World Development Vol. 38, No. 3, pp. 333–344, 2010
Ó 2009 Elsevier Ltd. All rights reserved
0305-750X/$ - see front matter
www.elsevier.com/locate/worlddev
doi:10.1016/j.worlddev.2009.05.010
Female Empowerment: Impact of a Commitment Savings Product
in the Philippinesq
NAVA ASHRAF
Harvard Business School, Boston, MA, USA
DEAN KARLAN
Yale University, New Haven, CT, USA
and
WESLEY YIN *
University of Chicago, IL, USA
Summary. — Female “empowerment” has increasingly become a policy goal, both as an end to itself and as a means to achieving other
development goals. Microfinance in particular has often been argued, but not without controversy, to be a tool for empowering women.
Here, using a randomized controlled trial, we examine whether access to and marketing of an individually held commitment savings
product lead to an increase in female decision-making power within the household. We find positive impacts, particularly for women
who have below median decision-making power in the baseline, and we find this leads to a shift toward female-oriented durables goods
purchased in the household.
Ó 2009 Elsevier Ltd. All rights reserved.
Key words — savings, microfinance, female empowerment, household decision making, commitment
or training which encourage separate assets. In theory, such
interventions could be unwound by adjustments to the control
over other assets in the household. Nevertheless, it is unknown
whether simply expanding access to products and training that
can directly impact financial control, and thus in turn affect
overall household power of women.
Using a randomized control trial, we implemented a program which provided a financial savings account whose use
was controlled by an individual and/or provided direct marketing to facilitate personalized savings goals. This program
did not necessarily increase income in the household (in fact,
we have no evidence that it did so); rather it offered individuals
a savings vehicle over which only the account holder has control.
Specifically, we designed and implemented a commitment
savings product with the Green Bank of Caraga, a rural bank
in the Philippines. Current bank clients were randomly chosen
to either (a) “savings commitment treatment” (SEED): receive
1. INTRODUCTION
Female “empowerment” has increasingly become a policy
goal, both as an end to itself and as a means to achieving other
development goals. 1 A growing literature on intra-household
bargaining finds that exogenous increases in female share of
income, interpreted as providing the female more power in
the household, lead to an allocation of resources that better reflect preferences of the woman (Duflo, 2003; Rangel, 2005).
This often leads to greater investment in education, housing,
and nutrition for children (Thomas, 1990, 1994; Thomas &
Strauss, 1995; Duflo, 2003). Many development interventions
have thus focused on transferring income as a way of inducing
empowerment (Adato, de la Brière, Mindek, & Quisumbing,
2000).
However, it is not clear in theory that transfers of income
alone to women can improve their status in the household. Marginal increases in income given to women may be bargained over
in the same way as the existing income, and are therefore not
guaranteed to lead to gains in bargaining power. 2 On a policy
level, microfinance proponents often argue that these empowerment mechanisms justify increased attention and financing to
microfinance institutions, and perhaps even subsidies (Hashemi, Schuler, & Riley, 1996; Kabeer, 1999). However, there
is little rigorous evidence that expanding financial access and
usage can promote female empowerment.
What may be more important than providing access to additional sources of income, or simply expanding access to finance, is giving control and property rights over allocated
money. 3 Household power could be increased directly by
interventions which lead women to have more control over
the existing assets. This could be done explicitly through financial accounts in her and only her name, or through marketing
q
This paper was formerly titled “Tying Husbands to the Mast: Impact of
a Commitment Savings Product in the Philippines.”
* We thank the Green Bank of Caraga for cooperation throughout this
experiment, John Owens and the USAID/Philippines Microenterprise
Access to Banking Services Program team for helping to get the project
started, Chona Echavez for collaborating on the field work, Robin Burgess, Pascaline Dupas, Larry Katz, Sendhil Mullainathan and Chris Udry
for comments, and Nathalie Gons, Tomoko Harigaya, Karen Lyons and
Lauren Smith for excellent research and field assistance. We thank the
National Science Foundation (SGER SES-0313877, CAREER SES-0547898), Innovations for Poverty Action, Russell Sage Foundation and the
Social Science Research Council for funding. All views, opinions, and
errors are our own. Final revision accepted: May 28, 2009.
333
334
WORLD DEVELOPMENT
an offer to open a “commitment” account accessible only by
them, and which does not mature until a pre-specified goal
is reached, 4 (b) “marketing treatment”: receive one-on-one
marketing about the importance of saving for a goal, or (c)
control: no household visit. The savings commitment device
could benefit those with self-control, but could also benefit
those with familial or spousal control issues. Indeed, the literature on household savings, and on informal savings devices in
particular, has emphasized motivations for both reasons
(Anderson & Baland, 2002; Gugerty, 2007).
Those who choose to open such accounts are likely fundamentally (and un-observably) different from those who do
not open such accounts, and thus a comparison of accountholders to non-account-holders would be plagued by a selection bias. By using a randomized control trial, and comparing
those who were offered the account to those who were not, we
are able to draw causal inference about the impact of the account itself (i.e., and not a self-selection bias in which impact
estimates are confounded by account openers being motivated
to save) on household dynamics.
We reported earlier (Ashraf, Karlan, & Yin, 2006) that after
one year individuals who were offered the product increased
their savings by 81% relative to a control group, and that in
accordance with the theoretical literature on hyperbolic preferences (Laibson, 1997; O’Donoghue and Rabin, 1999) and
dual-self models (Fudenberg & Levine, 2005; Gul & Pesendorfer, 2001, 2004), time-inconsistent individuals were the ones
most likely to demonstrate a preference for this commitment.
Using two new sources of data, a follow-up survey collected
after one year and administrative bank data collected after two
and a half years, we examine here the impact of this commitment savings product on both self-reported decision-making
processes within the household and the subsequent household
allocation of resources. We find positive impacts, particularly
for women who have below median decision-making power in
the baseline, and we find this leads to a shift toward femaleoriented durables goods purchased in the household.
This paper proceeds as follows. Section 2 describes the commitment savings product and the experimental design. Section
3 presents the empirical results on household decision making
and self-perception of savings behavior. Section 4 concludes
with a discussion of the theoretical mechanisms through which
this impact may have occurred.
2. INTERVENTION AND EXPERIMENTAL DESIGN
(a) The SEED account
We designed and implemented a commitment savings product called a SEED (Save, Earn, Enjoy Deposits) account with
the Green Bank of Caraga, a small rural bank in Mindanao,
Philippines. The SEED account requires that clients commit
not to withdraw funds that are in the account until they reach
a goal date or amount but does not explicitly commit the client
to deposit funds after opening the account. The SEED accounts are individual accounts, even if the participants were
married. There are three critical design features to the account,
one regarding withdrawals and two regarding deposits. First,
individuals restricted their rights to withdraw funds until they
reached a specific goal. Clients could restrict withdrawals until
a specified month when large expenditures were expected, for
example, the beginning of school, Christmas, a particular celebration, or when business needs arose. Alternatively, clients
could set a goal amount and only have access to the funds
once that goal was reached (e.g., saving a quantity of money
Table 1. Clients’ specific savings goals
Frequency
Percent
97
42
21
20
8
48.0
20.8
10.4
9.9
4.0
4
4
3
2
1
202
2.0
2.0
1.5
1.0
0.5
100.0
Date-based goals
Amount-based goals
Total
140
62
202
69.3
30.7
100.0
Bought Ganansiya box
Did not buy Ganansiya box
Total
167
35
202
82.7
17.3
100.0
Christmas/birthday/celebration/graduation
Education
House/lot construction and purchase
Capital for business
Purchase or maintenance of machine/
automobile/appliance
Agricultural financing/investing/maintenance
Vacation/travel
Personal needs/future expenses
Did not report reason for saving
Medical
Total
known to be needed for a new roof). The clients had complete
flexibility to choose which of these restrictions they would like
on their account. Once the client had made the decision they
could neither change it, nor could they withdraw from the account until they met their chosen goal amount or date. 5 After
the goal is reached, the SEED client, not his or her spouse,
could withdraw the funds. All clients, regardless of the type
of restriction they chose, were encouraged to set a specific savings goal as the purpose of their SEED savings account. SEED
marketers insisted that the client herself or himself, and not
another household member, set the goal. 6 Table 1 shows a list
of the savings goals selected broken down by percentage of the
group that selected them.
The savings goal was written on the SEED form used to
open the account, as well as on a “Commitment Savings Certificate” that was given to the client to keep. Forty-eight percent of clients reported wanting to save for a celebration,
such as Christmas, birthday, or fiesta. 7 Twenty-one percent
of clients chose to save for tuition and education expenses,
while 20% of clients chose business and home investments as
their specific goals.
The bank offered each SEED client a locked box (called a
“ganansiya” box) for a small fee in order to encourage deposits. This locked box is similar to a piggy bank: it has a small
opening to deposit money and a lock to prevent the client from
opening it. In our setup, only the bank, and not the client, had
a key to open the lock. Thus, in order to make a deposit, clients need to bring the box to the bank periodically. Of the 202
clients who opened SEED accounts, 167 opted for this box.
This feature can be thought of as a mental account with a
small, physical barrier; the box is merely a mechanism that
provides individuals a way to save their small change. Individuals put loose change or bills on an occasional basis, hence
making “deposits” that normally would be too small to warrant a trip to the bank. These small daily “deposits” keep cash
out of one’s (and others’) pocket; eventually, once enough
money accumulates in the box, the client deposits the funds
at the bank. The barrier, however, is largely psychological;
the box is easy to break and hence is a weak physical commitment at best. 8
Other than providing a possible commitment savings device,
no further benefit accrued to individuals with this account.
The interest rate paid on the SEED account was identical to
the interest paid on a normal savings account (4% per annum).
FEMALE EMPOWERMENT
(b) The experimental design and data collection
Our sample for the field experiment consists of 4,001 adult
Green Bank clients who have savings accounts in one of two
bank branches in the greater Butuan City area, and who have
identifiable addresses. We randomly chose 3,125 of 4,001 bank
clients to interview for our baseline survey. We then performed
a second randomization to assign these individuals to three
groups: commitment-treatment (T), marketing-treatment
(M), and control (C) groups. One-half the sample was randomly assigned to T, and a quarter of the sample each were
randomly assigned to groups M and C. We verified at the time
of the randomization that the three groups were not statistically different in terms of preexisting financial and demographic data. Of the 3,125, 1,776 were located by the survey
team and then completed a survey. Table 2 provides summary
statistics, broken down by treatment and control groups. See
Ashraf et al. (2006) for analysis that shows that the treatment
and control groups were observably statistically similar at the
time of the baseline.
