Skip to main content

The effects of access to credit on productivity: separating technological changes from changes in technical efficiency


Improving productivity among farm enterprises is important, especially in low-income countries where market imperfections are pervasive and resources are scarce. Relaxing credit constraints can increase the productivity of farmers. Using a field experiment involving in Bangladesh, we estimated the impact of access to credit on the overall productivity of rice farmers, and disentangled the total effect into technological change (frontier shift) and technical efficiency changes. We found that relative to the baseline rice output per decimal, access to credit resulted in, on average, approximately a 14 percent increase in yield, holding all other inputs constant. After decomposing the total effect into the frontier shift and efficiency improvement, we found that, on average, around 11 percent of the increase in output came from changes in technology, or frontier shift, while the remaining 3 percent was attributed to improvements in technical efficiency. The efficiency gain was higher for modern hybrid rice varieties, and almost zero for traditional rice varieties. Within the treatment group, the effect was greater among pure tenant and mixed-tenant farm households compared with farmers that only cultivated their own land.

This is a preview of subscription content, access via your institution.

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5


  1. We define productivity as yield per unit of land (kilogram of rice per decimal of land). A decimal (also spelled decimel) is a unit of area in India and Bangladesh approximately equal to 1/100 acre (40.46 m2); 247 decimal = 1 hectare.

  2. As per the taxonomy presented by Harrison and List (2004).

  3. The largest NGO in Bangladesh.

  4. The conventional production function approach does not allow us to separate technological change (frontier shift) and efficiency improvements from the overall productivity effect. We use the stochastic frontier model because it allows us to decompose these two effects, we use this approach as a tool to answer our research question.

  5. Hossain et al. (2019), Hossain et al. (2016), and Malek et al. (2015) examine the impact of the BCUP program on asset holdings, aggregate welfare, and wage employment.

  6. Numerous other experimental field studies, documented in Banerjee et al. (2015), Banerjee (2013), and Roodman and Morduch (2014), have examined how the availability of microcredit affects other important outcomes, such as business size and profits (Banerjee et al. 2015), income composition (Banerjee et al. 2015), stock of household durables (Attanasio et al. 2015), occupational choice, business scale, and risk management (Banerjee et al. 2015), female decision-making power (Angelucci et al. 2015), and happiness and trust (Angelucci et al. 2015).

  7. De Janvry et al. (2017) reviewed all of the recent experimental field studies on agricultural inputs in developing countries. We focus on credit as an input and its impact on productivity and technical efficiency. The impact of other production inputs on productivity has also been examined in relation to inputs such as credit (de Mel, McKenzie and Woodruff 2008, McKenzie and Woodruff 2008), capital (Karlan et al. 2015), labor (Shearer 2004), information (Beaman and Magruder 2012), monitoring (Nagin et al. 2002), and managerial practices (Bloom and Van Reenen 2010, Karlan and Valdivia 2011; Drexler et al. 2014).

  8. Chakravarty et al. (2019) examine the role of human capital constraints on productivity; Nikolov and Jimi (2018) review the role of informational constraints.

  9. Credit has been shown to affect the risk-taking behavior of producers (Boucher et al. 2008; Eswaran and Kotwal 1989), thereby affecting technology choices and adoption by farmers. The timing of the investment decision can also play an important role in one’s risk preferences (Nikolov 2018).

  10. Although we rely on an exogenous change in the price of borrowing as a result of the fact that the treatment group obtains access at a subsidized rate, other studies have examined exogenous changes in other aspects of microcredit programs such as microcredit access (Banerjee et al. 2015; Crépon et al. 2015), loan maturity (Karlan and Zinman 2008), and loan eligibility (Karlan and Zinman 2009).

  11. In this study, we use the term “farm households” interchangeably with “poor rice farmers”.

  12. Tenant farm households are farms that cultivate other people’s land, either through sharecropping or renting, or both.

  13. The timely and repeated use of pesticides is very important in ensuring higher returns from modern hybrid rice varieties.

  14. The BCUP program included complementary extension services in the initial years. However, BRAC ceased to provide extension services in 2012 because of high attrition rates and high recovery costs (Hossain et al. 2019).

