33
therefore, we simply test the mean differences beween those two groups.13 The results provided in Table 5 show that only the average years of household head’s education is weakly statistically significantly different at the 10% level. Although we do not strongly claim that these two groups are the same, we may safely say that they are sufficiently similar. Given this similarity, the probit estimation results shown in Tables 3–4 could be interpreted as correlates of individual-level acceptance, valid for the entire sample including individuals belonging to groups that rejected uptake of the credit scheme.
34
large-scale loans are preferred even by the ultra-poor, who are usually believed to be risk-averse and who demand small-scale loans. Although the overall uptake of in-kind credit is significantly lower than equivalently-valued cash credit, the ultra-poor are more likely to accept the in-kind offer than the moderately poor. Indeed, a key to attracting the ultra-poor is to provide a grace period in the repayment schedule, irrespective of whether credit is provided in cash or in kind. It is also found that when offered, in-kind (cow) credit was more likely to be accepted if a potential borrower had previous experience of livestock rearing, indicating the necessity of supplementary training for the ultra-poor. This paper provides evidence that a typical microcredit offer with a one-year maturity period without a grace period is less attractive for the ultra-poor. Our results suggest the possibility that microfinance institutions can expand their outreach to the ultra-poor by offering them longer maturity loans with convenient grace periods, without compromising loan repayment schedules.
As a thorough study of the suitability of long maturity loans with a grace period for the ultra-poor in developing countries, this paper lacks an analysis of the impact of contract designs on borrower repayment behavior and their welfare indicators.
While our field observations indicate that repayment rates have not substantially differed across the treatment arms, and some clients with a grace period contract have even voluntarily started saving to smooth future repayments, we cannot judge at this moment whether the large loans with a grace period benefit both MFIs and their clients.
As the data collection remains on-going in the field, these issues will be analyzed in more detail after appropriate data becomes available. Another remaining issue is understanding within-group dynamics of members that led to group rejection. The results shown in this paper are reduced-form, with little insight into this issue. Modeling
35
interactions among members, and theoretically and empirically analyzing the case in northern Bangladesh also remain for future studies.
36 References
Banerjee, Abhijit, Esther Duflo, Rachel Glennerster, Cynthia Kinnan. (2014). “The Miracle of Microfinance: Evidence from a Randomized Evaluation.” forthcoming in American Economic Journal: Applied Economics.
Banerjee, Abhijit V. and Andrew F. Newman. (1993). "Occupational Choice and the Process of Development." Journal of Political Economy, 101(2): pp.274-298.
Barkat, Abul, Prosanta Kumar Roy, Md. Shahnewas Khan. (2007). “Charland in Bangladesh: Political Economy of Ignored Resource.” Pathak Samabesh and Samata, Dhaka, ISBN.984-8120-67-X
Copestake, James, Martin Greeley, Susan Johnson, Naila Kabeer, and Anton Simanowitz. (2005). Money with a Mission, Volume 1: Microfinance and Poverty Reduction. Warwickshire, UK: ITDG Publishing.
Crepon, Bruno, Florencia Devoto, Esther Dulfo William Pariente. (2014). “Estimating the Impact of Microcredit on Those Who Take it Up: Evidence from a Randomized Experiment in Morocco.” forthcoming in American Economic Journal: Applied Economics.
Cuong, Nguyen Viet. (2008). “Is a Governmental Micro-Credit Program for the Poor Really Pro-Poor? Evidence from Vietnam.” Developing Economies, 46(2):
pp151–187.
De Mel, Suresh, David McKenzie, and Woodruff Christopher. (2008), “Returns to Capital in Microenterprises: Evidence from a Field Experiment”, Quarterly Journal of Economics, 123 (4): pp. 1329-1372.
Duong, Pham Bao, and Yoichi Izumida. (2002). “Rural Development Finance in Vietnam: A Microeconometric Analysis of Household Surveys.” World Development, 30 (2):pp. 319–335.
37
Fafchamps, Marcel, David McKenzie, Simon Quinn, Christopher Woodruff. (2014).
“Microenterprise Growth and the Flypaper Effect: Evidence from a Randomized Experiment in Ghana.” Journal of Development Economics, 106: pp.211-226.
Field, Erica, and Rohini Pande. (2008). “Repayment Frequency and Default in Microfinance: Evidence from India.” Journal of the European Economic Association 6 (2-3): pp. 501–09.
Field, Erica, Rohini Pande, John Papp, and Natalia Rigol. (2013). “Does the Classic Microfinance Model Discourage Entrepreneurship Among the Poor? Experimental Evidence from India.” American Economic Review, 103(6): pp. 2196-2226.
