Long-Run Price Elasticities of Demand for Credit:
Evidence from a Countrywide Field Experiment in Mexico 1
Dean Karlan
Yale University
Innovations for Poverty Action
M.I.T. Jameel Poverty Action Lab
NBER
Jonathan Zinman
Dartmouth College
Innovations for Poverty Action
M.I.T. Jameel Poverty Action Lab
NBER
August 2016
Abstract
We use randomized interest rates, offered across 80 geographically distinct regions for 29
months by Mexico’s largest microlender, to sketch the adjustment from a price change to
a new equilibrium. Demand is elastic, and increasingly so over time; e.g., the dollarsborrowed elasticity increases from -1.1 in year one to -2.9 in year three. Credit bureau
data does not show evidence of crowd-out. The lender’s profits increase, albeit noisily,
starting in year two. But competitors do not respond by reducing rates. These findings,
together with other results, suggest that informational frictions are important, and that
Compartamos Banco furthered its “double bottom line”, of improving social welfare
subject to a profitability constraint, by cutting rates.
1
[email protected], [email protected]. Thanks to Kerry Brennan, Angela Garcia Vargas, Matt Grant,
Kareem Haggag, Martin Sweeney and Rachel Strohm for excellent project management and research assistance, and
to Alissa Fishbane, Braulio Torres and Anna York from Innovations for Poverty Action for leadership of IPAMexico. Thanks to Abhijit Banerjee, Esther Duflo, Jake Kendall, Melanie Morten, David Roodman, Chris Snyder
and participants in seminars at NBER Household Finance, M.I.T./Harvard and NYU for comments. Thanks to
CGAP, in particular Richard Rosenberg, and the Bill and Melinda Gates Foundation for funding support. Thanks to
the management and staff of Compartamos Banco for their cooperation. Authors retained complete intellectual
freedom to report and interpret the results. Any opinions, errors or omissions are those of the authors.
1
Loan pricing is a strategic lever that drives profitability, market size, and market share.
Understanding consumer and competitor responses to price changes is of course critical for firms
seeking to maximize profits. It is also increasingly important for other parties—including
policymakers, donors and investors-- keen to see firms maximize subject to a “double-bottom
line” by pricing “social products” (like basic financial, health, and housing services) as low as
possible while still maintaining profitability, thus expanding outreach.
Understanding responses to price changes requires evidence on the responses of both
customers and suppliers, over time horizons that allow for adjustment dynamics to any new
equilibrium. We track consumer and competitor responses to randomly assigned, market-level
price cuts by the largest microlender 2 in Mexico. We estimate responses over 12 to 29 month
horizons, and explore market frictions that slow the adjustment to a new equilibrium.
Credit elasticities have been difficult to identify. Indeed, the experiment in this paper came
about because a leading microcredit policymaker, the World Bank-based Consultative Group to
Assist the Poor (“CGAP”), re-examined its prior belief of price-inelastic demand (Rosenberg
2002) after seeing experimental but short-run estimates from South Africa (Karlan and Zinman
2008). CGAP introduced us to Compartamos Banco (“CB”), a for-profit, publicly-traded bank
with a double-bottom line. CB was planning to cut rates but was unsure how much to cut, and
wanted to conduct a randomized test in order to better understand consumer and competitor
responses.
The nature of any market frictions that slow the adjustment to a new equilibrium is also
poorly understood, in part because it is rare to have all four key ingredients required for
convincing inference: a well-identified shock, data on both consumer and firm responses,
measured over appropriately long horizons. Understanding frictions is critical for properly
specifying the constraints facing firms and consumers in equilibrium, and/or along adjustment
paths.
2
Microlending is typically defined as the provision of small-dollar loans to (aspiring) entrepreneurs, although there
is some policy and practitioner debate over the definition. For recent reviews, see Karlan and Morduch (2009),
Armendariz de Aghion and Morduch (2010), and Banerjee (2013).
2
CB worked with us to randomize the interest rate offered on its core group lending product,
Crédito Mujer. We randomized at the level of 80 distinct geographic regions 3 throughout
Mexico, covering 130 bank branches, thousands of borrowing groups, and tens of thousands of
borrowers. “Treatment” branches implemented permanent 20 percentage point (pp) reductions in
the annual interest rate (on a base of roughly 100% APR), while control branches implemented
permanent 10pp reductions. Crucially, CB left these prices in place for 29 months, permitting
estimation of dynamics and long-run responses.
Demand and competitor responses can be attributed to price changes per se under the
assumption that no other changes are correlated with treatment assignment; e.g., we assume that
screening, monitoring, and marketing did not differ across treatment and control groups. The
screening and monitoring assumptions are supported by the finding that the experimentally lower
interest rate does not affect delinquency rates. The marketing assumption seems plausible given
study protocols, the fact that both treatment and control regions experienced rate cuts (20pp vs.
10pp) relative to CB’s prior rate, and that CB’s senior management did not want to adulterate the
experimental design. Another benefit of CB changing prices in all regions is that this reduces the
likelihood that competitors viewed any price cut as temporary. The main drawback to our design
is that the price variation is not rich enough to identify a global optimum; we can only identify
comparative statics between two points on the demand curve.
We identify four types of responses to CB’s price cuts. First we estimate demand elasticities
and how they change over time, using CB administrative data on loans and borrower
characteristics. We find elasticities that are towards the strong end of previous estimates even in
the short-run (e.g., a -1.1 for pesos borrowed over the first 12 months). More critically perhaps,
we find that lower prices bring in many new borrowers and that elasticities increase strongly
over time (reaching e.g., -2.9 for pesos borrowed during the final six months of the experiment).
Next we use credit bureau data to explore whether elasticities with respect to CB credit are upper
or lower bounds on net demand elasticities. We find no evidence of crowd-out—if anything, the
3
Each region comprises a distinct geographic market per CB’s market definition, corresponding roughly in most
cases to a municipality.
3
point estimates suggest crowd-in.
4
The lack of crowd-out on the consumer side—and
isomorphically, the lack of business-stealing on the competitor side—squares with several other
pieces of evidence suggesting that credit constraints bind (Sections 1 and 3-H). Then we examine
effects on CB’s costs and profitability, finding that the compositional shift toward new borrowers
induced by lower prices is costly for CB to manage in the short-run, but that ultimately the lower
price is more profitable. Indeed, based on preliminary results from this study, CB instituted the
lower-rate across its entire operation shortly after the experiment ended. Accounting sensibly for
the net present value of new clients—not just new loans—proves to be key in evaluating the
impacts on profits. Finally, we use mystery shopping data to measure competitor terms and find
little evidence that competitors respond (differentially) to CB’s lower prices, despite evidence
that many competitors operate in close proximity. This null result is imprecisely estimated
however. 5
What do these results imply? The elasticity results suggest that pricing can be an important
targeting lever for a social good, even one for which many key actors held priors of inelastic
demand (besides CGAP, see e.g., discussion in Armendariz de Aghion and Morduch (2010)).
Moreover our big differences between short- and long-run elasticities highlight the importance of
taking dynamics into account when making policy, business, or modeling decisions. For
example, although the lower interest rate produced substantially greater loan volume and
outreach (new clients) right away, profits did not increase until year two, meaning that CB only
saw benefits on both sides of its double bottom line when evaluating over a longer horizon than
any prior study of credit demand elasticities (details below).
4
These results square with those from a separate we conducted with CB on the impact of their loans. In Nogales, an
area not included in the interest rate study, we randomized the rollout of the lending program and also find that CB
did not crowd-out other sources of formal credit, despite substantial use of credit from close competitors (Angelucci,
Karlan, and Zinman 2015).
5
We note two additional caveats regarding our estimates of competitor non-response. First, we have only a single
snapshot, during year two of the experiment, and consequently might be missing longer-run effects. Of course that
sort of slow adjustment would be noteworthy as well. Second, it may be the case that competitors respond uniformly
in all regions. That pattern of response would also be noteworthy, as it would suggest that there are informational,
technological, and/or other frictions that discourage competitors from responding in a more granular way to CB’s
differential pricing across regions.
4
What can we learn about the frictions that mediate adjustment from a shock to equilibrium—
in this case a market leader’s pricing change? Putting our four main results together with other
pieces, we speculate that informational frictions are key on both sides of the market.
On the demand side, we start by casting doubt on some other candidate frictions. Contractual
frictions might be important in other settings, but here loan cycles are short enough (16 weeks) to
allow for much more rapid adjustment than we see over the 120-week span of the experiment.
Capital adjustment costs might be important in other settings, but our findings on the impacts of
CB’s expansion into a new geographic area, not covered by the interest rate experiment, show
that marginal CB borrowers expand their microbusinesses with inventory but not with lumpy
investments (Angelucci, Karlan, and Zinman 2015). It seems more likely that it takes time for
borrowers to learn about the new rates (lender practices probably contribute to this, as noted
below), including learning at the borrowing-group level about risk gradients with respect to price
under joint liability.
On the supply side, we speculate that information frictions also dampen competitor
responses, and more broadly create inertia at baseline prices. Prices are rarely featured in lender
advertising, or presented non-transparently, as further evidenced by the fact that CB staff did not
know, with any precision, what their competitors were charging (hence the need for our mystery
shopping effort). Competitors received especially weak signals that CB cut rates because they
did not actually lose clients (as evidenced by our results from the credit bureau data). Forecasting
profitability (at a new price level) is complex and replete with uncertainty, given e.g., the
dynamics of customer acquisition and retention. And our conversations with CB and many other
lenders across the globe indicate that installment lenders view experimenting with lower rates as
quite risky because it shifts customer reference points, making it very difficult to return to higher
rates in the event that lower ones prove unprofitable. This “no turning back” mechanism jibes
with our finding from a South Africa microloan market that borrowers are much more elastic
with respect to rate increases than decreases (Karlan and Zinman 2008).
These potential information frictions also help us characterize the industrial organization of
the Mexican microcredit market, which has hallmarks of many credit markets worldwide:
prevalent credit constraints and substantial price dispersion despite many competitors offering
5
similar products in close proximity (Zinman 2015). 6 We speculate that price dispersion persists,
in part, because competitors face substantial risks and information gaps in evaluating whether it
would be valuable to change prices. At the same time, prevalent credit constraints may contribute
to market power that cushions the blow for lenders who have not (yet) found their optimal price,
as evidenced by our finding that CB’s growth in lower-rate regions did not come at the expense
of competitors, and may have even produced some crowd-in.
What do our results reveal about the welfare implications of CB’s lower rate? Textbook
qualitative analysis (surplus under the demand curve) suggests that total welfare increases
substantially, particularly over the long-run as the demand curve rotates out and quantities
increase even more at the lower price relative to the higher one. Coupled with evidence that
credit constraints bind and total borrowing increases for consumers (the no crowd-out result
described above), and that profits increase for CB, it seems reasonable to infer that welfare
increases (although not as much is it would if competitors cut prices as well). Quantitatively
estimating welfare changes would require accounting for many potentially important features—
e.g.., the shape of the demand curve, the source and magnitude of credit constraints, credit being
an intermediate good, returns to that intermediate good being uncertain and/or mediated by
behavioral factors, etc.— that are not measured or identified in this study. 7 In the same vein,
inputting our elasticities into micro-founded macro models would require substantial additional
data and theory on how these parameters relate to tastes, technology, policy, market structure,
etc.
Methodologically, we show that one can combine a field experiment with various sources of
administrative data to identify dynamics in consumer and competitor responses to a market
leader’s pricing change, and that these dynamics and the frictions underlying them are
empirically important. This adds to a small number of market-level field experiments where
competitor responses (as well as consumer responses) are outcomes of interest (Busso and
Galiani 2014).
6
Proximity here, as is often the case, takes the form of geography: CB and its competitors often have branches on
the same street. In other markets (e.g., U.S. credit cards), proximity takes the form of competitors direct-marketing
to the same customer base.
7
See Zinman (2014) for a review of many different models of credit markets and their welfare implications.