  15. For simplicity, we have not modeled risk in this study.

  16. As per the rules of the Microcredit Regulatory Authority (MRA) of Bangladesh Bank, NGOs can charge up to a maximum of 27 per cent interest on declining balances through their microfinance operations.

  17. The eligibility criteria for the BCUP program were: 1) The farmer has a National ID card; 2) The age of the farmer is between 18 and 60 years; 3) The education level of the farmer is no higher than SSC; 4) The farmer must have been a permanent resident of the area for at least three years; 5) The farmer has at least three years of prior experience in farming; 6) The land holding must be between 33 decimals and 200 decimals; 7) The farmer cannot be an MFI (Micro Finance Institution) member; and 8) The farmer must be willing to accept credit from BCUP.

  18. The sub-district (upzila) is an administrative unit in Bangladesh. There are 491 sub-districts in Bangladesh.

  19. A few branches in the southern region were exceptions. For the southern region branches, GIS mapping (see Figure A1 in the Appendix) was undertaken and the results were forwarded to the program administrators so that they could continue to expand the number of treatment intervention branches within the appropriate areas while avoiding incursions into the control areas. Because the BCUP program administrators were aware of the status of each branch in the study, it was unlikely that the program officers would disburse loans in a control branch (Malek et al. 2015).

  20. Described in Section 2.1.

  21. Willingness to accept credit is measured by a ‘Yes’ or ‘No’ answer in response to the question of whether a respondent is inclined to accept credit from the BCUP program.

  22. We adopted a simple random sampling method to select households from each village. The survey covered 4,301 households, of which 2,155 were in treatment areas and 2,146 were in control areas.

  23. Following the baseline survey, we forwarded the list of treatment branches to the BRAC-BCUP administrators, whereupon BRAC launched the BCUP program in the treatment branches. The program organizers visited all the villages to locate potential borrowers based on the eligibility criteria.

  24. For balancing checks, we restricted our sample to the rice producing farm households surveyed in 2012 (3,292 households).

  25. The results of the balancing tests by rice variety are presented in Appendix A (Tables A1 and A2).

  26. In column 2 of Table 3, we present results of the regression of attrition in the follow-up survey on treatment dummy and household covariates. and find no evidence that treatment assignment is statistically significantly related to household attrition status.

  27. Rice is a major crop produced in Bangladesh, and almost all of the farm families in the country grow rice. Rice is cultivated on 75 percent of the country’s cropland (Ganesh-Kumar et al. 2012), and is the primary source of income and employment for nearly 15 million farm households in Bangladesh (Bangladesh Bureau of Statistics 2008).

  28. It is important to note that our analysis only covers a partial equilibrium effect and does not capture first-order general equilibrium effects. Moreover, the coverage of the BCUP credit program is not large enough to create a village-level effect. As noted in Section 2.1, the BCUP program uses the VO as the platform for service delivery. Members are grouped into teams of five, and three to eight teams consisting of 15 to 40 members form a village-level tenant farmer association. BCUP program administrative data from 2012 suggest that sometimes the number of participants in a village is insufficient to form an association, and so two or three villages must be combined. Therefore, although the theoretical maximum number of BCUP participants from a village can be as many as 40, in reality the number is much lower, and is not a large proportion of the total number of farm households in a village.

  29. Note that coefficient of ln L in specification (3) can be expressed as RTS−1. A positive coefficient indicates RTS of land is greater than 1 and a negative coefficient means RTS of land is less than 1.

  30. One might argue that the effect of credit access on the production frontier operates through inputs: credit enables poor farmers to use pesticides and fertilizer, and buy modern seed varieties in a timely manner, thereby affecting the production frontier. However, the relationship might be linear for some inputs and nonlinear for others. For simplicity, we are trying to find the overall effect of credit access. Therefore, we add credit access as a separate factor in the production frontier (that is, γ1Zi) rather than examining the effect of credit through inputs.

  31. In this approach, the estimation results do not impose any distributional assumption on ui(Zi). However, the major drawback of this approach is that the inefficiency effect cannot be separated from the noise (Zi) if the inefficiency is i.i.d. (a function of Zi).

