ISSN (Print) - 0012-9976 | ISSN (Online) - 2349-8846

A+| A| A-

Formal Financial Outreach in Rural India

The conventional wisdom is that opening up new branches is the best way to extend outreach of the formal financial sector to rural areas. Using a survey data set, this paper challenges the conventional view by concluding that relatively resource-rich rural households from distant locations availed multiple loans from formal lenders rather than the households located closer to them as often believed.

To date, the focus of the literature on financial development has been on two strands. The first strand of inquiry revolves around identifying the determinants of access to financial services from formal institutions. The second strand of empirical investigation evaluates the impact of access to the formal financial sector. Formal financial institutions include both public and private commercial banks, regional rural banks,1 village cooperatives, and their higher tier bodies. However, it is increasingly being accepted by policymakers that access to financial services differs significantly from the outreach of financial services. An economic agent with access to financial sources may not fully utilise these sources due to the high transaction cost involved in doing so (Sriram 2007). While access is defined as the likelihood of receiving the financial service of one’s choice, outreach indicates the extent of usage of the financial service by a household (Beck et al 2007).

Outreach was traditionally measured by the ratio of private sector credit to the gross domestic product (GDP). In contrast, Beck et al (2007) proposed additional measures of outreach that include per capita loan accounts and per capita deposit accounts. They surveyed regulatory bodies spread over 99 countries to present country-wise indicators of financial outreach. In another study, Beck et al (2008) consolidated the country-wise barriers to financial sector outreach. Barriers across 62 countries were captured under three broad categories, namely physical access, affordability, and eligibility. Physical access represents locations of branches; affordability includes the minimum balance required to open a savings account or the minimum amount of loan extended by the formal lender and processing fees paid to avail a financial product; and eligibility exhibits the number of documents one needs to submit along with the application. They found that government-owned as well as small-sized banks imposed a higher barrier on customers and limited the outreach of the formal financial sector. Beck et al (2007, 2008) argued that these macro indicators closely reflect harder-to-collect household micro-level data on formal financial sector outreach. Noted exception includes Brewer et al (2014) who explored the choice between single and multiple lenders among the Kansas farms and found that risky farms are more likely to carry banking relationship with multiple lenders to avail loan at the lowest possible cost.

In an attempt to fill this gap in the policy literature on formal financial sector outreach, this study uses micro-level data collected from 600 households spread over 14 villages across six Indian states. Here, the research objective is to assess the determinants of financial sector outreach in the context of a developing country. Following Beck et al (2007), this study captures financial sector outreach as the number of loans per rural household. To be precise, the study aims to answer the following research question: What factors determine the number of loans negotiated by households from formal financial sector and why some households remain out of their coverage? The study assumes that a higher number of loans per household indicates a deeper outreach and wider use of credit facilities from the formal financial sector. The result would be of great importance to the policymakers in crafting appropriate formal credit delivery system in rural locations of developing countries. According to a World Bank survey (Basu 2006), even after a century-long intervention of opening new branches in unbanked areas and offering credit at below-market clearing price, only 21% of borrower households in rural India enjoyed access to the formal credit delivery channel. Thus, this poor coverage of the formal financial sector may be an outcome of the formal lender extending credit to only those who could prove their credit worthiness in terms of asset holdings and profit potential. Following this, the study hypothesises that borrowers with better asset holdings and higher profit potential would receive multiple loans from formal lenders.

Scholars like Fletschner et al (2010) cited relatively higher transaction costs in terms of travel expenses for multiple visits and opportunity cost of wage loss that discouraged households to access financial services from formal lenders in Peru. However, the cross-country data on the barriers to bank access compiled by Beck et al (2008) showed that in India, customers availed banking facilities at multiple locations, namely head-offices, branches, and branch-like outlets. Furthermore, with regard to affordability, the minimum amount of loan in India was 28.79% of GDP per capita, against the world average of 76.84%. Moreover, fees for consumer loans as a percentage of the GDP per capita stood at 1.19%, lower than the world average of 1.58%. On the eligibility front, in India, on average, banks took 4.17 days to process a loan application, which is at par with the world’s average of 4.29 days. This shows that in India, credit facilities were available at multiple locations at affordable terms and the waiting time for a customer is quite low. This information did not support the argument that prospective borrowers in India were transaction-cost rationed.

The contribution of this paper is that unlike earlier studies that focused on identifying differential characteristics of rural households that affected their sectoral choice of credit, here this study explores what enables a rural household to avail multiple loans from institutional lenders. The results of this study revealed that resource-rich households, even at distant locations, enjoy greater outreach of the formal financial sector. This finding is important because it is in contrast to the established belief in the policy domain that opening more bank branches and taking the delivery window closer to the target group is not only necessary but also sufficient to ensure a steady flow of credit in rural areas.

