Prices or Knowledge? What Drives Demand For Financial Services in Emerging Markets?

THE JOURNAL OF FINANCE • VOL. LXVI, NO.
6 • DECEMBER 2011
Prices or Knowledge? What Drives Demand

for Financial Services in Emerging Markets?
SHAWN COLE, THOMAS SAMPSON, and BILAL ZIA∗
ABSTRACT
Financial development is critical for growth, but its microdeterminants are not well
understood. We test leading theories of low demand for financial services in emerging
markets, combining novel survey evidence from Indonesia and India with a field
experiment. We find a strong correlation between financial literacy and behavior.
However, a financial education program has modest effects, increasing demand for
bank accounts only for those with limited education or financial literacy. In contrast,
small subsidies greatly increase demand. A follow-up survey confirms these findings,
demonstrating that newly opened accounts remain open and in use 2 years after the
intervention.
FINANCIAL DEVELOPMENT IS WIDELY recognized as an important determinant

of economic growth, with a large literature examining the determinants of
the supply of banking and financial intermediation services (Levine (2005)).
Yet the determinants of the demand for financial services are much less well
understood, particularly in emerging market countries.
An important feature of emerging markets is the size of the informal sector.
Recent estimates place the size of the informal economy at 14% of GDP in
China, 23% in Indonesia, and 24% in India, against 8% in the United States
(Buehn and Schneider (2009)). In 76 emerging market countries, the average
size of the informal sector is almost 36% of GDP.1 Arguably, drawing these
individuals and firms into the formal financial sector would be one of the fastest
ways to foster financial development in emerging markets.
Two leading views may explain limited demand for formal financial services.
First, because these services involve high fixed costs and hence are expensive to
∗ Harvard Business School, Harvard University, and the World Bank, respectively. The au-
thors thank the World Bank Jakarta office for its assistance and support throughout this project.
The authors also thank seminar participants at the NBER Corporate Finance meetings, Oxford
University, the NEUDC Tufts Conference, the World Bank Global Seminar on Financial Literacy
and Consumer Protection, the World Bank Finance Seminar, and the OECD-Bank Indonesia Inter-
national Conference on Financial Education, and Anna Lusardi, David McKenzie, Jeremy Shapiro,
and Jeremy Tobacman for helpful comments. Financial support from the World Bank and HBS
Division of Research and Faculty Development is greatly appreciated.
1 Our baseline surveys find that 55% of the rural sample from India has savings in a nonbank
institution while 64% borrow from informal sources. Similarly, nationally representative figures
from Indonesia show that 51% of the population saves in nonbank institutions and 52% borrows
informally; nearly 20% of households in Indonesia borrow and save exclusively in the informal
sector.
1933
1934 The Journal of FinanceR
provide, it may be the case that low income individuals do not demand formal
financial services at market prices. Indeed, there is evidence that informal
savings, credit, and insurance markets function reasonably well in emerging
markets,2 with the benefits of formal financial market participation simply
not exceeding the relatively large fixed transactions costs associated with such
products (Beck, Demirguc-Kunt, and Peria (2007)). An alternative view argues
that limited financial literacy serves as an important barrier to demand for
services—if individuals are not familiar or comfortable with certain products,
they will not demand them.
While not mutually exclusive, these two views have significantly different
implications for the development of financial markets around the world, and
suggest quite different actions for financial institutions, governments, and in-
ternational organizations seeking to expand financial services use.
This paper aims to test the above theories. To do so, we conduct novel surveys
measuring household financial literacy and demand for financial services in
India and Indonesia. The survey in Indonesia represents the first nationally
representative household survey on financial literacy in a developing country.
We supplement the survey data with a randomized field experiment among
unbanked households in Indonesia to directly test the role and relative im-
portance of financial literacy and prices in determining demand for bank-
ing services. An intervention offering a financial education program on bank
accounts is randomly assigned to half of 564 unbanked households identified
by our survey team. Orthogonal to this treatment, individuals are randomly
offered small subsidies, ranging from U.S. $3 to $14, for opening a bank ac-
count. This design allows us to directly compare the effects of financial literacy
education versus price subsidies.
We find that financial literacy education has no effect on the probability
of opening a bank savings account for the full population, although it does
have an impact among those with low initial levels of education and financial
literacy. Modest financial subsidies, in contrast, have large effects, significantly
increasing the share of households that open a bank savings account within
the subsequent 2 months. Specifically, an increase in subsidy from $3 to $14
increases the share of households that open a bank savings account from 3.5%
to 12.7%, an almost three-fold increase.
Follow-up analysis conducted 2 years after the intervention shows that bank
accounts are “sticky”—those that were originally offered high subsidies are,
2 years later, significantly more likely to have used bank accounts in the past
year to deposit, withdraw, send, or receive funds. These long-run findings con-
firm our main short-run findings: financial literacy education alone does not
lead to greater demand for financial services in the general population, as
the share of individuals who opened a bank account in the 2 years after the
intervention is no different in the treatment group versus the control group.
The paper proceeds as follows. Section I discusses the motivation for the
study and the context in which the field experiment took place. Section II
2 See, for example, Besley, Coate, and Loury (1993) and Townsend (1994).
Prices or Knowledge? 1935
describes how we measure financial literacy and details the level of financial
literacy in our samples. In Section III, we explore the factors that predict finan-
cial literacy, and in Section IV we describe how financial literacy is related to
use of, and demand for, financial services. Sections V and VI describe the exper-
iment design and results, respectively, and Section VII discusses our follow-up
survey results. We conclude in Section VIII.
I. Motivation and Context

The role of financial literacy has received increasing attention in both the de-
veloped and developing world. For example, in the United States, in January
2008 the government set up the President’s Advisory Council on Financial
Literacy, which is charged with promoting programs that improve financial ed-
ucation at all levels of the economy and increase access to financial services.3
In the developing world, the Indonesian government declared 2008 “the year of
financial education” with a stated goal of improving access to and use of finan-
cial services by increasing financial literacy.4 Similarly, in India the Reserve
Bank of India launched an initiative in 2007 to establish Financial Literacy
and Credit Counseling Centers throughout the country that offer free financial
education and counseling to urban and rural populations.5
Much of this attention has been motivated by a compelling body of evidence
based on household surveys in developed countries that demonstrates a strong
association between financial literacy and household well-being. Households
with low levels of financial literacy tend not to plan for retirement (Lusardi
and Mitchell (2007a)), borrow at higher interest rates (Lusardi and Tufano
(2008), Stango and Zinman (2009)), acquire fewer assets (Lusardi and Mitchell
(2007b)), and participate less in the formal financial system relative to their
more financially literate counterparts (van Rooji, Lusardi, and Alessie (2007),
Hogarth and O’Donnell (1999)). In response to this evidence, financial literacy
programs have been advanced as a low-cost intervention with the potential to
improve household financial decision making and ultimately increase savings
and welfare.
There is less work in emerging markets. One exception is Cole et al. (2010),
who study demand for a rainfall insurance product in India. They find lim-
ited demand for the product, which, at least theoretically, appears to be quite
attractive. They also find no effect of a very modest, 5-minute long financial
literacy module on demand for the product.
The first substantive contribution of this paper is to measure the level and
predictors of financial literacy, and the relationship between financial literacy
3 See http://www.treasury.gov/offices/domestic-finance/financial-institution/fin-education/coun-
cil/index.shtml (accessed February 11, 2009). As an indication of the U.S. government’s resolve
to improve financial literacy, it named April 2008 Financial Literacy Month.
4 See http://www.oecd.org/document/3/0,3343,en 2649 34853 40660803 1 1 1 1,00.html
(accessed February 11, 2009).

5 See http://www.rbi.org.in/scripts/PublicationDraftReports.aspx?ID=526 (accessed
February 11, 2009).

and demand for financial services, in two of the most populous countries in the
world. We conduct two large household surveys in India and Indonesia and find
strong relationships between financial literacy and financial behavior.
Of course, as with any observational study, it is possible that other factors
explain some or all of the observed relationships. For example, individuals
with lower levels of financial literacy may have lower levels of education, be
less interested in financial matters, be poorer, or have different discount rates.
To measure causal relationships, we implement a field experiment in
Indonesia. We study one of the most basic, but perhaps most valuable, fi-
nancial services: bank savings accounts. We choose to study savings accounts
for several reasons. First, for households, a bank savings account can be an
efficient savings technology, secure from theft, and often paying interest, as
well as a means of sending and receiving payments. A savings account also
allows customers to build a relationship with the bank, potentially facilitating
eventual access to credit and other financial services. This may in turn improve
household welfare. Indeed, in the United States, the federal government and
individual states have passed legislation intended to draw individuals into the
banking system by establishing “lifeline” savings accounts and by providing in-
centives to retail banks to operate in underserved areas (Washington (2006)).
Transactions and savings accounts are the first and most obvious way in which
household participation in the formal financial sector begins.
Note that by conducting our field study in Indonesia, we consider a setting
in which financial literacy may be one of the most important barriers to access.
This may be explained in part by low educational expenditures: measured as a
share of GDP, education expenditures in Indonesia are the lowest in the world
(UNESCO (2007)). However, in contrast to many developing countries where
access to banking infrastructure is difficult, the Indonesian banking system
has wide geographical reach. Moreover, Indonesian banks have traditionally
offered savings accounts with low minimum deposits that are designed to serve
the needs of low income customers. The minimum balance in a savings account
with the nation’s largest bank, Bank Rakyat Indonesia (BRI), is only U.S. $0.53
and interest is paid on balances greater than $1.06.6 This compares to a per
capita income of approximately $1,918. Yet only 41% of the total population
and 32% of rural Indonesian households have a bank savings account.
To evaluate the importance of financial literacy, we randomly select half
of the unbanked households in our sample and offer them a 2-hour financial
literacy education session on how banks work and the benefits of opening a bank
savings account. To understand cost sensitivity, we offer unbanked households
subsidies ranging in value from $3 to $14 if they open a bank savings account.
While financial literacy has received increasing attention worldwide, our
paper is the first to systematically test the impact of a financial literacy train-
ing program in the developing world using randomized evaluation. In terms
of access to financial services, Indonesia and India are fairly representative.
6 See http://www.bri.co.id/english/layanan/simpanan.aspx?id=12 for terms of the savings prod-

uct (accessed February 11, 2009).
According to recent estimates from Beck, Demirguc-Kunt, and Peria (2007)

and Kendall, Mylenko, and Ponce (2010), India ranks 24th (out of 98 countries)
in bank branch penetration, with 22.6 branches per 1,000 square km, and
Indonesia ranks 38th with 10.0 branches per 1,000 square km. The United
States ranks 39th with 9.81 branches per square km. In terms of value of
deposits to GDP, India ranks 56th (out of 113 countries) with a 51.0% ratio,
Indonesia ranks 72nd with a 41.8% ratio, and the United States ranks 73rd
with a 39.8% ratio.
In the developed world, the most convincing evidence on the role of financial
education using a randomized evaluation comes from Duflo and Saez (2003),
who conduct an experiment at a U.S. university. The authors sent letters (at
random) to staff, encouraging the staff to attend an employee benefits fair.
The authors find that enrollment in retirement plans increased significantly
in the departments in which letters were received. The size of the effect, how-
ever, is quite small, at an increase of approximately 1.25 percentage points.
A related paper by Karlan and Valdivia (2010) studies the efficacy of offering
a business training program to female microentrepreneur clients of a bank in
Peru. While the content of the course falls outside the standard definitions
of financial literacy, the spirit was similar: provide education for individuals
making household decisions. They find that the treatment resulted in higher
repayment and client retention rates but had no impact on business income or
assets. Similarly, Bertrand and Morse (2010) look at the effect of financial lit-
eracy education intended to suppress demand for payday lending in the United
States and find that a treatment that emphasizes the dollar cost of repeated
borrowing is effective in reducing the probability that an individual renews a
payday loan.
This paper is also related to the literature on financial market development,
surveyed in great detail by Beck, Demirguc-Kunt, and Honohan (2008). Most
closely related to the present study, Beck, Demirguc-Kunt, and Peria (2007)
study household and firm use of banking services around the world and find that
GDP, institutional quality, and ownership structure are important predictors
of the use of financial services.
II. Measuring Financial Literacy and Financial Decisions

