Essay Writing Service

1 Star2 Stars3 Stars4 Stars5 Stars (No Ratings Yet)

Rural Microcredit Accessibility by Households: Empirical Evidence from Pakistan

do not necessarily reflect the views of

Rural Microcredit Accessibility by Households: Empirical Evidence from Pakistan
This paper examines the factors influencing the accessibility of microcredit by rural households in Pakistan. The empirical analysis utilizes logit model, with empirical data collected through interviewing 600 households, including both non-borrower and borrower households, a survey conducted between January 2017 and March 2017 in Punjab Province of Pakistan.  A total of eleven household-level factors are defined as the determinants of households’ access to microcredit, including education level, household size, income, and others. In addition to these, results indicate that rural households’ accessibility to credit can also be diminished by the supply-side factors (e.g., loan processing time, interest rates and collaterals). The empirical analysis establishes a positive relationship between households’ credit demand and access to credit. The paper thus concludes that households should be encouraged to raise capital requirements (for example, create investment opportunities in on/off farm activities) to increase their demand for credit, which can enhance their access to microcredit. In addition, microcredit institutions such as ZTBL (Zarai Taraqiati Bank Limited) should improve their lending schemes and micro loan products to better suit the diversified needs of the rural population.
Keywords: Microcredit, Accessibility, Rural household, Logit Model, Pakistan
1. Introduction
Microcredit has been progressively accepted as a powerful tool to help poor people invest and break out of the ‘vicious cycle’ of poverty because it has the potential to improve user’s incomes and savings, and consequently, enhance capital accumulation and reinforce high incomes (Atieno 2001Montgomery 2005). Despite the importance of credit in helping the poor to improve their welfare, poor people are excluded from formal financial systems. Such exclusion ranges from partial exclusion in developed countries to full or nearly full exclusion in less developed countries (LDCs) (Brau and Woller 2004Haq 2008). Traditional financial institutions (FIs) are reluctant to serve the poor mainly because these people fail to meet the selection criteria such as the requirement of physical collateral set by MFIs. At one end of the spectrum lie studies that have concluded that microfinance is a positive and effective mechanism for poverty reduction (Bateman 2008Imai and Azam 2012). To address the problems of adverse selection and moral hazards arising from asymmetric information between banks and borrowers, banks usually attach collateral requirements to loans. Collateral is used to assist in determining credit worthiness, as well as solving the incentive and enforcement problems (Klein, Meyer et al. 1999Imai, Arun et al. 2010Karlan and Zinman 2011). The collateral requirement becomes more stringent when the borrower is resource-poor. The perceived high risks and costs arising from processing and servicing unsecured small loans make MFIs shy away from financing the poor, mainly due to the concern of financial viability.
Like most Asian developing countries, the majority of the poor population in Pakistan dwell in rural areas.  There are 191.71 million people living in Pakistan, 29.5% ‘relatively’ poor people surviving on less than a dollar per day in rural Pakistan, with the rural population living in both ‘absolute poverty’ and ‘relative poverty’ accounting for 30 million of the total rural population. Moreover, rural incomes are just 30% of the urban average, which presents a wide gap between the rural and urban living standards (Pakistan Bureau of Statistics 2016). Inability to acquire formal credit support has constrained poor farmers’ ability to expand their production and improve their living conditions. However, farmers need credit support to meet their living needs, including the purchase of durable goods, daily consumption, and festivals and ceremonies. More importantly, affordable agricultural credit allows farmers to adopt new technologies, which provide them with potential economic opportunities to improve production and income. Failing to access formal credit, most farmers have to resort to informal borrowing which is typically offered at higher interest rates. Despite the high-interest rates charged by informal lenders, approximately 50–60% of rural households in Pakistan still rely on informal credit for their consumption and production (Williams, Shahid et al. 2016). However, the high interest charged on informal loans has increased farmers’ indebtedness and further kept most of these households trapped in poverty.
Microcredit was introduced into Pakistan as part of the government’s poverty alleviation strategies in the mid-2000s, aiming to ameliorate rural poverty through a financially sustainable approach. In terms of different providers, Pakistan microcredit programs can be categorized into three types. The first type includes experimental microcredit projects provided by NGOs and quasi-official organizations, aiming to explore the feasibility, operating capabilities and policy implications of microcredit in Pakistan; the second type focuses on poverty alleviation carried out by government agencies; the third type Centre’s on minimizing credit constraint in the rural areas of Pakistan and is operated by rural financial institutions such as ZTBL (Zarai Taraqiati Bank Limited) (Kouser and Saba 2012). Since implementing microcredit programs in 2000, ZTBLs have quickly expanded their microcredit activities with an extensive network in rural areas and has taken the leading role in popularizing and formalizing microcredit in Pakistan.