The novelty of this work lies in the application of a Poisson regression model on a household level data to identify the determinants of formal financial sector outreach, as this is one of the large-sample empirical investigations, apart from Brewer et al (2014), to use count data model on primary data set to explore financial sector outreach and more specifically in a developing country context. However, this study differs from Brewer et al (2014) in the following ways: First, while Brewer et al (2014) deal with a farm’s borrowing relationship with multiple lenders, this study deals with multiple loans from a single formal lender. This is because, in India, the borrower has to obtain a no dues certificate from a formal lender to approach another formal lender. Hence, simultaneous borrowing from multiple formal lenders is a remote possibility and the same is the case in this sample. Second, while Brewer et al (2014) deal with the sample farms from the Kansas Farm Management Association of the United States, this study looks at farm households from rural India—a developing nation. And there exists significant difference in farm characteristics as well as lending practices between the developed and developing nations.

The remainder of this paper is organised in the following manner. First, it offers a brief review of the existing literature. Then it goes on to describe the data. Later, the study presents the empirical model and a discussion of the results. Finally, the study concludes and highlights the policy implications.

Literature Review

Following Stiglitz and Weiss (1981), researchers argue that formal lenders in fear of adverse selection and moral hazard limit access to credit among a selected few. To overcome the restricted access imposed by formal lenders, a prospective borrower needs to build up credit history with small test loans to win the confidence of these formal lenders, or provide marketable collateral of lender’s interest and arrange for third-party guarantee to safeguard lender’s interest in case of default (Besley 1994). Poor people, unable to arrange these collaterals, often remain unsuccessful to access financial services from the formal sector, even if the venture carries an adequate prospect of becoming a profitable initiative (Barslund and Tarp 2008; Mohan 2006; Varghese 2005). Even if a borrower could obtain access to formal credit sources, they can only meet a part of their financial requirements for entrepreneurial activities, leaving a large portion of their consumption need unmet by formal lenders (Fisher and Sriram 2002: 40). This inaccessibility of formal finance or partial fulfilment of credit requirements from the formal lender creates a spillover of unmet demand (Bell et al 1997) into both the informal (for example, moneylenders, middlemen, input dealers, output traders, friends, and relatives) and semiformal (commonly known as microfinance) credit markets. Researchers, such as Sarap (1990), Pal (2002), Sahu et al (2004), and Pal and Laha (2015) in India; Yadav et al (1992) in Nepal; Zeller (1994) in Madagascar; Mohieldin and Wright (2000) in Egypt; Pham and Lensink (2007) in Vietnam; Guirkinger (2008) in Peru; Zhang (2008) in China; and Johnson and Nino-Zarazua (2011) in Kenya and Uganda, employed limited dependent models to assess a rural household’s likelihood of obtaining access to formal credit sources over informal ones. Evidence across countries suggests that access to financial services from formal creditors is confined within the domain of resource-rich households. Further, among the rural households that could get access to formal financial institutions, Pal and Laha (2014) found that distribution of credit offtake was highly skewed towards the resource-rich borrowers.

Theory suggests that access to financial services at affordable terms and conditions for households and firms augments a country’s growth and fosters its economic development (Girma and Shortland 2008). Studying the loan data of 253 small- and medium-sized loanees of an Indian bank, Banerjee and Duflo (2004) found that these firms expanded their businesses once they obtained access to credit facilities. Burgess and Pande (2005) showed that access to formal sources of financial credit substantially contributed to poverty alleviation in India. Klapper et al (2006) found that access to financial services promoted the entry of new firms as well as supported the growth of small enterprises. In the same line, experimental evidence documented by De Mel et al (2008) in Sri Lanka and McKenzie and Woodruff (2008) in Mexico indicated that micro-entrepreneurs who randomly received grants to procure inputs experienced a higher return of 5% to 20% compared to the control group. Kaboré (2009) suggested that targeted credit for rural micro-entrepreneurs carries significant potential for financial uplift of the rural poor. Similarly, Gatti et al (2012), using the annual data of 18 Organisation for Economic Co-operation and Development nations spread over 1980 to 2004, found that competition in the credit market had significantly reduced unemployment. Upon evaluating the Million Baht Village Fund in Thailand, one of the largest government-backed microfinance programmes worldwide, Kaboski and Townsend (2012) found that the programme of increasing credit flow among the villagers increased investment in agriculture, pushed up wages, and augmented income growth in agrarian centres. In the Indian context, Kumar et al (2017) had found corroborative evidence that institutional credit plays a pivotal role in augmenting income of the farming community.

A similar result was also noted by Abu-Bader and Abu-Qarn (2008) and Yang and Yi (2008) that documented a significant causal effect of access to credit on economic growth in Egypt and Korea, respectively. In the same line, Bittencourt (2012) found that the availability of finance encouraged entrepreneurs to invest, thus pushing the economic growth in the Latin American nations. Jedidia et al (2014) also found that access to credit had augmented the economic growth in Tunisia. Similarly, working on the panel data of 40 countries during 1989–2011, Durusu-Ciftci et al (2017) found significant impact of credit on economic growth.