In this section, we describe the Indonesian and Indian household surveys
from which we obtain our measures of financial literacy. We describe how we
measure financial literacy and present summary statistics from the surveys.
Both surveys focus on households’ financial sector participation and were cus-
tom designed by the authors in conjunction with partner organizations. To the
best of our knowledge, the Indonesian results are the first nationally represen-
tative measure of financial literacy in a developing country.
The Indonesian data were collected as part of the World Bank’s Access to
Finance survey conducted in collaboration with the World Bank Jakarta of-
fice. The Access to Finance survey is a nationally representative household
survey designed to measure use of, and attitudes toward, financial services in
Indonesia. Stratified sampling was used to select 112 villages, and from each
village 30 households were randomly selected to participate in the survey, for a
total sample size of 3,360 households. All Indonesian survey statistics reported
in this paper are corrected for appropriate sampling weights. The survey took
place between July and December 2007. Summary statistics are provided in
the Internet Appendix.7
We complement the Indonesian survey results with data from India using
questions from a household survey administered in the state of Gujarat in 2006.
Because we designed both survey instruments, the questions are comparable
across countries. Despite the strikingly different context (India is far poorer
than Indonesia), we find notable similarities both in the variables that predict
financial literacy and in the relationship between financial literacy and demand
for financial products.
The survey in India took place in March and April 2006 as a baseline sur-
vey for a study on weather insurance. The survey covers 15 households in
each of 100 villages located in three districts of India around Ahmedabad, the
capital of Gujarat,8 and focuses primarily on poor, subsistence agricultural la-
borers. While the sample is not representative of India or Gujarat, the selected
households live in similar circumstances and have comparable educational
backgrounds to households throughout much of rural India.
Both surveys use a measure of financial literacy that is very close to the work
of Lusardi and Mitchell (2006), who, to our knowledge, are the first to use a
large-scale survey of financial literacy. We ask four questions: (i) “Suppose you
borrow Rp. 100,000 from a money lender at an interest rate of 2% per month,
with no repayment for 3 months. After 3 months, do you owe less than Rp.
102,000, exactly Rp. 102,000, or more than Rp. 102,000?” (ii) “If you have Rp.
100,000 in a savings account earning 1% interest per annum, and prices for
goods and services rise 2% over a 1-year period, can you buy more than, less
than, or the same amount of goods in 1 year as you could today, with the money
in the account?” (iii) “Is it riskier to plant multiple crops or one crop?” We also
added one new question: (iv) “Suppose you need to borrow Rp. 500,000. Two
people offer you a loan. One loan requires you to pay back Rp. 600,000 in 1
month. The second loan requires you to pay back in 1 month Rp. 500,000 plus
15% interest. Which loan represents a better deal for you?”9
Table I presents the financial literacy results. Measured financial literacy is
low, especially in India. The mean share of correct answers is 52% in Indonesia
and 34% in India. It should be noted that all questions were multiple choice, two
with two possible answers and two with three possible answers. Thus, random
7 An Internet Appendix for this article is available online in the “Supplements and Datasets”
section at http://www.afajof.org/supplements.asp.
8 The survey served as a baseline for Cole et al. (2010), who study a weather insurance inter-
vention. The survey was conducted prior to any intervention.

9 For the Indian survey, the amounts used were Rs. 100 for questions (i) and (ii) and Rs. 500
for question (iv).

Table I
Financial Literacy, Cognitive Ability, and Discount Rates
This table reports levels of financial literacy among household survey respondents in India and Indonesia. The Indonesian sample is nationally
representative. The means are given for households above and below the median per capita expenditures, and for households above and below the
median cognitive ability. The column to the right of the comparison columns indicates whether the difference in means is statistically significant. ***
indicates that the difference is statistically significant at the 1% level, ** at the 5% level, and * at the 10% level.
India Indonesia
Per Capita Per Capita
Expenditure Cognitive Ability Expenditure Cognitive Ability
Below Above Below Above Below Above Below Above

All Median Median Median Median All Median Median Median Median
Compound interest % Correct 59% 55% 63%∗∗∗ 33% 80%∗∗∗ 78% 69% 86%∗∗∗ 56% 89%∗∗∗
% Do not know 30% 15%
If savings earns 1% and inflation % Correct 25% 21% 28%∗∗∗ 14% 33%∗∗∗ 61% 51% 70%∗∗∗ 37% 74%∗∗∗
is 2%, after 1 year is buying power % Do not know 38% 16%
greater, less, or the same?
Prices or Knowledge?
Is one crop safer than multiple crops? % Correct 31% 30% 32% 26% 34%∗∗∗ 28% 24% 31%∗∗∗ 23% 30%∗∗∗
% Do not know 6% 4%
Borrowing 500,000, repaying 600,000 % Correct 24% 24% 23% 11% 34%∗∗∗ 44% 39% 49%∗∗∗ 30% 52%∗∗∗
versus paying 15 % interest % Do not know 24% 14%
All questions taken together % Correct 34% 33% 36%∗∗∗ 21% 45%∗∗∗ 52% 46% 59%∗∗∗ 37% 61%∗∗∗
All questions taken together Avg. Score 1.38 1.31 1.45∗∗∗ 0.83 1.80∗∗∗ 2.10 1.83 2.36∗∗∗ 1.46 2.45∗∗∗
(out of 4)
N 1,496 749 747 622 843 3,360 1,680 1,680 1,412 1,948
1939
guessing would yield an average score of 42%, which is in fact higher than the
average score in India, where many respondents answer “do not know” rather
than guess. The percentage of “do not know” responses ranges from 6% to 38%
in India and from 4% to 16% in Indonesia. The inflation question elicits the most
“do not know” responses in both countries. Looking at individual questions, a
majority of people in both surveys respond correctly to the compound interest
question (59% in India and 78% in Indonesia). Discerning interest rate versus
lump sum loan repayments seems to be most difficult for Indian respondents
(only 24% correct), whereas the diversification question is difficult to answer
in both settings (31% correct in India and 28% in Indonesia).
In terms of the distribution of scores, in India 26% of respondents do not
answer a single question correctly, 29% answer one question correctly, 29%
answer two questions correctly, 13% answer three questions correctly, and only
3% answer all questions correctly. The comparative figures from the Indonesia
survey are 12%, 21%, 32%, 28%, and 7%, respectively. In the United States, the
average score on the first three questions is 65%. The corresponding scores for
India and Indonesia are 38% and 55%, respectively. Throughout the paper, we
use the total score from all four questions as our measure of financial literacy;
the variable thus ranges from zero to four.
In addition to financial literacy, the surveys capture other household charac-
teristics that may be important determinants of financial behavior. Cognitive
ability is evaluated using a series of eight mathematics questions: the mean
share answered correctly is 81% in Indonesia and 62% in India.10 Almost all re-
spondents correctly answer the simplest question (“what is 4+3”) while many
have difficulty with multiplication (“3 times 6”) and division (“one-tenth of
400”). Because respondents were not allowed to ask their friends or neighbors
for help, it is reasonable to think that in situations in which collaboration is
possible, they would perform better when answering these questions. While
these exact questions have not been asked in the United States, Lusardi (2008)
finds similar abilities in the United States: 84% of U.S. respondents perform
a percentage calculation correctly, while 56% correctly divide proceeds from a
lottery among five winners.
We proxy for household discount rates by eliciting the minimum amount a
household would be willing to accept in 1 month in lieu of a Rp. 80,000 pay-
ment today.11 Consistent with other evidence, respondents report relatively
high discount rates: the average elicited monthly discount rate is 36% in In-
donesia and 21% in India. To measure risk aversion, we follow Binswanger
(1980) and use actual lotteries for real amounts of money. In Indonesia, respon-
dents were offered a choice between receiving Rp. 2,000 for certain or playing
a lottery that paid Rp. 5,000 with probability 12 and Rp. 0 with probability 12 .
10 We do not distinguish between cognitive ability and numeracy skills in our analysis. See Lang
et al. (2005), Dohmen et al. (2010), and Gerardi, Goette, and Meier (2010) for survey questions that
can separate the two measures.
11 We calculate discount rates using answers to hypothetical questions of the form: “Would you
prefer to receive Rp. 80,000 today, or Rp. X in 1 month.” For India, the ordering was reversed and
respondents were asked to choose between Rs. X today and Rs. 10 in 1 month.
Thirty-six percent of households choose the safe bet. We code these households
as risk averse.12 In India, respondents are coded as risk averse if they opt
to receive Rs. 2 for certain rather than play a lottery that paid Rs. 5 with
probability 12 and Rs. 0 with probability 12 . Nineteen percent of Indian house-
holds meet this definition of risk aversion.13
The surveys also allow us to proxy for the extent to which respondents view
events as outside of their control. For Indonesia, we measure fatalism as the
proportion of the following statements with which the respondent either agrees
or strongly agrees: (i) “I have little control over what will happen to me in
my life.” (ii) “Good things tend to happen to other people, not to me or my
family.” (iii) “I have a hard time saving money, even though I know I want
to save money.” The average value of fatalism is 60%. For India, we measure
fatalism using the extent to which respondents agreed with the first two of
these statements. The average value is 53%.
Finally, the surveys collect standard data on household demographics and
expenditures. While household wealth is an important concept, in practice it
is difficult to measure using household surveys, particularly among low in-
come households whose main assets (real estate, livestock) may not be easy to
price. Accordingly, we follow the standard convention and focus on measured
per capita household expenditures. The Internet Appendix demonstrates that
the Indian sample is more rural, less educated, and much poorer than the
Indonesian sample. The average household size in the Indian sample is 5.9,
twice as large as in Indonesia. Further, in India the entire sample is rural,
compared to 58% in Indonesia. Though low by developed country standards,
the Indonesian sample also exhibits substantially higher levels of education
than the Indian sample. For instance, 80% of respondents in Indonesia com-
pleted primary school compared to 41% in India. In the Indian sample, mean
monthly per capita household expenditures (which includes consumption, but
not investment spending) is less than one-third of the Indonesian level, while
average annual reported household income is $674 in India and $1,315 in
Indonesia.
In the Internet Appendix, we present summary statistics on households’ use
of financial services. Bank accounts are uncommon in both locations. Only
12% of Indian and 41% of Indonesian households report having a bank ac-
count. In the United States, in contrast, approximately 88% of adults have a
bank account (FINRA (2009)). However, 29% of Indonesian households that
do not report having a bank account indicate that they did have an account
12 This test is also a test of a behavioral anomaly, namely, “small-stakes risk aversion” described
by Rabin and Thaler (2001).

