
0.021 0.018
(0.021) (0.021)
# peers w/treatment
0.024 0.032
(0.010)** (0.010)***
# peers w/placebo
0.015 0.007
(0.011) (0.014)
Baseline expectations
0.401 0.074 0.402 0.075
(0.056)*** (0.032)** (0.057)*** (0.032)**
Constant
-0.047 -0.070 -0.028 -0.093
(0.046) (0.048) (0.048) (0.047)
R2 0.16 0.06 0.16 0.06
N 1,093 1,141 1,093 1,141
* p<0.1; ** p<0.05; *** p<0.01; Screening site fixed effects not reported; Robust standard errors in parenthesis
16/10/2014 23

24. Treatment and Placebo Effects on Future-Oriented
Behaviour
Treatment effects on time allocation - work
(1) (2) (3) (4)
Treatment
-3.33 -3.83
(12.61) (12.67)
Placebo
-8.46 -8.85
(16.91) (16.865)
# peers w/treatment
18.48 24.91
(7.97)** (10.6)**
# peers w/placebo
-9.63 -9.47
(8.49) (6.73)
Baseline time
allocation -Work
0.69 0.61 0.70 0.60
(0.02)*** (0.03)*** (0.02)*** (0.03)***
Constant
138.1 167.7 156.0 141.7
(79.4)* (81.8)** (87.2)* (74.9)*
R2 0.3 0.18 0.3 0.18
N 1,280 1,317 1,280 1,317
* p<0.1; ** p<0.05; *** p<0.01; Screening site fixed effects not reported; Robust standard errors in parenthesis
16/10/2014 24

25. Treatment and Placebo Effects on Future-Oriented
Behaviour
Treatment effects on time allocation - leisure
(1) (2) (3) (4)
Treatment
19.97 19.84
(12.93) (13.04)
Placebo
26.39 26.83
(12.79)** (13.01)**
# peers w/treatment
-0.74 -10.87
(6.60) (6.28)*
# peers w/placebo
1.79 3.35
(6.52) (5.71)
Baseline time
allocation -Work
0.31 0.31 0.31 0.31
(0.03)*** (0.03)*** (0.03)*** (0.03)***
Constant
522.6 549.2 519.8 560.2
(34.2)*** (33.3)*** (34.9)*** (33.8)***
R2 0.07 0.07 0.07 0.07
N 1,284 1,322 1,284 1,322
* p<0.1; ** p<0.05; *** p<0.01; Screening site fixed effects not reported; Robust standard errors in parenthesis
16/10/2014 25

26. Treatment and Placebo Effects on Future-Oriented
Behaviour
Table 10 – Treatment effects on savings behaviour
Savings Savings Deposits Deposits Withdrawals Withdrawals
(1) (2) (3) (4) (5) (6)
Treatment
186.9 63.5 33.7
(106.4)* (22.3)*** (17.2)*
Placebo
126.1 17.9 4.6
(95.1) (20.6) (9.3)
# peers with treatment
34.3 -37.1 -5.4
(85.9) (11.9)*** (6.7)
# peers with placebo
-28.2 -7.0 7.7
(46.9) (10.5) (6.2)
Baseline savings
0.741 0.658 -0.011 0.137 0.004 0.016
(0.578) (0.513) (0.011) (0.198) (0.017) (0.019)
Constant
-105.3 -24.9 80.9 24.1 55.2 32.7
(314.9) (271.4) (47.9)* (21.8) (108.1) (16.4)**
R2 0.16 0.29 0.05 0.04 0.03 0.02
N 1,258 1,288 1,258 1,288 1,258 1,288
* p<0.1; ** p<0.05; *** p<0.01; Screening site fixed effects and controls for age, age², gender and education not
reported; Robust standard errors in parenthesis
16/10/2014 26

27. Treatment and Placebo Effects on Future-Oriented
Behaviour
Table A1 - Direct and indirect treatment effect on Locus of Control
LOC
others
LOC others LOC internal LOC
internal
LOC chance LOC
chance
Treatment
-0.027 0.083 -0.030
(0.051) (0.038)** (0.044)
Placebo -0.015 -0.027 -0.028
(0.050) (0.039) (0.043)
# peers w/treatment
-0.056 -0.016 -0.059
(0.028)** (0.020) (0.023)**
# peers w/placebo
-0.002 -0.018 0.023
(0.028) (0.023) (0.025)
Baseline LOC
0.196 0.212 0.089 0.098 0.166 0.144
(0.031)*** (0.030)*** (0.030)*** (0.030)*** (0.027)*** (0.025)***
Constant
1.720 1.736 2.726 2.684 1.900 1.911
(0.120)*** (0.127)*** (0.124)*** (0.123)*** (0.105)*** (0.104)***
R2 0.05 0.06 0.03 0.02 0.05 0.05
N 1,341 1,372 1,342 1,373 1,341 1,374
* p<0.1; ** p<0.05; *** p<0.01; Screening site fixed effects not reported; Robust standard errors in parenthesis
16/10/2014 27

28. Treatment and Placebo Effects on Future-Oriented
Behaviour
Table A2 - Direct and indirect treatment effect on Perception of Poverty
Poverty due to Fate Poverty Structural Poverty Individual
Treatment
-0.108 0.033 0.088
(0.048)** (0.038) (0.042)**
Placebo
-0.005 0.058 0.072
(0.048) (0.037) (0.042)*
# peers
w/treatment
-0.048 -0.046 -0.011
(0.027)* (0.021)** (0.024)
# peers
w/placebo
0.008 -0.012 -0.005
(0.029) (0.023) (0.025)
Baseline
percept poverty
0.060 0.028 0.111 0.052 0.058 0.083
(0.031)* (0.031) (0.033)*** (0.030) (0.032)* (0.030)***
Constant
2.397 2.506 2.465 2.723 2.907 2.869
(0.116)*** (0.120)*** (0.120)*** (0.114)*** (0.124)*** (0.119)***
R2 0.02 0.02 0.02 0.03 0.03 0.03
N 1,339 1,368 1,337 1,368 1,339 1,370
* p<0.1; ** p<0.05; *** p<0.01; Screening site fixed effects not reported; Robust standard errors in parenthesis
16/10/2014 28

29. Observations
 "Weak" treatment, but:
 Documentaries affected aspirations, expectations, time
allocation, savings behaviour, and perceptions more
than the placebo even 6 months after treatment;
 Direct and, even more visible, indirect (group) effects
are detected – more of an aspiration window story
rather than a role model one;
 It is not obvious why some effects are direct (savings)
while others are indirect (time allocation);
 Further analysis;
 Expanding coverage – Malawi, Pakistan via IFPRI;
16/10/2014 29

30. 16/10/2014 30