With the implementation of microcredit, Pakistan has boosted lending to farmers in recent years. In 2000, the Pakistan Poverty Reduction Fund (PPAF) was launched to fight poverty through the provision of funds through NGOs and community organizations. Initially, PPAF has signed agreements with five partner organizations to disburse 5 billion PKR (Pakistani rupee) over the next five years. These partner organizations include the Aga Khan Rural Support Program, Pakistan Family Planning Association, the National Rural Support Project, Taraqee Trust and the Kashf Foundation (Khan, Shirazi et al. 2009Ghalib, Malki et al. 2011). Under the agricultural lending support from State Bank of Pakistan (SBP), the main funding source for ZTBLs’ microcredit programs, ZTBLs have substantially developed their microcredit programs and evolved as the largest microcredit providers serving the grassroots level in rural Pakistan. However, in spite of the strong efforts made by the Pakistani government to facilitate credit access in rural areas, there is evidence to show that a large number of poor households are regarded as marginalized in their villages and do not have access to microcredit because of their weak social and economic conditions. In addition, women in rural Pakistan are still disadvantaged in accessing any form of formal credit, including microcredit, frequently having to use their husbands’ names to apply for microcredit loans (Mahmood 2011). A relevant question arises: What kind of factors, household-level or institution-level, are likely to influence rural households’ access to microcredit in Pakistan? Unfortunately, few empirical studies have been conducted to address this question.
This paper aims to empirically analyze households to identify the key factors affecting the access to microcredit. The microcredit program studied in this paper is carried out by the ZTBL, the largest microcredit provider in Pakistan. Outperforming the programs operated by NGOs and government agencies in terms of outreach and financial sustainability, ZTBL’s microcredit program is the most prevalent type in rural Pakistan. The remainder of the paper is organized as follows. Section 2 provides an overview of the rural credit market in Pakistan. Section 3 discusses the research methods and data collection. The empirical results are discussed in Section 4, with concluding remarks presented in Section 5.
2. The rural microcredit market in Pakistan
The rural microcredit market in Pakistan shares features with those found in many other developing countries: the market is fragmented, where formal and informal credit sources coexist; formal credit is highly regulated but difficult to access by rural households; informal lending is more readily accessed but always appears to be clandestine and considered illegal (Jia, Heidhues et al. 2007). In addition to a re -vitalization of the cooperative credit system and the introduction of commercial banks to agricultural lending, the mandate of the Agricultural Development Bank of Pakistan (ADBP), established in 1961, was also changed to cater to the credit needs of small farmers and the non-farm sector in rural areas. The ADBP has five loan windows, i.e., Development Loans, Production Loans, Agri-business Loans, Cottage Industry Loans, and Off-farm Income Generating Activity Loans. The ADBP was also persuaded by the government to evolve numerous special credit schemes to improve the access of special groups to credit (Qureshi and Hai 1995).
However, the RFIs have been heavily criticized for being unable to satisfy the various credit needs of rural households in Pakistan. Access to institutional credit by rural households remains constrained. Such constraint can be partly attributed to the insufficient credit supply by RFIs in rural Pakistan. The credit insufficiency mainly arises from the lack of RFIs providing financial services to farmers. As a result, the ZTBL is only RFI serving the grassroots of rural society with the provision of financial services. There are only 461 ZTBLs branches across the country and the credit supply by ZTBLs is inadequate to meet the considerable credit demand required by the enormous rural households in Pakistan (Baig and Bilal 2017).
Apart from the shortage of credit, households’ accessibility to formal credit has been severely weakened by the lending terms and procedures set by the RFIs. The collateral requirement is the most frequently noted obstacle that prevents poor households from accessing formal credit.
The land is always a preferred form of collateral in formal agricultural lending. However, the lack of land ownership equals the lack of proper collateral, which makes formal credit inaccessible to Pakistan’s farmers (Saqib, Kuwornu et al. 2017). In addition to the lack of appropriate collateral, the high borrowing costs borne by the Pakistani farmers keep them away from formal credit. Other than loan interest, farmers’ borrowing costs consist of the time spent on traveling and on loan applications, gifts and kickbacks to loan officers, and the membership fees (Saqib, Ahmad et al. 2016). The long and complicated loan application procedures have often dampened farmer’s willingness since they tend to jeopardize productive investment opportunities when quick credit is required (Hussain and Thapa 2016). It is also quite common for loan applicants to invite loan officers to banquets and/or give kickbacks directly to loan officers for loan approvals. In the case of ZTBLs, households have to pay membership fees (usually $ 4-6) to ZTBLs before they can lodge their loan applications (Khan, Ishaq et al. 2017).