13 These methods of eliciting time and risk preferences have received some validation. Chabris
et al. (2008), for example, find that elicited time preferences weakly predict individual behav-
iors such as exercising and smoking. Binswanger et al. (1980) report that elicited measures of
risk aversion correlate well with agricultural risk-taking in a sample quite similar to our Indian
sample. It must be acknowledged, however, that these measures are not perfect; in particular,
limited financial literacy may confound the measure if individuals do not, for example, have a firm
understanding of probability.
at some point in the past. Approximately half (51%) of Indonesian households

have savings with a nonbank institution, but only 13% have advanced savings
instruments such as certificates of deposit or mutual funds. In total, 68% of
Indonesian households own a savings product of some form.
On the loan side, 25% of Indonesian households have a formal sector loan,
while only 13% of Indian households have such a loan. Informal credit is more
common in both countries, with 64% of Indian households and 52% of Indone-
sian households having loans from microfinance institutions, moneylenders, or
other informal sources. The most common source of informal loans in Indonesia
is family and friends.
One surprising result is the familiarity with, and use of, insurance in the
Indian sample. Two-thirds of households have some form of insurance policy.
This is likely attributable to the fact that SEWA, a local microfinance institu-
tion in Gujarat oriented toward helping poor women, makes health insurance
policies available to its members. In contrast, crop insurance, which must be
separately obtained, is comparatively rare. Even in Indonesia, almost half of
the households report having an insurance policy. One-third of the population
has health insurance, while 26% has asset or homeowner’s insurance.
III. What Predicts Financial Literacy?

A breakdown of financial literacy performance by household expenditures
and cognitive ability is given in Table I. Within samples, the share of the pop-
ulation answering each question correctly shows substantial variation by per
capita expenditures and cognitive ability. Splitting the samples by household
per capita expenditures we see that households with high per capita expendi-
tures do significantly better than households with low per capita expenditures
on most questions. Similarly, dividing the samples by cognitive ability, we find
that the upper half of the distribution does significantly better on all questions.
In fact, the differences between the low and high cognitive ability subsam-
ples are on average more than twice as large as the differences based on per
capita expenditures, suggesting that cognitive ability may play an important
role in determining financial literacy. This finding is consistent with Cole and
Shastry (2009), who find close relationships between cognitive ability and fi-
nancial behavior in the United States.
While the connection between household expenditures and financial literacy
has been long documented, the relationship between cognitive ability and fi-
nancial literacy, though not surprising, is less well understood. Christelis et al.
(2006) describe the relationship between cognitive ability and portfolio choice
in European households, and find that higher cognitive ability households are
more likely to invest directly in stocks.
In Table II, we take a more systematic approach and regress our measure
of financial literacy on a variety of individual characteristics. The regression
confirms that both greater per capita expenditures and higher human capital,
as measured by level of schooling or cognitive ability, are associated with signif-
icantly higher levels of financial literacy in Indonesia. We also find that rural
Table II
Predictors of Financial Literacy
This table reports the results from OLS regressions predicting measured financial literacy among household survey respondents in India and
Indonesia. Financial literacy is measured by a series of questions about compounding, interest rates, and risk diversification. The Indonesian sample
is nationally representative and weighted by sampling weights. The Indian regressions are unweighted. Only select coefficients are shown here; full
regression results are available in the Internet Appendix. Standard errors, clustered at the village level, are given in parentheses beneath each point
estimate. ∗∗∗ indicates statistical significance at the 1% level, ∗∗ at the 5% level, and ∗ at the 10% level.
Financial Literacy Score
India Indonesia
Dependent variable: (1) (2) (3) (4) (5) (6) (7) (8)
∗ ∗ ∗ ∗ ∗∗ ∗
Per capita expenditure 0.073 0.079 0.080 0.051 0.074 0.087 0.071 0.100∗ ∗
(0.040) (0.041) (0.041) (0.043) (0.040) (0.042) (0.042) (0.047)
Rural household −0.152∗ ∗ ∗ −0.195∗ ∗ ∗ −0.196∗ ∗ ∗
(0.051) (0.053) (0.053)
Female −0.077 −0.090 −0.096 −0.074 −0.110∗ ∗ −0.123∗ ∗ −0.130∗ ∗ −0.135∗ ∗ ∗
(0.059) (0.061) (0.061) (0.061) (0.050) (0.052) (0.051) (0.051)
Age 0.022∗ ∗ 0.027∗ ∗ 0.027∗ ∗ 0.020∗ 0.021∗ ∗ 0.020∗ ∗ 0.022∗ ∗ 0.012
(0.011) (0.011) (0.011) (0.011) (0.010) (0.010) (0.010) (0.010)
HH has nonfarm enterprise −0.065 −0.040 −0.041 −0.096 0.112∗ ∗ 0.129∗ ∗ 0.136∗ ∗ ∗ 0.114∗ ∗
(0.105) (0.108) (0.107) (0.108) (0.051) (0.052) (0.050) (0.054)
Married −0.030 −0.040 −0.046 −0.032 −0.079 −0.111 −0.094 −0.075
(0.080) (0.082) (0.083) (0.080) (0.076) (0.079) (0.076) (0.077)
Muslim 0.048 0.076 0.074 0.187∗ −0.073 0.010 0.010 −0.104
(0.094) (0.096) (0.097) (0.104) (0.102) (0.109) (0.106) (0.155)

Completed primary school −0.007 −0.034 −0.035 0.143∗ ∗ 0.165∗ ∗ 0.127∗ 0.128∗ 0.070
(0.063) (0.064) (0.064) (0.068) (0.067) (0.068) (0.068) (0.071)
Completed high school 0.201 0.254 0.253 0.148 0.022 −0.019 −0.020 −0.072
(0.228) (0.243) (0.239) (0.196) (0.066) (0.071) (0.069) (0.071)
Cognitive ability 0.223∗ ∗ ∗ 0.226∗ ∗ ∗ 0.225∗ ∗ ∗ 0.187∗ ∗ ∗ 0.234∗ ∗ ∗ 0.233∗ ∗ ∗ 0.224∗ ∗ ∗ 0.191∗ ∗ ∗
(0.013) (0.013) (0.013) (0.014) (0.017) (0.018) (0.018) (0.019)
Risk averse −0.037 0.026 −0.075 −0.062
(0.068) (0.065) (0.055) (0.056)
Interested in financial matters 0.022 0.050
(0.062) (0.062)
Saves enough (self-reported) −0.057 −0.101∗
(0.050) (0.052)
Village fixed effects No No No Yes No No No Yes
N 1,450 1,369 1,369 1,369 3,057 2,818 2,818 2,818
1943
households and households with a female head exhibit lower levels of financial
literacy, while households that own a nonfarm enterprise have higher financial
literacy. With respect to age, financial literacy is quadratic and peaks at around
40 years old. Neither discount rates nor risk aversion predict financial literacy.
Household per capita expenditures and cognitive ability are also positively
correlated with financial literacy in India, but, surprisingly, there is no system-
atic relationship between education and financial literacy. As in Indonesia, age
is quadratic and peaks at around 45 years old.
One striking result is that households with a more fatalist worldview exhibit
consistently lower financial literacy, both in India and Indonesia, even after
controlling for a host of other characteristics. This result, found as well in
the United States (Cole and Shastry (2009)), may suggest that households
who believe that outcomes are predetermined feel less motivated to invest in
understanding how to make decisions that improve their well-being.
The regressions also allow us to quantify and compare the effects of per
capita expenditures and cognitive ability, two of the most important predictors
of financial literacy. The estimates from Column 2 indicate that in our Indian
sample, a 1-standard-deviation increase in household per capita expenditures
predicts a 0.05-standard-deviation increase in the financial literacy score. In
contrast, a 1-standard-deviation increase in cognitive ability is associated with
a 0.50-standard-deviation increase in the financial literacy score. In Indone-
sia, the corresponding magnitudes, based on the estimates in Column 6, are
0.05 and 0.37 standard deviations, respectively. In both samples, cognitive abil-
ity has a substantially stronger association with financial literacy than does
household expenditures.
IV. What Does Financial Literacy Predict?

A compelling body of evidence demonstrates a strong association between fi-
nancial literacy and household well-being in developed countries. The Internet
Appendix shows how use of financial services varies with household character-
istics in our Indian and Indonesian samples. Higher household expenditures
predicts greater use of bank accounts and formal credit in both countries,
but predicts greater use of informal credit and insurance in Indonesia only.
The results for human capital are mixed. Education is positively associated
with the use of bank accounts and formal credit in both countries and with
insurance in Indonesia, but is negatively associated with informal credit use
in both countries. Higher cognitive ability predicts greater insurance use in
both countries and greater use of formal credit in Indonesia, but is otherwise
insignificant.
In both countries, none of the household preference indicators consistently
predict the use of financial services. In Indonesia, a high discount factor is
associated with lower use of both formal and informal credit, while risk averse
households are more likely to have a bank account or a formal loan. Fatalism
is associated with lower use of bank accounts in Indonesia, but higher use of
insurance in India.
Higher financial literacy is significantly associated with greater use of bank

accounts in Indonesia and insurance in India, even after including a host of
controls. The coefficients on the borrowing regressions are positive but in-
significant. Although financial literacy is a significant predictor of the use of
bank accounts in Indonesia, the magnitude of the estimates suggests it is a
less important predictor than expenditure levels. The estimates from Column
2 indicate that a 1-standard-deviation increase in financial literacy is associ-
ated with a 2.2-percentage-point increase in the probability of having a bank
account, while a 1-standard-deviation increase in household expenditures is
associated with a 14.9-percentage-point increase.
A. Demand for Financial Products