6
Our study also contributes to the literature on elasticities of demand for consumer credit and
microcredit in various ways. We use random price variation at the level of large geographic
units, which internalizes any within-region feedback effects of price changes on learning,
competitive responses in credit and product markets, and input markets. We also track
competitor responses, and explore the mechanisms underlying dynamic adjustment. Relatedly,
leaving the randomized rates in place for 29 months allows more time for consumer learning and
competitor response than the direct mail experiments from South Africa and Mexico used in
Karlan and Zinman (2008) and Ponce et al (2014); indeed, our literature review—which includes
non-RCTs and studies set in wealthier countries—does not find any studies with a time horizon
longer than 12 months. (For details of each study, see Table 1a and 1b; Gross and Souleles 2002;
Alessie, Hochguertel, and Weber 2005; Attanasio, Goldberg, and Kyriazidou 2008; Karlan and
Zinman 2008; Bengtsson and Pettersson 2012; Dehejia, Montgomery, and Morduch 2012; Alan
and Loranth 2013; Bhutta and Keys forthcoming; Ponce, Seira, and Zamarripa 2014; DeFusco
and Paciorek 2015).
1. The Market Setting
A. Compartamos Banco and its Target Market
CB has long been the largest microlender in Mexico, 8 and it had well over one million
borrowers during our study period. CB was founded in 1990 as a nonprofit organization and was
converted to a commercial bank in 2006. It went public in 2007 and has a market capitalization
of about $3 billion as of this writing. During our study period a super-majority of CB clients
borrowed through Crédito Mujer, the group microloan product studied in this paper.
Crédito Mujer nominally targets women who have a business or self-employment activity or
intend to start one. Empirically, all borrowers are women, but a companion paper uses survey
data to estimate that only about 52% are “microentrepreneurs” (Angelucci, Karlan, and Zinman
2015). Borrowers tend to lack the income and/or collateral required to qualify for loans from
banks and other “upmarket” lenders.
8
According to Mix Market, http://www.mixmarket.org/mfi/country/Mexico.
7
B. CB’s Cost Structure, Loan Terms and Competition
Crédito Mujer loan amounts range from M$900-M$24,000 pesos (during our sample period
13 pesos, denoted M$, = $1US), with larger amounts subsequently available to members of
groups that have successfully repaid prior loans. 9 Borrowers typically start with a loan of about
M$4,000 and take successively larger loans over time. Loan repayments are due over 16 equal
weekly installments, and are guaranteed by the group (i.e., joint liability). Aside from these
personal guarantees, there is no collateral. Interest rates are marketed as monthly, “add-on” rates
of 3.0 to 4.5%, 10 excluding value-added tax and “required” simultaneous savings. 11 We convert
these add-on rates to portfolio-weighted APRs and estimate that CB charged APRs of roughly
100% at baseline. Market research, detailed below, indicates that CB’s pricing falls at the low
end of the market for comparable loans. 12
Prior to this study, CB had set prices based on its cost structure and anecdotal information
about competitive factors (we describe the former here, and the latter in the next paragraph). Its
most important costs are labor and of course the cost of funds. Labor costs loom large because,
as detailed below and in sub-section D, marketing and group administration (including
monitoring) is handled by field staff. Staff receive a base salary with a performance bonus based
on key factors such as revenue and repayment rates (the formula and factors changed, and also is
considered proprietary). New groups and borrowers are particularly costly because they require
more bank effort (in set-up and monitoring) for lower gross payouts (small loan amounts and
9
Also, beginning in weeks 3 to 9 of the second loan cycle, clients in good standing can take out an additional,
individual liability loan, in an amount up to 30% of their joint liability loan.
10
An add-on rate is calculated over the original loan amount and does not adjust for declining balances as an Annual
Percentage Rate (APR) does. CB rates vary only based on city and a borrowing group’s loan cycle, with newer
groups paying the highest rates and well-performing and more-experienced groups paying lower rates.
11
Borrowers are supposed to make an upfront deposit totaling 10% of their loan amount into a personal savings
account, and contribute at least 10 pesos weekly for the remainder of the loan cycle. This “forced savings”
component at either zero or low interest is not claimed as collateral, but rather meant to instill a “culture of regular
deposits” and generate a signal of the client’s ability to generate and manage cash flow that can be used to evaluate
creditworthiness for future loans. The forced savings is not necessarily held by CB; i.e., the effective APR paid by
the borrower may be higher than the effective APR earned by CB. Anecdotally, neither CB staff nor group members
vigorously enforce the savings requirement. Mexican law does not require advertisements or disclosures to include
savings requirements or value-added tax in APR calculations.
12
See David Roodman’s work http://blogs.cgdev.org/open_book/2011/02/compartamos-in-context.php for more
details on microloan pricing in Mexico. His APRs are higher than ours because he includes VAT and forced savings,
and compounds more frequently.
8
higher delinquency/default rates). In fact, the typical borrower is barely profitable at first but
then becomes profitable with subsequent loans (sub-section D). The cost of funds reported to us
by CB averaged about 9.5% during our sample period. Given CB’s access to financial markets it
is effectively unconstrained in supplying credit on the margin. 13
Market research paints a picture of a microloan market that was competitive, but imperfectly
so, during our study period. CB embarked on this study because it had already decided to cut
rates-- based on its perception of competitive pressures-- but was undecided about how much to
cut. CB field staff easily identified nearby close competitors for the mystery shopping exercise
detailed in Section 3-I. Two key results emerge from this exercise. The first speaks to the
“competitive” part of “imperfectly competitive”-- nearly every CB branch 14 identified at least
three close competitors (with “closeness” defined as being both geographically proximate and
offering a similar product), with many branches identifying more than three. The second key
result speaks to the “imperfectly” part-- there is substantial price dispersion even within a set of
close competitors. 15 Taken together with evidence of binding liquidity constraints for consumers,
the price dispersion reinforces that how consumers respond to CB’s rate cut is an open question
(e.g., whether cheaper CB debt complements, or substitutes for, existing debt), as is the question
of how competitors respond.
C. Targeting, Marketing, Group Formation, and Screening 16
Crédito Mujer groups range in size from 10 to 50 members. When CB enters a new market,
loan officers typically target self-reported female entrepreneurs and promote the Crédito Mujer
product through diverse channels, including door-to-door promotion, distribution of fliers in
public places, radio, and promotional events. As loan officers gain more clients in new areas,
they promote less frequently and rely more on clients to recruit other members.
13
The fact that we find CB’s lower rate drawing in many new borrowers (Section 3-B) is consistent with CB being
unconstrained, because if CB were constrained it presumably would ration new borrowers first, since existing
borrowers are far more profitable, as noted above and detailed in Section 3-D).
14
A branch is the main operating unit for field staff, as described in Section 2.
15
This sort of price dispersion is not so surprising, in light of similar evidence from various product markets in the
U.S. and elsewhere (Zinman 2015).
16
This sub-section also appears in Angelucci, Karlan, and Zinman (2015).
9
When a group of about five women—half of the minimum required group size—expresses
interest, a loan officer visits the partial group at one of their homes or businesses to explain loan
terms and process. These initial women are responsible for finding the rest of the group
members. The loan officer returns for a second visit to explain loan terms in greater detail and
complete loan applications for each individual. All potential members must be at least 18 years
old and must present a proof of address and valid identification to qualify for a loan. Business
activities (or plans to start one) are not verified; rather, CB relies on group members to screen out
any uncreditworthy applicants. In equilibrium, potential members who apply are rarely screened
out by their fellow members, since individuals who would not get approved are not recruited and
do not to tend to seek out membership.
CB reserves the right to reject any applicant put forth by the group but relies heavily on the
group’s endorsement. CB does pull a credit report for each individual and automatically rejects
anyone with a history of fraud. Beyond that, loan officers do not use the credit bureau
information to reject clients, as the group has responsibility for deciding who is allowed to join.
Applicants who pass CB’s screens are invited to a loan authorization meeting. Each applicant
must be guaranteed by every other member of the group to get a loan. Loan amounts must also
be agreed upon unanimously. Loan officers moderate the group’s discussion and sometimes
provide information on credit histories and assessments of individuals’ creditworthiness.
Proceeds from authorized loans are disbursed as checks to each client.
D. Group Administration, Loan Repayment, and Collection Actions
Each group elects a treasurer who collects payments from each group member at each weekly
meeting. The loan officer is present to facilitate and monitor proceedings but does not touch the
money. If a group member does not make her weekly payment, the president (and loan officer)
will typically solicit and encourage “solidarity” pooling to cover the payment and keep the group
in good standing. All payments are placed in a plastic bag that CB provides, and the Treasurer
then deposits the group’s payment at either a nearby bank branch or convenience store. 17
17
Compartamos has partnerships with six banks (and their convenience stores) and two separate convenience stores.
The banks include Banamex (Banamexi Aquí), Bancomer (Pitico), Banorte (Telecomm and Seven Eleven), HSBC,
Scotiabank, and Santander. The two separate convenience stores are Oxxo and Chedraui.
10
Beyond the group liability, borrowers have several other incentives to repay. Members of
groups with arrears are not eligible for another loan until the arrears are cured. Members of
groups that remain in good standing qualify for larger subsequent loan amounts and lower
interest rates. CB also reports individual repayment history for each borrower to the Mexican
Official Credit Bureau. Loans that are more than 90 days in arrears after the end of the loan term
are sent to collection agencies.
Late payments are common, but default is rare: in our data, we find a 90-day group
delinquency rate of 9.8%, but the ultimate default rate is only about 1%.
2. Study Design
The research team (IPA) worked with CB to identify 80 distinct geographic areas (“regions”
for the purpose of the study) throughout Mexico, for the purpose of randomly assigning interest
rates (Figure 1). Regions are the size of a small metropolitan area, a moderate-sized city, or a
large district in the largest cities. The CB operating unit within a region is a “branch” (actually
more like a regional or sub-regional office); the mean number of branches per region is 1.65. IPA
then assigned each of the 80 regions (and all branches within each region) to either “low-rate” or
“high-rate”. Regions were defined specifically to maximize compliance with the experimental
design by providing ample geographic distance between branches offering different rates. We
also verify that our main results are similar after dropping eight branches, from five regions, in
the densely populated Mexico City metropolitan area (Appendix Table 9). 18
The interest rate changes applied only to Crédito Mujer. CB did not implement any
additional operational or strategic changes that mirrored our treatment assignment, so any
responses we observe are demand-driven, i.e. pure price effects, with the assumption that
employees did not reallocate their time towards marketing, and thus away from other activities
such as enforcement. 19
18
We define the Mexico City metropolitan area using the Metropolitan Area of the Valley of Mexico.
One risk to any pricing design which relies on employees to promote or sell products or services (versus a direct
marketing campaign) is that employees may change their behavior and exert more effort in one treatment relative to
the other. In our study, we conjecture that such an experimenter demand effect is not a relevant risk, as individual
employees did not interact with borrowers from both treatments. Therefore, they would have had to exert more or
less effort on sales relative to other activities, and there is little reason to think that the optimal allocation of their
19
11
The motivation for randomizing at the region level, as opposed to a more granular level like
branches or groups, is twofold. The first is to allow for any consumer learning and competitive
response to take place at the level of a geographic market; i.e., we allow for within-market
spillovers. Second, region-level assignment facilitates compliance with the randomization in a
group lending setting, by ensuring that contiguous groups or contiguous branches (which would
normally draw some borrowers from overlapping geographic areas) are assigned the same rate.
Table 2 summarizes various baseline (April 2007) averages for low- and high-rate regions
and checks for balance on observables. Panel A presents borrower characteristics: education,
age, number of children, number of dependents, and marital status. Panel B presents loan
volume, both to all borrowers and to groups generally targeted by policymakers, MFIs, and
donors (new borrowers and borrowers with low education; we also consider relatively poor
borrowers below but lack pre-treatment data on income or wealth). Panel C presents loan
characteristics: APR, loan amount, group size, and number of groups. Delinquency data is absent
here because we lack the requisite pre-treatment data. Panel D covers market (i.e., regional)
characteristics, focusing on CB's market share as measured using credit bureau data. Overall,
Table 2 suggests that the randomization successfully generated similar treatment and control
groups.