335
Next, we trained a team of marketers hired by the partnering bank to go to the homes and/or businesses of the clients in
the commitment-treatment group, to stress the importance of
savings to them—a process which included eliciting the clients’
motivations for savings and emphasizing to the client that
even small amounts of saving make a difference—and then
to offer them the SEED product. We were concerned, however, that this special (and unusual) face-to-face visit might
in and of itself inspire higher savings. 9
To address this concern, we created a second treatment, the
“marketing” treatment. We used the same exact script for both
the commitment-treatment group and the marketing-treatment group, up to the point when the client was offered the
SEED savings account. For instance, members of both
treatment groups were asked to set specific savings goals for
themselves, write those savings goals into a specific “encouragement” savings certificate, and talk with the marketers
about how to reach those goals. However, members of the
marketing-treatment group were neither offered nor allowed
to open the SEED account. The bank staff was trained to
Table 2. Summary statistics
Total
Completed baseline survey
Completed follow-up survey
Baseline
Female, proportion
Married, proportion
Household decision-making power index 1
Household decision-making power index 2
Household decision-making power index 1 (married female)
Household decision-making power index 2 (married female)
Total savings at Green Bank, MIS
Total household savings
Total household informal savings
Savings in shared accounts (client is not the principal user)
Formal savings of other household members
Followup
Household decision-making power index 1
Household decision-making power index 2
Household decision-making power index 1 (married female)
Household decision-making power index 2 (married female)
All
(1)
Control
(2)
Treatment
(3)
Marketing
(4)
3,125
1,776
1,629
803
469
428
1,553
842
771
769
465
430
0.595
0.773
1.209
(0.422)
0.004
(0.812)
1.264
(0.401)
0.026
(0.799)
509.974
(506.408)
5,428.758
(15,781.820)
967.125
(4,641.664)
211.739
(2,784.990)
1,212.963
(7,365.828)
0.624
0.806
1.225
(0.423)
0.024
(0.799)
1.288
(0.385)
0.091
(0.739)
536.489
(515.373)
5,894.524
(16,279.700)
968.960
(5,697.623)
335.801
(3,533.014)
1,143.356
(7,212.905)
0.601
0.767
1.220
(0.416)
0.019
(0.808)
1.271
(0.399)
0.036
(0.803)
504.440
(500.692)
5,764.304
(18,305.750)
1,078.983
(4,988.806)
202.528
(2,885.735)
1,445.227
(8,639.445)
0.558
0.753
1.171
(0.432)
0.045
(0.834)
1.220
(0.424)
0.076
(0.856)
493.505
(507.773)
4,363.517
(8,852.169)
764.733
(2,171.288)
104.767
(1,426.876)
865.791
(4,462.855)
1.103
(0.286)
0.001
(0.775)
1.168
(0.273)
0.079
(0.779)
1.090
(0.289)
0.048
(0.799)
1.140
(0.266)
0.003
(0.773)
1.117
(0.285)
0.040
(0.766)
1.193
(0.270)
0.159
(0.771)
1.093
(0.282)
0.027
(0.763)
1.152
(0.284)
0.017
(0.789)
F statistic
(5)
0.136
0.151
0.190
0.480
0.275
0.167
0.423
0.262
0.531
0.475
0.415
0.270
0.203
0.068
0.036
Standard deviations are reported in the parentheses. Household decision making power indices are composed from answers to “Who decides” on the
following nine domains: what to buy at the market, expensive purchases, giving assistance to family members, family purchases, recreational use of the
money, personal use of the money, number of children, schooling of children, and use of family planning. The value for each item takes zero if the decision
making is done by spouse, one if the decision making is done by the couple, and two if decision making is done by the respondent. Index 1 is the equally
weighted mean of an individual’s responses across the nine decision categories; index 2 is the first factor of an individual’s responses across the nine
categories. The factor index (2) is created only for those who have no missing response to the nine questions on household decision-making power,
and thus removes all individuals without children. Analytical results throughout do not change if index 1 is calculated with the same sample restriction as
index 2.
336
WORLD DEVELOPMENT
refuse SEED accounts to members of the marketing-treatment
and control groups, and to offer a “lottery” explanation: clients were chosen at random through a lottery for a special trial
period of the product, after which time it would be available
for all bank clients. Green Bank reported that this happened
on fewer than ten occurrences. 10
After one year, we conducted a follow-up survey on each of
the participants. We completed follow-up surveys on 92% of
those in the baseline. Those in the treatment group were
equally likely to complete a follow-up survey as those in the
marketing or control group. This survey contained three sections: (1) inventory of assets, in order to measure whether
the impact on savings represented a net increase in savings
or merely a crowd-out of other assets, whose results are reported in a separate paper (Ashraf, Karlan, & Yin, 2008);
(2) impact on household decision making and savings attitudes; and (3) impact on economic decisions, such as the purchase of durable goods, health, and consumption.
Table 3. Impact on the aggregate household decision-making power.
Sample: Individuals who have children and whose spouses/partners live in the
same household
Index 1 (mean)
Level
(1)
Panel A: All
Treatment
Marketing
Constant
Observations
R-squared
Panel B: Female
Treatment
Marketing
3. IMPACT ON HOUSEHOLD DECISION MAKING
AND SELF-PERCEPTION OF SAVINGS BEHAVIOR
(a) Household decision-making power
We first examine whether being offered the SEED account
changed the decision-making roles in the household. In the
follow-up survey, we ask questions regarding family planning,
financial, and consumption decisions in order to ascertain the
structure of spousal or familial control within married households. For each decision category, we record whether the principle decision-maker is the respondent, the spouse, or both.
Responses are assigned values of two, zero, and one, respectively. We construct two decision-making indices from the
nine decision categories: (1) equally weighted mean of each response given, and (2) a linear combination, determined
through a factor analysis, of the individual responses to each
question (Pitt, Khandker, & Cartwright, 2006). The nine
categories refer to decisions on what to buy at the market,
expensive purchases, giving assistance to family members,
family purchases, recreational use of the money, personal
use of the money, number of children, schooling of children,
and use of family planning. 11
Table 3 shows the impact of treatment assignment on household decision making. Household decision making comprises
control over the following decisions: what to buy at the market, purchase of expensive items, giving assistance to family
members, family purchases, recreational use of the money,
personal use of the money, number of children, schooling of
children, and use of family planning.
Panel A provides the results for the full sample, Panel B for
married women and Panel C for married men. 12 The strongest
results are for married women. 13 We find that assignment to
the treatment group leads to a 0.14 standard deviation increase in the first (equally weighted) decision-making index
(Table 3, Panel B, Column 1), and a 0.25 standard deviation
increase in the second (factor-analysis) decision-making index
(Table 3, Panel B, Column 3). 14 In Table 4, we separately analyze the impact on women who began the year below (above)
the median decision-making power. We find that the average
effect is largely driven by increases in decision-making ability
for women who were below the baseline median (comparing
Panels A and B in Table 4)—a fact consistent with initially
less-empowered women experienced the largest gains in decision-making ability through increased financial savings and
control over committed assets. In contrast, we find no such
Constant
Observations
R-squared
Panel C: Male
Treatment
Marketing
Constant
Observations
R-squared
Change
(2)
Index 2 (factor)
Level
(3)
Change
(4)
0.029
(0.018)
0.012
(0.021)
0.778***
(0.028)
1,184
0.14
0.040
(0.028)
0.052
(0.033)
0.138***
(0.021)
1,184
0.00
0.107**
(0.053)
0.054
(0.061)
0.061
(0.043)
1,114
0.12
0.124*
(0.064)
0.102
(0.076)
0.080
(0.050)
1,114
0.00
0.056**
(0.023)
0.023
(0.027)
0.793***
(0.040)
643
0.16
0.073**
(0.034)
0.071*
(0.042)
0.147***
(0.025)
643
0.01
0.198***
(0.069)
0.087
(0.085)
0.032
(0.054)
600
0.15
0.241***
(0.080)
0.192*
(0.103)
0.090
(0.060)
600
0.01
0.001
(0.029)
0.018
(0.032)
0.791***
(0.039)
541
0.10
0.002
(0.047)
0.030
(0.052)
0.125***
(0.037)
541
0.00
0.006
(0.083)
0.041
(0.091)
0.105
(0.069)
514
0.09
0.019
(0.103)
0.012
(0.115)
0.068
(0.084)
514
0.00
Robust standard errors in parentheses. Dependent variable: index of
household decision-making power on what to buy at the market, expensive purchases, giving assistance to family members, family purchases,
recreational use of the money, personal use of the money, number of
children, schooling of children, and use of family planning. The value for
each item takes zero if the decision making is done by spouse, one if the
decision making is done by the couple, and two if decision making is done
by the respondent. See notes under Table 1 for the exact definition of each
index. Regressions in columns (1) and (3) control for the household
decision-making power in the baseline (August 2003).
*
Significant at 10%.
**
Significant at 5%.
***
Significant at 1%.
treatment effect for married men (Table 4, Panel A, Columns
5–8). We find that marketing has a smaller, but still significant,
effect on changes in decision-making indices, suggesting that
the encouragement of savings alone had a positive effect on
self-reported decision-making power of women in the household. 15
Next, we examine whether the increased reported decision
making led to a difference in the types of goods purchased
for the household. By increasing the assets available for lumpy
purchases, the mere presence of the SEED account may increase female decision-making power in the household and
hence increase the likelihood that the household acquires female-oriented durables. Naturally, if the account is held in
the women’s name this effect should be even stronger.
We use three categories for expenditures: house repair, female-oriented durables 16 (washing machines, sewing machines,
electric irons, kitchen appliances, air-conditioning units, fans, and
stoves), and other durables (vehicles/motorcycles, entertainment,
FEMALE EMPOWERMENT
337
Table 4. Impact on aggregate household decision-making power, by gender. Sample: Individuals who have children and whose spouses/partners live in the same
household
Female
Index 1 (mean)
Level
(1)
Change
(2)
Male
Index 2 (factor)
Level
(3)
Index 1 (mean)
Index 2 (factor)
Change
(4)
Level
(5)
Change
(6)
Level
(7)
Change
(8)
Panel A: Household decision-making power below median in baseline
0.094**
0.291***
Treatment
0.089***
(0.032)
(0.039)
(0.097)
Marketing
0.023
0.061
0.123
(0.040)
(0.050)
(0.117)
0.075**
0.124
Constant
0.800***
(0.068)
(0.030)
(0.090)
Observations
322
322
303
R-squared
0.08
0.02
0.07
0.341***
(0.102)
0.223*
(0.131)
0.233***
(0.080)
303
0.03
0.018
(0.036)
0.051
(0.040)
0.751***
(0.056)
296
0.06
0.021
(0.047)
0.075
(0.051)
0.105***
(0.037)
296
0.01
0.041
(0.102)
0.133
(0.117)
0.128
(0.101)
284
0.07
0.025
(0.115)
0.132
(0.128)
0.296***
(0.095)
284
0.00
Panel B: Household decision-making power above median in baseline
Treatment
0.026
0.022
0.111
(0.032)
(0.037)
(0.098)
Marketing
0.027
0.019
0.068
(0.037)
(0.048)
(0.120)
0.342***
0.115
Constant
0.879***
(0.103)
(0.027)
(0.096)
Observations
321
321
297
R-squared
0.04
0.00
0.03
0.109
(0.103)
0.045
(0.137)
0.380***
(0.078)
297
0.00
0.027
(0.049)
0.030
(0.053)
0.954***
(0.137)
245
0.01
0.015
(0.058)
0.027
(0.062)
0.440***
(0.047)
245
0.00
0.061
(0.137)
0.092
(0.145)
0.123
(0.139)
230
0.00
0.004
(0.149)
0.027
(0.157)
0.579***
(0.122)
230
0.00
Robust standard errors in parentheses. Dependent variable: index of household decision-making power on what to buy at the market, expensive
purchases, giving assistance to family members, family purchases, recreational use of the money, personal use of the money, number of children, schooling
of children, and use of family planning. The value for each item takes zero if the decision making is done by spouse, one if the decision making is done by
the couple, and two if decision making is done by the respondent. See notes under Table 1 for the exact definition of each index. Regressions in columns (1)
and (3) control for the household decision-making power in the baseline (August 2003).