  32. For a production-type stochastic frontier model with the composite error vi − ui(Zi), ui(Zi) ≥ 0 and vi distributed symmetrically around zero, the residuals from the corresponding OLS estimation should skew to the left (that is, negative skewness) regardless of the distribution function of ui(Zi) in the model estimation after pretesting. Thus, a test of the null hypothesis of no skewness can be constructed using the OLS residuals. If the estimated skewness has the expected sign, the rejection of the null hypothesis provides support for the existence of one-sided error.

  33. Inputs are in log form and include land (decimals), labor (days), ploughing land in preparation for planting (number of times), seed (kilograms), irrigation (hours), fertilizer (kilograms), and pesticide (number of times).

  34. Another reason why the sum of the two effects (from OLS) does not necessarily equal γ estimated using the maximum likelihood method is that the maximum likelihood method uses distributional assumptions, while OLS does not.

  35. If we estimate our frontier model (Eqs. 610) with a modification of Eq. (10), where credit access (Zi) is used as a determinant of the noise term, we find that the error/noise variance is on average higher for the households with credit access compared to the control group. And, when we compute the variance of the composed error \(\left[ {v_i - u_i\left( {Z_i} \right)} \right]\) as \(\sigma _u^2\left( {Z_i} \right) + \sigma _v^2(Z_i)\), then we find it smaller for Zi = 1, which is primarily due to the statistically significant negative effect of Zi on inefficiency.

  36. The findings of Table 4 show that rice variety choice is an outcome of the treatment status. This endogenous selection into rice variety can be a mediating mechanism of the productivity change differences by credit availability. Therefore, the results in Table A3 cannot be interpreted as the causal effect of credit availability on the productivity of a given rice variety.

  37. Mean baseline inefficiency is around 17 percent (estimated using Eqs. 610 and baseline data), which implies that before obtaining access to credit, farmers lose around 17 percent, on average, of their potential rice output through inefficiency.

  38. It is also tempting to consider that unmeasured or poorly measured inputs will show up as efficiency. However, because of the experimental design, this potential measurement bias is likely to be the same in both the treatment and control groups, and thus will be cancelled out.

  39. From columns 7 and 8, it can be seen that γ1 = 13.06, δj = 3.45. The mean for pure tenant farm households is 0.32, which implies that 32 percent of farm households in the sample only cultivate other people’s land, therefore the effect of the treatment assignment is (13.06 + (3.45*0.32)) = 14.16.


  • Anang BT, Bäckman S, Sipiläinen T (2016) Agricultural microcredit and technical efficiency. J Agric Rural Dev Trop Subtrop 117(2):189–202

    Google Scholar 

  • Angelucci M, Karlan D, Zinman J (2015) Microcredit impacts: evidence from a randomized microcredit program placement experiment by compartamos banco. Am Econ J: Appl Econ 7(1):151–182

    Google Scholar 

  • Attanasio O, Augsburg B, De Haas R, Fitzsimons E, Harmgart H (2015) The impacts of microfinance: evidence from joint-liability lending in Mongolia. Am Econ J: Appl Econ 7(1):90–122

    Google Scholar 

  • Banerjee AV (2013) Microcredit under the microscope: what have we learned in the past two decades, and what do we need to know? Annu Rev Econ 5(1):487–519

    Google Scholar 

  • Banerjee A, Duflo E, Glennerster R, Kinnan C (2015) The miracle of microfinance? Evidence from a randomized evaluation. Am Econ J: Appl Econ 7(1):22–53

    Google Scholar 

  • Banerjee A, Karlan D, Zinman J (2015) Six randomized evaluations of microcredit: introduction and further steps. Am Econ J: Appl Econ 7(1):1–21

    Google Scholar 

  • Bangladesh Bureau of Statistics. (2008) Report on agricultural censuses, 1960, 1983-84, 1996, 2008 (1st ed.). Ministry of Planning, Bangladesh Bureau of Statistics, Dhaka, Bangladesh

  • Beaman L, Magruder J (2012) Who gets the job referral? evidence from a social networks experiment. Am Econ Rev 102(7):3574–3593

    Google Scholar 

  • Bélanger D, Li XU (2009) Agricultural land, gender and kinship in rural China and Vietnam: a comparison of two villages. J Agrarian Change 9(2):204–230

    Google Scholar 

  • Blancard S, Boussemart J, Briec W, Kerstens K (2006) Short-and long-run credit constraints in french agriculture: A directional distance function framework using expenditure-constrained profit functions. Am J Agric Econ 88(2):351–364.