The experiment engineered prices that were about 10 percentage points lower (in APR units)
in low-rate regions. Starting May 15, 2007, CB implemented this variation by offering
differential cuts from pre-treatment prices. Low-rate regions got 20 percentage point cuts from
pre-treatment rates (which averaged about 100% APR, as shown in Table 2 Panel C), whereas
high-rate regions got 10 percentage point cuts. 20 CB presented these prices to (prospective)
borrowers as “permanent” in the sense of the “new normal”: these were not promotional rates.
time across activities shifted. If the design involved randomizing within the employee’s coverage, then this may
have been more of an issue (e.g., in Giné and Karlan (2014), this was the case, in testing group versus individual
liability, but with each individual credit officer having both treatment and control villages).
20
Compartamos advertises and administers interest rates in add-on, monthly terms, and prices each group into one
of three tiers based on tenure and past performance. The randomization assigned low-rate regions to tiered pricing of
3.0%/3.5%/4.0% monthly, and high-rate regions to 3.5%/4.0%/4.5% monthly. We convert these monthly rates to
balance-weighted APRs. Rates applied only to loans originated on or after May 15, 2007, but the amount of lock-in
on previous loans is minimal given the loan term of 16 weeks and treatment duration of 130 weeks (our analysis
horizon varies from 52 to 130 weeks).
12
CB kept these rates in place for 29 months and then implemented the low rate everywhere, along
with other pricing changes.
We measure price sensitivities by comparing various outcome measures in low- vs. high-rate
regions for up to 29-months post-treatment (i.e, post- differential rate cuts). In some cases we can
use pre-treatment data as well. The next section details our specifications and results.
3. Empirical Specifications and Results
A. Main Specification
Our primary analysis employs an OLS specification for five outcomes: number of loans, loan
amount, revenues, costs, and net present value of client relationships. The specification is:
(1) Yrt = α + β1(LowRater*Postt) + ρRr + τT + ε
Y is an outcome, measured for region r in month-year t. α is the constant. The variable of interest
here is the interaction term—which equals one if and only if the observation is from a low-rate
region in the post-treatment period—and β1 identifies price sensitivity. R is the LowRate main
effect. T is a vector of month-year dummies (e.g., separate dummies for June 2007 and June
2009), and these absorb the Post main effect. ε is the error term. Throughout the paper we cluster
standard errors at the unit of randomization: the region.
Our primary analysis starts by presenting the results from (1), for each of the five outcomes,
in Table 3. Appendix Table 1 through Appendix Table 5 report specification robustness checks
for each of the five outcomes. Column 1 in each of these Appendix Tables shows results with
fixed effects for both month-year and region. Column 2 shows fixed effects for month-year with
controls for the baseline values of the outcome variable (in lieu of the LowRate indicator variable
or branch fixed effects). Column 3 shows results for the log of the dependent variable. Columns
4-7 show results for winsorizing or truncating the dependent variable.
The price elasticity of demand, reported at the bottom of Table 3 and other tables, is defined
as the percentage change in quantity demanded divided by the percentage change in price. We
calculate the former, for each specification, by dividing the coefficient of interest by the mean of
13
Yrt across all high-rate (control-group) regions over the entire post-treatment period. We calculate
the latter, again over the entire post-treatment period, by dividing the average, balance-weighted
APR difference between high- and low-rate regions, and then dividing that difference by the
average, balance-weighted APR in high-rate regions.
As noted at the outset, we are able to identify price elasticities of demand under the
assumption of no impact on supply-side decisions as a result of treatment. For example, it must
be the case that the 20pp interest rate cut does not induce differential screening or monitoring as
opposed to the 10pp rate cut. Our finding that delinquencies are not affected by the interest rate
(see below) supports this assumption. It must also be the case that treatment and control regions
did not receive different marketing (if marketing affects demand); this assumption seems
reasonable given the study protocols (no operational changes other than price), and the fact that
both treatment and control regions experienced rate cuts relative to CB’s prior rates. 21
B. Elasticities of Number of Loans Disbursed
We find a statistically significant increase of 200 loans disbursed by CB (<-> taken by
borrowers) per month in the low-rate regions, compared to the high rate regions (Table 3 Column
1). The implied elasticity is -1.4. Appendix Table 1 shows that other specifications yield similar
estimates.
The interest rate sensitivity increases significantly from year-to-year (Table 3 Column 2),
from about -0.8 in year one to -1.5 in year two to -2.2 in year three. The p-values on the three
pairwise differences are 0.08, 0.01, and 0.02. The finding that elasticities increase over time is
consistent with borrower learning and/or adjustment costs.
Figures 2a and 2b plot cumulative distribution functions (CDFs) of loan counts to explore
effects throughout the distribution of regions as ordered by loan volume, separately for treatment
and control. Figure 2a plots the average monthly loan count per region over the post-treatment
period (i.e., one observation per region, so each CDF plots 40 data points). This suggests that
there is a treatment effect for the larger regions (starting around the 40th percentile) but not for
21
For example, suppose that any substantial price change triggers informative advertising. In that case both our
treatment and control regions would get the additional advertising, because both sets of regions got price cuts
relative to Compartamos’ prior rates.
14
smaller ones. Figure 2b takes the same approach, but counts only the final post-treatment month
(and hence focuses more on the long-run). Here we see treatment effects starting around the 20th
percentile of loan count volume.
C. Elasticities of Amount Borrowed from Compartamos Banco
Table 3 Columns 3 and 4 present estimates of price sensitivities of the amount lent by
(borrowed from) CB, again measured as region-month flows. This measure of demand combines
the extensive and intensive margins. Given the treatment effects on the extensive margin (Table
3, Columns 1 and 2) it is not straightforward to back out effects on the intensive margin alone,
since the lower price may be drawing in borrowers with different elasticities and we do not have
separate instruments for borrowing at all and for amount borrowed.
We find a statistically significant increase of about two million pesos per month in low-rate
compared to high-rate regions, on a base of about 6.5m. The implied point elasticity is -1.89.
Appendix Table 2 shows that other specifications yield similar estimates.
As with loan count, the price elasticity of dollars borrowed increases from year-to-year
(Table 3 Column 4): from -1.18 to -1.94 to -2.87. The p-values on the three pairwise differences
are 0.05, 0.003 and 0.01.
Figures 3a and 3b plot cumulative distribution functions (CDFs) of amounts lent to explore
effects throughout the distribution of regions, separately for treatment and control. Figure 3a
plots the average lending per month per region over the post-treatment period (i.e., one
observation per region, so each CDF plots 40 data points). This suggests that there is a treatment
effect for the larger regions (starting around the 40th percentile) but not below. Figure 3b takes
the same approach, but counts only the final post-treatment month (and hence focuses more on
the long-run). Here we see treatment effects starting around the 10th percentile of peso volume.
D. Elasticities of Revenue, Costs and Net Present Value of Client Relationships
A key question for equilibrium is whether the lower interest rate was profitable, or at least
sustainable, for CB. Table 3 Columns 5-10 analyze revenues, costs and the net present value of
the client relationship.
15
First we examine revenue (Columns 5 and 6). The coefficients on the treatment effect are
positive but statistically distinguishable from zero. This is true for the main specifications as well
as all robustness specifications in Appendix Table 3. Note that there are offsetting effects here:
the lower interest rate induces substantially more lending (Table 3 Columns 1-4) but of course
generates less gross revenue ceteris paribus. Here the net effect from an accounting perspective
(which does not deal with confidence intervals) is positive.
Table 3 Columns 7 and 8 report treatment effects on costs, which include chargeoffs. The
point estimates are all positive— the lower interest rate increases costs by slightly more than the
increase in revenues (except in year three), although costs are more precisely estimated than
revenues. Appendix Table 4 shows similar results for the other specifications. Appendix Table 6
suggests that about 40% of the cost increase is due to personnel. Appendix Table 7 shows that
personnel costs fall in the average number of members per group and increase with the number
of groups. Appendix Table 8 shows some evidence that the lower rate increases the number of
groups and group size (Column 1-4), but that loan officers do not handle significantly more
groups or clients. In all, we find little evidence of economies of scale given operations and firm
structure at the time of the experiment. Innovations with respect to lending technology, such as
mobile devices with customized software for credit officers (that are now being implemented by
many lenders, including CB), may help lower marginal costs and capture economies of scale in
the future (and were put in place after the study ended).
The final two columns of Table 3 report treatment effects on the NPV of clients. NPV moves
beyond realized profits to expected profits and is the key profitability metric from CB’s
perspective because it more fully accounts for the lower interest rate’s changes on portfolio
composition. Specifically, it is important to account for the dynamic that new clients are less
profitable than more seasoned clients. Thus as the new clients drawn in by the lower interest rate
“age”-- into lower default rates, large loan sizes, and lower marginal costs-- they may become
profitable over horizons even longer than the 29-month analysis period for realized profits.
We estimate NPV by first estimating the profitability of each loan depending on the cycle
(first loan, second loan, etc.) and on the treatment status. Next we “age” each client through a
life-cycle of loan cycles based on the data for likelihood of repeat borrowing (conditional on
16
treatment status). The probability of repeat borrowing starts at 59.7% and 60.8% for the first to
second loan for control and treatment groups, respectively, and the increases almost perfectly
monotonically for the first nine loans until it reaches 67.8% and 67.6%, respectively. We then
aggregate these estimates to the region-month level:
(2) NPV of portfoliort = Σclients Σ[t,T] (Still client probabilitycrt * (loan revenuecr – costcr – loss to
defaultcr)) / discount ratetT-t
where r and t index region and month as before, T represents the number of future periods
(period = month, from t=1 to T=224, with 224 being the most number of months anyone
continuously borrows in the data), and c represents tenure of the client. Our NPV measure
probably produces conservative estimates of the treatment effect on profits because it does not
account for cross-sells or referrals (both of which should increase monotonically in the number
of clients and hence be higher in the treatment group).
The coefficients on the NPV treatment effect are positive but statistically distinguishable
from zero (Table 3 Columns 9 and 10). This is true for the main specifications as well as nearly
all robustness specifications in Appendix Table 5. Figure 4 explores the dynamics by plotting the
NPV estimates month-by-month, separately for treatment versus control. The control group starts
off higher (because of the higher interest rate), but treatment group surpasses control at about
month 9, with the gap widening as time proceeds. Indeed, after seeing the preliminary results
over the 29-month horizon, which ends in October 2009, CB implemented the lower interest rate
in the control group regions in 2010. CB made this change roughly in concert with implementing
more granular risk-based pricing across their entire operation, which unfortunately renders moot
using post-experiment data to examine any convergence, since a lot changed in their pricing after
our experiment ended.
So why had not CB cut interest rates earlier? After all, given weakly increasing profits,
elastic demand, and a “double bottom line” of helping clients to “generate social and economic
value,” as well as the possibility of increased cross-sells, the lower interest rate has turned out to
be beneficial for CB ex-post. We speculate that costly experimentation may discourage some
lenders from deriving their demand curves. In addition to direct costs, there may be a risk to
cutting rates if demand curves develop kinks at current market rates that make subsequent
17
increases quite costly. 22 We discuss some implications of costly experimentation for market
equilibria and future research in the conclusion.
E. Do Lower Rates Improve “Outreach”?
Next, in Table 4, we explore whether the lower interest rate increased take-up by groups that
are often the focus of “outreach” intended to expand access to microcredit: new borrowers
(Columns 1 and 2), those with less education (Columns 3 and 4), and the (relatively) poor
(Columns 5 and 6). 23 We re-estimate the specifications used in Table 3. The estimates are
imprecise, but the overall pattern is consistent with the lower rate expanding access to these
groups: each of the 12 points estimates is positive, and for each of the groups we see the pattern
of increasing price sensitivity over time, suggesting that lower rates do improve outreach over
the long-run.
F. Mechanisms: What explains the increase in price-sensitivity over time?
What explains the strong growth in demand-sensitivity over time? (Or, isomorphically, why
does it take substantial time to reach steady-state price sensitivity?) Leading possibilities include
capital adjustment costs, contractual frictions, and information frictions. We start by casting
doubt on the importance of the first two mechanisms in our setting. Regarding capital
adjustment, although Angelucci, Karlan, and Zinman (2015) do find evidence that marginal CB
borrowers in Nogales (a region not included in the interest rates experiment because CB was not
operating there yet) expand their microbusinesses over a two-year period, their evidence suggests
that expansion takes the form of additional inventory rather than fixed investments. Regarding
contractual frictions, recall that CB loan cycles are short enough (16 weeks) to allow for much
more rapid adjustment than we actually see over the 120-week span of the experiment.