*
Significant at 10%.
**
Significant at 5%.
***
Significant at 1%.
and recreational goods). Table 5 finds no significant impacts
on the choice and/or quantity of durables purchased in the
household in aggregate, nor broken down by gender. Table
6 analyzes the same dependent variables, but separately for
those above and below the median in terms of household
decision-making power at the baseline. We find that both
the number of items purchased and the total expenditures of
consumer durables traditionally associated with female use
in the Philippines increase for married women who were below
the median in the pre-existing bargaining power. This effect is
smaller, and not statistically significant, for married women
above the median. This finding is consistent with the impact
on decision-making ability for the purchases of personal items
and durable goods. We do not, however, find that married
households where the women are below the median in decision-making ability increase expenditures on other non-female
specific durables. Likewise, we do not find any effect for men
offered SEED, either in aggregate (Table 3, Panel C) or for those
above or below the median in household decision-making
power (Table 4, Columns 5–8, Panels A and B).
Taken together, the presence of both direct impact on selfreported decision-making measures, and a greater composition of female-oriented durables, suggest that women who
were offered the commitment savings product indeed increased
their power within their household.
In Tables A2 and A3 we evaluate the additional effect of the
commitment savings product above and beyond the marketing
treatment for both self-reported decision-making measures
and household purchases. Indeed, the results suggest that for
women the SEED product increased both measures of empowerment above and beyond the marketing treatment, however
the differences are not statistically significant.
(b) Self-perception of savings behavior
In the follow-up survey, we included several qualitative questions about personal savings habits and attitudes. In earlier research we found that time-inconsistent women were more
likely than time-consistent women to take up the SEED product, but that no such differential was found for men. 17 Here
we examine whether there are heterogeneous treatment effects
on savings attitudes and practices for men versus women and
time-inconsistent versus time-consistent clients. Table 7 presents
four outcomes, using an ordered probit specification. For each
outcome, the respondent was asked whether they strongly agree,
agree, are neutral, disagree or strongly disagree with a specific
statement. First, we ask about savings practices: (1) (Columns
1 and 2) “Although my income is low, I am a disciplined saver,”
(2) (Columns 3 and 4) “I never save,” and (3) (Columns 5 and 6)
“When I have a little cash, I spend it rather than save it.”
We find no aggregate effect, although we do find that timeinconsistent women who were offered the SEED account report being more likely to be a disciplined saver, less likely to
never save, and less likely to report spending rather than saving extra cash. This indicates that at least in their perception,
the SEED account helped them overcome their self-control
problem and led to improved savings practices (in earlier research, we do not find that the time-inconsistent women actually save more than the time-consistent women). In addition,
the marketing condition may have had an independent effect
on women’s perceptions of their efficacy in financial decisions
(Column 5, Panel B).
The final statement (Columns 7 and 8) is “I often find that I
regret spending money. I wish that when I had cash, I was
better disciplined and saved it rather than spent it.” Being
338
WORLD DEVELOPMENT
Table 5. Impact on consumer durables. OLS, probit. Sample framework: Those whose spouses are living in the same house.
House repair
Panel A: All
Treatment
Marketing
Panel B: Females
Treatment
Marketing
Probit
(1)
Total number
(2)
Cost
(3)
Probit
(4)
Total number
(5)
Cost
(6)
0.007
(0.033)
0.018
(0.038)
172.201
(1,611.810)
1,393.116
(1,648.315)
7,615.907***
(1,299.894)
1,181
0.00
0.019
(0.032)
0.035
(0.036)
0.009
(0.062)
0.017
(0.072)
0.495***
(0.047)
,183
0.00
48.293
(312.882)
144.558
(475.376)
1,997.997***
(242.252)
1,183
0.00
0.015
(0.030)
0.011
(0.034)
0.006
(0.042)
0.024
(0.047)
0.305***
(0.034)
1,183
0.00
2,293.060
(1,529.312)
2,493.613
(1,543.340)
6,095.462***
(1,344.654)
1,183
0.00
2,758.632
(1,960.731)
1,133.261
(1,875.305)
6,761.989***
(1,289.453)
641
0.01
0.023
(0.043)
0.023
(0.051)
0.086
(0.086)
0.038
(0.104)
0.489***
(0.060)
642
0.00
504.622
(433.285)
56.553
(508.971)
1,947.878***
(297.011)
642
0.00
0.002
(0.040)
0.029
(0.048)
0.050
(0.052)
0.043
(0.058)
0.261***
(0.036)
642
0.00
2,146.550
(2,340.491)
1,731.438
(2,401.692)
6,230.154***
(2,032.658)
642
0.00
3,137.328
(2,759.733)
2,010.130
(2,942.709)
8,796.324***
(2,534.068)
540
0.00
0.012
(0.049)
0.043
(0.052)
0.086
(0.090)
0.071
(0.103)
0.504***
(0.077)
541
0.00
519.682
(456.142)
315.665
(805.930)
2,066.774***
(406.126)
541
0.00
0.032
(0.044)
0.055
(0.047)
0.080
(0.071)
0.107
(0.077)
0.365***
(0.062)
541
0.00
2,453.800
(1,739.883)
3,165.144*
(1,764.869)
5,910.628***
(1,555.118)
541
0.01
1,181
0.026
(0.045)
0.020
(0.053)
Constant
Observations
R-squared
Panel C: Males
Treatment
Marketing
641
0.016
(0.051)
0.016
(0.056)
Constant
Observations
R-squared
Other durables
Cost
(2)
Constant
Observations
R-squared
Female-oriented durables
Probit
(1)
540
1,183
642
541
1,183
642
541
Robust standard errors in parentheses. Female-oriented durables consist of washing machines, sewing machines, electric iron, kitchen appliances, air
conditioners, fans, and stoves. Other durables include vehicles, motorcycles, and entertainment items (i.e., CD players, TV, and radio). Marginal effects
reported for probit specifications.
*
Significant at 10%.
***
Significant at 1%.
assigned to treatment makes individuals more likely to report
feeling regret over their spending and savings decisions. 18
Note that only 28% of those offered SEED took up, and of
those only about one-third regularly used the account. Hence
it follows that although SEED helped 10% of the treatment
group save more (and generate an overall positive intent-totreat effect), the mere offer of the SEED account generated,
on average, a feeling of remorse. Perhaps those who did not
take up and use felt remorse, and those who did take up
and use did not feel remorse, but the average effect is an increase in remorse because of the relative size of these two
groups. Perhaps a second marketing would have been more
successful than the first, if the first offer made individuals more
aware of their inability to save as much as they would like.
4. CONCLUSION
Even when husbands appropriate their wives’ loans, microcredit is thought to empower women in household decisionmaking processes (Mizan, 1993). Policymakers frequently cite
these arguments as a key motivation for targeting microfinance and microsavings interventions to women. On the other
side, some have argued that microfinance usage and the subsequent need to repay (e.g., in order to protect her reputation
amongst her peers) may subjugate women to the power of
their spouses, hence potentially increasing domestic violence
(Rahman, 1999). Evidence (albeit weak) points both ways,
and naturally may depend largely on the region-specific economic and social setting. 19 The effects of microcredit and,
more generally, microfinance, which includes savings and/or
insurance products, on female empowerment remain unclear,
in large part because studies of it tend to suffer from a pronounced selection bias in the type of women who access microcredit (Pitt et al., 2006).
Using a randomized controlled trial, we evaluate the impact
of a commitment micro-savings account. We find that the
commitment product positively impacts both household decision-making power for women (i.e., the household is more
likely to buy female-oriented durables), self-perception of savings behavior (time-inconsistent females report being more
disciplined savers), as well as actual consumption decisions
regarding durables goods.
The offering of the commitment savings product could
change household dynamics through several mechanisms.
First, the commitment product could have affected bargaining
power through the various forms of control (both legal and
normative/psychological) over decisions to withdraw and to
roll-over balances. A second person may still apply pressure
to influence withdrawal decisions, or exert pressure on other
margins in response to the account, and unwind the control
gained by the account. Nonetheless, in restricting legal control
to one individual, the product creates a formal barrier to second persons that the account holder can use in bargaining. 20
FEMALE EMPOWERMENT
339
Table 6. Impact on consumer durables. OLS, Probit. Sample Framework: Those whose spouses are living in the same house.
House repair
Probit
(1)
Cost
(2)
Female-Oriented Durables
Total number
(3)
Other Durables
Cost
(4)
Total number
(5)
Cost
(6)
Panel A: Females with household decision-making power below median in baseline
Treatment
0.027
2,480.870
0.192*
(0.063)
(2,133.872)
(0.108)
Marketing
0.081
1,149.406
0.126
(0.075)
(1,676.488)
(0.142)
0.386***
Constant
5,206.818***
(1,276.748)
(0.069)
Observations
322
322
322
R-squared
0.01
0.01
1,456.938**
(654.295)
600.512
(786.664)
1,518.750***
(359.206)
322
0.01
0.006
(0.073)
0.052
(0.088)
0.273***
(0.058)
322
0.00
3,887.597
(4,109.914)
4,446.125
(3,691.585)
8,037.500**
(3,550.889)
322
0.01
Panel B: Females with household decision-making power above median in baseline
Treatment
0.080
3,247.131
0.008
(0.063)
(3,231.059)
(0.131)
Marketing
0.048
625.615
0.036
(0.077)
(3,433.478)
(0.148)
0.580***
Constant
8,130.540***
(2,145.179)
(0.094)
Observations
319
319
320
R-squared
0.00
0.00
403.082
(552.084)
702.348
(586.010)
2,325.510***
(458.549)
320
0.00
0.092
(0.075)
0.029
(0.077)
0.250***
(0.046)
320
0.00
623.256
(2,436.893)
926.486
(3,346.618)
4,639.690**
(2,202.953)
320
0.00
Panel C: Males with household decision-making power below median in baseline
Treatment
0.006
4,114.137
0.080
(0.066)
(4,284.529)
(0.122)
Marketing
0.052
3,657.542
0.014
(0.072)
(4,618.274)
(0.148)
0.468***
Constant
9,718.987**
(4,083.798)
(0.105)
Observations
296
296
296
R-squared
0.01
0.00
741.921
(619.640)
841.101
(1,316.247)
2,072.152***
(569.847)
296
0.01
0.092
(0.103)
0.212**
(0.102)
0.405***
(0.089)
296
0.02
2,878.840
(2,561.748)
4,822.457**
(2,415.286)
6,301.975***
(2,352.200)
296
0.02
Panel D: Males with household decision-making power above median in baseline
Treatment
0.030
1,795.457
0.100
(0.079)
(2,829.019)
(0.132)
Marketing
0.093
104.123
0.177
(0.087)
(2,980.016)
(0.143)
0.552***
Constant
7,517.544***
(2,156.450)
(0.113)
Observations
244
244
245
R-squared
0.00
0.01
259.666
(666.850)
288.920
(836.159)
2,059.448***
(568.124)
245
0.00
0.058
(0.094)
0.023
(0.114)
0.310***
(0.082)
245
0.00
1,881.499
(2,182.161)
1,172.725
(2,466.193)
5,377.586***
(1,813.668)
245
0.00
Robust standard errors in parentheses. Female-oriented durables consist of washing machines, sewing machines, electric iron, kitchen appliances, air
conditioners, fans, and stoves. Other durables include vehicles, motorcycles, and entertainment items (i.e., CD players, TV, and radio).