While much has been written on the impact of financial repression on finan-
cial development (e.g., La Porta et al. (1998)), many countries around the world
are liberalizing financial markets. However, mere entry may not be sufficient
to spur financial development if demand for the products is very limited. In this
section, we present some of the first survey evidence on demand for a range of
financial products.
In Panel A of Table III we explore demand for financial products. Data for this
section and the remainder of the paper are available for the Indonesian sample
only. We ask respondents if they are interested in three financial products that
have been identified as potentially beneficial in increasing household savings.
First, we ask about a commitment savings product similar to the one described
in Ashraf, Karlan, and Yin (2006a). This product allows clients to deposit money
at any time, but to withdraw only after a certain savings target has been met or
a specified time period has passed. Christmas savings clubs in the United States
are one example of this product. Approximately 43% of households express
interest in such a product. Second, we ask whether the household is interested
in deposit collection services. Deposit collection services have been shown to
increase savings in the Philippines (Ashraf, Karlan, and Yin (2006b)). Interest
in this product is lower, at 25%. Finally, we ask if households are interested in
retirement savings accounts. We find that 50% of households express interest
in such products.
To better understand barriers to the use of bank accounts, we ask respondents
whether they would open a bank account if account fees were reduced. Of the
unbanked, 37% report that they would open a bank account if fees were halved;
that figure rises to 58% if fees were eliminated. These responses are particularly
striking given that the basic account (described in Section V.A below) offered
by a government-owned bank charges no fees, suggesting that households have
a limited understanding of banking services.
Panel B of Table III explores which household characteristics predict inter-
est in the three financial products we consider. Interest in all three products is
increasing in financial literacy and household expenditures, even after includ-
ing a range of household controls and village fixed effects. Financial literacy is
therefore a strong and consistent predictor of demand for financial services.
1946
Table III
Demand for Financial Products, Indonesia
This table reports demand for financial products by household survey respondents in Indonesia. The sample is nationally representative. Panel A
gives average reported demand for each service, while Panel B reports OLS regressions relating individual characteristics to product demand. Only
select coefficients are shown here; full regression results are available in the Internet Appendix. Standard errors, clustered at the village level, are
given in parentheses beneath each point estimate. ∗∗∗ indicates statistical significance at the 1% level, ∗∗ at the 5% level, and ∗ at the 10% level.
Panel A: Summary Statistics
Indonesia
Sample Mean N
Demand for savings products

Interested in commitment savings product All 43% 3,360
Interested in using deposit collector All 25% 3,359
Interested in retirement savings product All 50% 3,360
Open account if fees cut 50% No bank account 37% 2,153
Open account if fees cut 100% No bank account 58% 2,153
Would attend financial literacy training No bank account 74% 2,153
Panel B: Determinants of Demand for Financial Products

The Journal of FinanceR
Commitment Savings Deposit Collector Retirement Savings Literacy Training

Demand for: (1) (2) (3) (4) (5) (6) (7) (8)
Financial literacy score 0.028∗∗∗ 0.025∗∗ 0.024∗∗∗ 0.026∗∗∗ 0.037∗∗∗ 0.033∗∗∗ 0.019∗ 0.014
(0.010) (0.010) (0.009) (0.010) (0.010) (0.011) (0.010) (0.011)
HH has bank account −0.012 −0.018 −0.051∗∗ −0.065∗∗∗ 0.087∗∗∗ 0.074∗∗
(0.026) (0.026) (0.020) (0.021) (0.025) (0.029)
Per capita expenditure 0.058∗∗∗ 0.043∗∗∗ 0.030∗∗ 0.025 0.073∗∗∗ 0.067∗∗∗ 0.061∗∗∗ 0.051∗∗
(0.015) (0.016) (0.014) (0.015) (0.017) (0.019) (0.021) (0.021)
(continued)
Table III—Continued
Panel B: Determinants of Demand for Financial Products
Commitment Savings Deposit Collector Retirement Savings Literacy Training

Demand for: (1) (2) (3) (4) (5) (6) (7) (8)
Female 0.007 0.009 −0.021 −0.013 0.031 0.030 −0.022 −0.025

(0.019) (0.021) (0.018) (0.017) (0.020) (0.019) (0.019) (0.020)
Age 0.005 0.005 0.003 0.004 0.003 0.002 0.010∗∗ 0.007∗
(0.004) (0.004) (0.003) (0.003) (0.004) (0.004) (0.004) (0.004)
HH has nonfarm enterprise 0.012 0.010 0.025 0.021 −0.044∗∗ −0.048∗∗ −0.022 −0.025
(0.020) (0.020) (0.018) (0.019) (0.018) (0.02) (0.022) (0.021)
Married 0.091∗∗∗ 0.085∗∗∗ −0.014 −0.034 0.005 −0.008 0.029 0.021
(0.024) (0.024) (0.026) (0.028) (0.025) (0.024) (0.034) (0.035)
Muslim 0.025 0.021 −0.020 −0.008 0.038 0.049 −0.042 −0.050
(0.049) (0.047) (0.036) (0.036) (0.046) (0.046) (0.059) (0.052)
Completed primary school 0.027 0.029 0.015 0.011 0.021 0.022 0.024 0.017
(0.025) (0.025) (0.024) (0.025) (0.028) (0.027) (0.025) (0.025)
Completed high school −0.017 −0.023 −0.057∗∗ −0.066∗∗ 0.008 −0.006 0.028 0.015
(0.024) (0.025) (0.026) (0.026) (0.026) (0.026) (0.030) (0.032)
Cognitive ability 0.007 0.002 −0.007 −0.010 −0.006 −0.012∗ 0.005 0.003
(0.006) (0.007) (0.007) (0.008) (0.007) (0.007) (0.007) (0.007)
Risk averse −0.037∗ −0.027∗ −0.030 −0.038
(0.020) (0.016) (0.023) (0.024)
Interested in financial matters 0.121∗∗∗ 0.096∗∗∗ 0.154∗∗∗ 0.070∗∗
(0.026) (0.023) (0.024) (0.033)
Saves enough (self-reported) 0.097∗∗∗ 0.102∗∗∗ 0.108∗∗∗ 0.092∗∗∗
(0.022) (0.020) (0.024) (0.021)
Village fixed effects Yes Yes Yes Yes Yes Yes Yes Yes
N 3,057 2,818 3,057 2,818 3,057 2,818 1,876 1,737
1947
There is no evidence of a robust effect of human capital on interest levels

for any of the three products. Households that have a bank account are less
interested in deposit collection services and more interested in retirement sav-
ings, but their interest in the commitment savings product is not significantly
different. Demand for the commitment savings and deposit collector products
is higher among households that are more patient and not risk averse. De-
mand for all three products is higher for households that have a fatalistic
outlook, are interested in financial matters, and report saving enough for the
future.
In the Internet Appendix, we examine self-reported attitudes toward the use
of financial services. The most common reasons cited for having a bank account
are: for security (53%), for predicted future needs (42%), for money transfers
(37%), and emergency needs (31%). Only 17% of respondents see having a
transactions account as a step toward borrowing from the bank.
When asked their reasons for not having a bank account, 92% of unbanked
households report that they do not have enough money. The second-most com-
mon answer, not knowing how a bank operates, is cited only by 32% of house-
holds. Interestingly, 29% of currently unbanked households report having had
an account at some point in the past. Among these households, 71% report that
they stopped using the account because they did not have enough money.
Just over half of households (54%) report that they are saving enough for the
future. Of those who answered “no,” lack of money is the most frequently cited
reason for insufficient savings (76%), with irregular income (31%) and failure
to control spending (23%) the second- and third-most common reasons.
We also ask about household demand for insurance. Among those without
insurance, not enough money is again the most frequent reason given (59%),
followed by not knowing about any insurance products (38%). Only 6% of house-
holds indicate that they do not have insurance because premiums are too ex-
pensive.
Finally, we ask households to describe the three most important financial
risks they face. Illness is the most common risk (79%), followed by loss of em-
ployment (56%) and loss of dwelling (33%). Conditional on owning a nonfarm
enterprise, 52% of households report concerns about business risk. Interest-
ingly, many of the reported risks (health, property loss, death, and vehicle
damage) are insurable, though most households choose not to insure them.
The data in Table III and the Internet Appendix provide support for the no-
tion that a financial literacy training intervention could increase the share of
households possessing a bank account. Lack of knowledge of how a bank works
is the second-most common reason for not having a bank account and is cited
by approximately one-third of households. The fact that only 31% of the popu-
lation reports knowing the requirements to open a bank account suggests that
knowledge may be a barrier to opening an account. Finally, 74% of households
without a bank account express interest in attending a free financial literacy
training session.
A challenge in interpreting observational regressions is that variables may
be collinear and hence causal links are not clear. This important caveat applies
to much of the literature on financial literacy. To draw out causal relationships,

we design a field experiment.
V. Experiment Design
This section describes the intervention we conducted in Indonesia to test
whether financial literacy acts as a barrier to opening a bank account. The
results of the experiment are analyzed in Section VI.
A. Financial Literacy Intervention

To study whether financial literacy training could stimulate demand for
financial services, we worked with an international nonprofit organization
in Jakarta, Microfinance Innovation Center for Resources and Alternatives
(MICRA). MICRA provides consulting and training programs to banks and
microfinance organizations in Indonesia.
MICRA developed a customized training session on bank accounts using
material adapted from a curriculum developed by the consortium consisting of
Microfinance Opportunities, Citigroup Foundation, and Freedom from Hunger.
The curriculum was designed for unbanked individuals, with the specific goal
of teaching households about bank accounts.
A challenge with financial education is that household preferences and cir-
cumstances vary, and not everyone may benefit from a particular financial be-
havior, making prescriptive education difficult (Lyons and Neelakantan (2008)).
To address this concern, our intervention focuses on a behavior that carries al-
most no cost: Bank Rakyat Indonesia, the country’s largest bank, offers a “SIM-
PEDES” account that, at the time of our intervention, requires a minimum
deposit of only Rp. 5,000 ($0.53) and charges no fees as long as an individual
deposits or withdraws money no more than four times per month. This account
pays no interest for deposit levels below Rp. 10,000 ($1.06), and pays increas-
ing interest rates for balances higher than this amount. Indonesian depositors
enjoy deposit insurance from the government.
Of course, if households have highly productive investment opportunities,
they may be loathe to save. Work by Rampini and Viswanathan (2010) suggests
that those who are capital constrained may in fact have less of an incentive to
manage risk than those who are somewhat wealthier. However, the value of a
buffer stock savings would certainly be high to extremely poor individuals (as
consumption approaches zero), particularly given the high cost of moneylender
credit. Recall as well from the baseline that 58% of households reported they
would open a bank account if there were no fees.
Moreover, while evidence suggests that some poor households have ac-
cess to high-return projects, it is also true that many poor households,
and even poor entrepreneurs, demonstrate quite low returns to capital
(de Mel, McKenzie, and Woodruff (2008)). Finally, we note that, even in a setting
with high returns to capital, access to savings technology could facilitate mak-
ing lumpy investments. In related work, Dupas and Robinson (2009) report a
randomized evaluation of savings accounts among a much poorer sample of in-

dividuals in Kenya. Despite the fact that the terms offered in the Kenyan bank
were much less favorable than those found in Indonesia (e.g., withdrawal fees),
Dupas and Robinson (2009) find that savings accounts improved investment
levels of self-employed individuals.
Working with MICRA, we identified individuals to serve as trainers who had
previous experience in financial sector work or education. The trainers were
given 2 days of specialized training related to the curriculum prior to the start
of the experiment. MICRA provided the training for the trainers. The salary
offered to the trainers was relatively high (200,000 Rp./hour); thus, the quality
of this intervention’s delivery is likely to be as good as or better than that of
any other large-scale intervention.
The financial literacy experiment took place in the 64 Access to Finance
survey villages on the island of Java. Thirty households were sampled in each
village, for a total of 64 × 30 = 1,920 households. Of these, 1,173 households did
not have a bank account at the time of the survey. After completing the Access
to Finance survey, each of these unbanked households was offered an opportu-
nity to participate in the experiment. If a respondent agreed to participate, he
or she was subsequently randomly assigned a financial incentive level and a fi-
nancial literacy training invitation status. The financial incentives offered were
Rp. 25,000, Rp. 75,000, and Rp. 125,000, with equal probability, for opening a
bank account within 2 months of the intervention. To receive the incentive,
the household was required to fill out a postage-paid mail-in form indicating
the participant’s name and bank account number. Upon receipt of this card,
the survey firm transferred the appropriate incentive amount to the respon-
dent’s account.
Independent of the incentive level, households were assigned to either treat-
ment or control for the financial literacy training program. Treatment house-
holds received from the surveyor a written invitation to attend a 2-hour finan-
cial literacy training session to be held in the village on a weekend. Households
that did not agree to participate in the experiment were eligible to receive invi-
tations to the financial literacy training, but, because we do not know if these
households decided to open a bank account, they do not form part of our exper-
imental sample. Half of the households (again randomly assigned) receiving a
financial literacy invitation were allowed to invite a friend to accompany them
to the session.14
In each of the 64 villages, a financial literacy training session was held
within 1 month of the date the survey was conducted. Invited households were
reminded about the training the day before it occurred.
Unfortunately, 23 villages had to be dropped from the sample because of
evidence that the surveyors were collaborating with households to ensure
14 The experimental plan initially called for a range of invitations designed to elicit the impor-
tance of peer effects. Operational limitations precluded any peer invitations in the first 14 villages
surveyed. In the subsequent villages, half of the treatment sample was offered an invitation for a
friend.
households received high incentives.15 This left a sample of 1,230 households,

of which 736 did not have bank accounts. Panel A of Table IV illustrates the
experiment sample composition.
The outcome of interest is whether a household opened a bank account. We
measure this based on financial incentive claims. After verifying the identity of
the claimant and the existence of a bank account, we were left with 49 claims
that came from eligible households that had indeed opened a bank account.
B. Summary Statistics and Checks of Randomization