Failing to secure credit support from formal financial institutions, the majority of poor farmers have to fall back on informal sources to meet their credit needs. Informal credit in Pakistan includes loans obtained from noncommercial sources such as friends, relatives, and acquaintances, and loans from private lending and borrowing organizations (PLBs), such as professional moneylenders, traders, pawnbrokers, and usurers. Such organizations are the dominant source of informal finance in rural Pakistan (Ghosh and Ray 2016). While farmers’ credit needs for daily consumption can be met by borrowing from their friends and relatives free of charge, the needs for production are largely met by borrowing from PLBs with high-interest rates (Khan, Qadeer et al. 2016). Compared to formal financing, informal financing possesses some advantages, such as close personal relationships with clients, flexibility, rapidity and low transaction costs, which make
informal finance either the exclusive or the preferred credit source in rural areas, despite exploitative interest rates (Farooq, Anwar et al. 2016). However, informal lenders normally depend on personal funds and the limited resources restrict the extent to which the informal lenders can effectively and sustainably satisfy the credit needs of their borrowers. The limited credit supply by informal lenders then leads to either severe credit constraints or usurious loans for some borrowers (Atieno 2001).
Informal finance remains controversial in Pakistan’s rural financial marketplace. On the one hand, there are opponents who traditionally regard informal finance as a violation of normal financial discipline, despite its contribution to meeting farmers’ urgent financial needs. The evidence supporting such argument is that the Chinese government never gives overt recognition to the legal existence of the informal sector, Thus the development of informal credit is generally clandestine and not under the government’s supervision (Saleem 2016). Opponents suggest excluding informal credit from rural financial markets by improving the lending operations of formal financial institutions to provide more loans in favor of rural households. This, they argue, is crucial in establishing a sound rural financial system and maintaining the sustainable development of Pakistan’s rural economy. However, advocates of informal finance contend that the existence of informal credit in Pakistan reflects the imperfections of Pakistan’s formal rural financing system, which is characterized as unable to meet the diverse capital demands of the rural households. If no changes are made in the current situation, the persistence of informal credit will be both necessary and rational in view of the credit facilities provided to the farmers (Kashif, Zafar et al. 2016).
3. Research method and data collection
3.1. Conceptual framework and empirical model
A household’s accessibility to credit can be defined as the ability to borrow from different sources of credit (Diagne 1999Diagne and Zeller 2001). Evans, Adams, Mohammed, and Norris (Evans, Adams et al. 1999) present a comprehensive conceptual framework for analyzing factors that affect households’ accessibility to microcredit in Bangladesh, in which both household-related factors and program-related factors are taken into account. Similarly, Maddala (2001) examines households’ accessibility to rural credit in Northern Nicaragua by analyzing both demand-side (households) factors and supply-side (lenders) factors. The previous studies show evidence that when examining access to credit, both household-level factors and institution-level factors should be taken into account. Following the literature.
This paper employs Evans, Adams et al. (1999) conceptual framework to investigate households’ accessibility to microcredit in rural Pakistan by focusing on the microcredit program implemented by the ZTBL (Zarai Taraqiati Bank Limited).
Household-related factors (such as income, occupation, age, education) are hypothesized to affect households’ demand for microcredit, which can directly influence households’ accessibility to microcredit. This is because households’ access to a certain type of credit can be conceptualized as a sequential decision making process that is initiated at the demand side (Zeller 1994). In addition to household-related factors, there are program-related (supply-side) factors influencing the households’ access to microcredit too. For example, Umoh (2006) argues that the inaccessibility to credit is generally created by the lending policies of financial institutions, which can be manifested by complicated application procedures, specified minimum loan amounts and prescribed loan purposes. In addition, some features unique to microcredit programs can also constrain households’ access to microcredit, including membership requirement, self-selected credit group, and group lending see for example (Maes and Foose 2006). Institutional incentives such as achieving repayment targets and ensuring program financial viability may induce lenders to shy away from lending to households that are or appear to be risky borrowers (Evans, Adams et al. 1999Maes and Foose 2006). Due to the supply-related factors, households that have a demand for microcredit may access microcredit or stay frustrated by denial. Therefore, household-related factors and program-related factors, singly or in combination, can work to impact households’ accessibility to microcredit.