On the contrary, working on a data set of 547 farm households in northern Peru, Guirkinger and Boucher (2008) found that limited access to credit reduced the value of agricultural output to the extent of 26%. Based on primary surveys, a similar trend has been reported by Kumar et al (2013) who found that 74% of Chinese households and 78% of Indian households were compelled to employ a lower level of agricultural input in crop production due to inadequate access to credit. McDermott et al (2014) had shown that the economies that are relatively weak in financial development had been impacted more adversely from natural calamities. In contrast, descending voices, such as Kochar (1999), argue that the lack of access to formal credit does not limit a rural household’s requirement of working capital.


A survey data set of 600 rural households from 14 villages spread over six Indian states was used in this study. The survey was conducted between May and December 2010. The six states include Chhattisgarh, Maharashtra, West Bengal, Andhra Pradesh, Tamil Nadu, and Gujarat. Data were collected under the research project “Assessing Policy Interventions in Agri-business and Allied Sector Credit versus Credit Plus Approach for Livelihood Promotion” undertaken by the Centre for Management in Agriculture at the Indian Institute of Management, Ahmedabad, with financial support from the Government of India’s Ministry of Agriculture and Farmers’ Welfare. For selecting the states, the study used two state-level indicator criteria, namely average population per rural bank branch as on March 2009 and the rural households’ extent of indebtedness as on June 2002.

The survey found that among 600 sample village households, 75 respondents did not have any loan during the corresponding period. Among 525 borrower households, 165 respondents did not avail of any credit facility from the formal financial sector, 323 respondents availed only one loan from formal financial sources, and 37 households availed two loans from formal lenders.

Tables 1A and 1B (p 101) report the descriptive statistics of the variables representing the characteristics of villages and households covered in this survey. Here, the study followed the coding pattern proposed by Walter et al (1987) for ordered independent variables.

Econometric Specification

On average, Indian banks were reported to be offering credit facilities at the nominal interest rate of approximately 11% per annum. Microfinance lenders are believed to be extending microloans at an interest rate ranging between 24% and 36% annually (Basu 2006), which is more than double the price charged by formal lenders. However, loans from informal lenders are made available at a much higher price, occasionally as high as 90% per annum (Ghate 2007). Therefore, a rational consumer would aim for a credit contract with a formal financial agency, as the nominal interest rate charged by formal lenders is much lower than the interest rate associated with loans from the semi-formal and informal sectors. Nevertheless, access to formal financial sources depends on the strength of the signalling capacity of the prospective borrower in terms of their asset holdings and profit potential. Hence, a borrower household from the economically marginalised section of the community with poor asset holdings and limited marketable collateral remained beyond the purview of formal lenders. In the next category, borrower households with relatively better signalling capacity—in terms of asset holdings as well as profit potential—may negotiate a loan from formal lenders. At the top of the hierarchy are borrower households with two loans from the formal sector. The borrower households classified by the number of loans from the formal sector fall into three finite, mutually exclusive, and exhaustive categories (Train 1986): without any loan from the formal sector, one loan from the formal sector, and two loans from the formal sector. The response variable-outreach of formal financial sector (FORMAL), a count data, ranges from zero to two.

Broadly, four alternative count regression models are available (Karazsia et al 2008): Poisson, negative binomial, zero-inflated Poisson (ZIP), and zero-inflated negative binomial (ZINB). The Poisson model fits discrete distribution involving count data (similar to linear regression and easily interpreted); and better if the mean and variance meet the equi-dispersion criterion (Cameron and Trivedi 1986) and the count-dependent variable is non-negative. Unlike the linear model, the Poisson model ensures that the predicted values are non-negative.

The negative binomial does not require equi-dispersion assumption (Cameron and Trivedi 1986) and thus is close to the Poisson model only if the former converges to the latter (implied by equi-dispersion). When the major source of over-dispersion emanates from the multitude of zero count, the ZINB model is more appropriate (Karazsia et al 2008).

The aim of the article is to estimate the probability of an event count. Given the non-negative integer ranging 0–2, the implementation of the Poisson model ensures best fit, provided the data exhibit equi-dispersion, which is a restrictive property.

Following Cameron and Trivedi (1998: 71–72), the study assumes that the count response yi, given the vector of covariates xi, follows the Poisson distribution independently with

The mean number of events per-period (mean parameter) is:

µ= exp (xi'β)

where, β represents a (k+1) × 1 parameter vector. The exponential of xi ensures that µi is non-negative.