Summary statistics for the experimental group are presented in the Internet
Appendix. Results are reported separately for all unbanked households who
agreed to participate in our experiment and for all unbanked households who
declined to participate. While, of course, we could not compel participation,
the take-up rate is relatively high, at 77%: 564 out of 736 households without
bank accounts choose to participate in the experiment. Households made this
decision prior to learning the precise details of the survey, including the size
of the incentive and whether they would receive a literacy invitation. We find
that rural households and older and unmarried respondents are less likely to
participate in the experiment, whereas respondents that are more educated,
more financially literate, and more interested in financial matters are more
likely to participate.
Turning to the summary statistics in the Internet Appendix, slightly more
than half of our experiment sample households are rural and half are female
headed. Further, respondents are, on average, in their early 40s, are over-
whelmingly married, and have attended some school. About 70% are employed
and 70% own their homes. The average financial literacy score, as measured
by questions asked in the Access to Finance Survey, is 50%, though 70% of the
sample claim they are interested in financial matters.
Panel B of Table IV provides a test of the randomization. We first present
mean differences between those invited to financial literacy training (274 out
of 564) and those who were not (290 out of 564), and then for those who were
offered the low (170), middle (190), or high (204) incentive. Column 3 tests the
hypothesis of equality of means between the invited and noninvited groups,
while Column 7 tests for equality of means across the assigned incentives. By
and large, the randomization appears successful as baseline characteristics do
not vary systematically by treatment status.
15 The survey was conducted in two waves. During wave one, which covered 48 villages,
the size of the incentive for participating households was chosen by the surveyor drawing
one of three colored balls from a bag. For four surveyors, a Pearson χ 2 test rejects the
hypothesis that the allocation of incentives was random. The 23 villages visited by these survey-
ors have been dropped from the sample. During wave two, incentive amounts were pre-assigned
to households. There is no evidence that the incentive amount affected households’ participation
decisions.
1952
Table IV
Experimental Sample, Indonesia
This table reports sample composition and tests of random treatment assignment for an experiment testing the effect of offering financial literacy
training and financial incentives on respondents’ decision to open a bank account. Panel A gives sample size and the mean of the outcome group by
treatment status. Panel B provides tests of random assignment. The p-values in column (3) report the statistical significance of a test for difference
between the mean of invited and noninvited individuals; the p-values in column (7) correspond to a joint test of significant differences between medium
and low, and high and low, incentive categories. Standard errors are adjusted for clustering at the village level. ∗∗∗ indicates statistical significance
at the 1% level, ∗∗ at the 5% level, and ∗ at the 10% level.
Panel A: Sample Composition
Opened Bank Account
N Percent N Percent
(1) (2) (3) (4)
Surveyed individuals 1,230

Of whom, no bank account 736 60%
Of whom, participated in experiment 564 77% 49 9%
Incentive treatment
The Journal of FinanceR
Low incentive ($3) 170 30% 6 4%

Medium incentive ($8) 190 34% 17 9%
High incentive ($14) 204 36% 26 13%
Literacy treatment
Invited to financial literacy training 274 49% 21 8%
Not invited to financial literacy training 290 51% 28 10%
(continued)
Table IV—Continued
Panel B: Test of Random Assignment
Invited Not Invited p-value Low Medium High p–value

(1) (2) (3) (4) (5) (6) (7)
Rural household 0.58 0.53 0.053∗ 0.57 0.53 0.55 0.591

Female 0.55 0.50 0.287 0.54 0.50 0.53 0.681
Age 41.84 40.55 0.302 40.76 40.72 41.95 0.554
Married 0.87 0.85 0.529 0.88 0.86 0.85 0.710
Muslim 0.97 0.99 0.102 0.99 0.98 0.98 0.662
Family size 2.73 2.82 0.446 2.73 2.76 2.82 0.756
Attended school 0.90 0.90 0.916 0.89 0.93 0.88 0.134
Log of consumption expenditure 17.26 17.32 0.332 17.18 17.33 17.35 0.213
Employed 0.68 0.69 0.792 0.65 0.67 0.72 0.367

Financial literacy score 0.46 0.51 0.039∗∗ 0.49 0.49 0.48 0.821
Cognitive / math skills score 0.79 0.80 0.408 0.78 0.80 0.79 0.727
Believe household saves enough 0.43 0.49 0.101 0.45 0.47 0.47 0.846
Interested in financial matters 0.72 0.72 0.867 0.69 0.73 0.73 0.626
1953
Table V
Experimental Results: The Effect of Financial Literacy Education
and Incentives on Bank Account Opening
This table reports the results from a randomized experiment measuring the effect of offering
financial literacy training and financial incentives on respondents’ decision to open a bank account.
The dependent variable is an indicator for whether the respondent opened a bank account. A linear
probability model is used. Standard errors, clustered at the village level, are given in parentheses
beneath each point estimate. ∗∗∗ indicates statistical significance at the 1% level, ∗∗ at the 5% level,
and ∗ at the 10% level.
(1) (2) (3) (4) (5) (6)
Financial literacy invitation −0.020 −0.022 0.022 0.029

(0.027) (0.028) (0.028) (0.034)
Incentive = 75,000 0.054∗∗ 0.048∗ 0.065∗ 0.066∗
(0.024) (0.026) (0.036) (0.037)
Incentive = 125,000 0.092∗∗∗ 0.088∗∗∗ 0.136∗∗∗ 0.137∗∗∗
(0.026) (0.029) (0.036) (0.033)
(Incentive = 75,000) ∗ Financial −0.021 −0.036
literacy invitation
(0.047) (0.052)
(Incentive = 125,000) ∗ −0.090 −0.101
Financial literacy invitation
(0.057) (0.062)
Constant 0.097∗∗∗ −0.444 0.035∗∗ −0.447 0.024 −0.455
(0.017) (0.306) (0.014) (0.308) (0.017) (0.303)
Household controls Yes Yes Yes
Observations 564 564 564 564 564 564
R2 0.001 0.068 0.018 0.082 0.023 0.089
VI. Experimental Results

The main experimental results are presented in Table V. Because the assign-
ment of incentives and invitations to financial literacy training were randomly
determined, unbiased estimates of the causal impact of each can be obtained
by estimating the following equation:16
Openi = α + β ∗ LitInvitei + εi , (1)
where Openi is a dummy variable indicating whether a household has opened a
bank account and LitInvitei is a dummy variable for whether the household was
invited to attend the training session. We focus initially on the reduced-form
relationship because it is difficult to compel people to attend a training session;
thus, the intention-to-treat estimate may be of greatest interest. Equation (1)
is therefore the reduced form.
The point estimate on LitInvitei in equation (1) is −0.02, with a standard
error of 0.027. Thus, the financial literacy program we offered appears to have
no effect on the likelihood that a client opens a bank account. Column 2 presents
16 We chose a linear probability model because the coefficients are simple to interpret. We obtain
very similar results from a marginal effects probit model.
the same results, but includes a set of household controls available from our
survey.17
Similarly, to determine the effect of incentives on opening an account, we
estimate
Openi = α + γ M ∗ MidPayi + γ H ∗ HiPayi + εi , (2)
where MidPayi indicates whether the household received an incentive of Rp.

75,000 and HiPayi indicates whether the household received an incentive of Rp.
125,000. The omitted category is the small incentive of Rp. 25,000. Standard
errors in all specifications are clustered at the village level.
The point estimates on MidPayi and HiPayi in equation (2) are large and
statistically significant. These estimates suggest that incentives have a large
effect on households’ decision to open a bank account. A household receiving the
middle incentive is 5.4 percentage points more likely to open a bank account
than a household receiving a low incentive. This represents a 150% increase
over the group offered the low incentive, of whom only 3.5% open accounts. The
effect of HiPay is even greater: the point estimate of 9.2 percentage points rep-
resents a 260% increase in the probability of opening a bank account compared
to the group receiving Rp. 25,000.
This effect is large. For example, we saw in Table V that a 1-standard-
deviation increase in log household expenditures is associated with a 14.9-
percentage-point increase in the likelihood of having a bank account. Moving
from the low to the high incentive has an effect equivalent to increasing house-
hold expenditures by two-thirds of a standard deviation.
Finally, we explore the possibility of an interaction between financial literacy
training and financial incentives with the following regression:
Openi = α + β ∗ LitInvitei + γ M ∗ MidPayi + γ H ∗ HiPayi
(3)
+ θ M ∗ MidPayi ∗ LitInvitei + θ H ∗ HiPayi ∗ LitInvitei + εi .
Columns 5 and 6 of Table V report results. We find no interaction effect: the in-
teraction point estimates are relatively imprecisely estimated but statistically
indistinguishable from zero. The main effect of incentives is unchanged.
Although our experiment results are quite strong when comparing low to
high incentives, the overall take-up of bank accounts is fairly low—among all
households who were offered an incentive, fewer than 10% open a bank ac-
count. Hence, other constraints, possibly social and behavioral barriers, may
be present. Cole et al. (2010), for example, find that lack of trust in financial
service providers acts as a barrier to take-up for insurance. However, trust may
not be as important in this setting as the survey company was well known. The
initial survey itself involved small payments to participants and the interven-
tion was conducted in cooperation with the local government.
17 The controls include household/respondent location, gender, age, marital status, religion,
family size, schooling, consumption, employment status, financial literacy score, cognitive ability,
and expressed interest in financial matters.
A. Heterogeneous Treatment Effects

