The Real Impact of Improved Access to Finance: Evidence from Mexico

Transcription

1 The Real Impact of Improved Access to Finance: Evidence from Mexico Miriam Bruhn Inessa Love GFDR Seminar February 14, 2012

2 Research Questions Does expanding access to finance to previously unbanked, low-income individuals reduce poverty? If yes, what are the channels? Does increased access to finance promote entrepreneurial activity?

3 Existing Evidence (1) Increased access to finance is associated with growth and poverty alleviation at the country level (World Bank, 2008, Honohan, 2004) Little micro level evidence showing that this relationship is causal Exception: Burgess and Pande (2005) show poverty reduction after rural banks open in India, but not clear what the channel is

4 Existing Evidence (2) US literature on banking deregulation shows positive impacts on firm creation and growth, and a reduction in inequality Jayarathne and Strahan (1996) Black and Strahan (2002) Cetorelli and Strahan (2006) Beck, Levine and Levkov (2008) Results probably due to increased bank competition, rather than reaching previously unbanked individuals

5 Impact of Microfinance Microfinance industry has expanded rapidly Recent studies examine its impact on poverty and microenterprise growth, e.g. Banerjee, Duflo, Glennester, and Kinnan (2010) Kaboski and Townsend (2009) Karlan and Zinman (2012) Mixed results: some find isolated increases in consumption, investment, or business growth

6 This Paper Exploit cross-municipality and cross-time variation in the opening of a bank for low-income individuals in Mexico to measure its effects on Entrepreneurial activity Employment Income The event: in October 2002, Grupo Elektra launched Banco Azteca, opening a total of 815 branches in all pre-existing Grupo Elektra stores

7 Preview of Results Number of informal businesses increased by 7.6 percent Employment went up by 1.4 percent Income increased by about 7 percent Larger effects for Individuals with below median income Municipalities that were relatively underserved by the banking sector before Azteca opened 7

8 Outline Banco Azteca Data and identification strategy Results Conclusion 8

9 A Bank for the Underserved Population Ricardo B. Salinas, Chairman of the Board of Grupo Elektra (Reuters, 2002) "Banco Azteca will improve access to goods and services for our people. A major impediment to the growth of the Mexican middle class has been the lack of access to credit, one of the main vehicles for personal financial improvement. Banco Azteca will demonstrate the importance of offering financial services to this under-served segment of the Mexican population."

10 Comparatively Low Access to Finance in Mexico % of households with access to a bank account Argentina Brazil Chile Colombia Mexico Peru US Source: World Bank Policy Research Report Finance for All (2008)

11 We changed banking, now it s your time to change Unique features of Banco Azteca Low documentation (collateral or co-signors instead of documents) 3,000 motorcycle-riding agents Extensive experience making small installment loans and large database of 4 mil. past clients Small loan size: $250-$500 (microfinance: $360) Synergies with retail stores lower operational costs Made it possible to reach the previously un-bankable population

12 Bank for

13 BUY A TOASTER, OPEN A BANK ACCOUNT Banco Azteca caters to the little guy--in appliance stores 13 January 2003 BusinessWeek 54, Number 3815 Pedro Rubio was in a bind. The 56-year-old carpenter needed to come up with thousands of pesos in notary fees to get legal title to his modest cinderblock house.. But Rubio, who earns the equivalent of $600 a month, had no proof of income and no bank account. So on a recent morning, he walked through his gritty Mexico City neighborhood to an Elektra appliance store. At the back, behind an aisle of microwave ovens, he sat down with a loan officer from a new bank, Banco Azteca. Unfazed by Rubio's worn jeans and unshaven face, the officer drew up an inventory of his possessions: TV, refrigerator, washing machine--all bought on credit at Elektra in the past three years. Accepting these as collateral, the bank approved Rubio's application within 24 hours. The nine-month, $200 loan carries a 48% annual interest rate, usurious by U.S. standards but not in Mexico, where the banking sector is still recovering from the effects of the 1994 peso crash. It's a little expensive,'' says Rubio. Still, he says he can swing the weekly $8 payments. In any event, he adds, I don't really have any other option.

14 Banco Azteca vs. Microfinance Institutions (MFIs) Azteca Compartamos Other MFIs Non-performing loans (%) Interest rate (%) Operating expenses over assets Note: Values are 2004 to 2011 averages. Source: Data on Azteca from the Bank Supervisory Commission (CNBV). Data on MFIs from MIXmarket.org.