Fischer, Greg and Maitreesh Ghatak. (2010). “Repayment Frequency in Microfinance Contracts with Present-biased Borrowers,” Economic Organisation and Public Policy Discussion Papers Series, 21.
Galor, Odedm, and Joseph Zeira. (1993). “Income Distribution and Macroeconomics.”
Review of Economic Studies, 60 (1): pp.35-52.
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:pp.65-83.
Hulme, David. (1999). “Client Dropouts from East African Microfinance Institutions.”
Mimeo.
Karlan, Dean., and Jonathan Zinman. (2011). “Microcredit in Theory and Practice:
Using Randomized Credit Scoring for Impact Evaluation.” Science, 332(6035): pp.
1278-1284.
Karlan, Dean, and Jonathan Zinman. (2012). “List Randomization for Sensitive Behavior: An Application for Measuring Use of Loan Proceeds.” Journal of
38 Development Economics, 98 (1): pp. 71–75.
Khandker, Shahid. (2005), “Micro-finance and poverty: Evidence using panel data from Bangladesh.” World Bank Economic Review, 19(2):pp. 263-286.
Kono, Hisaki, and Kazushi Takahashi. (2010). “Microfinance Revolution: Its Effects, Innovations, and Challenges.” Developing Economies, 48(1): pp.15-73.
Lancaster, Kelvin J. (1966), “A New Approach to Consumer Theory,” Journal of Political Economy, 74 (2): pp.132-157.
Lonborg, Jonas H, and Ole D. Rasmussen. (2014). “Can Microfinance Reach the Poorest: Evidence from a Community Managed Microfinance Intervention.” World Development, 64: pp. 460-472.
Lybbert, Travis, and Christopher B. Barrett. (2010). “Risk Taking Behavior in the Presence of Nonconvex Asset Dynamics.” Economic Inquiry, 49(4):pp. 982-988.
Maes, Jan P, and Larry R Reed. (ed). (2012) State of the Microcredit Summit Campaign Report 2012: Washington D.C: Microcredit Summit Campaign.
Marks, Malcolm, and Catherine Vignon. (2008). “Char Life, Livelihoods, and Development in Floodplain Bangladesh”. CLP Publications.
McKenzie, David, and Christopher Woodruff. (2008). “Experimental Evidence on Returns to Capital and Access to Finance in Mexico,” World Bank Economic Review, 22(3):pp.457-482.
Mood, Alexander M., Franklin A. Graybill, and Duane C. Boes. (1974). Introduction to the Theory of Statistics, Second ed., New York: McGraw-Hill.
Morduch, Jonathan. (1999), “The Microfinance Promise”, Journal of Economic Literature, 37(4): pp.1569-1614.
Morduch, Jonathan, Shamika Ravi, and Jonathan Bauchet. (2013). “Substitution Bias
39
and External Validity: Why an Innovative Anti-Poverty Program Showed No Net Impact.” Working Paper, Financial Access Initiative.
Navajas, Sergio, Mark Schreiner, Richard Meyer, Claudio Gonzalez-Vega, and Jorge Rodriguez Meza. (2000). “Microcredit and the Poorest of the Poor: Theory and Evidence from Bolivia.” World Development, 28(2):pp.333–346.
Roodman, David, and Jonathan Morduch. (2014) “The Impact of Microcredit on the Poor in Bangladesh: Revisiting the Evidence.” Journal of Development Studies, 50(4): pp.583-604.
Shonchoy, Abu. (2014). Seasonality and Microcredit: The Case of Northern Bangladesh.
Tokyo: Springer.
Shonchoy, Abu, and Takashi Kurosaki. (2014). “Impact of Seasonality-adjusted Flexible Microcredit on Repayment and Food Consumption: Experimental Evidence from Rural Bangladesh.” IDE Discussion Paper No. 460, IDE-JETRO.
Siliki, Anne-Claire. (2012). “Why People Dropout from Microfinance Institutions? Case Study of and MFI in Mali.” Solvary Brussel School Mimeo.
Udry, Christopher, and Santosh Anagol. (2006). “The Return to Capital in Ghana.”
American Economic Review, 96(2): pp. 383-393.
40
Figure 1: Satellite Image of Chars located in Northern Bangladesh (Note: Blue dots indicate the points where GPS coordinates were measured)
41
Figure 2: Timeline of Interventions and Surveys
Source: Prepared by the authors. The blue panels show events regarding interventions, red panels show events regarding surveys and the green panels show events regarding sample selection .