Information frictions seem more likely to be important. The lack of transparency in pricing
and advertising practices in the microcredit market may well slow the spread of new information
22
Karlan and Zinman (2008) find extremely elastic demand to rate increases but not decreases, an asymmetry the
lender there had anticipated.
23
We define a loan to a new borrower as one disbursed to someone who had not borrowed from Compartamos in
any previous month. We measure educational attainment for each borrower using Compartamos application data.
We measure poverty likelihood for each borrower using application data that Compartamos starting collecting in
June 2007. The poverty score formula is from Schreiner (2006).
18
about prices. Relatedly, it seems reasonable to assume substantial search costs given evidence
from other credit markets (Zinman 2015) and the substantial price dispersion we document in
this market (sub-section H below). Joint liability could further dampen the price response
initially if groups underestimate any advantageous selection produced by the lower rate and/or
simply are somewhat risk-averse with respect to guaranteeing new loans; i.e., there may be
learning about the price-risk gradient by groups.
G. Does Delinquency Fall with Price?
We now turn to an analysis of whether the lower interest rate led to lower average
delinquency (Table 5, Column 1 and 4). We have already considered this question implicitly in
our costing and NPV exercises (Table 3), but looking directly at delinquency has value for
identifying frictions due to asymmetric information and/or liquidity constraints (Karlan and
Zinman 2009). We also examine whether interest rates have delinquency effects that vary with
measures of the prevalence of new borrowers in a region-month (Table 5, Columns 2, 3, 5, and
6), under the hypothesis that any asymmetric information problem may be relatively severe for
borrowers who have not contracted with CB before. Our outcome of interest in Panel A is the
proportion of groups, in a region-month, that are behind on their repayments.
Our main empirical model and sample is the same as for demand estimation, with two
exceptions. First, CB did not track delinquency systematically prior to the experiment, so we
lack pre-treatment data. Second, we count only groups that could possibly be late in the
denominator; i.e., we exclude groups that are too new to be delinquent. For the serious
delinquency measure this excludes 163 region-months that are comprised entirely of groups with
loans disbursed in the previous four months.
Table 5 Columns 1-3 examine moderate delinquency—any lateness—and show little
evidence of price response. The point estimates are all negative, however (consistent with lower
rates mitigating asymmetric information), and the confidence intervals do not rule out
economically meaningful effect sizes relative to the base rates at the region-month level, which
are 0.14 for all groups, 0.15 for groups with >75% new members, and 0.19 for new groups.
Columns 4-6 show a similar pattern of results for severe delinquency (later than 90 days). Again
19
the point estimates are all negative, with confidence intervals that contain large effects relative to
the base rates of 0.10 to 0.16. We do not find evidence of time-varying effects (results not
tabulated to conserve space).
H. Rate Cuts in Equilibrium: Evidence on Net Elasticities from Credit Bureaus
Credit bureau data enables us to examine another important aspect of general equilibrium
responses to interest rate changes: net elasticities that take into account any crowd-out or crowdin of credit from other lenders. We have data from two bureaus: the Mexican Official Credit
Bureau and the Circulo Bureau. Both bureaus allow us to focus only on loans that are
comparable to CB’s loans. 24 The Official Bureau has more comprehensive coverage (compare
the means in Column 1 and 2 in Table 6), but has the disadvantage of including CB loans in the
region-level data we were able to obtain. 25 The Circulo Bureau has the advantages of including
Azteca, a main competitor of CB, and of excluding CB loans. 26
The bureau data (Table 6) has three main differences from the CB data used to estimate our
main results on elasticities of demand for CB debt. First, we were only able to obtain two
snapshots from the bureaus: April 2007 (one month before the start of the pricing experiment),
and December 2008. In all we have data from 79 regions at both points in time, for a total of 158
region-month observations. As such, Table 6 reports estimates from a specification with monthyear fixed effects and an indicator for treatment branches, with the parameter of interest being
the coefficient on Post*LowRate. Second, the Official Bureau data lacks loan amounts and the
number of borrowers, so our main measure of aggregate demand is the number of loans. Third,
the credit bureau captures stocks, not flows. Given these differences, Table 6 Column 3 also
reports a comparable estimate for the stock of CB loans measured from CB’s data, using the
24
We include the following lending institution types: Bank, Bank loan PFAE, Credit line from bank, Non-bank loan
for PFAE, Financial non-bank loan for PFAE, Personal finance company, Medium market commerce, Medium
market NF service, and Credit Union.
25
The credit bureaus would only provide us with data at the level of the municipality*lending institution type, so we
cannot identify individual lenders or borrowers. To aggregate to regions we exclude the 3% of municipalities that
straddle our regional boundaries, and the 16.5% of municipalities with <20 Compartamos clients. The latter
exclusion improves power by focusing on the parts of regions where Compartamos actually has a presence (regions
are large enough that Compartamos does not necessarily operate throughout the entire region).
26
Another difference between the two credit bureaus is that the Circulo enables us to filter on loan type rather than
lending institution type: we count only “Personal Loans.”
20
same months and specification. There is a marginally significant treatment effect of about 800
loans in the CB data, with an implied elasticity of -1.6.
The main inference we can make from the credit bureau data is a lack of strong evidence for
crowd-out. Focusing on the Circulo data (Column 1), which excludes CB loans, we see a positive
point estimate on LowRate*Post, with a p-value of 0.16, suggesting that crowd-in is more likely
than crowd-out. Crowd-in can result if there are both liquidity constraints and non-convexities in
investment (Banerjee and Newman 1993), and Angelucci, Karlan, and Zinman (2015) also finds
some evidence that CB lending increases generate crowd-in. Column 2 (from the Official
Bureau) also shows a large, positive point estimate that is much larger than the direct effect on
CB borrowing (compare to Column 3), but this point estimate is imprecisely estimated as well
(p-value = 0.2).
Several other data points buttress the interpretation that the lack of crowd-out on the
consumer side (isomorphically, the lack of business-stealing on the lender side) is due to binding
credit constraints. We do not find differential results in study regions where staff identified more
versus fewer nearby competitors (Table 6 Columns 4-6), as one would expect if a pure market
power/segmentation story were driving the results. Nor did CB staff have any difficulty
identifying multiple nearby competitors for the mystery shopping exercise (sub-section I below),
again suggesting that the market is competitive. Borrower-level survey data, from the Nogales
region, shows substantial overlap in borrowing from CB and close competitors, and also that
when CB expanded credit supply in Nogales, the additional credit taken by borrowers did not
crowd-out other formal sources of credit (Angelucci, Karlan, and Zinman 2015). 27 Note that we
cannot simply measure overlap within the interest rate sample, for two reasons: 1) the Nogales
region is not in our interest rate sample because CB entered there subsequent to the interest rate
experiment; 2) the credit bureau data for our interest rate sample is aggregated to the
municipality level. It may also be the case that CB’s lower rate, which ends up being toward the
low end of the market (Table 7), draws new borrowers into the market who had reservation
prices at or slightly below pre-experiment prices.
27
The AEJ sample is comprised of people in the Nogales region who are eligible for Compartamos credit and
reached for an endline survey approximately two years after Compartamos entered the region.
21
I. Rate Cuts in Equilibrium: Evidence on Competitors’ (Non-)Responses
Our last bit of evidence on the general equilibrium effects comes from estimates of whether
and how competitors responded to CB’s lower interest rate (Table 8). The data for this exercise
come from a standard “mystery shopping” market research exercise. We worked with CB’s
senior management to lay out some simple protocols for the bank’s branch staff to gather data on
interest rates and loan terms: 1) collect data for at least the top three competitors in each region;
2) collect data by having someone pose as a prospective client seeking a 4,000 peso loan (a
typical loan size for a new microcredit client); 3) collect data at two or more points in time, first
in May 2007 (just prior to the start of the study), and then one at least one year after the start of
the experiment (although some branch managers reported back as early as April 2008); 4) collect
data on the same competitors in the pre- and post-periods, and also on new competitors in the
post-period as merited by any changes in the competitive environment.
These mystery shopping trials produced 616 observations on competitor loan offers, with 201
from the pre-treatment period and 415 from the post-treatment period. 28 Observations are
generated by 107 branches in 80 regions. 76 of these branches and 72 of these regions are present
both pre-treatment and post-treatment. 29 The 616 offer-level observations come from 106 unique
competitors, with the top five competitors comprising 48% of the observations. The number of
competitors reported per branch*reporting date (N=183) ranges from 1 to 10, with a mean and
median of 3, and a standard deviation of 1.4. Seventy-one percent of the observations are for the
requested loan amount of 4,000 pesos, and 48% of the observations are for the standard CB loan
term (maturity) of 16 weeks (94% are for 52 weeks or less). 30
Our main objective here is to estimate whether competitors change their prices, and so we
calculate the Annual Percentage Rate (APR) from the periodic rate supplied by CB branches in
28
Many branches reported on multiple dates and/or multiple loan offers (different loan amounts and maturities)
from the same competitor on the same date, most likely in response to email reminders sent by senior management
that were actually targeting a handful of straggling branches that had not yet reported any mystery shopping data
during the post-treatment period.
29
Results are unchanged if we limit the simple to the 76 branches, or the 72 regions, present in both the baseline and
endline mystery shopping data.
30
Controlling flexibly for loan amount and term does not change the results.
22
the raw data. 31 Starting from our sample of 616 observations, 3 lack information needed to
calculate the APR, and 5 observations reported non-positive interest rates (suggesting data
entry/capture errors), leaving us with 608 usable observations. The mean APR is 154% and the
median is 116% (both somewhat higher than CB’s APRs 80-100%), with the substantial
dispersion; e.g., a standard deviation of 150 percentage points. (See also Table 7, which shows
dispersion across competitors in Column 2, and to a lesser extent within them in Columns 3 and
4).
Table 8 reports estimates of competitor responses from a specification with month-year fixed
effects and control for treatment branches, with standard errors again clustered on region. 32 In
Table 8 Columns 1-5 the dependent variable is a function of the APR. Column one shows that
our point estimate of the competitor response in level terms (recall that post*treat estimate the
differential effect in treatment vs. control areas) is actually positive, and large: 36pp (se =
25.7pp). We can rule out a matching price cut (10pp) with 92% confidence. But Columns 2 and 3
caution that our estimate of the level effect may be unduly influenced by APR outliers: our point
estimates on log(APR) and median(APR) are essentially zero. These null effects are imprecisely
estimated, and do not reject the hypothesis of a matching price cut (p-values of 0.29 and 0.20).
Columns 4 and 5 use the minimum APR observed (level or log) at the branch*reporting date
level, and again find imprecisely estimated zeros for the treatment effect. These results suggest
that the welfare gains (to borrowers) of CB’s price cut are somewhat muted by the lack of similar
cuts from competitors.
Columns 6 and 7 use the competitor count (level or log) at the branch*reporting date level,
and suggest weakly negative effects. This would be consistent with CB’s price cut driving out
higher-cost competitors.
All told, the results in Table 8 suggest that competitors responded by exiting if at all, 33
although our estimates are imprecise and hence do not rule out economically meaningful
31
The raw data also includes other loan terms including the periodic installment period, allowing us to verify that
the periodic is in fact the correct rate to use in calculating the APR.
32
Clustering at the competitor or the competitor-region level produces somewhat smaller standard errors.
33
We also estimate treatment effects on loan maturity and loan amount and do not find evidence of effects on either,
although again these nulls are imprecisely estimated.
23
responses. We note two additional caveats re: our estimates of competitor non-response. First,
we have only a single snapshot, during Year 2 of the experiment, and consequently might be
missing longer-run effects. Of course that sort of slow adjustment would be noteworthy as well.
Second, it may be the case that competitors respond uniformly in all regions (although the “Post”
period coefficient suggests otherwise). Of course that pattern of response would be noteworthy
as well, as it suggests that there are informational, technological, and/or other frictions that
discourage competitors from responding in a more granular way to CB’s differential pricing
across regions.