This paper attempts to measure households’ accessibility to microcredit by empirically examining the influence of household factors on the probability of securing micro loans from the ZTBLs. Data used in the empirical analysis includes primary data collected from a rural household survey in Pakistan (data collection is discussed in the subsequent section). The influence of institution-level factors (i.e., supply side factors) on households’ accessibility to microcredit is examined descriptively with qualitative information collected from the household survey. Formal loans from ZTBL are always provided with collateral requirements while informal loans offered by PLBs are normally charged at exploitative interest rates. In addition, it was found in the survey that households usually are not willing to borrow from friends and/or relatives for interest-free loans because they feel indebted and may have to reciprocate to the lender in the future.
Therefore, this paper assumes that rural households in Pakistan prefer microcredit to other credit types such as formal credit and informal credit when they need to borrow, due to the merits of microcredit such as small, medium and large loans to very poor people for self-employment and affordable interest rates (ZTBL’s micro loans are provided at commercial rates).
Previous studies have identified a variety of household-level factors that influence households’ ability to access a certain type of credit. For example, Mohamed (2003) conducted an empirical study examining the accessibility to formal and quasi-formal credit by farmers in Zanzibar, where socio-economic characteristics of rural households such as age, gender, education attainment, and income level are identified as determinants affecting farmers’ access to formal credit. In addition
to age, gender, and education level, Okurut (2006) found that household characteristics such as residence location, family size, and household expenditure also have significant effects on households’ access to different types of credit (formal, semiformal and informal) in South Africa. Maddala (2001) further pointed out that household access to networks of recommendation and information plays a crucial role in obtaining formal credit by households.
In our study, household variables encompass household demographics (such as age and gender), socio-economic factors (such as income level and assets ownership) and other household-related factors (such as attitude towards debt and ability to access other sources of credit). Table 1 presents the definitions of variables used in the empirical model.
The empirical approach used to analyze accessibility to microcredit from the perspective of rural households is based on binary choice models which describe the probability of households’ choice between two mutually exclusive alternatives (accessing or not accessing) according to their evaluations of the utilities of these two choices (Train Kenneth 2003Umoh 2006). Let Un (Yn, Xn) be the utility function of household n, where Yn is a dichotomous variable denoting whether the household has access to microcredit (1 if yes; 0 otherwise); Xn is a vector of household characteristics. The household will choose to borrow from a microcredit program if such a choice implies a higher level of utility compared to not borrowing:
U1n (Y= 1, Xn) > U0n (Yn = 0, Xn) [or U1n (Yn = 1, Xn) U0n (Yn = 0, Xn) > 0]               (1)
Consequently, the probability that household n chooses to access microcredit can be written as:
P(Y= 1) = Pr(U1n >U0n)         (2)
Logit and probit are two binary choice models commonly used in analyzing households’ accessibility to credit in the literature. For example, Chaudhary and Ishfaq (2003) and Maddala (2001) employed the logit model to examine the relative importance of household factors in determining the probability of accessing different types of credit, while Okurut (2006) and Umoh (2006) opted for the probit model for their empirical analyses. Both logit and probit models provide consistent, efficient, and asymptotically normal estimates, and yield very similar prediction results in empirical work. Instead of trying to determine the household’s choice.
This paper utilizes the observed information of household’s choice (borrow or not borrow) and household’s characteristics to estimate the probability of the household’s choice conditional on the household characteristics using the logit model, owing to the merits possessed by the model such as good approximation to the normal distribution and analytical convenience (Ben-Akiva and Lerman 1985Train Kenneth 2003). The empirical model is specified as follows:
Pn (Yn = 1) = 1
1[1+exp⁡α+βXn ]        (3)
where: α is a constant term; β is a vector of coefficients for the independent variables Xn; Yn is the dependent variable if the household has secured microcredit from ZTBL equal to 1 and 0 otherwise; Pn is the estimated probability of a household having access to microcredit.
Table 1. Description of variables used in a logit model.