*
Significant at 10%.
**
Significant at 5%.
***
Significant at 1%.
Second, a commitment savings account could establish a
norm within the household that the funds are to be used for
certain purposes. Any norms created by the commitment savings account might not be unwound by ex-post reallocation of
resources. Duflo and Udry (2003) find that crop revenues in
Cote d’Ivoire are labeled as either male, female, or family,
and shocks to one “mental account” remain in that account
and are not reallocated fully ex-post. The mere labeling of this
account as the wife’s provided her with additional power to
allocate those funds, which did not in turn crowd-out the allocation of other funds.
Third, it may also be the case that the woman actually got
more control of liquid funds. Many who took up the savings
product made use of a lock-box. These individuals were thus
able to keep small amounts aside, giving the person the power
to make decisions about the accumulated savings. Particularly
given the small amount of individual deposits, it is possible
that accumulations in this account were generated without
other household members being aware of the amount being
saved (although note that the treatment effect on savings volume was not stronger for women than it was for men).
Fourth, the commitment savings treatment (or the marketing treatment, which had a positive but insignificant statistically impact on savings (Ashraf et al., 2006)), could have
encouraged savings in general. The increased savings by woman could signal her outside option in case of a breakdown
of marriage. Female savings in this setting functions as the female wage rate in previous cooperative bargaining models
(Pollak, 2005). Although plausible in theory, note that the savings amounts here were small enough such that this theory is
likely only true for marriages on the margin of breakdown.
Greater savings or the opening of a non-joint savings account
raises the threat point in bargaining, representing what could
be earned in a non-cooperative outcome.
340
WORLD DEVELOPMENT
Table 7. Impact on savings attitude. Ordered probit.
Dependent variable
Panel A: All
Treatment
Marketing
Although my income
is low, I’m a
disciplined saver
(3)
(4)
(5)
(6)
(7)
(8)
0.025
(0.069)
0.057
(0.078)
0.053
(0.080)
0.073
(0.091)
0.147
(0.126)
0.300*
(0.156)
0.050
(0.175)
1,626
0.104
(0.072)
0.105
(0.085)
0.021
(0.083)
0.064
(0.098)
0.252*
(0.138)
0.303*
(0.165)
0.152
(0.195)
1,626
0.095
(0.065)
0.084
(0.075)
0.051
(0.077)
0.105
(0.090)
0.109
(0.115)
0.163
(0.146)
0.064
(0.161)
1,626
0.181***
(0.066)
0.070
(0.074)
0.160**
(0.078)
0.102
(0.088)
0.043
(0.120)
0.082
(0.149)
0.102
(0.161)
1,626
0.136
(0.103)
0.160
(0.123)
0.310**
(0.158)
0.395**
(0.196)
0.040
(0.225)
968
0.049
(0.093)
0.148
(0.112)
0.069
(0.107)
0.082
(0.132)
0.308*
(0.173)
0.389*
(0.209)
0.209
(0.246)
968
0.104
(0.081)
0.214**
(0.099)
0.005
(0.097)
0.209*
(0.123)
0.216
(0.136)
0.339*
(0.180)
0.018
(0.199)
968
0.130
(0.084)
0.118
(0.096)
0.065
(0.128)
0.007
(0.135)
0.128
(0.213)
0.133
(0.263)
0.249
(0.283)
658
0.199*
(0.116)
0.077
(0.131)
0.155
(0.133)
0.066
(0.148)
0.196
(0.222)
0.200
(0.266)
0.080
(0.312)
658
0.084
(0.110)
0.073
(0.118)
0.123
(0.126)
0.000
(0.134)
0.118
(0.212)
0.168
(0.255)
0.285
(0.279)
658
0.257**
(0.109)
0.010
(0.117)
Marketing time inconsistent, baseline
Marketing
1,629
0.021
(0.088)
0.176*
(0.103)
Time inconsistent, baseline
Treatment time inconsistent, baseline
Marketing time inconsistent, baseline
Observations
Panel C: Male
Treatment
Marketing
970
0.105
(0.112)
0.066
(0.118)
Time inconsistent, baseline
Treatment time inconsistent, baseline
Marketing time inconsistent, baseline
Observations
I often regret
spending, I wish I
was more
disciplined to save
(2)
Treatment time inconsistent, baseline
Panel B: Female
Treatment
When I have a little
cash, I spend it rather
than save
(1)
Time inconsistent, baseline
Observations
I never save
659
1,629
970
659
1,629
970
659
1,629
970
659
0.153
(0.101)
0.184
(0.118)
0.069
(0.140)
0.072
(0.180)
0.216
(0.203)
968
0.170
(0.121)
0.001
(0.134)
0.014
(0.241)
0.344
(0.277)
0.066
(0.288)
658
Robust standard errors in parentheses. Dependent variables are categorical, indicating how strongly the respondent agrees to each statement. The variable
equals one if the respondent strongly disagree, two if somewhat disagree, three if neutral, four if somewhat agree, and five if strongly agree.
*
Significant at 10%.
**
Significant at 5%.
***
Significant at 1%.
Finally, even in the absence of an actual increase in savings, the simple act of having a bank staff member come to
one’s door and encourage one to set savings goals could in
itself have increased a sense of “locus of control.” The
presence of the bank staff member may offer an external
social reinforcement of the account holder’s preferences
for how deposits are to be spent. This is akin to the second mechanism detailed above, but works through the
marketing process, not the design features of the savings
product itself.
Our results suggest that both the marketing process and control over the asset through the product design seem important—although the product design effect is somewhat larger,
we do not have the sample size to distinguish well between
the two treatments. We do find, however, that the package
of increased control over assets and direct encouragement
via marketing to take control of goal-setting and savings
caused a significant increase in empowerment for women,
compared to a control group that did not receive any special
asset or marketing.
Through continued experimentation, we can learn more
about the factors that drive savings decisions in the household
and thus also how to best design savings products that help
individuals reach goals such as asset building and consumption smoothing. We also need continued measurement of
how products impact household decision making, and how
household decision-making affects the efficacy of different savings products.
The results here suggest that commitment features, in
particular loss of liquidity combined with sole control of the
FEMALE EMPOWERMENT
account, appeal to those with self-control and have positive
impacts on female decision-making power. These are not
contradictory findings, but rather point out that a simple design feature such as a restriction on withdrawals or encourag-
341
ing savings through marketing or door-to-door deposits, can
benefit both those in search of self control devices as well as
those who desire to have more decision-making power in the
household.
NOTES
1. See, for example, Engendering Development (World Bank, 2001). By
“female empowerment” we mean increasing the bargaining power of the
woman within the household, manifested through increased influence in
household decisions and through household outcomes that greater reflect
her preferences.
2. See Garikipati (2008) as an example of other work posing a similar
question with respect to credit. In that work, the author finds that women
with longer durations in a lending program do not experience higher levels
of empowerment. Further work to separate selection and tenure effects
from the impact of credit would help to link those findings to ours to
understand whether the results are inconsistent or not.
3. Anderson and Eswaran (in press) find that income needs to be in the
control of women – not just generated by them – in order to impact their
bargaining power in the household. The relevant threat point in their
context, as in ours where divorce is uncommon, is non-cooperative
behavior.
4. The commitment savings product also incorporated the option to keep
a locked box (for which only the bank had the key) into which cash and
coins could be deposited.
5. Exceptions are allowed for medical emergency, in which case a
hospital bill is required, for death in the family, requiring a death
certificate, or relocating outside the bank’s geographic area, requiring
documentation from the area government official. The clients who signed
up for the SEED product signed a contract with the bank agreeing to these
strict requirements. After six months of the project, no instances occurred
of someone exercising these options. For the amount-based goals, the
money remains in the account until either the goal is reached or the funds
withdrawn or the funds are requested under an emergency.
6. SEED marketers reported instances of household visits in which the
husband tried to influence the goal-setting process. Typically the marketers then asked that only the wife give her goal and this was recorded, but
at no point did the marketer make an issue out of the goal setting process.
Green Bank prohibits spouses from being able to withdraw from each
others’ accounts, unless the account was explicitly opened as a joint
account. No SEED accounts were opened as joint accounts.
7. Fiestas are large local celebrations that happen at different dates
during the year for each barangay (smallest political unit and defined
community, on average containing 1,000 individuals) in this region.
Families are expected to host large parties, with substantial food, when it
is their barangay’s fiesta date. Families often pay for this annual party
through loans from local high-interest-rate money-lenders.
8. To facilitate deposits, clients also were offered automatic transfers
from a primary checking or savings account into the SEED account. This
feature was not popular. Many clients reported not using their checking or
savings account regularly enough for this option to be meaningful. Even
though preliminary focus groups indicated demand for this feature, only 2
of the 202 clients opted for automated transfers.
9. Because individuals were randomly selected, marketers were trained to
ask only for that person and ensure that the individual was the one setting
goals and, in the case of SEED, opening the account (i.e., the privilege
went to the individual, not to their spouse or others in the household, even
if they wanted to be the ones setting the goals (as happened in the case of a
few husbands).
10. In only one instance an individual in the control group opened a
SEED account. This individual is a family member of the owners of the
bank and hence was erroneously included in the sample frame. Due to the
family relationship, the individual was dropped from all analysis.
11. See Pitt, Khandker, and Cartwight (2006) for a discussion of
alternative constructions of a household decision making index. Our
results are robust to summing across the measures, and to specifications
that measure changes, rather than controlling for baseline levels as we
report in the text. Furthermore, since the factor analysis drops observations for which any answer is missing, we also examine the first measure of
equal weights but omit all observations for which any one answer is
missing. Results for the equally weighted mean index do not change on
this smaller sample of individuals.
12. This applies to married women whose spouses live at home with
them. Fifty-three of 696 married women had no spouse in the house in
both baseline and follow-up; 24 of 541 married men had no spouse
during both surveys. These married individuals were not included in our
analysis.
13. If we were to examine each question individually, we would find
positive impacts of the SEED treatment for women on 8 of the 9 variables
in the decision making index, two of which were significant at the 90%
level. The significant results are from survey questions asking who makes
the primary decision on expensive purchases in the household and the
number of children to have. Positive results were also found for the use of
family planning, giving assistance to other family members, buying items
for personal use, spending money on personal recreational (movies, liquor,
gambling), family purchases, working outside the household, and schooling for children.
14. The treatment effect in terms of standard deviations is calculated by
dividing the point estimates on the coefficient of interest (0.056 and 0.198
from Table 2) by the standard deviation for the dependent variable of each
index for married women, as found in Table 1.
15. In Table A2 we test the impact for married women for each of the nine
household decision categories that comprise the indices used in Table 2.
16. These goods were classified as female-oriented durables after
consultations with qualitative and quantitative social science researchers
at the Research Institute for Mindanao Culture (RIMCU) at Xavier
University, and conversations in focus group discussions.