The expected number of events per-period (conditional mean) is:

E(yi|xi) = µ= exp (xi'β); the conditional variance equates the conditional mean:

V(yi|xi) = E(yi|xi) = µi

The marginal effect, proportional change in conditional mean E(yi|xi) of βj in response to a one-unit change in the jth regressor, is:

The study estimates the parameters using the maximum likelihood method. The log-likelihood function:

In this study, out of 600 sample households, 75 are non-borrowers, while 525 are borrowers. Among the 525 village households that have access to a credit market, 37 households (that is, 7%) availed two credit facilities from the formal sector during the period 2009–10, while 323 households (that is, 61.5%) negotiated one single loan from the formal sector during the corresponding period. The remaining 165 borrower households did not obtain any loan from formal lenders. Thus, the count data model includes a total of 525 observations, with a dependent variable (FORMAL) categorised into three mutually exclusive, exhaustive, and finite categories (Train 1986: 4).

Following Cameron and Trivedi (1986), the likelihood-ratio test is:

-2(LPoisson - LNegative Binomial) = -2 (-485.73-(-485.73)) = 0,

This satisfies the equi-dispersion requirement.

Now, following Cameron and Trivedi (1998: 71–72), the Poisson regression model is:

y= β1+ β2AGRI + β3EDU + β4FININS++ β5PUCASTE +
β6MEDUDUM + β7FAMTOT + β8LANDD2 + β9MTDUM + β10MIGRATEDUM + β11CGARH + β12MAHA + β13WB + β14 AP + β15TN + ε,

y= 0,1,2 ... (1)

where, CGARHMAHAWBAP, and TN are state dummies of Chhattisgarh, Maharashtra, West Bengal, Andhra Pradesh, and Tamil Nadu (Gujarat state is included as base in the analysis), ε is the error term, and other regressors as explained in Tables 1A and 1B. As the statewise survey was conducted in a village or cluster of villages within a single agroclimatic zone, the study assumes that the within state-surveyed villages would not display much variation in the village-level characteristics. Hence, we have accounted the variation with state-level dummies instead of using village fixed effects.


A borrower from a village situated close to agricultural infrastructure, an educational institution, and a formal financial institution may be more likely to be covered by institutional lenders. Since rural households from villages closer to agricultural infrastructure would carry higher business potential, those households are expected to borrow to expand their crop production or other enterprises. Those households are also expected to be preferred over the households staying farther from the agricultural infrastructure. While proximity to agricultural infrastructure increases business potential, education ensures a wider variability of occupational choice in the non-farm sector (Mishra and Moss 2013) and reduces dependence on agriculture, which is much more prone to seasonality and weather vagaries. Thus, villages that are located in close proximity to educational facilities are expected to become the preferred choice for formal sector lenders. Proximity to a formal financial agency is expected to reduce the cost of supervising the loan contract by formal financial institutions and would minimise transaction costs to both the debtor and the creditor (Burgess and Pande 2005; Witte et al 2015). Thus, the lesser the distance of households from such infrastructure, the more likely are households to have obtained an institutional credit facility. The coefficient of AGRI, EDU, and FININS are expected to carry “-” signs for the borrower household with a single loan from the formal sector as well as the borrower household with two loans from the formal sector.

Further, following Pal (2002), the study assumes that villages with a higher percentage of upper-caste Hindus are expected to be resource-rich households, as, in India, traditionally, resources were held by Hindu communities and may evolve to be an obvious choice of formal lenders. Thus, the coefficient of PUCASTE was expected to show “+” sign.

An educated borrower is expected to be more aware about the loan schemes of the bank and also conversant with the associated procedures (Sahu et al 2004; Stephens and Barrett 2011). Hence, the coefficient of MEDUDUM may carry a “+” sign. Similarly, familiarity with key social leaders, such as the village administrator, bank officials, and extension agents—that is, a household’s strength of social networking—may be a crucial factor in negotiating a loan contract from a formal financial agency and the coefficient of FAMTOT is expected to carry a “+” sign.

Since loan, in rural areas, is provided to farming households who avail credit facilities based on their operational landholding, it that acts as security from the creditor’s perspective (Sarap 1990; Barslund and Tarp 2008; Stephens and Barrett 2011). The creditor remains assured that an agricultural household with larger operational landholding would carry a higher production potential and greater likelihood of timely repayment of loan. Similarly, the holding of mortgageable assets would act as a signal of household’s creditworthiness. Thus, the coefficients associated with LANDD2 and MTDUM are expected to carry a positive sign.

Concerns regarding possible endogeneity in the model app­lication could be as follows. It could be that provided a farmer with less than one hectare of land wants to avail two loans from the formal financial channel, it decides previously to buy one additional hectare of land to strengthen its signalling capacity and hence introduce endogeneity. However, the study did not find any such instance while conducting the survey and during interactions with the sample households. Hence, the possibility of this kind of endogeneity has been ruled out from the model.

Households with a history of migration may not be a preferred set for formal lenders as it would become difficult for the lender to monitor the credit usage in case of a migratory borrower and thus following Sarap (1990), the study also expects that the coefficient associated with MIGRATEDUM may carry “-” sign for the borrowers of the formal sector.