Char selection [April-May 2012]
Detailed information collection of each village belong to the identified Chars
(May-June 2012)
Village selection (80 villages in total) (June 2012)
PRA method to create the wealth status of all residents in each village
(July 2012)
Random offer for microcredit memebrship to eligible villagers.
Credit group formation, with 20 persons in each group . (August 2012)
Baseline survey of 80 groups, 1600 households (September-October 2012)
Announcement of randomization, both at the village level and household level
(November 2012)
Continue or discontinue the microcredit memebrship (Decemeber 2012- April 2013)
42
Figure 3: Randomization design
Randomly selected 80 villages (out of 128) on Chars to form microcredit groups. Each group has 20 members, 14 ultra-poor (UP) and 6 moderately poor
(MP) households, who are randomly given offer to have the group membership.).
Baseline survey to 1600 households
(4)
20 groups randomly selected for IK-GP
treatment group, within each group,
10 persons are selected randomly for credit (7 UP and
3 MP) 400 houeholds 400 houeholds 400 houeholds 400 houeholds (3)
20 groups randomly selected
for LC+GP treatment group, within each group,
10 persons are selected randomly
for credit (7 UP and 3 MP) 400 houeholds 400 houeholds 400 houeholds 400 houeholds (2)
20 groups randomly selected
for LC treatment group, within each
group, 10 persons are selected randomly for credit
(7 UP and 3 MP) 400 houeholds 400 houeholds400 houeholds 400 houeholds (1)
20 groups randomly selected
for RC treatment group, within each
group, 10 persons are selected randomly for credit
(7 UP and 3 MP) 400 houeholds 400 houeholds 400 houeholds 400 houeholds
43 Table 1. Characteristics of sample households and balance test
Total RC LC LC+GP IK+GP
(1) (2) (3) (4)
Treatment (=1) 0.500 0.500 0.500 0.500 0.500 0.000 0.000 0.000
(0.500) (0.501) (0.501) (0.501) (0.501) (0.035) (0.035) (0.035)
Ultrapoor (=1) 0.700 0.700 0.700 0.700 0.700 0.000 0.000 0.000
(0.458) (0.459) (0.459) (0.459) (0.459) (0.032) (0.032) (0.032)
Total HH income ('0000taka) 7.289 7.003 7.355 7.824 6.975 -0.353 -0.821** 0.028 -0.353 -0.821 0.028
(3.760) (3.307) (3.173) (4.754) (3.544) (0.229) (0.290) (0.242) (0.415) (0.610) (0.353)
Agricultural income ('0000taka) 0.018 -0.008 0.047 0.001 0.033 -0.054* -0.008 -0.041 -0.054 -0.008 -0.041
(0.376) (0.239) (0.481) (0.033) (0.523) (0.027) (0.012) (0.029) (0.030) (0.012) (0.026)
Livestock and poultry income ('0000taka) 0.169 0.132 0.192 0.166 0.184 -0.060 -0.035 -0.052 -0.060 -0.035 -0.052
(0.488) (0.355) (0.544) (0.498) (0.532) (0.032) (0.031) (0.032) (0.053) (0.053) (0.060)
Non-farm enterprise ('0000taka) 0.306 0.264 0.149 0.449 0.361 0.115 -0.185 -0.098 0.115 -0.185 -0.097
(1.405) (1.207) (0.848) (1.883) (1.464) (0.074) (0.112) (0.095) (0.085) (0.140) (0.107)
Wage income ('0000taka) 6.759 6.577 6.932 7.173 6.356 -0.355 -0.596* 0.220 -0.355 -0.596 0.220
(3.870) (3.444) (3.308) (4.911) (3.562) (0.239) (0.300) (0.248) (0.429) (0.646) (0.357)
Non-income ('0000taka) 0.037 0.038 0.036 0.035 0.040 0.002 0.003 -0.002 0.002 0.003 -0.002
(0.133) (0.102) (0.124) (0.167) (0.132) (0.008) (0.010) (0.008) (0.011) (0.012) (0.013)
Poverty (=1) 0.558 0.547 0.530 0.555 0.598 0.018 -0.008 -0.050 0.018 -0.007 -0.050
(0.497) (0.498) (0.500) (0.498) (0.491) (0.035) (0.035) (0.035) (0.057) (0.062) (0.054)
Experience of livestock production (=1) 0.476 0.435 0.525 0.482 0.460 -0.090* -0.048 -0.025 -0.090 -0.048 -0.025
(0.500) (0.496) (0.500) (0.500) (0.499) (0.035) (0.035) (0.035) (0.059) (0.066) (0.058)
# cattle owned 0.456 0.422 0.448 0.568 0.385 -0.025 -0.145* 0.037 -0.025 -0.145 0.038
(0.950) (0.906) (0.967) (1.072) (0.833) (0.066) (0.070) (0.062) (0.131) (0.124) (0.123)
Value of assets ('0000taka) 0.221 0.196 0.209 0.273 0.204 -0.012 -0.077* -0.008 -0.012 -0.077 -0.008
(0.441) (0.274) (0.262) (0.722) (0.331) (0.019) (0.039) (0.022) (0.033) (0.044) (0.042)
Sample mean Difference in mean at the
household level
Difference in aggregate mean at the group level (1)-(2) (1)-(3) (1)-(4) (1)-(2) (1)-(3) (1)-(4)
44 Table 1. (cont’d) Characteristics of sample households and balance test
Note: The difference is statistically significant at the 1% ***, 5% **, and 10% * level.