17. Individuals defined as present-biased time-inconsistent when in
hypothetical time preference questions in the survey, they revealed a
higher discount rate for tradeoffs between now and 30 days than tradeoffs
between 6 months and 7 months. We measured this by posing questions
about two hypothetical situations involving winning a raffle cash prize. In
the first, respondents are asked whether they would like to receive the
winnings now or a larger amount of money in 30 days. In the second
situation, respondents are asked to choose between receiving the winnings
in 6 months or a larger amount in 7 months.
342
WORLD DEVELOPMENT
18. Interestingly, agreeing with this statement is also correlated with
being time-inconsistent when answering hypothetical time preference
questions.
19. Recent evidence from a randomized controlled trial in South Africa
finds no impact from access to credit on household decision making
(Karlan & Zinman, 2007). See Chapter 7 of Armendariz de Aghion and
Morduch (2005) for more discussion on this.
20. Particularly, the threat of roll-overs, combined with illiquidity, may
enhance bargaining power, even in the absence of any positive savings
impact.
REFERENCES
Adato, M., de la Brière, B., Mindek, D., & Quisumbing, A. (2000). The
impact of PROGRESA on women’s status and intrahousehold relations.
IFPRI working paper.
Anderson, S., & Baland, J. M. (2002). The economics of roscas and intrahousehold resource allocation. Quarterly Journal of Economics, 117(3),
963–995.
Anderson, S., & Eswaran, M. (in press). What determines female
autonomy? Evidence from Banglades. Journal of Development Economics.
Armendariz de Aghion, B., & Morduch, J. (2005). The economics of
microfinance. Cambridge: MIT Press.
Ashraf, N., Karlan, D., & Yin, W. (2006). Tying Odysseus to the Mast:
Evidence from a commitment savings product in the Philippines.
Quarterly Journal of Economics, 121(2), 673–697.
Ashraf, N., Karlan, D., & Yin, W. (2008). Tying Odysseus to the Mast
for a long time: Challenges of sustaining commitment. Working
paper.
Duflo, E. (2003). Grandmothers and granddaughters: Old age pension and
intra-household allocation in South Africa. World Bank Economic
Review, 17(1), 1–25.
Duflo, E., & Udry, C. (2003). Intrahousehold resource allocation in Cote
d’Ivoire: Social norms, separate accounts and consumption choices. MIT
working paper.
Fudenberg, D., & Levine, D. (2005). A dual self model of impulse control.
Working paper.
Garikipati, S. (2008). The impact of lending to women on household
vulnerability and women’s empowerment: Evidence from India. World
Development, 36(12), 2620–2643.
Gugerty, M. K. (2007). You can’t save alone: Commitment and rotating
savings and credit associations in Kenya. Economic Development and
Cultural Change, 55(2), 251–282.
Gul, F., & Pesendorfer, W. (2001). Temptation and self-control.
Econometrica, 69(6), 1403–1436.
Gul, F., & Pesendorfer, W. (2004). Self-control and the theory of
consumption. Econometrica, 72(1), 119–158.
Hashemi, S., Schuler, S. R., & Riley, A. P. (1996). Rural credit programs
and women’s empowerment in Bangladesh. World Development, 24(4),
635–653.
Kabeer, N. (1999). Conflicts over credit: Re-evaluating the empowerment
potential of loans to women in rural Bangladesh. World Development,
29(1), 63–84.
Karlan, D. & Zinman, J. (2007). Expanding credit access: Using
randomized supply decisions to estimate the impacts. Working
paper.
Laibson, D. (1997). Golden eggs and hyperbolic discounting. Quarterly
Journal of Economics, 112(2), 443–477.
Mizan, A. (1993). Women’s decision making power in rural Bangladesh: A
study of the Grameen Bank. The Grameen Bank: Poverty relief in
Bangladesh. Boulder: Westview.
O’Donoghue, T., & Rabin, M. (1999). Doing it now or doing it later. The
American Economic Review, 89(1), 103–121.
Pitt, M. M., Khandker, S. R., & Cartwright, J. (2006). Empowering
women with micro-finance. Evidence from Bangladesh. Economic
Development and Cultural Change, 54(4), 791–831.
Pollak, R. (2005). Bargaining power in marriage: Earnings, wage rates and
household production. Working paper.
Rahman, A. (1999). Microcredit initiatives for equitable and sustainable development who pays?. World Development, 27(1), 67–
82.
Rangel, M. (2005). Alimony rights and intrahousehold allocation of
resources: Evidence from Brazil. Harris school working paper series
05.5.
Thomas, D. (1990). Intra-household resource allocation: An inferential
approach. Journal of Human Resources, 25, 635–664.
Thomas, D. (1994). Like father, like son; like mother, like daughter:
Parental resources and child health. Journal of Human Resources,
29(4), 950–989.
Thomas, D., & Strauss, J. (1995). Human resources: Empirical models of
household decisions. In Behrman J.R., & T. N. Srinivasan (Eds.),
Handbook of Development Economics (Vol. IIIA, pp. 1885–2023).
Amsterdam: North Holland.
World Bank (2001). Engendering development: Through gender equality
in rights, resources and voice. A World Bank Policy Research Report.
Washington, DC and New York: World Bank and Oxford University
Press.
APPENDIX
See Tables A1–A4.
Table A1. Qualitative feedback from SEED account holders
Frequency
Those that did not withdraw: reason for not withdrawing
Argued with spouse
Bad bank service/bank is far
Could not save
Damaged passbook
Destroyed ganansiya box
Did not need money
Did not like terms/low interest
Forgot about it
Inconvenience
Money stolen (7)/lost (1)
Never joined/not a member
Nobody collected
Not interested
Not to term
Rolled over
Total
1
3
43
1
2
1
3
13
8
9
5
2
1
51
3
149
Those that withdrew: spent SEED money on
Fiesta
Children’s schooling
Other/did not say
Add to capital of business/sari-sar
Birthday (own, child, grandchild, missus, etc)
Child is giving birth
Children’s graduation
Christmas
Contruction of house/repair of kitchen
Everyday needs/necessities/groceries
Medical treatment
Reached time goal (3 months)
Refrigerator
Supplement mothers budget
Total
7
6
4
2
5
1
2
3
2
4
2
1
1
2
42
Spent money on original goal
Spent money on different goal from original
26
14
FEMALE EMPOWERMENT
343
Table A2. Impact on household decision making, components. Ordered probits. Sample: Women whose spouses/partners are living in the same house
Dependent
variable
What to buy
in market
(1)
Personal
use
(6)
Recreation
(4)
Assist family
members
(5)
0.023
(0.110)
0.117
(0.131)
641
0.143
(0.113)
0.046
(0.125)
642
decision-making power below median in baseline
0.175
0.010
0.409**
(0.162)
(0.164)
(0.162)
0.148
0.165
0.192
(0.181)
(0.182)
(0.187)
321
321
321
Panel C: Females with household decision-making power above median in baseline
0.033
Treatment
0.005
0.037
0.297*
(0.171)
(0.148)
(0.159)
(0.151)
Marketing
0.169
0.020
0.178
0.048
(0.205)
(0.184)
(0.207)
(0.186)
Observations
321
321
318
320
Panel A: Female
Treatment
Marketing
Observations
0.004
(0.117)
0.026
(0.134)
641
Panel B: Females with household
Treatment
0.005
(0.162)
Marketing
0.154
(0.182)
Observations
320
Expensive
purchases
(2)
Number of
children
(3)
Family planning
0.203*
(0.109)
0.060
(0.128)
642
0.217*
(0.114)
0.139
(0.137)
639
(7)
Family
purchase
(8)
Schooling
for children
(9)
0.013
(0.118)
0.124
(0.137)
643
0.112
(0.107)
0.062
(0.120)
642
0.174
(0.111)
0.115
(0.138)
641
0.162
(0.125)
0.220
(0.151)
609
0.323**
(0.158)
0.316*
(0.174)
321
0.243
(0.167)
0.238
(0.183)
322
0.229
(0.152)
0.282*
(0.171)
321
0.237
(0.164)
0.150
(0.191)
320
0.065
(0.199)
0.123
(0.228)
306
0.002
(0.160)
0.174
(0.179)
321
0.222
(0.170)
0.130
(0.213)
321
0.022
(0.152)
0.143
(0.169)
321
0.136
(0.155)
0.127
(0.197)
321
0.328*
(0.168)
0.509**
(0.210)
303
Robust standard errors in parentheses. All regressions in this table control for the initial household decision-making power in the baseline. The value for
each item takes zero if the decision making is done by husband, one if the decision making is done by the couple, and two if decision making is done by
wife.
*
Significant at 10%.
**
Significant at 5%.
Table A3. Impact on the aggregate household decision-making power (marketing and treatment groups only). Sample: Individuals who have children and
whose spouses/partners live in the same household
Index 1 (mean)
Panel A: All
Treatment
Constant
Observations
R-squared
Panel B: Female
Treatment
Constant
Observations
R-squared
Panel C: Male
Treatment
Constant
Observations
R-squared
Index 3 (factor)
Level
(1)
Change
(2)
Level
(5)
Change
(6)
0.022
(0.020)
0.822***
(0.034)
813
0.12
0.005
(0.031)
0.091***
(0.025)
813
0.00
0.055
(0.054)
0.008
(0.044)
809
0.10
0.022
(0.070)
0.022
(0.057)
809
0.00
0.040
(0.027)
0.865***
(0.051)
430
0.13
0.002
(0.042)
0.070**
(0.036)
430
0.00
0.115
(0.078)
0.052
(0.066)
427
0.12
0.049
(0.098)
0.102
(0.083)
427
0.00
0.012
(0.028)
0.827***
(0.044)
383
0.08
0.018
(0.046)
0.110***
(0.036)
383
0.00
0.036
(0.075)
0.064
(0.059)
382
0.08
0.030
(0.098)
0.057
(0.078)
382
0.00
Robust standard errors in parentheses. Dependent variable: Index of household decision-making power on what to buy at the market, expensive
purchases, giving assistance to family members, family purchases, recreational use of the money, personal use of the money, number of children, schooling
of children, and use of family planning. The value for each item takes zero if the decision making is done by spouse, one if the decision making is done by
the couple, and two if decision making is done by the respondent. See notes under Table 1 for the exact definition of each index.
**
Significant at 5%.
***
Significant at 1%.
344
WORLD DEVELOPMENT
Table A4. Impact on consumer durables (marketing and treatment groups only). Sample framework: Those whose spouses are living in the same house
House repair
Panel A: All
Treatment
Panel B: Females
Treatment
Binary
(3)
Total number
(4)
Cost
(5)
Binary
(6)
Total number
(7)
Cost
(8)
0.011
(0.034)
1,565.317
(1,391.052)
6,222.791***
(1,013.413)
857
0.00
0.016
(0.033)
0.026
(0.067)
0.479***
(0.054)
858
0.00
96.265
(454.382)
2,142.554***
(408.977)
858
0.00
0.003
(0.030)
0.019
(0.041)
0.281***
(0.032)
858
0.00
200.554
(1,050.847)
3,601.848***
(757.422)
858
0.00
3,891.893*
(2,008.677)
5,628.728***
(1,361.465)
453
0.01
0.001
(0.047)
0.048
(0.105)
0.527***
(0.085)
454
0.00
561.176
(519.888)
1,891.324***
(413.268)
454
0.00
0.031
(0.044)
0.006
(0.059)
0.304***
(0.046)
454
0.00
415.112
(1,726.796)
4,498.716***
(1,279.057)
454
0.00
1,127.198
(1,852.180)
6,786.194***
(1,495.551)
404
0.00
0.032
(0.046)
0.015
(0.083)
0.432***
(0.069)
404
0.00
835.347
(726.221)
2,382.439***
(695.914)
404
0.00
0.024
(0.043)
0.027
(0.058)
0.258***
(0.046)
404
0.00
711.343
(1,142.098)
2,745.484***
(834.237)
404
0.00
857
0.005
(0.048)
Constant
Observations
R-squared
Panel C: Males
Treatment
453
0.032
(0.049)
Constant
Observations
R-squared
Other durables
Cost
(2)
Constant
Observations
R-squared
Female-oriented durables
Binary
(1)
404
858
454
404
858
454
404
Robust standard errors in parentheses. Female-oriented durables consist of washing machines, sewing machines, electric iron, kitchen appliances, air
conditioners, fans, and stoves. Other durables include vehicles, motorcycles, and entertainment items (i.e., CD players, TV, and radio).