Results and Discussions

The study estimated the model parameters following the full information maximum likelihood that was implemented here employing the genmod procedure of the Statistical Analysis Software 9.2. The estimation of the Poisson regression model is presented in Table 2 (p 103).

With reference to the Poisson regression, as expected, the study finds that the households residing in villages closer to the agricultural infrastructure such as markets, office of extension, service providers, agriculture input retailers, and dealers of farm machinery have evolved to be associated with borrowing multiple loans from the formal lenders. This seems plausible as proximity to agribusiness hubs enhances the business potential of the peasant farmers that in turn raises their demand for loan. On the contrary, a formal lender would find value in extending multiple loans to these farmers considering their high profit potential as well as repayment possibilities. A coefficient of -0.0924 on AGRI indicates that the reduction of the distance of the village of the sample household by one km from the agricultural infrastructure increases the expectation (or mean) number of loans from the formal financial sector increases by a factor of e-0.0924 = 0.9117. The marginal effect of xi (distance kilometer) on μi (expected number of loans from the formal sector when the farmer is from a village at a distance of i km from the agricultural infrastructure) for a one unit decrease in distance from the agricultural infrastructure increases the estimated count by a factor of 0.9117.

Households belonging to the villages in proximity to the educational institutions, as expected, have also evolved as the preferred group for availing multiple loans from the formal lenders. Proximity to the educational institutions would expose households to a wider occupational choice in the non-farm sector. Income from non-agricultural sources would reduce their dependence on agriculture and at the same time would increase the chances of on-time repayment of the loan. A coefficient of -0.1205 on EDU indicates that the reduction of the distance of the village of the sample household by one more km from the educational institution increases the expectation (or mean) number of loans from the formal financial sector increases by a factor of e-0.1205 = 0.8864. The marginal effect of xi (distance km) on µi (expected number of loans from the formal sector when the farmer is from a village at a distance of km from the educational infrastructure) is that for every one-unit decrease in distance from the educational infrastructure, the estimated count increases by a factor of 0.8864.

With reference to the Poisson regression, all the explanatory variables, except FININS, were as expected and statistically significant at the conventional level. The only exception observed was that borrowers from distant locations were also favoured by institutional lenders. The positive sign of FININS’s coefficient is in contrast to the established wisdom that households in proximity to the formal financial institutions would show greater financial inclusion (Burgess and Pande 2005; Witte et al 2015). This finding implies that even an agricultural household residing closer to a formal lender may be excluded from its services unless it holds marketable assets and better production resources. A coefficient of 0.1258 on FININS indicates that a farming household with higher asset holding and better profit potential which is even residing farther by one more km from the formal financial institution, the expectation (or mean) number of loans from the formal financial sector increases by a factor of e0.1258 = 1.1340. The marginal effect of xi (distance i km) on µi (expected number of loans from the formal sector when the farmer is from a village at a distance of km from the formal financial institution) is that for every one-unit increase in distance from the formal financial institution, the estimated count increases by a factor of 1.1340.

Peasant households residing in villages with high percentage of upper-caste Hindus, as expected, are found to be borrowing multiple loans from the formal lenders. As expected, the superior asset holding in terms of cultivable land of the borrowers from the Hindu community made them a preferred choice of the formal creditors. Working on village survey data from the Indian states of Andhra Pradesh and Maharashtra, Pal (2002) reported that households that belonged to upper-caste communities had a higher probability of accessing loans from the formal financial sector. Caste status also evolved to be a significant variable in affecting a household’s probability of accessing formal sector credit in an empirical investigation conducted by Sahu et al (2004) in Odisha.

An educated borrower, as expected, has turned to be a better prospect for availing multiple loans from the formal lenders. Education besides opening the scope of alternative employment opportunity in the non-farm sector also makes a prospective borrower more familiar with the processes involved in availing credit facilities from the formal lenders. Sahu et al (2004) found a similar result in Odisha. Educational quali­fication of class 10 or over was found to have a positive effect on the household’s likelihood of borrowing from formal lenders in Odisha.

An agricultural household familiar with a larger number of key social leaders, such as the head of village administration, government extension officers, headmaster of local school, and officials of financial institutions, has evolved to be the preferred choice of formal lenders for extending multiple loans. A coefficient of 0.1262 on FAMTOT indicates that for every one-unit increase in the number of key social leaders known to the household, the expectation (or mean) number of loans from the formal financial sector increases by a factor of e0.1262 = 1.1344.

A minimum holding of one hectare of land by a peasant household has come out as a significant determinant for availing of one or more loans (that is, two loans) from the formal lenders. On the demand side, household with larger acreage may require a higher amount of money to meet their production expenses. On the supply side, peasant households with a minimum of one hectare of cultivable land may be preferred by lenders over the farmers with less than one hectare of land due to greater revenue and higher profit potential associated with large-sized farm lands. Examining the characteristics of the formal financial sector borrowers, Sarap (1990) also found a similar result where 44.38% of formal credit was received by the households owning large tracts of land; these borrower households could meet 87% of their credit demand from the formal sector itself.