    Article  Google Scholar 

  • Bloom N, Van Reenen J (2010) Why do management practices differ across firms and countries? J Econ Perspect 24(1):203–224

    Google Scholar 

  • Boucher SR, Carter MR, Guirkinger C (2008) Risk rationing and wealth effects in credit markets: theory and implications for agricultural development. Am J Agric Econ 90(2):409–423

    Google Scholar 

  • Bravo-Ureta BE, Solís D, López VHM, Maripani JF, Thiam A, Rivas T (2007) Technical efficiency in farming: a meta-regression analysis. J Product Anal 27(1):57–72

    Google Scholar 

  • Bruhn M, Karlan D, Schoar A (2010) What capital is missing in developing countries? Am Econ Rev 100(2):629–633

    Google Scholar 

  • Brümmer B, Loy J (2000) The technical efficiency impact of farm credit programmes: a case study of northern Germany. J Agric Econ 51(3):405–418

    Google Scholar 

  • Chakravarty S, Lundberg M, Nikolov P, Zenker J (2019) Vocational training programs and youth labor market outcomes: Evidence from Nepal J Devel Econ 136(January):71–110

    Google Scholar 

  • Chirkos AY (2014) The impact of microfinance on living standards, empowerment and poverty alleviation of the poor people in Ethiopia, A case study in ACSI. Res J Financ Account 5(13):43–66

    Google Scholar 

  • Coelli T (1995) Estimators and hypothesis tests for a stochastic frontier function: a Monte Carlo Analysis. J Product Anal 6(3):247–268

    Google Scholar 

  • Cornia GA (1985) Farm size, land yields and the agricultural production function: an analysis for fifteen developing countries. World Dev 13(4):513–534

    Google Scholar 

  • Crépon B, Devoto F, Duflo E, Parienté W (2015) Estimating the impact of microcredit on those who take it up: Evidence from a randomized experiment in Morocco. Am Econ J: Appl Econ 7(1):123–150

    Google Scholar 

  • Davidova S, Latruffe L (2007) Relationships between technical efficiency and financial management for Czech Republic farms. J Agric Econ 58(2):269–288

    Google Scholar 

  • De Janvry A, Sadoulet E, Suri T (2017) Field experiments in developing country agriculture. In: Banerjee AV. In: Duflo Esther (Ed.) Handbook of economic field experiments, 1st ed. Elsevier, Amsterdam, Netherlands, p 427–466

    Google Scholar 

  • De Mel S, McKenzie D, Woodruff C (2008) Returns to capital in microenterprises: evidence from a field experiment. Q J Econ 123(4):1329–1372

    Google Scholar 

  • Drexler A, Fischer G, Schoar A (2014) Keeping it simple: financial literacy and rules of thumb. Am Economic J: Appl Econ 6(2):1–31

    Google Scholar 

  • Duflo E (2006) Poor but rational? In: Banerjee AV, Bénabou R, Mookherjee D (Ed.) Understanding poverty, 1st ed. Oxford University Press, New York, NY, p 367–378

    Google Scholar 

  • Eswaran M, Kotwal A (1989) Credit as insurance in agrarian economies. J Dev Econ 31(1):37–53

    Google Scholar 

  • Fan S, Chan‐Kang C (2005) Is small beautiful? farm size, productivity, and poverty in asian agriculture. Agric Econ 32(s1):135–146

    Google Scholar 

  • Ganesh-Kumar A, Prasad SK, Pullabhotla H (2012) Supply and demand for cereals in Bangladesh. International Food Policy Research Institute, Washington, DC, p 2010–2030

    Google Scholar 

  • Guirkinger C, Boucher SR (2008) Credit constraints and productivity in Peruvian agriculture. Agric Econ 39(3):295–308.

    Article  Google Scholar 

  • Hadley D, Shankar B, Thirtle C, Coelli T (2001) Financial exposure and farm efficiency: evidence from the England and Wales dairy sector. Paper presented at the Annual Meetings of the American Agricultural Association, Chicago, IL

  • Harrison GW, List JA (2004) Field experiments. J Economic Lit 42(4):1009–1055

    Google Scholar 

  • Hossain M, Md M, Md H, Md R, Md A (2019) Agricultural microcredit for tenant farmers: evidence from a field experiment in Bangladesh. Am J Agric Econ, (In press).