*
Significant at 10%.
***
Significant at 1%.
Available online at www.sciencedirect.com
American Economic Journal: Applied Economics 2016, 8(2): 35–64
http://dx.doi.org/10.1257/app.20150023
The Returns to Microenterprise Support among the
Ultrapoor: A Field Experiment in Postwar Uganda†
By Christopher Blattman, Eric P. Green, Julian Jamison, M. Christian
Lehmann, and Jeannie Annan*
We show that extremely poor, war-affected women in northern
Uganda have high returns to a package of $150 cash, five days of
business skills training, and ongoing supervision. Sixteen months
after grants, participants doubled their microenterprise ownership and incomes, mainly from petty trading. We also show these
ultrapoor have too little social capital, but that group bonds, informal insurance, and cooperative activities could be induced and had
positive returns. When the control group received cash and training
20 months later, we varied supervision, which represented half of the
program costs. A year later, supervision increased business survival
but not consumption. (JEL I38, J16, J23, J24, L26, O15, Z13)
T
he World Bank, the United Nations, and the United States government have
made the eradication of extreme poverty by 2030 a central development goal.1
Since the world’s poor often live in economies with few firms, anti-poverty programs often try to foster self-employment. This includes farm enterprises such as
raising livestock for sale, and nonfarm enterprises such as trading or retail. But can
the extreme poor be expected to start and sustain such microenterprises? And what
constraints hold them back?
* Blattman: Columbia University School of International and Public Affairs (SIPA) and National Bureau of
Economic Research (NBER), 420 W 118 Street, Suite 819, New York, NY 10027 (e-mail: chrisblattman@columbia.
edu); Green: Duke Global Health Institute, Box 90519, Durham, NC 27708 (e-mail: eric.green@duke.edu); Jamison:
Global INsights Initiative, The World Bank, 1818 H Street NW, Washington, DC 20433 (e-mail: julison@gmail.
com); Lehmann: University of Brasilia, Department of Economics, Campus Universitário Darcy Ribeiro, Brasília,
DF, 70910-900, Brazil (e-mail: christianlehmann0@gmail.com); Annan: International Rescue Committee, 122
East 42nd Street, Suite 1407, New York, NY 10168 (e-mail: jeannie.annan@rescue.org). Association of Volunteers
in International Service (AVSI) implemented the program and we thank Jackie Aldrette, Fabio Beltramini, Ezio
Castelli, Filippo Ciantia, Francesco Frigerio, John Makoha, Francesca Oliva, Federico Riccio, Samuele Rizzo, and
Massimo Zucca for collaboration. For comments we thank Abhijit Banerjee, Theresa Betancourt, Gustavo Bobonis,
Nathan Fiala, Don Green, Nathan Hansen, Dean Karlan, Bentley MacLeod, David McKenzie, several anonymous
referees, and seminar participants at George Washington University (GWU), Harvard University, Massachusetts
Institute of Technology (MIT), United States Agency for National Development (USAID), the World Bank, and
Yale University. A Vanguard charitable trust and the World Bank’s Learning on Gender and Conflict in Africa
(LOGiCA) trust fund funded the research. This article does not necessarily represent the views of the World Bank,
the Consumer Financial Protection Bureau, or the US government. For research assistance we thank Filder Aryemo,
Natalie Carlson, Samantha DeMartino, Mathilde Emeriau, Sara Lowes, Lucy Martin, Godfrey Okot, Richard Peck,
Alex Segura, Xing Xia, and Adam Xu through Innovations for Poverty Action.
† Go to http://dx.doi.org/10.1257/app.20150023 to visit the article page for additional materials and author
disclosure statement(s) or to comment in the online discussion forum.
1
“Extreme poverty” refers to earning less than the $1.25 per day international poverty line. See Burt, Hughes,
and Milante (2014) for a discussion of the goals.
35
36
American Economic Journal: applied economics
April 2016
Two in five of the world’s extreme poor are projected to live in fragile and
conflict-affected states by 2030, yet rigorous evidence on what works in these
settings is sparse.2 To help fill this gap, this paper studies a relatively common
approach to relieving extreme poverty—transfers of human and physical capital—
but to a postwar population: the most marginalized people living in small villages in
northern Uganda, following a 20-year war.
A humanitarian organization, the Association of Volunteers in International
Service (AVSI), identified 1,800 poor people, mostly women, in 120 war-affected
villages, and tried to help them start very small but sustainable retail and trading
enterprises. AVSI’s Women’s INncome Generating Support (WINGS) program provided people grants of $150 (about $375 in purchasing power parity, or PPP, terms),
along with five days of business skills training and planning, plus ongoing supervision to help implement the plan. The grant was 30 times larger than the beneficiaries’ baseline monthly earnings.
An abundance of evidence argues that the average poor person has high returns
to capital and is held back in part by poor access to credit and insurance, and that
capital transfers and insurance products help grow microenterprises and incomes.3
Most of this evidence, however, comes from people who already have businesses
or were selected for their business aptitude.4 It’s not clear if it applies to the most
marginalized and “ultrapoor”—the people with the lowest incomes, no capital, and
limited social networks—especially after war.5
The WINGS program has parallels to “graduation” style programs delivered
to hundreds of thousands of ultrapoor households globally. Graduation programs
give a bundle of temporary income support, livestock, livestock training, access to
microfinance, supervision, and life-skills education. On balance, these programs
have been successful: several studies show substantial shifts from casual labor to
farm self-employment, and 10 to 40 percent increases in household consumption
or earnings compared to control groups, lasting at least two to four years (Bandiera
et al. 2013; Banerjee et al. 2015). The WINGS program differed from these other
ultrapoor programs in several dimensions, however, including: the postwar setting;
fewer program components; the focus on petty business; and providing cash rather
than livestock.6 WINGS was also focused on young women.
2
See Burt, Hughes, and Milante (2014) for population projections. For reviews of the evidence see Blattman
and Miguel (2010) and Puri et al. (2014).
3
For example, see Udry and Anagol (2006); de Mel, McKenzie, and Woodruff (2008); Banerjee and Duflo
(2011); Karlan, Knight, and Udry (2015); Fafchamps et al. (2014); Blattman, Fiala, and Martinez (2014).
4
For example, Blattman, Fiala, and Martinez (2014) see high returns to a group-based cash transfer in northern
Uganda. But the program targeted young adults with much higher levels of education and existing business plans
for relatively high-skill microenterprises. That program also excluded the two most conflict-affected districts, where
WINGS was implemented.
5
On the one hand, returns to capital or other inputs could be greater on the extensive margin than the intensive
one. Indeed there is growing evidence that poor households use cash to start new enterprises and earn high returns,
although little of this evidence comes from the poorest of the poor (Macours, Premand, and Vakis 2012; Gertler,
Martinez, and Rubio-Codina 2012; Blattman, Fiala, and Martinez 2014; Bianchi and Bobba 2013). Returns could
also be high in a newly stable political equilibrium, as neoclassical models of growth predict (Blattman and Miguel
2010). On the other hand, the ultrapoor could have low returns to capital, for instance because they lack crucial
inputs such as education or business experience, or because they are vulnerable to expropriation within or outside
the family.
6
Many in the aid community fear cash can be seized, wasted, or cause harm. They could be right. Besides
the lack of other important inputs, extreme poverty has also been associated with cognitive deficits that impede
Vol. 8 No. 2
Blattman et al.: Microenterprise Support For Ugandan UltraPoor
37
We evaluated WINGS by assigning the targeted people to either receive the program immediately or a year-and-a-half later, randomizing at the village level. Given
the extreme setting, AVSI was reluctant to have a permanent control group—a common concern in humanitarian settings, and one reason humanitarian evaluations are
rare. Thus, our design evaluates impacts a few months before the 60 control villages
entered the program.
We also tested the role of social capital in business success: could social capital
be fostered, and would it increase the returns to grants? In poor rural villages, social
networks are a main source of business advice, cooperation, and informal finance.7
For instance, in microcredit, growing evidence suggests that group lending is helpful not because of joint liability, but rather because it builds social capital and promotes risk-pooling.8
To test this, in half of the treatment villages, AVSI returned a couple of months
after the grants (after individual businesses had already been started) to encourage the participants to form self-help groups, and offered three additional days of
training in working together. The curriculum focused on developing organizational
structures, decision-making processes, leadership, and helping them form a rotating
savings and credit association (ROSCA).
Sixteen months after grants, the standard WINGS program (without group
encouragement) led to large changes in occupation and incomes. Thirty-nine percent of the control group had a nonfarm business, and this rose to 80 percent among
WINGS participants. Employment rose from 15 to 24 hours per week, and cash
earnings rose about PPP $1 a day. Since the average person in the control group
earned less than $1 a day, the program doubled earnings. As a result, a conservative
estimate of household consumption rose by almost a third, to roughly PPP $1.25
per day. Annualized, this impact corresponds to a PPP $465 increase in nondurable
consumption—about a quarter of the PPP $1,946 standard program cost.
For program participants, the gains were mainly economic. There was little evidence of changes in physical health, mental health, financial autonomy, or domestic violence. Outside the household, however, the program increased self-reported
social support and community participation. Participants also reported an increase
in resentment and verbal abuse from some neighbors, however, perhaps due to jealousy, or because they posed competition for preexisting traders.
The group encouragement, meanwhile, increased the frequency and intensity of
group activities. We see no impact on consumption after 16 months, but program
participants who received group encouragement reported double the earnings of
those that did not. Interestingly, this was not because their petty trading businesses
were larger, more likely to survive, or more profitable. Rather, the evidence suggests
that groups spurred informal finance as well as labor-sharing and cooperative cash
investment and raise the risk of temptation spending (e.g., Bertrand, Mullainathan, and Shafir 2004). Among the
poorest women, moreover, traditional norms could pressure them to share cash, make it easy to expropriate them,
or hinder their business growth (Field, Jayachandran, and Pande 2010; Duflo 2012). This is the fundamental reason
that AVSI designed WINGS to include training and supervision.
7
See Fafchamps (1992), Foster and Rosenzweig (1995), Murgai et al. (2002).
8
Feigenberg, Field, and Pande (2013) show that encouraging social interaction via group meetings reduces
default on individual loans in India. Giné and Karlan (2014) also show individual liability has little impact on
default in the Philippines.