Variable name             Variable type                                                  Variable description

AGE   Continuous                Age of household head (in years)
GEND   Binary         Gender of household head (1 = male, 0 = female)
HHSZ   Continuous  Household size (in numbers)
EDU   Continuous  Educational level of household head (in years)
INCOME   Continuous  Household annual income (in 1000 PKR(a))
ASSET   Continuous  Total value of household assets(b) (in 1000 PKR)
FARMSZ  Continuous  Size of household farmland (in acres)
EDR    Continuous   Ratio of household members without income to household income
SELFEMPL   Binary    Household head’s involvement in self-employment (1 = yes, 0
OFFICIAL   Binary    Family member working as village or township officials (1 = yes,
0 otherwise)
SHAREHLD   Binary    Household owning shares of ZTBL (1 = yes, 0 otherwise)
SAV    Binary    Household savings with ZTBL (1 = yes, 0 otherwise)
Other variables
ALTER    Binary    Access to other credit sources (1 = yes, 0 otherwise)
ATTITUD   Binary    Attitude towards debt (1 = averse, 0 otherwise)
DIST       Distance between household residence and ZTBL branch office
DIST1    Binary    1 = within 5 kms, 0 otherwise
DIST2    Binary    1 = between 6 and 10 kms, 0 otherwise
DIST3    Binary    1 = more than 10 kms, 0 otherwise

a PKR (Pakistani rupee). 
b The household asset values exclude house values and farmland values.