What could explain non-response by competitors? A variety of factors suggest that lenders
(like consumers) face informational frictions that hinder rapid and precise responses to
competitor behavior. Contracting data is not readily available (hence the motivation for the
mystery shopping trials), so competitors may not have even recognized CB’s (differential) price
cuts. Competitors received especially weak signals because they did not actually lose clients
(Sections 3-H. and 3-J.). Price-setting is fraught by the difficulty of forecasting profitability, as
indicated by our work in Section 3-D and the fact that CB chose to engage in costly
experimentation to help it pin down its preferred price. A closely related problem is that
experimentation is more costly than it appears at first glance: our conversations with CB and
other lenders around the world indicate a belief that past prices serve as reference points for
customers, and hence that experimenting with lower rates is quite risky because it endogenously
makes “turning back” to higher rates unprofitable. This inference jibes with our finding from
South Africa that borrowers are much more elastic with respect to rate increases than decreases
(Karlan and Zinman 2008).
J. Why no business stealing?
Putting together the competitor and consumer results, we ask why borrowers do not
substitute away from competitors if CB offers lower prices and its competitors do not? One
possibility is product differentiation, although we view this as unlikely given the evidence of
substantial price dispersion even amongst suppliers offering very similar products (Table 7).
Another is search and/or switch costs, which have been linked to price dispersion in at least one
other major small-dollar credit market (Stango and Zinman forthcoming). Yet another, and
24
complementary, plausible explanation is liquidity constraints: if borrowers had excess demand at
pre-treatment market rates, they might borrow more on the margin, from CB, without reducing
their inframarginal borrowing.
IV. Conclusion
We study the long-run (up to 29-month) effects of a 10 percentage point interest rate
reduction by the largest microlender in Mexico, using a field experiment implemented at the
level of 80 distinct geographic regions. Demand for CB loans is quite elastic, and gets more so
over time. For example, the average elasticity of amount borrowed over the 29 months is about 1.9, with a Year 1 elasticity of about -1.1, and a Year 3 (months 25-29) elasticity of about -2.9.
There is no strong evidence of crowd-out in credit bureau data. The lower rate is profit-neutral
for CB, and the bank has maintained the lower rate post-experiment, confirming that the lower
rate has been sustainable for CB. We find no evidence that competitors responded by cutting
rates.
These findings suggest several avenues for future research. First, understanding why
elasticities grow over time—e.g., the relative importance of learning vs. adjustment costs—is
important for modeling and policy analysis. A closely related line of inquiry is whether and why
elasticities differ for group versus individual lending products. Another closely related line of
inquiry would unpack relationships between elasticities of demand for credit and for saving.
Estimates of the elasticities of demand for savings are as low as zero (Hall 1988; Karlan and
Zinman 2014), and most are strictly below the elasticities we find here.
There is also much to learn about microlending production functions as well. Our results
suggest an absence of commonly-assumed economies of scale, and it would be useful to know
what prevents CB (and, presumably, other microlenders) from capturing scale economies. Is it
something inherent to the group lending model? Labor market frictions? A longer transition path
to scale economies than our study window; i.e., perhaps it takes time to re-optimize fixed costs to
handle the increased loan volume?
Obtaining more precise estimates of aggregate credit demand and net elasticities is also
important. Our finding of no crowd-out even though competitors do not price-match raises
possibilities that switch costs have big effects on market outcomes, and/or that liquidity
25
constraints work in surprising ways: the standard story is that liquidity constraints make agents
price-inelastic, but perhaps this is true only of responses to price increases.
The possibility of asymmetric responses to price increases and decreases raises another
possibility that fits with our results: multiple equilibria borne of costly experimentation along the
profit-maximizing frontier. Holding profits constant (as turned out to the case in our experiment),
a lower interest rate would seem to be at least weakly better for a lender like CB, since the lower
rate delivers social benefits from the largely beneficial average impacts of increased access to
microcredit (Angelucci, Karlan, and Zinman 2015), in addition to public relations benefits and
increased opportunities for cross-sells. In this sense, a natural follow-up question is why firms
(including CB) do not charge lower rates to begin with. One possibility worth exploring is that
experimenting with lower rates is risky: a lower rate may reset customer expectations of a
fair/market rate, and create a kink in the demand curve. Karlan and Zinman (2008) find evidence
along these lines in South Africa. If this dynamic holds, then cutting rates may reduce or
eliminate the option to increase rates in the future (e.g., if it turns out that the lower rate was not
as profitable as the initial rate). In this case policymakers might consider interventions to spur
learning about pricing.
26
References
Alan, Sule, and Gyongyi Loranth. 2013. “Subprime Consumer Credit Demand: Evidence from a
Lender’s Pricing Experiment.” Review of Financial Studies 26 (9): 2353–74.
Alessie, R., S. Hochguertel, and G. Weber. 2005. “Consumer Credit: Evidence from Italian
Micro Data.” Journal of the European Economic Association 3 (1): 144–78.
Angelucci, Manuela, Dean Karlan, and Jonathan Zinman. 2015. “Microcredit Impacts: Evidence
from a Randomized Microcredit Program Placement Experiment by Compartamos
Banco.” American Economic Journal: Applied Economics 7 (1): 151–82.
doi:10.1257/app.20130537.
Armendariz de Aghion, Beatriz, and Jonathan Morduch. 2010. The Economics of Microfinance.
2nd ed. Cambridge, MA: MIT Press.
Attanasio, Orazio, Pinelope Goldberg, and Ekaterini Kyriazidou. 2008. “Credit Constraints In
The Market For Consumer Durables: Evidence From Micro Data On Car Loans.”
International Economic Review, International Economic Review, 49 (2): 401–36.
Banerjee, Abhijit. 2013. “Microcredit Under the Microscope: What Have We Learnt in the Last
Two Decades, What Do We Need to Know?” Annual Review of Economics 5: 487–519.
Banerjee, Abhijit, and Andrew Newman. 1993. “Occupational Choice and the Process of
Development.” Journal of Political Economy 101: 274–98.
Bengtsson, Niklas, and Jan Pettersson. 2012. “The Outreach and Sustainability of Microfnance:
Is There a Tradeoff?” Working Paper.
Bhutta, Neil, and Benjamin Keys. forthcoming. “Interest Rates and Equity Extraction During the
Housing Boom.” American Economic Review.
Busso, Matias, and Sebastian Galiani. 2014. “The Causal Effect of Competition on Prices and
Quality: Evidence from a Field Experiment.” Cambridge, MA.
http://www.nber.org/papers/w20054.pdf.
DeFusco, Anthony, and Andrew Paciorek. forthcoming. “The Interest Rate Elasticity of
Mortgage Demand: Evidence from Bunching at the Conforming Loan Limit.” American
Economic Journal: Economic Policy.
Dehejia, Rajeev, Heather Montgomery, and Jonathan Morduch. 2012. “Do Interest Rates Matter?
Credit Demand in the Dhaka Slums.” Journal of Development Economics 97 (2): 437–49.
Giné, Xavier, and Dean S. Karlan. 2014. “Group versus Individual Liability: Short and Long
Term Evidence from Philippine Microcredit Lending Groups.” Journal of Development
Economics 107 (March): 65–83. doi:10.1016/j.jdeveco.2013.11.003.
Gross, David B, and Nicholas S Souleles. 2002. “Do Liquidity Constraints and Interest Rates
Matter for Consumer Behavior? Evidence from Credit Card Data.” The Quarterly
Journal of Economics 117 (1): 149–85.
Hall, Robert E. 1988. “Intertemporal Substitution in Consumption.” Journal of Political
Economy 96: 339–57.
Karlan, Dean, and Jonathan Morduch. 2009. “Access to Finance.” In Handbook of Development
Economics, edited by Dani Rodrick and M. R. Rosenzweig. Vol. 5. Elsevier.
Karlan, Dean, and Jonathan Zinman. 2008. “Credit Elasticities in Less-Developed Economies:
Implications for Microfinance.” American Economic Review 98 (3): 1040–68.
doi:10.1257/aer.98.3.1040.
27
———. 2009. “Observing Unobservables: Identifying Information Asymmetries with a
Consumer Credit Field Experiment.” Econometrica 77 (6): 1993–2008.
———. 2014. “Price and Control Elasticities of Demand for Savings.” Working Paper.
Ponce, Alejandro, Enrique Seira, and Guillermo Zamarripa. 2016. “Borrowing on the Wrong
Credit Card: Evidence from Mexico.”
Rosenberg, Richard. 2002. “Microcredit Interest Rates.” Consultative Group to Assist the Poor
Occasional Paper (1).
Stango, Victor, and Jonathan Zinman. forthcoming. “Borrowing High vs. Borrowing Higher:
Price Dispersion and Shopping Behavior in the U.S. Credit Card Market.” Review of
Financial Studies.
Zinman, Jonathan. 2014. “Consumer Credit: Too Much or Too Little (or Just Right)?” Journal of
Legal Studies 43 (S2 Special Issue on Benefit-Cost Analysis of Financial Regulation):
S209–37.
———. 2015. “Household Debt: Facts, Puzzles, Theories, and Policies.” Annual Review of
Economics 7 (1): 251–76. doi:10.1146/annurev-economics-080614-115640.
28
Table1a.Summaryofpapersestimatingpricesensitivityofdemandforconsumercredit
paper
country
product
study
year(s)
sample
datasource(s)
potential/new
andexisting
clients
singlelender,
bureaus
multiplelenders,
existingaccounts
bureau
applicationsand
singlelender
loans
Consumer
autoloan
Expenditure
contracts
Survey
singlelender,
formerclients
bureaus
existingclients
singlelender
potential/new
andexisting
singlelender
clients
KarlanandZinman(thispaper)
Mexico
microloans
2007-09
GrossandSouleles(2002QJE)
USA
creditcards
1995
Alessieetal(2005JEEA)
Italy
consumerloans
1995-99
Attanasioetal(2008IER)
USA
carloans
1984-95
KarlanandZinman(2008AER)
BengtssonandPettersson(2012wp)
SouthAfrica
Tanzania
consumermicroloans
microloans
2003
2010-11
Dehejiaetal(2012JDE)
Bangladesh
microloans
1999-2001
AlanandLoranth(2013RFS)
UK
creditcards
2008
existingaccounts
singlelender
BhuttaandKeys(forthcomingAER)
USA
2ndmortgages
1999-2010
creditbureau
Mexico
USA
creditcards
firstmortgages
2005
1997-07
existingclients
seenote
creditbureau
singlelender,
bureau
vendors
Ponceetal(2016wp)
DeFuscoandPaciorek(forthAEJ:Pol)*
* Sample is "transactions of single-family homes with positive first loan amounts… within [MSAs] in California." Given that the data source
(DataQuick) covers deed transfers and property assessment records, we infer that these transactions are home purchases , not refinances.
Noestimatesofelasticitydynamics.Preferredestimatesofmeansemi-elasticityisabout-0.03.
30
Table1b.Summaryofpapersestimatingpricesensitivityofdemandforconsumercredit
identificationstrategy
directmail
price
shortest short-run longest
long-run balance
only?
variesacross
horizonfeatured elast horizonfeatured elast
shifting?
RCT
no
80regions 12months
-1.1
29months
-2.9
0
KarlanandZinman(thispaper)
lenderpricing
GrossandSouleles(2002QJE)
yes
24,000accts 2months
-0.8
9months
-1.3
about1/3
rules
nationallaw
loantypeand
Alessieetal(2005JEEA)*
no
seenote seenote seenote seenote allloans
change
size
nationallaw
Attanasioetal(2008IER)**
no
time
seenote seenote seenote seenote
n/a
change
53,810
KarlanandZinman(2008AER)
RCT
yes
-0.32
n/a
n/a
0
individuals 2-6weeks
195
not
BengtssonandPettersson(2012wp)
RCT
yes
n/a
n/a
n/a
individuals 6months reported
lenderprice
no
3branches 3months
-0.39
12months
-1.04
n/a
Dehejiaetal(2012JDE)
change
39,883
AlanandLoranth(2013RFS)***
RCT
no
3months seenote 3months
n/a
n/a
accounts
BhuttaandKeys(forthcomingAER)****
monetarypolicy
no
time
seenote seenote seenote seenote about1/10
961
Ponceetal(2014wp)
RCT
yes
-0.57
3months
-0.84
0
individuals 1month
DeFuscoandPaciorek(forthAEJ:Pol)*
RD
no
loanamount seenote seenote seenote seenote about1/3
*Noestimatesofelasticitydynamics.Estimatesofmeanelasticityinfullsamplerangefrom-1.1to-1.8.