38
American Economic Journal: applied economics
April 2016
cropping. Group formation also seems to have mitigated the resentment and abuse
from neighbors.
Ideally, we could have randomly varied all program components and measured
their returns. This was not possible. But, following the main evaluation, we used
the entry of the control group into the program to investigate the marginal effect of
the most expensive component: supervision. The supervisory visits provided substantive advice as well as pressure to implement the business plan, but were more
than twice as costly as the grant. When the control group received WINGS after 20
months, we randomized them individually to receive the business training and grant
plus: no supervisory visits; two visits (to provide commitment to invest); or five
visits for both commitment and substantive business advice.
A month after grants, but before any potential visits, expecting a visit had an
ambiguous effect on business investment: those assigned to supervision increased
investment by some measures and decreased by others. A year later, the effect of
supervision on incomes is also ambiguous: nondurable consumption is marginally lower among those assigned to visits, earnings are marginally higher, but neither effect is statistically significant. Supervision, however, did increase business
survival.
Altogether, these results come with caveats. First, the control group knew they
were on the waitlist, and so anticipation of treatment could affect their behavior.
Second, all measures are self-reported. Experimenter demand is a risk, but given the
size of impacts (and the absence of noneconomic impacts, where we might expect
experimenter demand) the bias seems unlikely to drive our results. Third, there is
mild randomization imbalance and attrition. Treatment effects, however, are robust
to corrections and to missing data scenarios. Finally, these are 16-month impacts
and given the fact that the control group entered the program, we cannot say whether
they persist.
Nonetheless, WINGS illustrates that the poorest may be able to start and sustain small enterprises, even in very small, fairly poor communities. Moreover, the
16-month consumption impacts of WINGS are almost identical to the one-year or
two-year impacts of livestock-based ultrapoor programs, although WINGS was
about half as costly. So far the livestock programs have longer term evidence in their
favor, and the sustainability of cash-centric programs to the very poorest is an open
question.9 Even so, studies of cash transfers to the non-extreme poor show sustained
or growing impacts after four to six years (Blattman, Fiala, and Martinez 2014; de
Mel, McKenzie, and Woodruff 2012b).
Cash should be much cheaper and easier to deliver than livestock or capital goods,
so if it stimulates employment as well as the accumulation of income-generating
assets it could affect how ultrapoor programs are designed and scaled. This warrants more investigation, especially in humanitarian settings where cash is becoming more common as it can be difficult to provide in-kind capital. Also important to
investigate are cost-effective forms of supervision and training.
9
Livestock programs sustain gains after two to four years, while ultrapoor cash transfer studies have 1 to 2 years
of evidence so far (e.g., Haushofer and Shapiro 2013; Macours, Premand, and Vakis 2012).
Vol. 8 No. 2
Blattman et al.: Microenterprise Support For Ugandan UltraPoor
39
Finally, the results support the view that social interactions encourage cooperation, and that such social capital delivers economic returns. Most social capital is
endogenously formed, and it’s unusual to have experimental variation in local bonds.
Echoing Feigenberg, Field, and Pande (2013) on microcredit, we see that a program
that simply encourages group and ROSCA formation can increase social interactions,
enhance social capital, increase risk-pooling and cooperation, and perhaps even raise
incomes. What’s striking is that these profitable social bonds did not form in the
absence of encouragement, and yet were provoked by a relatively short training. It
implies the poor may be social capital constrained as well as credit constrained, and
external intervention seems to help overcome barriers to collective action.
I. Setting and Study Participants
Uganda as a whole is a poor but stable and growing country. National income grew
roughly 6.5 percent per year for the two decades prior to this study (Government of
Uganda 2007). A long-running, low-level insurgency in northern Uganda, however,
meant that most of the north was left out.
From 1987 to 2006, small bands of rebels conscripted, abused, and stole from
civilians in northern Uganda, especially the Kitgum and Gulu districts. Equally devastating was the Ugandan government’s decision to fight the insurgency by forcibly
moving nearly the entire rural population of Kitgum and Gulu—about two million
people—into dozens of displacement camps. The camps were often no more than
a few miles from people’s rural homes, but people generally could not access their
farmland during the war. Most households lost everything—livestock, homes, savings, and household durables—as a result.
By 2006 the rebels were mostly defeated or pushed out of the country, and by
2007 the government permitted displaced people to return home and rebuild. The
north’s economy began growing quickly, aided by an increase in demand from
a newly peaceful Sudan. The government started a number of large-scale development programs to help the north catch up to the rest of the country. Even so,
northern Uganda had some of the lowest standards of living in the world. By 2007,
two-thirds of households were unable to meet basic needs and lived mainly on food
aid (Government of Uganda 2007).
By 2009, when this study began, most people had rebuilt their homes and had
begun farming again. Food distribution and other emergency relief had ended. Most
rural villagers, however, were still desperately impoverished.
A. Study Sites and Participants
AVSI identified 120 villages in the two most war-affected districts, Kitgum
and Gulu. Most villages ranged in size from 350 to 1,000 people, with an average
population of 699 (about 100 households). The study villages represented about a
quarter of the population of the six rural subcounties where AVSI worked.10
10
AVSI actively worked in six subcounties—Odek, Lakwana, and Lalogi in Gulu and Omiya Anyima,
Namokora, and Orom in Kitgum. These have 252 total villages: 84 in Gulu; 168 in Kitgum. AVSI excluded from
40
American Economic Journal: applied economics
April 2016
AVSI held community meetings to describe the program and asked communities
to nominate 20 marginalized villagers, asking that three-quarters be women aged
14 to 30. AVSI staff interviewed all nominees and selected 10 to 17 participants per
village, excluding relatives of leaders and the least poor.
Table 1 describes the 120 villages and all 1,800 study participants, based on a
baseline survey of participants and each community leader. Twenty-six percent of
villages had a weekly market, and while on average there were three shops or kiosks
selling goods, the median village had none. Most goods were imported from the
district capital and retailed by a handful of shop owners.
Outside of traditional occupations (e.g., subsistence agriculture and some casual
labor), main work opportunities came from petty trade and retail, cottage production
(e.g., bricks, charcoal), livestock rearing, and cash crops. These farm and nonfarm
microenterprises required few new skills or education, but they were capital-intesive
and had fixed costs of entry.
The average participant in the program was female, 27 years old, and had 2.8 years
of education. Half were married or partnered. They reported an average of 15 hours
of work a week in the past month, mainly in their own agriculture. Just 3 percent did
any petty trade or business.
In general they were poor with no access to finance. Average cash earnings in the
month before the survey were 8,940 Ugandan shillings (UGX) ($4.47 at 2009 market exchange rates). Formal insurance was unknown, and almost no formal lenders
were present in the north at the outset of this study in 2008. Only 9 percent of the
sample were members of a village savings and loans group. On average they had
UGX 4,860 ($2.42) in cash savings and a nearly equal amount in debts, usually
from family and friends. Just 4 percent said they could get a loan of $50, which
is unsurprising because of high transaction costs and the near absence of informal
or formal lending institutions. Formal loan terms seldom extended beyond three
months, moreover, with annual interest rates of 100 to 200 percent. Because of high
fees, real interest rates on savings were typically negative. Given the startup costs
of microenterprise, this absence of credit and insurance was a major barrier to entry.
Effects of the War.—The war affected and displaced everyone in the study sample, impoverishing all. Until about a year before the program, everyone in the village had lived in a nearby displacement camp for at least three years, with no access
to farmland, during which their lands became overgrown and their houses destroyed.
At the time of the WINGS program, households were reestablishing agriculture for
the first time since at least 2003.
One in five people in our sample were abducted into the armed group at some
point, usually only for a few days to carry looted goods. Long stays with the armed
group were less common, and only 5 percent of the sample became fighters or were
forced to marry a rebel commander. Abduction and conscription, however, were not
the sample villages with fewer than 80 households. AVSI then chose program villages proportional to parish population, whereby more populous parishes would have a higher number of program villages (A parish is an administrative unit within the subcountry with five to ten villages). Official population figures did not exist and estimates were
based on 2008 data from AVSI and the United Nations. We estimate participant households in treatment villages
were less than 2 percent of households in the subcounty.
Vol. 8 No. 2
Blattman et al.: Microenterprise Support For Ugandan UltraPoor
41
Table 1—Descriptive Statistics and Randomization Balance for Select Covariates
Variable
Means, full sample
Treatment
Control
(Observations = 896) (Observations = 904)
Balance test
Difference
p-value
(1)
(2)
(3)
(4)
Demographics
Age
Female
Married or living with partner
Single-headed household
Highest grade reached at school
Forcibly recruited into rebel group
Carried gun within rebel group
Forcibly married within rebel group
27.02
0.86
0.46
0.51
2.82
0.20
0.03
0.03
27.63
0.86
0.50
0.47
2.75
0.25
0.04
0.03
−0.62
−0.01
−0.05
0.04
0.07
−0.05
−0.01
−0.00
0.17
0.72
0.26
0.17
0.70
0.03
0.45
0.63
Lagged outcomes
Any nonfarm self-employment
Average work hours per week
Agricultural
Nonagricultural
Average hours of chores per week
No employment hours in past month
Monthly cash earnings, 000s UGX
Durable assets (consumption), z-score
Durable assets (production), z-score
Number of community groups
Member of a savings group
Total savings, 000s UGX
Total debts, 000s UGX
Activities of daily life, z-score
Symptoms of distress, z-score
Quality of family relationships, z-score
Any community maltreatment, past year
0.03
14.57
11.27
3.29
34.88
0.23
8.54
−0.67
−0.53
0.48
0.08
4.24
4.24
−0.06
0.09
−0.09
0.19
0.04
16.19
13.36
2.83
34.25
0.18
9.32
−0.59
−0.50
0.58
0.11
5.47
4.08
−0.04
−0.09
0.09
0.16
−0.01
−1.62
−2.09
0.46
0.63
0.05
−0.78
−0.07
−0.02
−0.10
−0.03
−1.23
0.15
−0.02
0.18
−0.19
0.03
0.17
0.12
0.02
0.25
0.68
0.07
0.26
0.05
0.48
0.04
0.07
0.20
0.82
0.75
0.02
0.00
0.11
Village-level covariates (Observations = 120)
Village population
Inverse distance to all villages
Inverse distance to treatment villages
Distance to capital (km)
Accessible by bus
Village has a market
Number of shops in village
Total NGOs in village
749.62
0.51
0.56
46.21
0.98
0.18
1.23
7.13
649.05
0.58
0.47
44.72
0.91
0.34
1.75
7.42
100.58
−0.07
0.09
1.48
0.08
−0.16
−0.52
−0.29
0.34
0.34
0.43
0.58
0.05
0.05
0.30
0.68
p-value from joint significance of 76 covariates
< 0.01
Notes: All variables denominated in UGX and hours were top-censored at the nintey-ninth percentile to contain outliers. The durable asset indexes (z-scores) are calculated so that they have mean zero and unit standard deviation for
the full sample over all survey waves, and hence the sign is negative at baseline. The differences in columns 3 and
4 come from OLS regressions of baseline covariates on an indicator for treatment plus a district fixed effect, with
robust standard errors clustered by village.
necessarily a source of relative poverty. Annan et al. (2011) used exogenous variation in conscription to identify the long-term effects. Social acceptance of former
conscripts was high, and most people were psychologically resilient.11 These findings held even for the longest-serving females and those who were forcibly married
11
Psychological distress is commonplace, but serious symptoms are concentrated in the minority exposed to the
most violence and with the least social support.