Eq. (3) represents the cumulative logit distribution function in a nonlinear form, which gives rise to difficulty in interpreting the coefficients. For the purpose of interpretation, it is normal to write the model in terms of log-odds ratio (Maddala 2001Abdou, El-Masry et al. 2007). With a logit transformation, the estimated model becomes a linear function of the explanatory variables,
which is expressed as follows:
logit[Pn (Yn = 1)] = log
Pn1Pn= α + βXn        (4)
where: α is a constant term; β is a vector of coefficients for the independent variables Xn; Xn is a vector of independent variables (see Table 1), including household’s demographics, socioeconomic characteristics, and other household-related factors.
3.2. Data collection
The data include primary measurements collected through a rural household survey which was conducted between January 2017 and March 2017 in Punjab Province in Pakistan. Punjab Province is one of the major agricultural provinces in Pakistan, where farmers are geographically distributed plain areas and produce various agricultural crops, aquatic products, and livestock. In addition, the rural population in Punjab Province comprises different ethnic minorities, such as the Christianity, Sikhs, and Hindus. The minority population in Punjab Province is around three million (PBS2016).
The sample was drawn from the rural areas in Punjab Province and included households from different ethnic groups and at different levels of income.
Therefore, studying rural households in Punjab Province allows for making similar comparisons between the Pakistani rural population based on characteristics such as income inequality and multi ethnic groups. There is a total of 253 ZTBL branches located in districts throughout Punjab and all have been engaged in micro-financing since ZTBL initiated microcredit program in the Province in 1984. According to the statistics from the ZTBL Punjab Head Office, The GOP (Government of Pakistan) supports the microfinance sector to stimulate agriculture, GOP provides the amount of 90.0 million PKR (8.58 million USD) to ZTBL. Among MFIs, ZTBL is the largest provider of microfinance in Pakistan. It provides credit facilities for more than half a million borrowers per year.
C:UsersRana WaqasDesktopPunjab-Pakistan-Map.png
Figure 1: Map of Punjab showing the 36 districts of the province
An organized survey questionnaire was used to elicit relevant household information, such as age, gender, household size, which is used in the logit model to identify key household-level factors that influence microcredit accessibility among rural populations. In addition, the questionnaire gathered qualitative information, such as knowledge of the ZTBL microcredit program, reasons for not applying for micro loans as well as for loan rejection, etc., for the purpose of investigating the influence of other factors (such as institutional factors) on households’ accessibility to microcredit.
A multi-stage stratified random sampling technique was applied to draw the household sample. In the first stage of the sampling process, sample districts selected on the basis of the availability of the ZTBL microcredit program. A list of districts obtained from ZTBL Punjab Head Office, indicating in which districts a ZTBL microcredit programs were available, as well as the program operation duration and geographic location of these districts. Due to time and resources constraints, only four districts selected from the 36 districts hosting the ZTBL microcredit program. The selected districts were Muzaffargarh, D G Khan, Rajanpur, and Lodhran basing on the percent population living below the poverty line.

Sampling Method
Randomly selected
5 villages from
each district
30 respondents
from each village
Total non-borrower and
borrower respondents
Total male and female
Figure 2: Schematic presentation of sampling frame for the study
Following the selection of sample districts, sample villages were selected. A total of five villages from each selected districts were randomly chosen from a list of villages (the list took the administrative office in each selected district). A total of 20 villages included. The selection of sample households was accomplished in the final stage of the sampling process. The household selection included two steps: the first was to select households that have accessed ZTBL’s microcredit (namely borrowers). Based on the borrower list obtained from each ZTBL branch office in the selected district, a total of 381 borrowers were randomly chosen to participate in the interview. Subsequent to the selection of borrowers, another 219 households who had never secured ZTBL’s microcredit (namely non-borrowers) were randomly selected from a list of rural households obtained from the village committee office in each selected village. Overall, 600 households were included in the sample and all respondents were heads of households.
4. Results and discussion
4.1. Characteristics of the rural households
Table 2 summarizes the household characteristics used in the analysis for the whole sample according to the status of respondents’ access to microcredit. Student’s t-test was used to determine whether the mean values of household variables between borrowers and nonborrowers were statistically different. Chi-square was used to test for relationships between the non-metric household variables and access to microcredit. The t-test results are not statistically significant at the ten percent level, except for ASSET. This demonstrates that the mean value of assets owned by nonborrowers is significantly higher than that of assets owned by borrowers. Moreover, rural households’ access to credit is intensely associated with EDU, GEND, FARMSZ, SAV, SELFEMPL, DIST, ATTITUD, and ALTER because the chi-square tests on these variables were all significant at the ten percent level or better. Table 2 provides a description of the sample. The influence of the household variables and their significance on households’ credit accessibility are analyzed with by the logit model, with results reported in Table 3.
Out of the 600 sampled household heads, 381 are microcredit borrowers of the ZTBL and mainly consist of males (see Table 2). The survey respondents are divided into three groups with respect to educational attainment, including without education, secondary school education or less, and post-secondary education. The data in Table 2 shows that the proportion of the borrowers with no education is only 3%, much lower than that for the non-borrowers (11%). Around 90% of the borrowers and 79% of the non-borrowers have acquired a secondary education or less (including primary, middle and high school). However, in terms of post-secondary education (college and university), the non-borrowers appear to be better educated than the borrowers (10% versus 7%).
Table 2. Profile of the respondents (borrowers and non-borrowers).