While there is no effect on the general population, it is possible that financial
literacy training is effective for particular subsets of the population. Because
the experiment was conducted in conjunction with the survey, we did not strat-
ify by education or levels of financial literacy when assigning treatment levels.
However, there is strong reason to believe that the effects of financial educa-
tion may vary based on individuals’ characteristics. Limited financial literacy
is likely a larger constraint for households with low levels of formal or financial
education, as information acquisition may be costlier or more difficult for those
who cannot read. Similarly, because the program was designed for individuals
with low levels of financial literacy, it may have been most effective among this
group. As can be seen in the Internet Appendix, both financial literacy and
education levels are positive predictors of having a bank account. This implies
that financial literacy training may have greater effects on individuals who are
less educated and less financially literate.
To test the above prediction, in Table VI we split the sample and explore the
possibility of heterogeneous treatment effects. In Columns 1 and 2, we interact
LitInvitei , MidPayi , and HiPayi with a dummy variable indicating whether the
respondent reports having no formal schooling:
Openi = α + δ ∗ NoSchooli + β ∗ LitInvitei + θ ∗ (NoSchooli ∗ LitInvitei )

+ γ M ∗ MidPayi + γ H ∗ HiPayi + κ M ∗ NoSchooli ∗ MidPayi (4)

+ κ H ∗ NoSchooli ∗ HiPayi + εi .
We find, as before, that for literate households, the invitation has no ef-
fect: the point estimate of β is indistinguishable from zero at −0.032. How-
ever, for households that report having received no schooling, we find that
the financial literacy training program has a substantial effect: the sum
β + θ is equal to 12.3 percentage points (Column 1); an F-test for the
joint significance of β + θ yields a p-value of 0.07. Approximately one-
tenth of the sample is illiterate. The coefficients κ M and κ H are negative,
with κ M weakly statistically significant. The hypotheses γ M + κ M = 0 and
γ H + κ H = 0 cannot be rejected at standard levels of significance, suggesting
that for this subgroup, financial incentives are not important determinants of
behavior.
As a second way of cutting the data, we test whether the effect varies with
initial levels of financial literacy. Columns 3 and 4 estimate equation (4)
with a main effect and interactions for whether an individual obtained a
score below the median score in the baseline financial literacy test replac-
ing the schooling terms. The point estimate of the effect of an invitation on
those with above-average financial literacy is negative but statistically in-
distinguishable from zero at −4.9 percentage points. The estimate of the ef-
fect of the program on low financial literacy households β + θ is 5.1%. The
hypothesis that this sum is zero can only be rejected at the 15% level of
Table VI
Experimental Results: Heterogeneous Effects of Financial Literacy
Education and Incentives on Opening of Bank Accounts
This table reports the results from a randomized experiment measuring the effect of offering
financial literacy training and financial incentives on respondents’ decision to open a bank account.
The dependent variable is an indicator for whether the respondent opened a bank account. Columns
(1) and (2) include main effects and interaction terms for households with no formal schooling.
Columns (3) and (4) include main effects and interaction terms for households who initially scored
below the median level of financial literacy. A linear probability model is used. Standard errors,
clustered at the village level, are given in parentheses beneath each point estimate. ∗∗∗ indicates
statistical significance at the 1% level, ∗∗ at the 5% level, and ∗ at the 10% level.
(1) (2) (3) (4)
Financial literacy invitation −0.032 −0.031 −0.049 −0.048

(0.029) (0.030) (0.034) (0.036)
Incentive = 75,000 0.061∗∗ 0.057∗∗ 0.06 0.051
(0.028) (0.029) (0.039) (0.040)
Incentive = 125,000 0.099∗∗∗ 0.091∗∗∗ 0.1∗∗∗ 0.098∗∗∗
(0.027) (0.030) (0.030) (0.034)
Unschooled −0.055 −0.067
(0.050) (0.068)
Unschooled ∗ Financial literacy 0.155∗∗ 0.139∗
invitation
(0.068) (0.071)
Unschooled ∗ Incentive = 75,000 −0.135∗ −0.131∗
(0.071) (0.072)
Unschooled ∗ Incentive = 125,000 −0.062 −0.036
(0.084) (0.093)
Below-median financial literacy −0.076∗∗ −0.056
(0.037) (0.050)
(0.044) (0.043)
Below-median financial literacy ∗ −0.016 −0.008
Incentive = 75,000
(0.060) (0.058)
Below-median financial literacy ∗ −0.024 −0.031
Incentive = 125,000
(0.049) (0.055)
Constant 0.05∗∗ −0.377 0.067∗∗ −0.377
(0.020) (0.325) (0.027) (0.331)
Household controls Yes Yes
Observations 564 564 564 564
R2 0.029 0.09 0.03 0.089
significance. The incentives have an effect for both subgroups: the point
estimate of the sum γ H + κ H is 7.6 percentage points, significant at the 10%
level.
These results suggest that an intervention delivered to the general popu-
lation will not produce significant effects. However, a training program that
targets individuals with low levels of education and financial literacy can in-
crease demand for financial services.
B. Treatment on Treated
Approximately 69% of respondents invited to attend the program do in fact
attend the training. An alternative method of estimating equation (1) is to use
the invitation for the program as an instrument for the endogenous indicator
of whether the individual attended.18 Under reasonable assumptions, this pro-
vides the effect of treatment on the treated, also known as the local average
treatment effect (Imbens and Angrist (1994)). These results are reported in
Table VII.
Given that there is no reduced-form relationship between the training in-
vitation and opening a bank account (Table V), it is not surprising that the
instrumental variable estimate of the effect of training is also zero (Columns 1
and 2). The size of the standard error increases somewhat, but we can still com-
fortably rule out an effect size equivalent to the large incentive. Columns 3 and
4 examine heterogeneous treatment effects using Invited as an instrument for
attending and Invited∗Unschooled as an instrument for Attended∗Unschooled.
The treatment effect for unschooled is still positive, though no longer statisti-
cally significant. In Columns 5 and 6, we repeat this exercise for respondents
below the median level of financial literacy. Here, we continue to find that the
financial literacy intervention has a large effect on households with low ini-
tial financial literacy. Respondents with below-median financial literacy are
20% more likely to open a bank account within 2 months if they attended the
financial literacy training session.
VII. Follow-Up Results

In January 2010, approximately 2 years after our intervention, we conducted
a brief follow-up survey to investigate whether households still had their bank
accounts open, and whether households had improved their savings habits. In
this section, we study whether these behaviors were correlated with any of our
treatments.
A. Follow-Up Sample Characteristics

Our primary purpose of conducting a follow-up survey was to verify whether
households who opened bank accounts immediately following our initial inter-
vention were still using them 2 years later, or whether they had simply allowed
them to lapse after collecting the subsidy. While it would have been preferable
to visit all households, budget constraints prevented this. As a compromise,
we conducted brief interviews of all households in villages in which at least
one household had opened a bank account in response to our initial study.
Thus, from the baseline sample of 564 households in 40 villages, our follow-up
18 There is no need to instrument the incentives offered as there was no endogenous take-up of
the incentives.
Table VII
Instrumental Variable Estimates of Experiment and Heterogeneous Treatment Effects
This table reports instrumental variable estimates of the effect of offering financial literacy training and financial incentives on respondents’ decision
to open a bank account. The dependent variable is an indicator for whether the respondent opened a bank account. Financial literacy attendance is
instrumented for with the assignment of a financial literacy invitation. Columns (1) and (2) include main effects. Columns (3) and (4) include main
effects and interaction terms for households with no formal schooling. Columns (5) and (6) include main effects and interaction terms for households
who initially scored below the median level of financial literacy. A linear probability model is used. Standard errors, clustered at the village level, are
given in parentheses beneath each point estimate. ∗∗∗ indicates statistical significance at the 1 % level, ∗∗ at the 5% level, and ∗ at the 10% level.
(1) (2) (3) (4) (5) (6)
Attended financial literacy program −0.033 −0.036 −0.056 −0.059 −0.081 −0.078
(0.049) (0.051) (0.050) (0.053) (0.056) (0.057)
Incentive = 75,000 0.053∗∗ 0.047∗ 0.06∗∗ 0.051∗ 0.057 0.049
(0.024) (0.025) (0.027) (0.029) (0.039) (0.038)
Incentive = 125,000 0.092∗∗∗ 0.088∗∗∗ 0.099∗∗∗ 0.089∗∗∗ 0.103∗∗∗ 0.101∗∗∗
(0.026) (0.027) (0.026) (0.029) (0.030) (0.034)
Unschooled −0.159 −0.166
(0.154) (0.153)
Unschooled ∗ Attended financial literacy program 0.544 0.489
(0.468) (0.403)
(0.113) (0.103)
(0.125) (0.107)
Below-median financial literacy −0.115∗ ∗ −0.084
(0.058) (0.060)
∗
Below-median financial literacy Attended financial literacy program 0.206∗ ∗ 0.172∗
(0.104) (0.094)
Below-median financial literacy ∗ Incentive = 75,000 −0.013 −0.006
(0.059) (0.056)
Below-median financial literacy ∗ Incentive = 125,000 −0.027 −0.032
(0.053) (0.056)
Constant 0.05∗∗ −0.404 0.058∗∗ −0.426 0.077∗ ∗ −0.391
(0.024) (0.312) (0.026) (0.331) (0.032) (0.317)
Observations 564 564 564 564 564 564
1959
sample comprises 394 households from 27 of the 40 villages.19 Because the

initial treatments were randomly assigned within the village, we are assured
(and we confirm) that we achieve a balanced sample in the follow-up: treat-
ment status is not correlated with observable socio-economic or demographic
characteristics.20
In the 27 follow-up villages, we were able to re-interview 349 out of the
394 households. The attrition is not correlated with any of the treatments or
with whether a bank account was opened previously. In addition, no attrition
was due to household refusal to answer. Rather, 85% of the attrition occurred
because the household had moved permanently, while the remainder was due
to death or debilitating illness. It is unlikely that our intervention affected
mortality rates.
B. Long-Run Effects
Regression results from the follow-up survey are reported in Tables VIII
through X. Table VIII investigates whether those households that reported
opening a bank account following our intervention still have their accounts
open 2 years later. The results show that households that received the highest
incentive are significantly more likely to still have their accounts open com-
pared to those that received the lowest incentive. These results are statistically
indistinguishable from the short-run results. Further, survey questions reveal
that of the households that still have their accounts open, 62% have used their
account in the last year to deposit, withdraw, send, or receive money.
A necessary feature of our study is that the subsidy payment offered for
opening a bank account following the study be time-limited. In a separate set
of regressions, we use as a dependent variable whether the household opened
a bank account at any point in time during the 2 years between the initial
treatment and the follow-up survey. The point estimates on financial literacy
invitation range from 1 percentage point to 7.6 percentage points, but are not
statistically significant (results available in the Internet Appendix).
Table IX next measures whether the heterogeneous effects of bank ac-
count opening are present in the long run. We find that the impact of fi-
nancial literacy training invitation remains significant for households below
the median level of initial financial literacy.21 The results based on school-
ing status are no longer significant, though the point estimates are for the
most part similar to the short-run estimates. In this regression, the coef-
ficient on Unschooled∗HighPay is negative and statistically significant even
with household controls,22 suggesting that uneducated households simply take
19 Much of the cost of surveying is the fixed travel and accommodation of visiting a village, and
thus it made sense to interview all study households in villages that were visited.
20 In an estimation model of bank accounts with village fixed effects, the 13 omitted villages
would not contribute to the identification of any parameters of interest.