**Noestimatesofelasticitydynamics.Estimateofmeanelasticityinfullsampleiszero.
***Noestimatesofelasticitiesordynamicsreported.Aggregate(fullsample)responsetopricechangeis"small".
****Noestimatesofelasticitydynamics.Estimateofmeansemi-elasticitywithrespecttocash-outrefis(juniorliens)isabout40%(10%).
*****Sampleis"transactionsofsingle-familyhomeswithpositivefirstloanamounts…within[MSAs]inCalifornia."Giventhatthedatasource(DataQuick)coversdeed
keysource
31
Table 2. Baseline (April 2007) Summary Statistics and Orthogonality
Means and Standard Errors
Low Rate High Rate
(1)
(2)
Difference in
Means
(1) - (2)
(3)
Panel A: Borrower characteristics
Proportion of clients with high school or more
Age
Number of children
Number of dependents
Number of married clients
Number of clients with high school or more
Number of clients without high school
0.503
(0.025)
40.222
(0.226)
3.072
(0.063)
1.986
(0.037)
510.718
(63.465)
395.179
(47.786)
462.051
(61.725)
0.457
(0.023)
40.191
(0.235)
3.183
(0.064)
1.970
(0.052)
518.179
(47.453)
353.077
(31.846)
467.923
(45.506)
0.046
(0.034)
0.031
(0.326)
-0.111
(0.089)
0.015
(0.064)
-7.462
(79.244)
42.103
(57.425)
-5.872
(76.686)
858.179
(99.863)
858.179
(99.863)
2,925.333
(359.782)
192.231
(19.161)
665.949
(87.198)
2,337.769
(324.065)
6,064.544
(840.700)
821.385
(69.634)
821.385
(69.634)
2,826.564
(249.812)
155.744
(10.935)
665.641
(62.002)
2,341.615
(228.227)
5,924.230
(638.784)
36.795
(121.744)
36.795
(121.744)
98.769
(438.006)
36.487
(22.062)
0.308
(106.994)
-3.846
(396.366)
140.314
(1055.851)
39
39
78
Panel B: Loan volume
Number of clients borrowing this month
Number of loans
Number of outstanding loans
Number of new clients borrowing this month
Number of retained clients borrowing this month
Number of outstanding loans of retained clients
Loan amount for all loans disbursed (M$ thousands)
N
Continued on next page
32
Table 2 (continued). Baseline Summary Statistics and Orthogonality
Means and Standard Errors
Low rate
(1)
High Rate
(2)
Difference in
Means
(1) - (2)
(3)
6.464
(0.250)
716.133
(75.249)
5,348.411
(793.515)
97.956
(1.709)
16.771
(0.282)
49.051
(5.345)
6.468
(0.277)
566.186
(47.316)
5,358.044
(604.132)
99.673
(1.967)
16.774
(0.281)
48.000
(3.966)
-0.004
(0.373)
149.947*
(88.889)
-9.633
(997.317)
-1.717
(2.606)
-0.003
(0.398)
1.051
(6.655)
1.535
(0.311)
1.106
(0.300)
0.430
(0.432)
Panel C: Loan characteristics
Loan amount (M$ thousands)
Loan amount for loans disbursed to new clients
Loan amount for loans disbursed to retained clients
APR (including VAT, not including forced savings)
Number of members in a group
Number of groups
Panel D: Market Share Characteristics
Log(Population per sq km in 2005)
Borrowers Per Capita: (Number of Compartamos Clients in April 2007)
/ (Population as of 2005 Mexico Census)
Market Proportion 1: (Number of Compartamos Clients in April 2007) /
(Number of Borrowers with Comparable Loans in Official Credit
Bureau Data in April 2007)
Market Proportion 2: (Number of Compartamos Clients in April 2007) /
(Number of Borrowers with Comparable Loans Recorded in Circulo
Credit Bureau Data in April 2007)
Market Proportion 3: (Number of Compartamos Clients in April 2007) /
(Number of Borrowers with Comparable Loans Recorded in Merged
Credit Bureau Data in April 2007)
Market Proportion 4: (Compartamos Loan Balance in $1,000s in April
2007) / (Comparable Loan Balance in Circulo Credit Bureau Data in M$
millions in April 2007)
N
0.029
0.029
-0.000
(0.002)
(0.002)
(0.003)
0.154
0.157
-0.003
(0.030)
(0.026)
(0.040)
0.985
1.193
-0.208
(0.144)
(0.173)
(0.225)
0.126
0.137
-0.011
(0.022)
(0.023)
(0.031)
339.511
423.418
-83.907
(54.178)
(69.709)
(88.287)
39
39
78
Unit of observation is the region for the time period of April 2007. Regions are the unit of randomization and comprised of 1-4 bank branches. Of the 80
regions in our study, two were not open before the start of the treatment period, yielding a total of 78 regions included in the table. Standard errors in
parentheses. Each client can only have one loan at a time. New clients are defined as those who get their first loan ever from Compartamos in that month.
Retained clients are defined as those who had already had a loan in previous months. Poverty likelihood variables and delinquency data is not available at
baseline. Population estimates are derived from the 2005 Mexican Census. We define "comparable loans" in the Mexican Credit Bureau data as loans from a
Bank, Bank loan PFAE, Credit line from bank, Non-bank loan for PFAE, Financial non-bank loan for PFAE, Personal finance company, Medium market
commerice, Medium market NF service, or Credit Union. In the Circulo data, "comparable" loans are those which institutions choose to report as "personal
loans." Compartamos loans are captured in the Mexican Credit Bureau, but not in the Circulo Credit Bureau. All municipalities included in the credit bureau
data have 20 or more Compartamos clients. * significant at 10%; ** significant at 5%; *** significant at 1%.
33
Table 3. Main outcomes
# loans
(1)
Treatment effect Low Interest Rate
Treatment effect Low Interest Rate (Year 1)
Treatment effect Low Interest Rate (Year 2)
Treatment effect Low Interest Rate (Year 3)
P-value: Treatment*1st year=Treatment*2nd year
P-value: Treatment*2nd year=Treatment*3rd year
P-value: Treatment*1st year=Treatment*3rd year
(2)
199.6**
(93.12)
Loan amount (M$1,000s)
(3)
(4)
1,960**
(777.3)
95.63*
(55.85)
218.1**
(109.3)
403.0**
(167.1)
0.080
0.009
0.022
Revenues (M$1,000s)
(5)
(6)
73.28
(92.99)
1,006**
(469.8)
2,135**
(916.7)
3,813***
(1,378)
0.054
0.003
0.012
Costs (M$1,000s)
(7)
(8)
106.5**
(44.76)
9.777
(58.92)
83.08
(109.4)
201.1
(160.8)
0.280
0.065
0.135
Net present value (M$1,000s)
(9)
(10)
972.6
(756.3)
57.57*
(29.35)
121.2**
(54.01)
187.9***
(68.65)
0.052
0.002
0.008
249.4
(336.8)
833.3
(714.4)
1,640
(1,190)
0.205
0.189
0.166
Treatment effect estimated by:
Low*Post Low*Post*Year Low*Post Low*Post*Year Low*Post Low*Post*Year Low*Post Low*Post*Year Low*Post Low*Post*Year
Control for low interest rate branch
y
y
y
y
y
y
y
y
y
y
Fixed effects for month-year
y
y
y
y
y
y
y
y
y
y
Fixed effects for region
n
n
n
n
n
n
n
n
n
n
Includes controls for values of dependent variable in pren
n
n
n
n
n
n
n
n
n
treatment months (March & April 2007)
Includes pre-treatment observations
y
y
y
y
y
y
y
y
y
y
Mean of dependent variable for pre-treatment period
885.8
885.8
6451
6451
1025
1025
361.3
361.3
8317
8317
Mean of dependent variable for high-rate regions
1246
1246
9116
9116
1279
1279
566.5
566.5
11286
11286
(treatment period)
Point elasticity
-1.41
-1.89
-0.50
-1.65
-0.76
Point elasticity year 1
-0.83
-1.18
-0.08
-1.11
-0.24
Point elasticity year 2
-1.45
-1.94
-0.54
-1.70
-0.61
Point elasticity year 3
-2.20
-2.87
-1.12
-2.40
-1.00
N
2,469
2,469
2,469
2,469
2,469
2,469
2,469
2,469
2,469
2,469
Each column reports results for an OLS regression of the outcome on the variables shown or summarized in the rows. Unit of observation is the region-month, and unit of randomization is the region, so standard
errors allow for clustering at the region level. Treatment was implemented on May 16, 2007. The month of May 2007 is excluded because the experiment started in the middle of the month. Post = 1 for the time
period of June 1, 2007 to October 31, 2009. Pre-treatment months are March and April 2007. * significant at 10%; ** significant at 5%; *** significant at 1%.
34
Table 4. Outreach
# loans to new clients
(1)
(2)
Treatment effect Low Interest Rate
45.02
(29.44)
Treatment effect Low Interest Rate (Year 1)
Treatment effect Low Interest Rate (Year 2)
Treatment effect Low Interest Rate (Year 3)
P-value: Treatment*1st year=Treatment*2nd year
P-value: Treatment*2nd year=Treatment*3rd year
P-value: Treatment*1st year=Treatment*3rd year
P-value: Treatment (new/low clients=old/high clients)
P-value: Treatment*1year (new/low ed/low inc clients =
old/high ed/high inc clients)
P-value: Treatment*2year (new/low ed/low inc clients =
old/high ed/high inc clients)
P-value: Treatment*3year (new/low ed/low inc clients =
old/high ed/high inc clients)
# loans to clients w/ low
education
(3)
(4)
# loans to clients w/ low
income
(5)
(6)
75.42
(47.84)
79.63
(122.7)
19.67
(19.29)
51.01
(35.10)
91.05*
(49.43)
43.10
(31.96)
77.32
(56.13)
147.9*
(78.49)
58.16
(107.9)
84.43
(130.0)
119.3
(146.3)
0.177
0.057
0.059
0.317
0.019
0.077
0.416
0.200
0.260
0.151
0.488
0.628
0.237
0.820
0.924
0.198
0.446
0.605
0.105
0.398
0.358
Treatment effect estimated by:
Low*Post Low*Post*Year Low*Post Low*Post*Year Low*Post Low*Post*Year
Control for low interest rate branch
y
y
y
y
y
y
Fixed effects for month-year
y
y
y
y
y
y
Fixed effects for region
n
n
n
n
n
n
Includes controls for values of dependent variable in pren
n
n
n
n
n
treatment months (March & April 2007)
Includes pre-treatment observations
y
y
y
y
y
y
Mean of dependent variable for pre-treatment period
184.9
184.9
494.6
494.6
n/a
n/a
Mean of dependent variable for high-rate regions (treatment
301.4
301.4
667.9
667.9
614.5
614.5
period)
Point elasticity
-1.31
-0.99
-1.14
Point elasticity year 1
-0.75
-0.67
-0.97
Point elasticity year 2
-1.35
-0.96
-1.14
Point elasticity year 3
-2.02
-1.57
-1.42
N
2,469
2,469
2,469
2,469
2,313
2,313
Each column reports results for an OLS regression of number of loans for new clients on the variables shown or summarized in the rows. Unit of
observation is the region-month, and unit of randomization is the region, so standard errors allow for clustering at the region level. Treatment was
implemented on May 16, 2007. The month of May 2007 is excluded because the experiment started in the middle of the month. Post = 1 for the time
period of June 1, 2007 to October 31, 2009. Pre-treatment months are March and April 2007. New clients are defined as those who get their first loan ever
from Compartamos in that month. * significant at 10%; ** significant at 5%; *** significant at 1%.