Source: Compiled from the microdata in the baseline survey (same as the following tables).
Household size 4.206 4.080 4.235 4.282 4.225 -0.155 -0.202 -0.145 -0.155 -0.202 -0.145
(1.483) (1.490) (1.523) (1.479) (1.435) (0.107) (0.105) (0.103) (0.163) (0.172) (0.153)
Dependency ratio 0.862 0.815 0.861 0.862 0.909 -0.045 -0.046 -0.094* -0.045 -0.046 -0.094
(0.616) (0.603) (0.635) (0.598) (0.625) (0.044) (0.042) (0.043) (0.049) (0.058) (0.058)
Head's age 38.583 38.925 38.042 38.672 38.690 0.883 0.252 0.235 0.883 0.253 0.235
(10.528) (10.529) (10.533) (9.878) (11.153) (0.745) (0.722) (0.767) (0.989) (1.121) (1.167)
Head is male (=1) 0.899 0.907 0.902 0.897 0.890 0.005 0.010 0.018 0.005 0.010 0.018
(0.301) (0.290) (0.297) (0.304) (0.313) (0.021) (0.021) (0.021) (0.035) (0.031) (0.027)
Head's years of schooling 0.748 0.498 0.877 0.660 0.958 -0.380** -0.163 -0.460** -0.380* -0.163 -0.460*
(2.150) (1.816) (2.248) (2.015) (2.445) (0.145) (0.136) (0.152) (0.181) (0.189) (0.216)
Years of current location 5.090 4.185 8.482 3.277 4.415 -4.298*** 0.907 -0.230 -4.297* 0.908 -0.230
(8.654) (8.214) (10.244) (7.369) (7.568) (0.657) (0.552) (0.558) (1.755) (1.188) (1.338)
Gaibandha (=1) 0.750 0.700 0.850 0.700 0.750 -0.150*** 0.000 -0.050 -0.150 0.000 -0.050
(0.433) (0.459) (0.358) (0.459) (0.434) (0.029) (0.032) (0.032) (0.133) (0.149) (0.145)
N 1600 400 400 400 400 800 800 800 40 40 40
45 Table 2A. Household-level uptake status by treatment arms and type of rejection
# of respondents
Uptake Individual rejection
Group rejection
Erosion and relocation
Total
RC (traditional) 226 54 80 40 400
LC (large w/o grace period) 347 13 40 400
LC+GP (large w grace period) 337 23 20 20 400
IK+GP (inkind) 301 79 20 400
Total 1211 169 140 80 1600
if treated
RC 107 33 40 20 200
LC 170 10 20 200
LC+GP 166 14 10 10 200
IK+GP 149 41 10 200
if control
RC 119 21 40 20 200
LC 177 3 20 200
LC+GP 171 9 10 10 200
IK+GP 152 38 10 200
46 Table 2B. Group-level uptake status by treatment arms and type of rejection
# of groups
Group-level uptake, distinguished by the number of members within each group who rejected individually
Group rejection
Erosion and relocation
Total 0 1 2 3 4 5 6 7 8 9
10 and more
Sub-total
RC 5 1 1 2 1 1 1 2 14 4 2 20
LC 10 5 1 2 18 2 0 20
LC+GP 12 1 2 1 1 1 18 1 1 20
IK+GP 2 7 1 1 1 2 2 1 2 19 0 1 20
Total 29 13 5 5 4 1 3 2 0 2 5 69 7 4 80
47
Table 3. Correlates of individual-level uptake decisions (including control households) Dep.var = Uptake dummy
Full sample RC LC LC+GP IK+GP
(1) (2) (3) (4) (5)
LC a 0.126***
(0.030)
LC+GP a 0.089***
(0.034)
IK+GP a -0.005
(0.043)
Treatment (=1) -0.036** -0.105* -0.029* -0.014 -0.001
(0.016) (0.056) (0.016) (0.024) (0.033)
Ultra-poor (=1) 0.044** 0.028 0.004 0.041* 0.110**
(0.018) (0.053) (0.016) (0.021) (0.051)