21 The F-test of the sum of LitInvite and Below Median Financial Literacy * LitInvite is signifi-
cant at the 10% level for the specification without household controls.
22 The F-test of the sum of HighPay and Unschooled * HighPay is not statistically significant.
Table VIII
Follow-Up Results: The Long-Run Persistence of Financial Literacy
Education and Incentives on Bank Account Opening
This table reports results from a follow-up survey 2 years after the financial literacy education
and incentives intervention, conducted among participants in villages where a household opened
a bank account immediately after the intervention. The dependent variable is an indicator for
whether a respondent opened a bank account immediately after the intervention and still has the
bank account 2 years later. The sample includes all households that were successfully interviewed
in the follow-up survey. A linear probability model is used. Standard errors, clustered at the village
level, are given in parentheses beneath each point estimate. ∗∗∗ indicates statistical significance
at the 1% level, ∗∗ at the 5% level, and ∗ at the 10% level.
(1) (2) (3) (4) (5) (6)
Financial literacy invitation −0.041 −0.044 0.011 0.019

(0.028) (0.029) (0.032) (0.041)
Incentive = 75,000 0.043 0.058∗ 0.063 0.09∗
(0.031) (0.034) (0.048) (0.048)
Incentive = 125,000 0.092∗∗∗ 0.088∗∗ 0.127∗∗∗ 0.123∗∗∗
(0.032) (0.035) (0.046) (0.047)
(Incentive = 75,000) ∗ −0.041 −0.068
Financial literacy
invitation
(0.069) (0.074)
(Incentive = 125,000) ∗ −0.083 −0.086
Financial literacy
invitation
(0.061) (0.068)
Constant 0.094∗∗∗ −0.831∗∗ 0.028∗ −0.872∗∗∗ 0.021 −0.866∗∗∗
(0.019) (0.319) (0.015) (0.300) (0.021) (0.300)
Observations 349 349 349 349 349 349
R2 0.006 0.1 0.02 0.111 0.028 0.119
advantage of the financial incentive and subsequently close their bank ac-
counts. This result has important implications for the desirability of subsidies
as a tool to expand financial access: financial incentives alone may not be suffi-
cient to draw uneducated households into the financial system as these house-
holds may simply claim the incentives without actually using the financial
services.
C. Effect on Savings Decision

An advantage of examining banking status is that it is easy to measure.
However, financial education often promotes asset accumulation as well. In
fact, one of the key messages in our financial literacy seminars is the importance
of savings for future expected and unexpected needs.
Our follow-up analysis examines household savings behavior. We elicit this
information by asking the question “Do you currently have any savings?” We
also ask households to report the level of savings. The latter variable, however,
Table IX
Long-Run Persistence: Heterogeneous Effects of Incentives
and Financial Literacy Education on Bank Account Opening
This table reports results from a follow-up survey 2 years after the financial literacy education
and incentives intervention, conducted among participants in villages where a household opened
a bank account immediately after the intervention. The dependent variable is an indicator for
whether a respondent opened a bank account immediately after the intervention and still has
the bank account 2 years later. Columns (1) and (2) include main effects and interaction terms for
households with no formal schooling. Columns (3) and (4) include main effects and interaction terms
for households who initially scored below the median level of financial literacy. The sample includes
all households that were successfully interviewed in the follow-up survey. A linear probability
model is used. Standard errors, clustered at the village level, are given in parentheses beneath
each point estimate. ∗∗∗ indicates statistical significance at the 1% level, ∗∗ at the 5% level, and
∗ at the 10% level.
(1) (2) (3) (4)
Financial literacy invitation −0.042 −0.038 −0.076∗ −0.071∗

(0.03) (0.03) (0.04) (0.04)
Incentive = 75,000 0.05 0.065∗ 0.055 0.058
(0.04) (0.04) (0.05) (0.05)
Incentive = 125,000 0.103∗∗∗ 0.096∗∗∗ 0.082∗∗ 0.073∗
(0.03) (0.04) (0.04) (0.04)
Unschooled −0.006 0.003
(0.05) (0.06)
Unschooled ∗ Financial literacy invitation 0.1 0.058
(0.07) (0.07)
(0.08) (0.09)
Unschooled ∗ Incentive = 125,000 −0.158∗∗ −0.122∗
(0.06) (0.07)
Below-median financial literacy −0.081∗ −0.106∗
(0.04) (0.06)
Below-median financial literacy ∗ Financial 0.14∗∗ 0.114∗
literacy invitation
(0.06) (0.06)
Below-median financial literacy ∗ Incentive = −0.056 −0.018
75,000
(0.06) (0.06)
Below-median financial literacy ∗ Incentive = −0.003 0.009
125,000
(0.06) (0.07)
Constant 0.045∗ −0.793∗∗ 0.072∗∗ −0.733∗∗
(0.02) (0.31) (0.03) (0.32)
Household controls Yes Yes
Observations 349 349 349 349
R2 0.037 0.119 0.044 0.126
is reported with significantly more noise (and refusals to answer) than the
simple question of whether the household has any savings.
Regression results in Table X show that, while there is no direct effect of
the financial literacy invitation, there is also no direct effect of the subsidies,
Table X
Long-Run Effects of Financial Literacy Education and Incentives
on Savings
This table reports results from a follow-up survey 2 years after the financial literacy educa-
tion and incentives intervention, conducted among participants in villages where a household
opened a bank account immediately after the intervention. The dependent variable is an indicator
for whether the household currently has any savings. The sample includes all households that
were successfully interviewed in the follow-up survey. A linear probability model is used. Stan-
dard errors, clustered at the village level, are given in parentheses beneath each point estimate.
∗∗∗ indicates statistical significance at the 1% level, ∗∗ at the 5% level, and ∗ at the 10% level.
(1) (2) (3) (4) (5) (6)
Financial literacy invitation 0.008 0.041 −0.104 −0.037

(0.040) (0.044) (0.064) (0.061)
Incentive = 75,000 0.031 −0.011 −0.023 −0.016
(0.060) (0.059) (0.068) (0.069)
Incentive = 125,000 0.007 −0.03 −0.105 −0.116
(0.057) (0.061) (0.083) (0.078)
(Incentive = 75,000) ∗ Financial 0.092 0.013
literacy invitation
(0.094) (0.100)
(Incentive = 125,000) ∗ 0.231∗∗ 0.205∗∗
Financial literacy invitation
(0.116) (0.105)
Constant 0.311∗∗∗ −0.888 0.303∗∗∗ −0.851 0.362∗∗∗ −0.92
(0.047) (0.590) (0.053) (0.586) (0.070) (0.612)
Observations 349 349 349 349 349 349
R2 0.000 0.177 0.001 0.176 0.011 0.188
suggesting that (relatively small) “high subsidies” were dissaved by households