35
Table 5. Delinquency: Proportion of Lending Groups Delinquent
Delinquency proportion (Loan payment late
>0 days)
Groups >75%
All groups New groups
new members
(1)
(2)
(3)
Delinquency proportion (Loan payment late
>90 days)
Groups >75%
All groups
New groups
new members
(4)
(5)
(6)
Panel A. Unit of observation = region*month
Low interest rate
Fixed effects for month-year
Mean of dependent variable for high-rate regions
(treatment period)
N
Panel B. Unit of observation = group*month
Low interest rate
Fixed effects for month-year
Mean of dependent variable for high-rate regions
(treatment period)
-0.00935
-0.0171
-0.00815
-0.0159
-0.0163
-0.0115
(0.0172)
(0.0202)
(0.0176)
(0.0150)
(0.0211)
(0.0172)
y
y
y
y
y
y
0.142
2,313
0.190
2,312
0.151
2,311
0.098
2,150
0.157
2,066
0.097
2,102
-0.0200
(0.0159)
-0.0261
(0.0205)
-0.0190
(0.0178)
-0.0196
(0.0145)
-0.0273
(0.0204)
-0.0274
(0.0215)
y
y
y
y
y
y
0.133
0.196
0.145
0.105
0.186
0.130
662,275
138,724
237,287
152,595
36,675
54,198
Each column reports results for an OLS regression of proportion of lending groups delinquent on the variables shown or summarized in the
rows. Unit of randomization is the region, so standard errors allow for clustering at the region level. Treatment was implemented on May 16,
2007. The month of May 2007 is excluded because the experiment started in the middle of the month. Post = 1 for the time period of June 1,
2007 to October 31, 2009. No pre-treatment data available. * significant at 10%; ** significant at 5%; *** significant at 1%
36
Table 6. Crowd-Out and Competition
Dependent Variable:
Number of Loans
Number of Loans
Number of Loans
Number of Loans
Number of Loans
Number of Loans
Circulo Bureau
No
(1)
12,232***
(1,801)
1,708
(1,180)
4,440
(3,101)
Official Bureau
Yes
(2)
70,837***
(8,862)
28,787
(35,077)
18,889
(15,096)
Compartamos
Circulo Bureau
No
(4)
Official Bureau
Yes
(5)
Compartamos
-9,058**
(3,485)
-2,040
(2,085)
-5,490
(5,709)
-53,902***
(16,094)
-48,133
(61,648)
-18,127
(27,040)
-37.38
(190.5)
134.7
(204.3)
-50.63
(297.2)
Data source:
Data include Compartamos?:
Post
Low Interest Rate
Post*Low Interest Rate
Above median competition*Post
Above median competition*Low Interest Rate
Above median competition*Post*Low Interest Rate
(3)
1,992***
(280.5)
148.3
(520.3)
788.3*
(427.0)
(6)
Mean of dependent variable for high-rate regions
10,694.52
136,778.24
4,382.48
10769
137921
1130
(pre and post treatment period)
Point Elasticity
-3.65
-1.21
-1.58
4.48
1.15
0.39
N
158
158
158
156
156
156
Each column reports results for an OLS regression of number of loans on the variables shown or summarized in the rows. Unit of observation is the region-month, and
unit of randomization is the region, so standard errors allow for clustering at the region level. All municipalities included in this analysis have 20 or more Compartamos
clients. Pre-treatment data is taken from April 2007 report. Post-treatment data is taken from December 2008 report.We define "comparable loans" in the Mexican Credit
Bureau data as loans from a: Bank, Bank loan PFAE, Credit line from bank, Non-bank loan for (PFAE), Financial non-bank loan for PFAE, Personal finance company,
Medium market commerice, Medium market NF service, or Credit Union. In the Circulo data, "comparable" loans are those which institutions choose to report as
"personal loans." We measure competition using the number of competitors reported by CB branch staff in the mystery shopping exercise detailed in Section 3-I. *
significant at 10%; ** significant at 5%; *** significant at 1%.
37
Table7.CompetitiveStructureandPriceDispersion
PanelA.Baseline
Top 5 competitors, and everyone else
Finsol
Credito Familiar
Banco Azteca
FINCA
Financiera Independencia
other
total N
PanelB.Endline
Top 5 competitors, and everyone else
Finsol
Credito Familiar
Banco Azteca
FINCA
Financiera Independencia
other
total N
proportion
of N
(1)
0.21
0.11
0.11
0.04
0.03
0.50
201
0.20
0.08
0.07
0.06
0.05
0.54
415
APR, in percentage points
median
25th
75th
(2)
(3)
(4)
107
105
116
151
109
214
126
116
148
138
134
144
170
170
251
108
72
139
198
109
131
135
138
215
110
105
114
101
121
176
82
410
116
154
157
142
296
159
Data come from mystery shopping trials conducted by Compartamos branches (see Section 3-I for details). Sample for
this analysis starts with 616 observations at the level of: (compartamos branch)*(competitor loan offer)*(date), where
date is pre- or post-treatment. For the APR distributions, we drop three observations that lack sufficient information to
calculate the APR, and five observations that report nonpositive interest rates. The comparable APR for CB is 100% at
baseline and 80-90% at endline.
38
Table 8. Competitor responses
LowRate*Post
LowRate
Post
R-squared
pval LowRate*Post =size of CB rate cut
mean or median(LHS)
sd(LHS)
unit of observation
Observations
Annual Percentage Rate (APR)
level pp, log, level pp,
level pp,
log,
mean
mean median
minimum
minimum
(1)
(2)
(3)
(4)
(5)
36.20
0.01
0.00
2.04
-0.02
(25.70) (0.11)
(7.73)
(11.58)
(0.13)
-16.46 -0.01
0.00
-0.91
0.04
(16.37) (0.08)
(6.35)
(8.81)
(0.12)
5.90
0.09
0.00
3.91
0.01
(18.33) (0.08)
(5.59)
(9.28)
(0.10)
0.01
0.01
0.00
0.00
0.08
0.29
0.20
0.30
0.55
154
4.81
116
90.8
4.37
150
0.64
150
48.6
0.57
branch*competitor offer*date
branch*date
608
608
608
183
183
Competitor Count
level,
log,
mean
mean
(6)
(7)
-0.44
-0.16*
(0.33)
(0.09)
0.02
0.02
(0.15)
(0.08)
1.22***
0.35***
(0.21)
(0.06)
0.13
0.10
3.14
1.05
1.40
0.45
branch*date
183
183
* 0.10 ** 0.05 *** 0.01. OLS regressions standard errors clustered on region, except for median regression in column 3. Data come from
mystery shopping trials conducted by Compartamos branches (see Section 3-I for details). Sample for this analysis starts with 616
observations at the level of: (compartamos branch)*(competitor loan offer)*(date), where date is pre- or post-treatment. For the APR
regressions, we drop three observations that lack sufficient information to calculate the APR, and five observations that report
nonpositive interest rates.
39
Appendix Table 1. Number of loans (alternate specifications)
Levels
Treatment effect Low Interest Rate
(1)
(2)
Logs
(3)
199.6**
(90.21)
190.2**
(82.59)
0.134**
(0.0673)
Winsorized
(99%)
(4)
Winsorized
(95%)
(5)
Truncated
(99%)
(6)
Truncated
(95%)
(7)
185.7**
(81.38)
167.0**
(78.49)
191.1**
(80.17)
178.2**
(70.47)
Treatment effect estimated by:
Low*Post
Low
Low
Low
Low
Low
Low
Control for low interest rate branch
n
n
n
n
n
n
n
Fixed effects for month-year
y
y
y
y
y
y
y
Fixed effects for region
y
n
n
n
n
n
n
Includes controls for values of dependent
variable in pre-treatment months (March & April
n
y
y
y
y
y
y
2007)
Includes pre-treatment observations
y
n
n
n
n
n
n
Mean of dependent variable for pre-treatment
885.8
n/a
n/a
n/a
n/a
n/a
n/a
period
Mean of dependent variable for high-rate regions
1246
1246
6.919
1244
1227
1223
1154
(treatment period)
Point elasticity
-1.41
-1.34
-1.18
-1.31
-1.20
-1.37
-1.36
N
2,469
2,313
2,313
2,313
2,313
2,289
2,197
Each column reports results for an OLS regression of number of loans on the variables shown or summarized in the rows. Unit of observation is
the region-month, and unit of randomization is the region, so standard errors allow for clustering at the region level. Treatment was implemented
on May 16, 2007. The month of May 2007 is excluded because the experiment started in the middle of the month. Post = 1 for the time period of
June 1, 2007 to October 31, 2009. Pre-treatment months are March and April 2007. * significant at 10%; ** significant at 5%; *** significant at
1%.
40
Appendix Table 2. Loan amount (M$1,000s, alternate specifications)
Levels
Treatment effect Low Interest Rate
(1)
(2)
Logs
(3)
1,979**
(759.4)
1,920***
(640.4)
0.188**
(0.0740)
Winsorized
(99%)
(4)
Winsorized
(95%)
(5)
Truncated
(99%)
(6)
Truncated
(95%)
(7)
1,890***
(629.9)
1,715***
(627.3)
1,944***
(620.6)
1,856***
(560.9)
Treatment effect estimated by:
Low*Post
Low
Low
Low
Low
Low
Low
Control for low interest rate branch
n
n
n
n
n
n
n
Fixed effects for month-year
y
y
y
y
y
y
y
Fixed effects for region
y
n
n
n
n
n
n
Includes controls for values of dependent
variable in pre-treatment months (March & April
n
y
y
y
y
y
y
2007)
Includes pre-treatment observations
y
n
n
n
n
n
n
Mean of dependent variable for pre-treatment
6451
n/a
n/a
n/a
n/a
n/a
n/a
period
Mean of dependent variable for high-rate regions
9116
9116
8.799
9076
8866
8864
8215
(treatment period)
Point elasticity
-1.91
-1.85
-1.65
-1.83
-1.70
-1.93
-1.98
N
2,469
2,313
2,313
2,313
2,313
2,289
2,197
Each column reports results for an OLS regression of loan amount on the variables shown or summarized in the rows. Unit of observation is the
region-month, and unit of randomization is the region, so standard errors allow for clustering at the region level. Treatment was implemented on
May 16, 2007. The month of May 2007 is excluded because the experiment started in the middle of the month. Post = 1 for the time period of June
1, 2007 to October 31, 2009. Pre-treatment months are March and April 2007. * significant at 10%; ** significant at 5%; *** significant at 1%.
41
Appendix Table 3. Revenues (M$1,000s, alternate specifications)
Levels
(1)
(2)
Treatment effect Low Interest Rate
71.50
(87.78)
64.56
(81.41)
Logs
(3)
Winsorized
(99%)
(4)
Winsorized
(95%)
(5)
Truncated
(99%)
(6)
Truncated
(95%)
(7)
0.0436
(0.182)
89.21
(198.1)
73.15
(178.5)
81.21
(78.73)
106.4
(73.97)
Treatment effect estimated by:
Low*Post
Low
Low
Low
Low
Low
Low
Control for low interest rate branch
n
n
n
n
n
n
n
Fixed effects for month-year
y
y
y
y
y
y
y
Fixed effects for region
y
n
n
n
n
n
n
Includes controls for values of dependent
variable in pre-treatment months (March & April
n
y
y
y
y
y
y
2007)
Includes pre-treatment observations
y
n
n
n
n
n
n
Mean of dependent variable for pre-treatment
1025
n/a
n/a
n/a
n/a
n/a
n/a
period
Mean of dependent variable for high-rate regions
1279
1279
6.861
1271
1246
1246
1160
(treatment period)
Point elasticity
-0.49
-0.44
-0.38
-0.62
-0.52
-0.57
-0.81
N
2,469
2,313
2,313
2,313
2,313
2,289
2,195
Each column reports results for an OLS regression of revenues on the variables shown or summarized in the rows. Unit of observation is the
region-month, and unit of randomization is the region, so standard errors allow for clustering at the region level. Treatment was implemented on
May 16, 2007. The month of May 2007 is excluded because the experiment started in the middle of the month. Post = 1 for the time period of
June 1, 2007 to October 31, 2009. Pre-treatment months are March and April 2007. * significant at 10%; ** significant at 5%; *** significant at
1%.