Unsurprisingly, the ability to offer marketable collateral is observed to positively associate with the availing of multiple loans by a peasant household as lenders would have chosen those households who are able to mortgage a marketable asset as collateral. A similar result was also noted by Sahu et al (2004) who revealed that households with more valuable assets, which could be offered as collateral, carried a higher probability of receiving a loan from formal lenders. As expected, borrowers who are of a non-migratory nature are found to be the preferred choice for extending multiple loans by the formal lenders.

The study applies Akaike’s information criterion (AIC) to the nine generated models to examine whether the explanatory variables in the Poisson regression influence the number of loans from the formal lenders (Table 3). It is found that the model proposed in equation (1) with complete set of variables provides the lowest AIC measure among all the models. The entropy or loss of information is the minimum when the complete set of regressors are applied. This confirms that the model outcome (Table 3) best explains the probability of the event count.

Next, to help assess the goodness-of-fit of the Poisson model in equation (1), the study conducts the goodness-of-fit chi-square test which assumes that deviance (that is, 228.4546 in this case) follows a chi-square distribution with degrees of freedom equal to the model residual (that is, 510). As the goodness-of-fit test is not statistically significant, the study concludes that the model fits reasonably well.

Cameron and Trivedi (2009) recommend using robust standard errors for the parameter estimates to control for mild violation of the distribution assumption that the variance equals the mean (that is, equi-dispersion). The robust standard errors for the Poisson regression coefficients are given in Appendix I (p 106). The study finds that though robust standard errors are marginally different from that of the original standard errors, estimates remain unchanged. This establishes that the Poisson regression model in equation (1) is robust for mild violation of the equi-dispersion criterion.

Conclusions and Policy Implications

This paper documents a household-level empirical investigation conducted to assess the determinants of financial sector outreach in 14 villages. The study found that households residing in villages that were close to critical infrastructures, such as educational institutions and agricultural infrastructure, avail of multiple loans from the formal sector. Furthermore, households belonging to the upper-caste Hindu community, holding a minimum of one hectare of operational land, having mortgageable assets, being familiar with key social leaders, and of a non-migratory nature availed of multiple loans from formal lenders. Interestingly, the study found that relatively resource-rich rural households with better asset holding and higher profit potential, even staying at distant locations from bank branches, had availed of one or two loans from formal lenders.

The findings of this study have the following policy implications: First, opening new branches may be necessary but not a sufficient condition to attain deeper financial outreach. With the withdrawal of service area approach, when the banks are not constrained to serve within a limited area, they are lending at distant locations given the loan is adequately covered with collateral and borrowers are found to be of higher profit potential.

Second, the limited financial sector outreach in rural areas may be an outcome of the information asymmetry between lenders and borrowers. Formal lenders were found to be extending multiple loans to village households, offering collateral in lieu of covering any possible default in the future. This seems possible as a large proportion of the rural branches of commercial banks and cooperatives were operated by a single person. Hence, rural bank officials claimed to be overburdened to canvass new loan proposals as well as monitor existing loan accounts (Reserve Bank of India 2011).

Third, the central bank of India may advise commercial banks to appoint independent agents, preferably local people with good social standing on payment basis, to help the bank branches in providing banking services to the rural poor. On the one hand, local banking agents would be in a better position to collect information regarding prospective borrowers, while on the other, they would be able to monitor loan accounts. Hence, both the jobs of screening and monitoring of loan accounts could be transferred to local banking agents who are closer to the target group.

Finally, offering credit with complementary services, such as input provision, output procurement, and extension services, may relax the formal lender’s demand for collateral.

In the future, this research can be extended to other developing nations and more indicators of financial sector outreach such as the number of deposit accounts per rural household.


Abu-Bader, S and A S Abu-Qarn (2008): “Financial Development and Economic Growth: The Egyptian Experience,” Journal of Policy Modeling, Vol 30, No 5, pp 887–98.

Banerjee, A V and E Duflo (2004): “Do Firms Want to Borrow More? Testing Credit Constraints Using a Directed Lending Program,” Discussion Paper 4681, Centre for Economic Policy Research.

Barslund, M and A Tarp (2008): “Formal and Informal Rural Credit in Four Provinces of Vietnam,” Journal of Development Studies, Vol 44, No 4, pp 485–503.

Basu, P (2006): Improving Access to Finance for India’s Rural Poor, Washington, DC: World Bank.

Beck, T, A Demirguc–Kunt and M S M Peria (2007): “Reaching Out: Access to and Use of Banking Services across Countries,” Journal of Financial Economics, Vol 85, No 2, pp 234–66.