HH size 0.007 0.036 0.002 -0.008 0.033*
(0.009) (0.026) (0.006) (0.011) (0.020)
Dependency ratio -0.011 -0.034 -0.003 -0.001 -0.059
(0.014) (0.049) (0.011) (0.018) (0.048)
Head's age 0.012** 0.037** 0.001 0.016* 0.002
(0.006) (0.019) (0.002) (0.008) (0.011)
Its squared/1000 -0.138** -0.461** -0.020 -0.183* -0.014
(0.062) (0.219) (0.028) (0.093) (0.136)
Head is male (=1) 0.032 -0.146*** -0.010 0.128 0.074
(0.036) (0.051) (0.016) (0.088) (0.066)
Head's years of schooling 0.006 0.006 0.003 -0.002 0.017
(0.004) (0.010) (0.005) (0.004) (0.012)
Years of current location -0.002 0.000 0.000 -0.002 -0.007
(0.001) (0.003) (0.001) (0.002) (0.005) Experience of livestock production (=1) 0.011 -0.014 0.013 0.029* -0.029 (0.018) (0.075) (0.014) (0.016) (0.037)
# cattle owned 0.016 0.091* 0.017 -0.002 -0.010
(0.010) (0.051) (0.010) (0.006) (0.028)
Value of assets (10 thousands taka) -0.004 -0.039 0.055 -0.016 -0.022
(0.019) (0.062) (0.038) (0.014) (0.049)
Gaibandha (=1) -0.016 0.110 -0.001 -0.031 -0.048
(0.032) (0.129) (0.015) (0.035) (0.073)
1,380 280 360 360 380
Notes: Estimated by probit, using the subsample of members whose groups accepted the credit scheme. The parameter estimate is significantly different from zero at the 1% ***, 5% **, and 10% * level, using Char-level clustered standard error. a The omitted category is the regular microcredit (RC).
48
Table 4. Correlates of individual-level uptake decisions (using treatment households only) Dep.var = Uptake dummy
All treated RC LC LC+GP IK+GP
(1) (2) (3) (4) (5)
LC a 0.137***
(0.036)
LC+GP a 0.118***
(0.037)
IK+GP a 0.018
(0.048)
Ultra-poor (=1) 0.063** -0.010 0.028 0.067* 0.119*
(0.027) (0.070) (0.027) (0.036) (0.064)
HH size -0.009 0.011 -0.002 -0.012 -0.013
(0.012) (0.040) (0.006) (0.009) (0.031)
Dependency ratio 0.015 0.089 0.000 0.018 -0.023
(0.022) (0.086) (0.017) (0.021) (0.070)
Head's age 0.020** 0.041 0.003 0.020** 0.011
(0.008) (0.031) (0.004) (0.008) (0.013)
Its squared/1000 -0.233** -0.477 -0.053 -0.221** -0.119
(0.095) (0.389) (0.047) (0.092) (0.147)
Head is male (=1) 0.117* -0.210*** -0.002 0.223* 0.272**
(0.068) (0.069) (0.025) (0.131) (0.131)
Head's years of schooling 0.009 0.005 0.003 -0.000 0.035*
(0.007) (0.015) (0.006) (0.004) (0.019)
Years of current location -0.002 -0.002 0.001 -0.002 -0.009
(0.002) (0.004) (0.001) (0.001) (0.006)
Experience of livestock production (=1) 0.061*** 0.068 0.023 0.051** 0.073**
(0.022) (0.086) (0.026) (0.025) (0.036)
# cattle owned 0.010 0.086* 0.019 -0.014 -0.046
(0.015) (0.048) (0.013) (0.010) (0.037)
Value of assets (10 thousands taka) -0.012 -0.020 0.105 -0.021** -0.078
(0.032) (0.110) (0.067) (0.008) (0.130)
Gaibandha (=1) -0.011 0.141 -0.022 0.014 -0.081
(0.039) (0.155) (0.018) (0.036) (0.076)
690 140 180 180 190
Notes: See Table 3.