42
Appendix Table 4. Costs (M$1,000s, alternate specifications)
Levels
Treatment effect Low Interest Rate
(1)
(2)
Logs
(3)
108.5**
(44.44)
80.26**
(35.96)
0.192*
(0.107)
Winsorized
(99%)
(4)
Winsorized
(95%)
(5)
Truncated
(99%)
(6)
Truncated
(95%)
(7)
145.6**
(70.35)
137.5**
(62.52)
84.71**
(33.57)
87.24***
(31.22)
Treatment effect estimated by:
Low*Post
Low
Low
Low
Low
Low
Low
Control for low interest rate branch
n
n
n
n
n
n
n
Fixed effects for month-year
y
y
y
y
y
y
y
Fixed effects for region
y
n
n
n
n
n
n
Includes controls for values of dependent
variable in pre-treatment months (March & April
n
y
y
y
y
y
y
2007)
Includes pre-treatment observations
y
n
n
n
n
n
n
Mean of dependent variable for pre-treatment
361.3
n/a
n/a
n/a
n/a
n/a
n/a
period
Mean of dependent variable for high-rate regions
566.5
566.5
6.216
564.6
555.1
555.1
522.5
(treatment period)
Point elasticity
-1.68
-1.24
-1.69
-2.27
-2.18
-1.34
-1.47
N
2,469
2,313
2,313
2,313
2,313
2,289
2,191
Each column reports results for an OLS regression of costs on the variables shown or summarized in the rows. Unit of observation is the regionmonth, and unit of randomization is the region, so standard errors allow for clustering at the region level. Treatment was implemented on May 16,
2007. The month of May 2007 is excluded because the experiment started in the middle of the month. Post = 1 for the time period of June 1, 2007
to October 31, 2009. Pre-treatment months are March and April 2007. * significant at 10%; ** significant at 5%; *** significant at 1%.
43
Appendix Table 5. Net present value (M$1,000s, alternate specifications)
Levels
(1)
(2)
Treatment effect Low Interest Rate
944.6
(708.6)
965.6
(696.7)
Logs
(3)
Winsorized
(99%)
(4)
Winsorized
(95%)
(5)
Truncated
(99%)
(6)
Truncated
(95%)
(7)
0.0692
(0.0681)
938.7
(680.2)
847.6
(613.8)
1,003
(658.5)
1,115*
(578.8)
Treatment effect estimated by:
Low*Post
Low
Low
Low
Low
Low
Low
Control for low interest rate branch
n
n
n
n
n
n
n
Fixed effects for month-year
y
y
y
y
y
y
y
Fixed effects for region
y
n
n
n
n
n
n
Includes controls for values of dependent
variable in pre-treatment months (March & April
n
y
y
y
y
y
y
2007)
Includes pre-treatment observations
y
n
n
n
n
n
n
Mean of dependent variable for pre-treatment
8317
n/a
n/a
n/a
n/a
n/a
n/a
period
Mean of dependent variable for high-rate regions
11286
11286
9.080
11258
11106
11078
10421
(treatment period)
Point elasticity
-0.74
-0.75
-0.61
-0.73
-0.67
-0.80
-0.94
N
2,469
2,313
2,313
2,313
2,313
2,289
2,194
Each column reports results for an OLS regression of net present value on the variables shown or summarized in the rows. Unit of observation is
the region-month, and unit of randomization is the region, so standard errors allow for clustering at the region level. Treatment was implemented
on May 16, 2007. The month of May 2007 is excluded because the experiment started in the middle of the month. Post = 1 for the time period of
June 1, 2007 to October 31, 2009. Pre-treatment months are March and April 2007. * significant at 10%; ** significant at 5%; *** significant at
1%.
44
Appendix Table 6. Personnel Costs
Treatment effect Low Interest Rate
Treatment effect Low Interest Rate (Year 1)
Treatment effect Low Interest Rate (Year 2)
Treatment effect Low Interest Rate (Year 3)
Personnel costs
(1)
(2)
43.18*
(21.96)
11.35
(14.73)
53.62*
(27.27)
93.91***
(34.98)
Number of Account Officers
(3)
(4)
1.400
(1.027)
0.353
(0.649)
1.843
(1.282)
2.828*
(1.600)
Treatment effect estimated by:
Low
Low* Year
Low
Low* Year
Control for low interest rate branch
n
n
n
n
Fixed effects for month-year
y
y
y
y
Fixed effects for region
n
n
n
n
Includes controls for values of dependent
variable in pre-treatment months (March & April
y
y
y
y
2007)
Includes pre-treatment observations
n
n
n
n
Mean of dependent variable for high-rate regions
253.175
13.83
(treatment period)
N
2,313
2,313
Each column reports results for an OLS regression of personnel costs or number of account officers on the
variables shown or summarized in the rows. Unit of observation is the region-month, and unit of
randomization is the region, so standard errors allow for clustering at the region level. Treatment was
implemented on May 16, 2007. The month of May 2007 is excluded because the experiment started in the
middle of the month. Post = 1 for the time period of June 1, 2007 to October 31, 2009. Pre-treatment
months are March and April 2007. All amounts are in Mexican pesos ($1,000s). * significant at 10%; **
significant at 5%; *** significant at 1%.
45
Appendix Table 7. Costing (in M$1,000s)
Total Costs
(1)
2.542
(5.865)
4.146***
(0.605)
Average number of members per group
Total number of lending groups
Personnel Costs
(2)
-6.172*
(3.612)
1.975***
(0.331)
Fixed effects for month-year
y
y
Fixed effects for region
n
n
Includes controls for values of dependent variable in prey
y
treatment months (March & April 2007)
Includes pre-treatment observations
n
n
Mean of dependent variable
639.041
296.485
Mean of average number of members per group
16.894
16.894
Mean of total number of lending groups
79.028
79.028
N
2,313
2,313
Each column reports results for an OLS regression of costs on the variables shown or summarized in the rows. Unit
of observation is the region-month, and unit of randomization is the region, so standard errors allow for clustering at
the region level. Treatment was implemented on May 16, 2007. The month of May 2007 is excluded because the
experiment started in the middle of the month. Average number of members per group and total number of lending
groups are flow variables. Total costs is the sum of all itemized operating costs reported per branch office (salaries,
benefits, bonuses, internal training, external training, and administrative expenses) plus charge-offs and 9.58%
(2007, 2008) or 6.31% (2009) of the outstanding balance as approximate cost of funds, accumulated over a month
and summed by region. All amounts are in Mexican pesos ($1,000s). The results for profits and total costs are
robust to changes in their definitions, such as definitions excluding the changes in the amount due after an account
becomes irrecuperable. * significant at 10%; ** significant at 5%; *** significant at 1%.
46
Appendix Table 8. Number of Groups and Members per Group
Treatment effect Low Interest Rate
Treatment effect Low Interest Rate (Year 1)
Treatment effect Low Interest Rate (Year 2)
Treatment effect Low Interest Rate (Year 3)
Total number of lending
groups
(1)
(2)
8.217
(5.022)
2.500
(2.847)
9.305
(5.997)
19.22**
(9.413)
Average number of members
per group
(3)
(4)
0.662*
(0.336)
0.587**
(0.288)
0.680*
(0.393)
0.796*
(0.441)
Average number of
groups per account
officer
(5)
(6)
0.178
(0.644)
0.491
(0.808)
-0.185
(0.808)
0.304
(0.866)
Average number of
clients per account
officer
(7)
(8)
17.28
(13.28)
21.61
(15.85)
11.43
(16.04)
21.01
(16.81)
Low* Year
Low
Low* Year
Low
Low* Year
Low
Low* Year
Treatment effect estimated by:
Low
Control for low interest rate branch
n
n
n
n
n
n
n
n
Fixed effects for month-year
y
y
y
y
y
y
y
y
Fixed effects for region
n
n
n
n
n
n
n
n
Includes controls for values of dependent variable in
y
y
y
y
y
y
y
y
pre-treatment months
Includes pre-treatment observations
n
n
n
n
n
n
n
n
Mean of dependent variable for high-rate regions
74.27
16.51
20.84
345.29
(treatment period)
N
2,313
2,313
2,313
2,313
Each column reports results for an OLS regression of the number of lending groups, the average number of members per group, average number of clients or average number
of groups per account officer on the variables shown or summarized in the rows. Unit of observation is the region-month, and unit of randomization is the region, so standard
errors allow for clustering at the region level. Treatment was implemented on May 16, 2007. The month of May 2007 is excluded because the experiment started in the middle
of the month. Average number of members per group and total number of lending groups are flow variables. * significant at 10%; ** significant at 5%; *** significant at 1%.
47
Appendix Table 9. Main outcomes (dropping Mexico City)
# loans
(1)
Treatment effect Low Interest Rate
238.5**
(97.44)
Treatment effect Low Interest Rate (Year 1)
Treatment effect Low Interest Rate (Year 3)
P-value: Treatment*1st year=Treatment*2nd year
P-value: Treatment*2nd year=Treatment*3rd year
P-value: Treatment*1st year=Treatment*3rd year
Low*Post
y
y
n
(4)
2,267***
(816.7)
115.9*
(58.46)
260.1**
(114.9)
478.6***
(174.8)
0.053
0.003
0.011
Treatment effect Low Interest Rate (Year 2)
Treatment effect estimated by:
Control for low interest rate branch
Fixed effects for month-year
Fixed effects for region
(2)
Loan amount
(M$1,000s)
(3)
Low*Post*Year
y
y
n
Revenues
(M$1,000s)
(5)
84.61
(99.79)
1,162**
(495.6)
2,471**
(965.6)
4,406***
(1,446)
0.035
0.001
0.006
Low*Post
y
y
n
(6)
Low*Post*Year
y
y
n
Costs
(M$1,000s)
(7)
104.1**
(47.79)
6.568
(63.23)
97.65
(117.2)
239.1
(171.4)
0.205
0.037
0.086
Low*Post
y
y
n
(8)
Low*Post*Year
y
y
n
Net present
value
(M$1,000s)
(9)
1,165
(801.7)
58.53*
(31.12)
113.0*
(57.64)
191.2**
(73.60)
0.118
0.001
0.012
Low*Post
y
y
n
(10)
Low*Post*Year
y
y
n
274.2
(354.9)
973.9
(761.8)
2,010
(1,256)
0.157
0.109
0.102
Low*Post
y
y
n
Low*Post*Year
y
y
n
Includes controls for values of dependent variable in pren
n
n
n
n
n
n
n
n
n
treatment months (March & April 2007)
Includes pre-treatment observations
y
y
y
y
y
y
y
y
y
y
Mean of dependent variable for pre-treatment period
924.2
924.2
6743
6743
1075
1075
359.9
359.9
8719
8719
Mean of dependent variable for high-rate regions
1272
1272
9319
9319
1307
1307
571.3
571.3
11534
11534
(treatment period)
Point elasticity
-1.65
-2.14
-0.57
-1.60
-0.89
Point elasticity year 1
-0.98
-1.33
-0.05
-1.12
-0.26
Point elasticity year 2
-1.69
-2.20
-0.62
-1.57
-0.91
Point elasticity year 3
-2.56
-3.25
-1.30
-2.41
-1.88
N
2,314
2,314
2,314
2,314
2,314
2,314
2,314
2,314
2,314
2,314
Each column reports results for an OLS regression of the outcome on the variables shown or summarized in the rows. Unit of observation is the region-month, and unit of randomization is the region, so standard
errors allow for clustering at the region level. Treatment was implemented on May 16, 2007. The month of May 2007 is excluded because the experiment started in the middle of the month. Post = 1 for the time
period of June 1, 2007 to October 31, 2009. Pre-treatment months are March and April 2007. * significant at 10%; ** significant at 5%; *** significant at 1%.
48
Figure 1. Randomized Pricing by Study Region
“high” rate
“low” rate
Bhutta and Keys (forthcoming AER)
Ponce et al (2016 wp)
DeFusco and Paciorek (forth AEJ: Pol)*
49
50
51
52