15 Non-Trivial Impact on Lending Volume While Elektra was offering installment loans even before Azteca opened, the amount of loans grew significantly after the event Access to cheaper capital deposit base and lower cost of capital due to bank status

16 Elektra / Azteca s Loan Portfolio over Time (Consumer + Business) 12,000 Grupo Elektra s loan portfolio (millions of pesos) 10,000 8,000 6,000 4,000 2, Quarters

17 Non-Trivial Impact on Banking Market Total number of savings accounts grew rapidly in municipalities with Azteca branches

18 Impact of Azteca Opening on Savings Accounts over Time Coefficients on Azteca Municipalities in Dec 2002 *Quarter Dummies 16,000 14,000 12,000 10,000 8,000 6,000 4,000 2, ,000-4, Quarter

19 Outline Banco Azteca Data and identification strategy Results Conclusion 19

20 Mexican National Employment Survey Data (ENE) Random sample of ~ 150,000 households 2000-II to 2004-IV (19 quarters in total, 10 before and 9 after Azteca opening) Households remain in the survey for five consecutive quarters Use only the 576 municipalities with a bank branch (comparability) 43% had an Azteca branch in 2002-IV

21 Variables Outcome variables Informal business owner dummy Formal business owner dummy Wage earner dummy Not Employed dummy Above minimum wage dummy Log monthly income +1 Controls Age, gender, marital status, and education dummies

22 Difference-in-Difference Methodology We exploit cross-municipality and crosstime variation in Azteca branches y ict = α + β + γ + δ * Azteca * Post + π * Z + ε c t c t ict ict Where Azteca = dummy for municipalities which had at least one Azteca branch in 2002-IV Post = dummy for after 2002-IV Y = individual outcome variable Z = individual controls

23 Map of Municipalities with Banco Azteca Branches and Other Bank Branches

24 Identification Issues (1) Are municipalities without Azteca a good counterfactual for municipalities with Azteca? Differences in levels across municipalities are not a direct concern since the methodology compares changes

25 Identification Issues (2) Difference in changes might bias our results Placement of branches not targeted based on expected demand since branches opened in ALL preexisting stores

26 Pre-Azteca GDP per Capita Level and Growth Rate (Census Data) Azteca Non-Azteca Difference (1) (2) (3) Average 1993 GDP per capita (Pesos) 26,359 12,354 14,005*** (26,163) (20,777) (2,037) Average 1998 GDP per capita (Pesos) 30,450 15,910 14,539*** (28,134) (27,353) (2,360) Average 1993 to 1998 GDP per capita *** growth rate (%) (45) (67) (5) Statistical significance levels: * 10%, ** 5%, *** 1%

27 Pre-Azteca Outcome Variables (Levels) Azteca Non-Azteca Difference (1) (2) (3) Informal business owner dummy *** (0.275) (0.345) (0.009) Formal business owner dummy *** (0.270) (0.248) (0.003) Wage earner dummy *** (0.500) (0.496) (0.011) Not Employed dummy ** (0.474) (0.479) (0.007) Log monthly income *** (3.935) (3.810) (0.092) Statistical significance levels: * 10%, ** 5%, *** 1%

28 Pre-Azteca Outcome Variables (Changes) Azteca Non-Azteca Difference (1) (2) (3) Informal business owner change (0.039) (0.055) Formal business owner change (0.026) (0.036) (0.001) Wage earner change *** (0.049) (0.069) (0.003) Not Employed change *** (0.047) (0.065) (0.003) Log monthly income change *** (0.361) (0.459) (0.018) Statistical significance levels: * 10%, ** 5%, *** 1%

29 Identification Issues (3) Based on pre-azteca trends, our results may be biased against finding a positive effect on employment and income Our estimates are on the conservative side

30 Reducing Potential Biases Group trends (municipalities with and without Azteca are allowed to have different time trends) Each municipality is allowed to have a different trend Graphical analysis