over the 2 years between the experiment and the follow-up survey. However,
the interaction between the high incentive and the financial literacy invitation
is large and statistically significant. Compared to households that received the
low incentive and no financial literacy invitation, households receiving both a
high incentive and a financial literacy invitation are more than 20 percentage
points more likely to report having savings. However, summing the relevant
coefficients (on invitation, high incentive, and the interaction) yields a reported
fraction of household saving of 38.5%, only 2.2 percentage points higher than
the mean for the households who received no invitation and the low incentive
(36.2%). An F-test of this sum cannot reject the hypothesis that it is zero.23
While this result shows some promise for our financial literacy program, we do
not have sufficient statistical power to detect an overall net effect on savings.
In particular, the sum of all relevant coefficients, although positive, is not
statistically significant.
Of the households that reported savings, 61% reported that they were saving
for emergencies, 34% for school fees, and 12% for business investment.
23 The F-test of joint significance has a p-value of 0.79.
VIII. Conclusion
Using new surveys from two of the most populous countries in the world, this
paper presents compelling evidence that financial literacy is an important pre-
dictor of financial behavior in emerging market countries. These correlations,
well documented in developed countries, have spurred governments, nonprof-
its, and firms to promote financial literacy as a means of expanding the depth
and breadth of the financial system.
The benefits of better financial literacy may be great. On a personal level,
individuals may save more, and better manage risk, by purchasing insurance
contracts. There may even be general equilibrium effects: increased demand by
households for financial services may improve risk-sharing, reduce economic
volatility, improve intermediation, and speed overall financial development.
In turn, this could facilitate competition in the financial services sector and,
ultimately, more efficient allocation of capital within society.
Despite the potential benefits of financial literacy, to date there is no credible
evidence on the effects of financial literacy programs. This paper reports the
first randomized evaluation of a carefully designed and delivered financial lit-
eracy training program. We find that the education program has modest effects,
stimulating demand for bank accounts among uneducated and less financially
literate households. A second intervention providing small subsidies for open-
ing an account demonstrates that, given proper incentives, many individuals
would open accounts even without financial literacy training. A follow-up study
conducted 2 years after the initial intervention shows that those who were origi-
nally offered the high incentives are significantly more likely to have used bank
accounts in the past year to deposit, withdraw, send, or receive funds.
Where does this study leave us? On the one hand, the survey data from
Indonesia and India demonstrate that financial literacy is an important cor-
relate of household financial behavior and well-being. Indeed, it is one of the
strongest and most consistent predictors of demand for financial services. These
results provide evidence that financial literacy is important and that educated
consumers make better financial decisions. In addition, we find that demand
for financial education is quite high: 69% of those invited choose to attend the
course.
On the other hand, our experimental results show that this financial educa-
tion program is not an effective tool for promoting the use of bank accounts.
It is useful to think about a simple cost-benefit analysis. Even if targeted to
those for whom the intervention is most effective, the program is not cost ef-
fective. The literacy training costs approximately U.S. $17 per head to deliver.
Among those with low levels of initial financial literacy (i.e., below-median
score on baseline financial literacy assessment), the training program increased
the share of households opening a bank savings account by approximately
five percentage points. Thus, inducing the opening of one bank account cost
$17/0.05 = $340. In contrast, for this same subsample, increasing the sub-
sidy from $3 to $14 led to a 7.6% increase in the probability of opening a
bank savings account, suggesting a cost per bank savings account opened of
$11/0.076 = $145. Thus, subsidies are almost 2.5 times more cost effective than
financial literacy education.
Of course, financial literacy may have additional value if it promotes asset
accumulation; a buffer stock of savings may be far more important than simply
having a bank account. Nevertheless, our evidence does not support the view
that low financial literacy is a severe impediment to demand for formal financial
services. Our study clearly demonstrates that prices matter both for opening of
bank accounts and for savings, and that individuals who open bank accounts
in response to incentives tend to keep them open for the long term. This finding
is consistent with the common practice in U.S. banks whereby banks offer cash
incentives or other gifts to those opening a new account.
The financial literacy program we evaluated was based on global best prac-
tices, using experienced, highly educated facilitators, and likely represents a
higher quality intervention than could be delivered on a mass scale. Neverthe-
less, we acknowledge that it was a short program, and that many respondents
reported in the baseline that they did not previously open a bank account be-
cause they had insufficient funds. The point estimate on the impact of financial
literacy on savings decisions is positive though statistically insignificant. We,
of course, cannot rule out the possibility that a more comprehensive and bet-
ter targeted education program could have positive, measurable impacts on
individuals’ lives.
Ultimately, however, our results suggest that financial deepening may be
more easily achieved through measures designed to reduce the price of finan-
cial services, for example, by promoting competition or low-cost technological
solutions such as mobile banking, rather than through large-scale financial lit-
eracy education. A carefully designed and targeted financial literacy program
that is more cost effective than a large-scale effort may serve as a valuable
complement to such financial reform.
REFERENCES
Ashraf, Nava, Dean Karlan, and Wesley Yin, 2006a, Tying Odysseus to the mast: Evidence from a
commitment savings product in the Philippines, Quarterly Journal of Economics 121, 635–672.
Ashraf, Nava, Dean Karlan, and Wesley Yin, 2006b, Deposit collectors, Advances in Economic
Analysis & Policy 6, 1–22.
Beck, Thorsten, Asli Demirguc-Kunt, and Patrick Honohan, 2008, Finance for All? Policies and
Pitfalls in Expanding Access (World Bank, Washington, DC).
Beck, Thorsten, Asli Demirguc-Kunt, and Maria Soledad Martinez Peria, 2007, Reaching out:
Access to and use of banking services across countries, Journal of Financial Economics 85,
234–266.
Bertrand, Marianne, and Adair Morse, 2010, Information disclosure, cognitive biases and payday
borrowing, Working paper, University of Chicago Booth School of Business.
Besley, Timothy, Stephen Coate, and Glenn Loury, 1993, The economics of rotating savings and
credit associations, American Economic Review 83, 792–810.
Binswanger, Hans P., 1980, Attitudes toward risk: Experimental measurement in rural India,
American Journal of Agricultural Economics 62, 395–407.
Binswanger, Hans, Dayanatha Jha, T. Balaramaiah, and Donald Sillers, 1980, The impacts of risk
aversion on agricultural decisions in semi-arid India, Working paper, ICRISAT.
Buehn, Andreas, and Friedrich Schneider, 2009, Shadow economies and corruption all over the
world: Revised estimates for 120 countries, Economics: The Open-Access, Open-Assessment
E-Journal 1, 1–53.
Chabris, Christopher F., David Laibson, Carrie L. Morris, Jonathon P. Schuldt, and Dmitry
Taubinsky, 2008, Individual laboratory-measured discount rates predict field behavior, Jour-
nal of Risk and Uncertainty 37, 237–269.
Christelis, Dimitris, Tullio Jappelli, and Mario Padula, 2006, Cognitive abilities and portfolio
choice, Discussion paper no. 5735, CEPR.
Cole, Shawn, Xavier Giné, Jeremy Tobacman, Petia Topalova, Robert Townsend, and James
Vickery, 2010, Barriers to household risk management: Evidence from India, Working paper,
Harvard Business School.
Cole, Shawn, and Kartini Shastry, 2009, Smart money: The effect of education, cognitive ability,
and financial literacy on financial market participation, Working paper no. 09-071, Harvard
Business School.
de Mel, Suresh, David McKenzie, and Christopher Woodruff, 2008, Returns to capital in microen-
terprises: Evidence from a field experiment, Quarterly Journal of Economics 123, 1329–1372.
Dohmen, Thomas, Armin Falk, David Huffman, and Uwe Sunde, 2010, Are risk aversion and
impatience related to cognitive ability? American Economic Review 100, 1238–1260.
Duflo, Esther, and Emmanuel Saez, 2003, The role of information and social interactions in
retirement plan decisions: Evidence from a randomized experiment, Quarterly Journal of
Economics 118, 815–842.
Dupas, Pascaline, and Jonathan Robinson, 2009, Savings constraints and microenterprise develop-
ment: Evidence from a field experiment in Kenya, NBER Working paper no. 14693, University
of California.
FINRA Investor Education Foundation, 2009, Financial Capability in the United States, http://
www.finrafoundation.org/resources/research/p120478, accessed April 1, 2010.
Gerardi, Kristopher, Lorenz Goette, and Stephan Meier, 2010, Financial literacy and subprime
mortgage delinquency: Evidence from a survey matched to administrative data, Working
paper no. 2010-10, Federal Reserve Bank of Atlanta.
Hogarth, Jeanne M., and Kevin H. O’Donnell, 1999, Banking relationships of lower-income families
and the governmental trend toward electronic payment, Federal Reserve Bulletin 86, 459–473.
Imbens, Guido W., and Joshua D. Angirst, 1994, Identification and estimation of local average
treatment effects, Econometrica 62, 467–475.
Karlan, Dean, and Martin Valdivia, 2011, Teaching entrepreneurship: Impact of business training
on Microfinance clients and institutions, The Review of Economics and Statistics, 93, 510–527.
Kendall, Jake, Nataliya Mylenko, and Alejandro Ponce, 2010, Measuring financial access around
the world, Policy research working paper no. 5253, World Bank.
La, Porta Rafael, Florencio Lopez-de Silanes, Andrei Shleifer, and Robert Vishny, 1998, Law and
finance, Journal of Political Economy 106, 1113–1155.
Lang, Frieder R., Dorothee Hahne, Stefanie Gymbel, Stefan Schroepper, and Katharina Lutsch,
2005, Erfassung des kognitiven leisungspotenzials und der “Big Five” mit computer-assisted-
personal-interviewing (CAPI): Zur reliabilitaet und validitaet zweier ultrakurzer tests und
des BFI-S, DIW Research Note, German Institute for Economic Research.
Levine, Ross, 2005, Finance and growth: Theory and evidence, in Philippe Aghion and Steven
Durlauf, eds. Handbook of Economic Growth (Elsevier Science, The Netherlands).
Lusardi, Annamaria, 2008, Financial literacy: An essential tool for informed consumer choice?
Working paper, Dartmouth College.
Lusardi, Annamaria, and Olivia S. Mitchell, 2006, Financial literacy and planning: Implications
for retirement well-being, Working paper no. 1, Pension Research Council.
Lusardi, Annamaria, and Olivia S. Mitchell, 2007a, Financial literacy and retirement prepared-
ness: Evidence and implications for financial education, Business Economics 42, 35–44.
Lusardi, Annamaria, and Olivia S. Mitchell, 2007b, Baby boomer retirement security: The roles of
planning, financial literacy, and housing wealth, Journal of Monetary Economics 54, 205–224.
Lusardi, Annamaria, and Peter Tufano, 2008, Debt literacy, financial experience and over-
indebtedness, Working paper, Dartmouth College and Harvard Business School.
Lyons, Angela, and Urvi Neelakantan, 2008, Potential and pitfalls of applying theory to the
practice of financial education, Journal of Consumer Affairs 42, 106–112.
Rabin, Matthew, and Richard Thaler, 2001, Anomalies: Risk aversion, Journal of Economic Per-
spectives 15, 219–232.
Rampini, Adriano, and S. Viswanathan, 2010, Collateral, risk management, and the distribution
of debt capacity, Journal of Finance 65, 2293–2322.
Stango, Victor, and Jonathan Zinman, 2009, Exponential growth bias and household finance,
Journal of Finance 64, 2807–2849.
Townsend, Robert, 1994, Risk and insurance in village India, Econometrica 62, 539–591.
UNESCO, 2007, Global Education Digest, http://www.uis.unesco.org, accessed May 1, 2009.
van Rooij, Maarten, Annamaria Lusardi, and Rob Alessie, 2007, Financial literacy and stock
market participation, NBER Working paper no. 13565, Dartmouth College.
Washington, Ebonya Lia, 2006, The impact of banking and fringe banking regulation on the
number of unbanked Americans, Journal of Human Resources 41, 106–137.

Prices or Knowledge? What Drives Demand For Financial Services in Emerging Markets?

Uploaded by

Copyright:

Available Formats

Prices or Knowledge? What Drives Demand For Financial Services in Emerging Markets?

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Prices or Knowledge? What Drives Demand For Financial Services in Emerging Markets?

Uploaded by

Copyright:

Available Formats

THE JOURNAL OF FINANCE • VOL. LXVI, NO.

Prices or Knowledge? What Drives Demand

SHAWN COLE, THOMAS SAMPSON, and BILAL ZIA∗

FINANCIAL DEVELOPMENT IS WIDELY recognized as an important determinant

I. Motivation and Context

(accessed February 11, 2009).

February 11, 2009).

6 See http://www.bri.co.id/english/layanan/simpanan.aspx?id=12 for terms of the savings prod-

According to recent estimates from Beck, Demirguc-Kunt, and Peria (2007)

II. Measuring Financial Literacy and Financial Decisions

vention. The survey was conducted prior to any intervention.

for question (iv).

Below Above Below Above Below Above Below Above

by Rabin and Thaler (2001).

at some point in the past. Approximately half (51%) of Indonesian households

III. What Predicts Financial Literacy?

(0.094) (0.096) (0.097) (0.104) (0.102) (0.109) (0.106) (0.155)

IV. What Does Financial Literacy Predict?

Higher financial literacy is significantly associated with greater use of bank

A. Demand for Financial Products

Panel A: Summary Statistics

Demand for savings products

Panel B: Determinants of Demand for Financial Products

Commitment Savings Deposit Collector Retirement Savings Literacy Training

Panel B: Determinants of Demand for Financial Products

Commitment Savings Deposit Collector Retirement Savings Literacy Training

Female 0.007 0.009 −0.021 −0.013 0.031 0.030 −0.022 −0.025

There is no evidence of a robust effect of human capital on interest levels

to much of the literature on financial literacy. To draw out causal relationships,

A. Financial Literacy Intervention

randomized evaluation of savings accounts among a much poorer sample of in-

households received high incentives.15 This left a sample of 1,230 households,

B. Summary Statistics and Checks of Randomization

Panel A: Sample Composition

Opened Bank Account

Surveyed individuals 1,230

Low incentive ($3) 170 30% 6 4%

Panel B: Test of Random Assignment

Invited Not Invited p-value Low Medium High p–value

Rural household 0.58 0.53 0.053∗ 0.57 0.53 0.55 0.591

Employed 0.68 0.69 0.792 0.65 0.67 0.72 0.367

(1) (2) (3) (4) (5) (6)

Financial literacy invitation −0.020 −0.022 0.022 0.029

VI. Experimental Results

Openi = α + γ M ∗ MidPayi + γ H ∗ HiPayi + εi , (2)

where MidPayi indicates whether the household received an incentive of Rp.

A. Heterogeneous Treatment Effects

Openi = α + δ ∗ NoSchooli + β ∗ LitInvitei + θ ∗ (NoSchooli ∗ LitInvitei )

(1) (2) (3) (4)

Financial literacy invitation −0.032 −0.031 −0.049 −0.048

VII. Follow-Up Results

A. Follow-Up Sample Characteristics

(1) (2) (3) (4) (5) (6)

sample comprises 394 households from 27 of the 40 villages.19 Because the

would not contribute to the identification of any parameters of interest.

(1) (2) (3) (4) (5) (6)

Financial literacy invitation −0.041 −0.044 0.011 0.019

C. Effect on Savings Decision

(1) (2) (3) (4)