— (2008): “Banking Services for Everyone? Barriers to Bank Access and Use Around the World,” World Bank Economic Review, Vol 22, No 3, pp 397–430.

Bell, C, T N Srinivasan and C Udry (1997): “Rationing, Spillover, and Interlinking in Credit Markets: The Case of Rural Punjab,” Oxford Economic Papers, Vol 49, No 4, pp 557–85.

Besley, T (1994): “How Do Market Failures Justify Interventions in Rural Credit Markets?” World Bank Research Observer, Vol 9, No 1, pp 27–47.

Bittencourt, M (2012): “Financial Development and Economic Growth in Latin America: Is Schumpeter Right?” Journal of Policy Modeling, Vol 34, No 3, pp 341–55.

Brewer, B E, C A Wilson, A M Featherstone and M R Langemeier (2014): “Multiple vs Single Lending Relationships in the Agricultural Sector,” Agricultural Finance Review, Vol 74, No 1, pp 55–68.

Burgess, R and R Pande (2005): “Do Rural Banks Matter? Evidence from the Indian Social Banking Experiment,” American Economic Review, Vol 95, No 3, pp 780–95.

Cameron A C and P K Trivedi (1986): “Econometric Models Based on Count Data: Comparisons and Applications of Some Estimators,” Journal of Applied Econometrics, Vol 1, No 1, pp 29–53.

— (1998): Regression Analysis of Count Data, Cambridge: Cambridge University Press.

Cameron, A C and P K Trivedi (2009): Microeconometrics Using Stata, College Station, TX: Stata Press.

De Mel, S, D McKenzie and C Woodruff (2008): “Returns to Capital in Microenterprises: Evidence from a Field Experiment,” Quarterly Journal of Economics, Vol 123, No 4, pp 1329–72.

Durusu-Ciftci, D, M S Ispir and H Yetkiner (2017): “Financial Development and Economic Growth: Some Theory and More Evidence,” Journal of Policy Modeling, Vol 39, No 2, pp 290–306.

Fisher, T and M S Sriram (2002): Beyond Micro-credit: Putting Development Back into Micro-Finance, New Delhi: Sage Publications (Vistaar Imprint).

Fletschner, D, C Guirkinger and S Boucher (2010): “Risk, Credit Constraints and Financial Efficiency in Peruvian Agriculture,” Journal of Development Studies, Vol 46, No 6, pp 981–1002.

Gatti, Donatella, Christophe Rault and Anne-Gaël Vaubourg (2012): “Unemployment and Finance: How do Financial and Labour Market Factors Interact?” Oxford Economic Papers, Vol 64, No 3, pp 464–89.

Girma, S and A Shortland (2008): “The Political Economy of Financial Development,” Oxford Economic Paper, Vol 60, pp 567–96.

Ghate, P (2007): “Consumer Protection in Indian Microfinance: Lessons from Andhra Pradesh and the Microfinance Bill,” Economic & Political Weekly, Vol 42, No 13, pp 1176–84.

Guirkinger, C (2008): “Understanding the Coexistence of Formal and Informal Credit Markets in Piura, Peru,” World Development, Vol 36, No 8, pp 1436–52.

Guirkinger, C and S R Boucher (2008): “Credit Constraints and Productivity in Peruvian Agriculture,” Agricultural Economics, Vol 39, No 3, pp 295–308.

Johnson, S and M Nino-Zarazua (2011): “Financial Access and Exclusion in Kenya and Uganda,” Journal of Development Studies, Vol 47, No 3, pp 475–96.

Jedidia, K B, T Boujelbène and K Helali (2014): “Financial Development and Economic Growth: New Evidence from Tunisia,” Journal of Policy Modeling, Vol 36, No 5, pp 883–98.

Kaboré, S T (2009): “Effectivité d’un Crédit Ciblé Aux Vauvres: Le Cas Des Microentreprises Rurales du Burkina Faso,” Canadian Journal of Development Studies, Vol 29, Nos 1–2, pp 215–33.

Kaboski, J P and R M Townsend (2012): “The Impact of Credit on Village Economies,” American Economic Journal: Applied Economics, Vol 4, No 2, pp 98–133.

Karazsia, B T and M H Van Dulmen (2008): “Regression Models for Count Data: Illustrations Using Longitudinal Predictors of Childhood Injury,” Journal of Pediatric Psychology, Vol 33, No 10, pp 1076–84.

Kochar, A (1999): “Does Lack of Access to Formal Credit Constrain Agricultural Production? Evidence from the Land Tenancy Market in Rural India,” American Journal of Agricultural Economics, Vol 79, No 3, pp 754–63.

Klapper, L, L Laeven and R Rajan (2006): “Entry Regulation as a Barrier to Entrepreneurship,” Journal of Financial Economics, Vol 82, No 3, pp 591–629.