31 Outline Banco Azteca Data and identification strategy Results Conclusion 31

32 Positive Impact on Informal Entrepreneurial Activity Dependent variable: Informal business owner dummy (1) (2) (3) Azteca*Post Dec ** ** ** (0.0027) (0.0028) (0.0027) R-squared No. of observations 4,728,268 4,728,268 4,728,268 Group time trend No Yes No Municipality time trend No No Yes Statistical significance levels: * 10%, ** 5%, *** 1% Effect size = 7.6 percent increase (0.0062/0.0821)

33 Positive Impact on Informal Entrepreneurial Activity (cont.) Average Infomal Business Dummy Quarter Municipalities without Azteca Municipalities with Azteca

34 No Impact on Formal Entrepreneurial Activity Dependent variable: Formal business owner dummy (1) (2) (3) Azteca*Post Dec * (0.0022) (0.0021) (0.0021) R-squared No. of observations 4,728,268 4,728,268 4,728,268 Group time trend No Yes No Municipality time trend No No Yes Statistical significance levels: * 10%, ** 5%, *** 1%

35 Why an effect on informal, but not formal businesses? Low documentation requirements make loans attractive for informal businesses Azteca s interest rates are relatively high 40 percent per year Formal businesses are likely to have lower cost sources of finance Other commercial banks, with yearly interest rates starting at 20 percent

36 Where does the increase in informal businesses come from? Use panel structure of the data Each individual is surveyed for 5 consecutive quarters Examine effect by past occupation Informal business owners Formal business owners Wage earners Not employed

37 Banco Azteca appears to promote survival of informal businesses Dependent variable: Informal business owner dummy Coefficient on Azteca*Post Dec 2002*Past : Informal business owner Formal business owner Wage earner Not employed *** (0.0148) (0.0107) (0.0034) (0.0033) Statistical significance levels: * 10%, ** 5%, *** 1%

38 What is the mechanism that promotes business survival? Increased access to credit Increased access to savings accounts Dupas and Robinson (2009) Providing savings accounts to microenterprise owners in Kenya promotes investment Save up for lumpy investments? Increased product demand from consumers who get Azteca credit

39 Positive Impact on Employment Dependent variable: Not employed dummy (1) (2) (3) Azteca*Post Dec *** *** (0.0032) (0.0034) (0.0034) R-squared No. of observations 4,728,268 4,728,268 4,728,268 Group time trend No Yes No Municipality time trend No No Yes Statistical significance levels: * 10%, ** 5%, *** 1% Effect size = 1.4 percent increase in employment (0.0098/0.6583)

40 Positive Impact on Income Dependent variable: Log (1 + Income) (1) (2) (3) Azteca*Post Dec ** *** (0.0240) (0.0286) (0.0275) R-squared No. of observations 4,533,848 4,533,848 4,533,848 Group time trend No Yes No Municipality time trend No No Yes Statistical significance levels: * 10%, ** 5%, *** 1% Effect size = ~ 7 percent increase in income

41 Sample Splits Study the effect on individuals with above vs. below median income Does Azteca help low income individuals? Examine effect by high vs. low pre- Azteca bank branches density Are the effects really driven by increased access to finance?

42 Stronger effects for individuals with below median income Dependent variable: Informal business Not employed owner dummy dummy Log (1+ income) Income (1) (2) (3) Above median (i) (No. of obs. 2,302,575) (0.0039) (0.0056) Below median *** *** *** (No. of obs. 2,231,273) (0.0034) (0.0058) (0.0337) (i) Everybody is employed. Statistical significance levels: * 10%, ** 5%, *** 1%

43 Stronger effects for municipalities with fewer banks pre-azteca Pre-Azteca branches per 100,000 people Dependent variable: Informal business Not employed owner dummy dummy Log (1+ income) (1) (2) (3) Above 75 th percentile (No of obs. 2,644,793) (0.0058) (0.0095) (0.0833) Below 75 th percentile *** *** *** (No. of obs. 2,083,475) (0.0030) (0.0041) (0.0317) Statistical significance levels: * 10%, ** 5%, *** 1%

44 GDP Growth in Azteca and Non-Azteca Municipalities Avg. GDP per capita growth rate (%) Start and end year for growth rate Municipalities with Azteca Municipalities without Azteca

45 Outline Banco Azteca Data and identification strategy Results Conclusion 45

46 Conclusions Increased availability of financial services to low income individuals has a positive impact on economic activity More informal businesses (decreased turnover) Increase in employment Higher income Cannot disentangle whether effects are due to increased credit, savings, or product demand