    Google Scholar 

  • Hossain M, Malek MA, Hossain MHR, Ahmed MS (2016) Impact assessment of credit program for the tenant farmers in Bangladesh: evidence from a field experiment. (Mimeo). University of Tokyo, Tokyo, Japan,

  • Hossain M, Bayes A (2009) Rural economy and livelihoods: Insights from Bangladesh, 1st ed. A.H. Development Pub. House, Dhaka, Bangladesh

    Google Scholar 

  • Hussain A, Thapa GB (2012) Smallholders’ access to agricultural credit in Pakistan. Food Secur 4(1):73–85

    Google Scholar 

  • Islam KZ, Sipilainen T, Sumelius J (2011) Influence of credit constraints on technical, allocative and cost efficiency in peasent farming in Bangladesh. Eur J Sci Res 56(2):229–243

    Google Scholar 

  • Islam KZ, Sumelius J, Bäckman S (2012) Do differences in technical efficiency explain the adoption rate of HYV rice? Evidence from Bangladesh. Agric Econ Rev 13(1):93

    Google Scholar 

  • Jack BK (2013) Market inefficiencies and the adoption of agricultural technologies in developing countries. Tufts University, Mimeo

    Google Scholar 

  • Karlan DS, Zinman J (2008) Credit elasticities in less-developed economies: implications for microfinance. Am Econ Rev 98(3):1040–1068

    Google Scholar 

  • Karlan D, Osei R, Osei-Akoto I, Udry C (2014) Agricultural decisions after relaxing credit and risk constraints Q J Econ 129(2):597–652.

    Article  Google Scholar 

  • Karlan D, Knight R, Udry C (2015) Consulting and capital experiments with microenterprise tailors in Ghana. J Econ Behav Organ 118:281–302

    Google Scholar 

  • Karlan D, Valdivia M (2011) Teaching entrepreneurship: impact of business training on microfinance clients and institutions. Rev Econ Stat 93(2):510–527

    Google Scholar 

  • Karlan D, Zinman J (2009) Expanding credit access: using randomized supply decisions to estimate the impacts. Rev Financial Stud 23(1):433–464

    Google Scholar 

  • Kondo T, Orbeta Jr, A, Dingcong C, Infantado C (2008) Impact of microfinance on rural households in the Philippines. IDS Bull 39(1):51–70

    Google Scholar 

  • Kumbhakar SC, Wang H, Horncastle AP (2015) A practitioner’s guide to stochastic frontier analysis using Stata, 1st ed. Cambridge University Press, New York, NY

    Google Scholar 

  • Liu Z, Zhuang J (2000) Determinants of technical efficiency in post-collective Chinese agriculture: evidence from farm-level data. J Comp Econ 28(3):545–564

    Google Scholar 

  • Malek MA, Ahasan A, Hossain MA, Ahmed MH, Reza MH (2015) Impact assessment of credit programme for the tenant farmers in Bangladesh, 3ie grantee final report. International Initiative for Impact Evaluation (3ie), New Delhi, India

  • Martey E, Dogbe W, Etwire PM, Wiredu AN (2015) Impact of farmer mentorship project on farm efficiency and income in rural Ghana. J Agric Sci 7(10):79

    Google Scholar 

  • McKenzie D, Woodruff C (2008) Experimental evidence on returns to capital and access to finance in Mexico. World Bank Econ Rev 22(3):457–482

    Google Scholar 

  • McKernan S (2002) The impact of microcredit programs on self-employment profits: do noncredit program aspects matter? Rev Econ Stat 84(1):93–115

    Google Scholar 

  • Montgomery H, Weiss J (2011) Can commercially-oriented microfinance help meet the millennium development goals? Evidence from Pakistan. World Dev 39(1):87–109

    Google Scholar 

  • Nagin DS, Rebitzer JB, Sanders S, Taylor LJ (2002) Monitoring, motivation, and management: the determinants of opportunistic behavior in a field experiment. Am Econ Rev 92(4):850–873

    Google Scholar 

  • Nikolov P (2018) Time delay and investment decisions: evidence from an experiment in Tanzania. Econ Bull 38(2):1124–1137