42
American Economic Journal: applied economics
April 2016
or bore children. Conscription also had little effect on women’s schooling and labor
market outcomes. Women’s options outside the armed group were not much better
than inside the armed group, since most would not have been schooled or accumulated capital. Conscripted men, however, were well behind their peers after the war,
because they missed out on opportunities to accumulate physical and human capital.
In short, the war was so destructive that few young people emerged with any
assets or schooling. At the time of the WINGS program, they were rebuilding their
livelihoods from almost nothing.
B. Comparison to Nonparticipating Villagers
In general, the program successfully targeted the villages’ poorest, but it’s worth
noting that almost all villagers were very poor by any measure. We do not have data
on nonparticipants at baseline. Twenty months after the start of the program, however, we surveyed 2,836 nonparticipant households in treatment and control villages
(about 25 from each village), and sought to interview two working age adults per
household, in order to measure spillovers.12 Table 2 reports summary statistics for
participants and nonparticipants in the control villages only, in order to compare
people in the absence of direct treatment effects. We distinguish between households that were and were not traders at baseline.
If we look at similar-aged adults in “nonparticipant” households, participants
have similar cash earnings, but 24 percent lower consumption, 0.63 standard deviations fewer durable consumption assets (e.g., house quality, furniture, and household items), and 0.22 standard deviations fewer production assets (e.g., livestock
or farm tools). Participants also have about half the education and 63 percent of
nonparticipants’ work hours. About a third of nonparticipant households have at
least one adult engaged in trading at baseline, and these tend to be wealthier than
average.
II. Intervention, Experimental Design, and Data
The WINGS program had four components:
Basic Skills Training.— Participants received five days of business skills training,
ending with the preparation of a simple business plan. Training was designed for the
illiterate and focused on business planning, sales, marketing, record-keeping, and
budgeting (see online Appendix A for the curriculum). Trainers reviewed plans with
the participants and returned unsatisfactory plans for revision. They encouraged
participants to consider high cash flow activities that would diversify their income
sources, especially petty trading and retailing.
12
Shortly before Phase 2 disbursement, we created village household lists, randomly sampled 25 nonparticipant
households from each village (excluding the roughly 15 participant households), and sought to interview two working age adults per household on their economic activities, plus household data on assets and expenditures. We also
collected village prices on a variety of goods. Nonresponse to the survey was only 5.5 percent.
Vol. 8 No. 2
Blattman et al.: Microenterprise Support For Ugandan UltraPoor
43
Table 2—Participants versus Nonparticipants (control villages at Phase 1 endline)
Nonparticipants ages 17–40, control villages
Covariate
Age
Years of education
Average weekly work hours
Agricultural weekly hours
Nonagricultural weekly hours
Reports any hours in petty business
Monthly cash earnings, 000s of UGX
Monthly household consumption, 000s of UGX
Durable assets (consumption), z-score
Metal roof
Number of goats
Number of bicycles
Number of mobile phones
Durable assets (production), z-score
Observations
Participants
(1)
Traders
(2)
Non-traders
(3)
All
(4)
28.10
2.81
15.02
9.68
5.47
0.16
15.76
108.38
−0.45
0.00
0.97
0.39
0.14
−0.21
29.35
5.58
31.08
21.11
9.98
0.26
23.45
175.05
0.64
0.03
1.62
0.77
0.58
0.30
28.55
4.48
21.93
16.80
5.13
0.07
10.14
134.04
0.06
0.00
1.22
0.60
0.35
−0.07
28.71
4.70
23.78
17.67
6.11
0.11
12.82
142.30
0.18
0.01
1.30
0.63
0.39
0.01
917
360
1,427
1,787
Notes: For work hours and income, we report household averages in nonparticipant households, restricting data to
adults aged 17–40. A household is coded as a trading household if at least one household respondent says he or she
regularly sold an item a year ago, and did not obtain that item from his or her own production, for any items in a list
of 35. Individual-level covariates come from a self-reported survey of all respondents. All variables denominated in
UGX and hours were top-censored at the ninety-nineth percentile to contain outliers.
Cash.—Once a plan was approved, the participant received a grant of 300,000
UGX or $150 at 2009 market exchange rates. The grant was framed as funds to
implement the business plan. AVSI delivered cash in two equal installments about
two and six weeks after training.
Supervision.—AVSI trainers traveled four to five times to the villages in the six
months after the grant to provide one-on-one advising and supervision.
Group Formation.—About two months after grants were disbursed, AVSI also
offered a three-day group dynamics training that encouraged participants in the village to form self-help groups that would exchange ideas for improving their petty
business and agriculture, organize savings and credit, and (to a lesser extent) collaborate or cooperate in activities such as marketing their produce or buying inputs.
The vast majority of the three days focused on providing information and skills
related to working effectively as a group, including: leader selection, group decision
making, communication and listening skills, and conflict resolution methods. It
applied these skills to the same topics that were the focus of the five-day business
skills training: business planning, saving, and record-keeping. The difference is
that the group dynamics training mostly focused on how they could adapt these
skills when working as a group. The final day focused on how to organize a savings group, including best practices and record-keeping. Other forms of business
cooperation, such as joint purchasing and collaborative marketing, were mentioned
in passing as advantages of working in groups, but these production economies of
scale received very little attention. Indeed, AVSI deliberately did not encourage
44
American Economic Journal: applied economics
April 2016
participants to form firms or cooperatives. This is one reason AVSI offered the
group training some weeks after the individual business plans, grants, and initial
investment decisions.
Groups decided on their own aims and organization, however, and at the end of
the training AVSI helped groups write a constitution that formalized the aims, leadership, and decision-making structure of the group. Online Appendix A describes
the curriculum.
On average, WINGS cost $860 per person at market exchange rates: $150 for
grants; $125 for targeting and disbursement; $124 for business training; $82 for
group dynamics training; $348 for five supervisory visits; and $31 in other costs.
This is equivalent to PPP $2,150.
A. Phase 1 Research Design
In Phase 1, we randomized 60 of the 120 villages to receive the WINGS program.
The other 60 were randomized to a waitlist group (Phase 2) to be treated at least 18
months later. The participants in the waitlist villages were aware of this treatment
status.
Within these 60 treatment villages in Phase 1, we randomized 30 to receive the
group dynamics training and 30 to no group encouragement. Participants in the
60 control villages were told they would receive the intervention in about 18 to 24
months, called Phase 2. Figure 1 illustrates the sample, design, and timing.
We randomized by public draw, to ensure village buy-in and transparency.13 The
draw resulted in a slight imbalance in baseline covariates, reported in Table 1.14
Treatment participants were slightly worse off, with lower durable assets, employment, literacy, savings group memberships, participation in armed groups, and family and community support. The villages they lived in were also less likely to have
a market. A test of joint significance of all covariates has a p-value of < 0.001. If
anything, this may lead us to understate treatment effects. To account for possible
bias, we will control for these covariates in all treatment effects estimates and show
robustness to difference-in-differences measures.
To evaluate Phase 1, we attempted to survey all 1,800 participants 20 months
after baseline, 16 months after the first grant (at the median). Attrition was minimal;
we tracked migrants to their current location and found 96.3 percent (not including
three who died). Attrition is generally not significantly correlated with treatment or
baseline covariates (see online Appendix B).15
13
We gathered village leaders in each district. They drew village names without replacement to be assigned to
Phase 1 or 2. The authors were present for the draw to ensure its validity. We randomized group dynamics training
via computer.
14
See online Appendix B for all 76 covariates, as well as balance tests for the group dynamics randomization. In
total, 24 percent of the (nonindependent) covariates have p < 0.10. In general, the group dynamics randomization
was balanced.
15
In addition to these survey data, we collected formal qualitative data to better understand program experiences, constraints, and mechanisms. Two Ugandan research assistants interviewed 32 randomly selected participants in eight villages three times during and after the program. They followed semi-scripted questionnaires and
recorded, transcribed, and translated all interviews and notes.
Vol. 8 No. 2
Blattman et al.: Microenterprise Support For Ugandan UltraPoor
02/09: Selected 120 villages (60 per district) and
communities nominated ~2,300 persons
03/09–04/09: Registered 1,800 clients in 120 villages
45
Persons deemed not
“vulnerable”
excluded from study
04/09–06/09: Baseline survey of 1,800 clients (100%)
06/09: 60 villages (896 clients) randomized to
receive training (06/09–08/09), cash (08/09–10/09),
and follow-up (10/09–10/10).
06/09: 60 villages (904 clients) randomized to
waitlist treatments
02/09–03/09: 30
villages also receive
group dynamics
training
30 villages receive core
program only
11/10–02/11: Surveyed 861 (96%) of clients (0
deaths, and 0 villages, and 35 people unfound)
01/11: 57 clients no
longer in village replaced
11/10–02/11: Surveyed 870 (96%) of clients (3
deaths, and 0 villages, and 31 people unfound)
904 clients (847 original Phase 2 clients and 57
replacements) receive training (03/11–05/11) and
cash grants (08/11–09/11)
318 clients:
No follow-up
300 clients:
1 to 2 followups
286 clients:
3 to 5 followups
09/11–10/11: Surveyed 858 (95%) of clients (3
deaths, and 0 villages, and 35 people unfound)
Performed by implementer (AVSI)
Performed by researchers
09/11 to 06/12: Follow-ups performed (94%
adherence rate)
06/12–08/12: Surveyed 868 (96%) of clients (1
death, and 0 villages, and 35 people unfound)
Figure 1. Description of the Study Sample and Experimental Design
B. Phase 2 Research Design
In Phase 2, participants in the 60 control villages received the WINGS program.16
We used this as an opportunity to evaluate the marginal impact of the highest-cost
component, supervision.17 The first supervisory visit or two was mainly to hold
grantees accountable for implementing their business plan. The later visits were
primarily to provide advice. Of the 904 people in Phase 2, we randomly assigned
16
Program changes were minor. AVSI increased grants to 360,000 UGX to account for inflation, and disbursed
grants in a single tranche for efficiency.
17
Of the 60 villages, 30 were also randomized to have spouses included in the training, and present at the grant
disbursement, described in a companion paper (Green et al. 2015).
46
American Economic Journal: applied economics
April 2016
individuals to receive the cash and basic training with one of three treatments: no
supervisory visits; one to two supervisory visits, focused principally on commitment
to invest; or all five supervisory visits, to provide commitment but also substantive
advice on business management and household bargaining.18
To evaluate impacts, we first surveyed Phase 2 participants about a month after the
grant, shortly before the first follow-up. We intended this short-run survey to assess
how participants’ actions and investments varied with expectations of any supervision. We surveyed them again about a year after the grants to study the impacts of
actual supervision. Again, attrition was low; we found 95 percent at the one-month
survey and 96 percent at the one-year survey.
III. Empirical Strategy
We estimate intent-to-treat (ITT) effects via the ordinary least squares (OLS)
regression:
(1) Yij = θTj + δT D jT + δT–D jA + Xijβ + εij,
where Y is an outcome for participant i in village j, T is an indicato...
Purchase answer to see full
attachment