Kumar, A, A K Mishra, S Saroj and P K Joshi (2017): “Institutional versus Non-institutional Credit to Agricultural Households in India: Evidence On Impact from A National Farmers’ Survey,” Economic Systems, Vol 41, No 3, pp 420–32,

Kumar, C S, C G Turvey and J D Kropp (2013): “The Impact of Credit Constraints on Farm Households: Survey Results from India and China,” Applied Economic Perspectives and Policy, Vol 35, No 3, pp 508–27.

Mohan, R (2006): “Agricultural Credit in India: Status, Issues and Future Agenda,” Economic & Political Weekly, Vol 41, No 11, pp 1013–21.

McKenzie, D J and C Woodruff (2008): “Experimental Evidence on Returns to Capital and Access to Finance in Mexico,” World Bank Economic Review, Vol 22, No 3, pp 457–82.

McDermott, T K J, F Barry and R S J Tol (2014): “Disasters and Development: Natural Disasters, Credit Constraints, and Economic Growth,” Oxford Economic Papers, Vol 66, No 3, pp 750–73.

Mishra, A K and C B Moss (2013): “Modeling the Effect of off-farm Income on Farmland Values: A Quantile Regression Approach,” Economic Modelling, Vol 32, No 5, pp 361–68.

Mohieldin, M S and P W Wright (2000): “Formal and Informal Credit Markets in Egypt,” Economic Development and Cultural Change, Vol 48, No 3, pp 657–70.

Pal, S (2002): “Household Sectoral Choice and Effective Demand for Rural Credit in India,” Applied Economics, Vol 34, No 14, pp 1743–55.

Pal, D and A K Laha (2014): “Credit Off-take From Formal Financial Institutions in Rural India: Quantile Regression Results,” Agricultural and Food Economics, Vol 2, No 9, pp 1–20.

— (2015): “Sectoral Credit Choice in Rural India,” Journal of Choice Modelling, Vol 14, No 1, pp 1–16.

Pham, T T T and R Lensink (2007): “Lending Policies of Informal, Formal and Semi-formal Lenders: Evidence from Vietnam,” Economics of Transition, Vol 15, No 2, pp 181–209.

Reserve Bank of India (2011): “RBI Central Board Meets Today at Agartala; Governor Meets State Government and Bankers,” Press Release, 19 May.

Sarap K (1990): “Factors Affecting Small Farmers’ Access to Institutional Credit in Rural Orissa, India,” Development and Change, Vol 21, No 2, pp 281–307.

Sahu, G B, S Madheswaran and D Rajasekhar (2004): Credit Constraints and Distress Sales in Rural India: Evidence from Kalahandi District, Orissa,” Journal of Peasant Studies, Vol 31, No 2, pp 210–41.

Sriram, M S (2007): “Productivity of Rural Credit: A Review of Issues and some Recent Literature,” International Journal of Rural Management, Vol 3, No 2, pp 245–68.

Stephens, E C and C B Barrett (2011): “Incomplete Credit Markets and Commodity Marketing Behavior,” Journal of Agricultural Economics, Vol 62, No 1, pp 1–24.

Stiglitz, J and A Weiss (1981): “Credit Rationing in Markets with Imperfect Information,” American Economic Review, Vol 71, No 3, pp 393–410.

Train, K (1986): Qualitative Choice Analysis: Theory, Econometrics, and an Application to Automobile Demand, Cambridge: MIT Press.

Varghese, A (2005): “Bank-moneylender Linkage As an Alternative to Bank Competition in Rural Credit Markets,” Oxford Economic Papers, Vol 57, No 2, pp 315–35.

Walter, S D, A R Feinstein and C K Wells (1987): “Coding Ordinal Independent Variables in Multiple Regression Analyses,” American Journal of Epidemiology, Vol 125, No 2, pp 319–23.

Witte, T, E A DeVuyst, B Whitacre and R Jones (2015): “Modeling the Impact of Distance between Offices and Borrowers on Agricultural Loan Volume,” Agricultural Finance Review, Vol 75, No 4, pp 484–98.

Yadav, S, K Otsuka and C C David (1992): “Segmentation in Rural Credit Financial Markets: The Case of Nepal,” World Development, Vol 20, No 3, pp 423–36.

Yang, Y Y and M H Yi (2008): “Does Financial Development Cause Economic Growth? Implication for Policy in Korea,” Journal of Policy Modeling, Vol 30, No 5, pp 827–40.

Zeller, M (1994): “Determinants of Credit Rationing: A Study of Informal Lenders and Formal Credit Groups in Madagascar,” World Development, Vol 22, No 12, pp 1895–1907.

Zhang, G (2008): “The Choice of Formal or Informal Finance: Evidence from Chengdu, China,” China Economic Review, Vol 19, No 4, pp 659–78.


Dear Reader,

To continue reading, become a subscriber.

Explore our attractive subscription offers.

Click here

Updated On : 17th Jul, 2022
Back to Top