    Google Scholar 

  • Nikolov P, Jimi N (2018) What factors drive individual misperceptions of the returns to schooling in Tanzania? Some lessons for education policy Appl Econ 50(44):4705–4723

    Google Scholar 

  • Rahman S, Hussain A, Taqi M (2014) Impact of agricultural credit on agricultural productivity in Pakistan: an empirical analysis. Int J Adv Res Manag Soc Sci 3(4):125–139

    Google Scholar 

  • Roodman D, Morduch J (2014) The impact of microcredit on the poor in Bangladesh: revisiting the evidence. J Dev Stud 50(4):583–604

    Google Scholar 

  • Shah AK, Shafir E, Mullainathan S (2015) Scarcity frames value. Psychol Sci 26(4):402–412

    Google Scholar 

  • Shah AK, Mullainathan S, Shafir E (2012) Some consequences of having too little. Science 338(6107):682–685

    Google Scholar 

  • Shearer B (2004) Piece rates, fixed wages and incentives: evidence from a field experiment. Rev Econ Stud 71(2):513–534

    Google Scholar 

  • Singh I, Squire L, Strauss J (1986) Agricultural household models: extensions, applications, and policy. The Johns Hopkins University Press, Baltimore, MD

    Google Scholar 

  • Stiglitz JE, Morshed A, Emran S (2006) Microfinance and missing markets. Columbia University, New York, NY

    Google Scholar 

  • Taylor TG, Drummond HE, Gomes AT (1986) Agricultural credit programs and production efficiency: an analysis of traditional farming in southeastern minas gerais, brazil. Am J Agric Econ 68(1):110–119

    Google Scholar 

  • Tiruneh A, Tesfaye T, Mwangi W, Verkuijl H (2001) Gender differentials in agricultural production and decision-making among smallholders in ada, lume, and gimbichu woredas of the central highlands of Ethiopia. 1st ed. Mexico: CIMMYT

  • Udry C (1996) Gender, agricultural production, and the theory of the household. J Political Econ 104(5):1010–1046

    Google Scholar 

  • World Bank (2004) World development report 2005: a better investment climate for everyone. Oxford University Press, New York, NY

Download references


The field experiment was registered at the AEA RCT Registry (AEARCTR-0004460) and additional project information is available at Data analysis and ongoing data collection received an IRB approval from the Binghamton University HSRRC (STUDY00000449). Matthew Bonci and Declan Levine provided outstanding research support. We thank Seema Jaychandran, Eric Edmonds, David McKenzie, David Lam, Jessica Goldberg, Sam Asher, Rachel Heath, Jack Willis, Ruixue Jia, David Canning, Livia Montana, James Berry, Morgan Hardy, Zoe McLaren, Gil Shapira, Jeremy Barofksy, Denni Tommasi, Susan Wolcott, Neha Khanna, Solomon Polachek, Leila Salarpour, Emir Malikov, Jorgen Harris, Maulik Jagnani, seminar participants at NBER’s 10th Entrepreneurship Bootcamp, Cornell University’s Economics and AEM Departments, the 2016 Agricultural and Applied Economics Annual Meeting, and the 2017 Pacific Development Conference (PacDev) for constructive feedback and helpful comments. We acknowledge financial support from The International Initiative for Impact Evaluation (3ie) and The Lois De Fleur International Innovation Fund at The State University of New York (at Binghamton). This paper was accepted when Mohammad Abdul Malek was at BRAC Research and Evaluation Division, Dhaka, Bangladesh.

Author information

Authors and Affiliations


Corresponding author

Correspondence to Plamen V. Nikolov.

Ethics declarations

Conflict of interest

The authors declare that they have no conflict of interest.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Jimi, N.A., Nikolov, P.V., Malek, M.A. et al. The effects of access to credit on productivity: separating technological changes from changes in technical efficiency. J Prod Anal 52, 37–55 (2019).

Download citation

  • Published:

  • Issue Date:

  • DOI:


  • Field experiment
  • Microfinance
  • Credit
  • Efficiency
  • Productivity
  • Farmers

JEL classification

  • G21
  • G31
  • L25
  • I38
  • E22
  • H81
  • Q12
  • D2
  • O12
  • O16