                                Non-borrower (N1 = 219)        Borrower (N2 = 381)        All respondents (N3 = 600)           Statistical test 
Count (n1) %to N1                 Count (n2) %to N2         Sub-total (N4 = n1 + n2) %to N4

Male   146  66.7   286  75   432  72              x2 = 4.86**
Female     73 33.3   95  25  168 28                   p = 0.027
Total     100.0    100.0    100.0
No education  24  11   12  3.0   36  6.5             x2 = 38.64***
Secondary school   173  79   342  90   515  85.8        p = 0.000
Post-secondary   22  10   27  7   49  8.2
Total     100.0    100.0    100.0
AGE   40.24    40.16    40.33                    t = -0.44
HHSZ    4.47    4.42    4.51              t = 0.2
Yes    43  19.6  143  37.5   186  31              x2 = 20.82***
No    176  80.4   238  62.5   414  69             p = 0.000
Total     100.0    100.0    100.0
INCOME (in PKR)   37,440    3,8153    37,830                              t = -0.31
Main income sources
Agriculture   133  60.7   329 86.3   462 77              x2 = 51.54***
Non-agriculture   86  39.3   52 13.7   138  23              p = 0.000
Total     100.0    100.0    100.0
ASSET (in PKR)   2,29,421    2,23,808    2,26,872                    t = 3.0**
Land holding status
Contracted   172  78.5   331  86.8   503 83.8              x2 = 7.13**
Leased    47  21.5   50  13.2   97  16.2            p = 0.008
Total     100.0    100.0    100.0
FARMSZ (in acre)
5 or less    179  81.7   285  74.8   464  77.3              x2 = 26.35***
More than 5   40  18.3   96  25.2   136  22.7            p = 0.000
Total     100.0    100.0    100.0
DIST (in kilometer)
1–5   111  50.6   232  61.8   343  57              x2 = 8.88**
6–10   76 34.4   119  31   195  32.5           p = 0.003
>10    32 15   27 7.2   59  10.5
Total     100.0    100.0    100.0
Yes    128  58.4   166  43.6   294  49              x2 = 12.31***
No    91 41.6   215  56.4   306  51            p = 0.000
Total     100.0    100.0    100.0
Yes    63  28.7   75 19.6   138  23             x2 = 6.47**
No    156  71.3   306  80.4   462  77             p = 0.011
Total     100.0    100.0    100.0
Yes    35  15.9   80  20.9  115  19.2                       x2 = 2.25**
No    184  84.1   301 79.1   485  80.8             p = 0.133
Total     100.0    100.0    100.0
Aversion    121  55.2   117 30.7   238  39.6                       x2 = 35.0***
Others    98  44.8   264  69.3   362  60.4           p = 0.000
Total     100.0    100.0   100.0
Yes    190  86.7   227  59.5   417  69.5                       x2 = 48.45***
No    29 13.3   154 40.5   183  30.5           p = 0.000
Total     100.0    100.0    100.0

Note: 1. See Table 1 for definitions of all variables included in this table. 
2. Data are based on the survey results and authors’ estimates.
3. Entries for variables AGE, HHSZ, INCOME and ASSET are mean values.
*10% significant level.
** 5% significant level.
*** 1% significant level.

Only a small portion (31%) of the respondents is engaged in self-employment. The results also recommend that the borrower respondents are more likely to take up self-business compared to the non-borrower respondents (37.5% versus 19.6%). The x2 test (x2 = 20.82, p < .05) indicates a strong association between households’ access to microcredit and self-employment engagement (see Table 2).
A total of 462 respondents (77%) rely on agriculture (crop farming, raising livestock, fishery, etc.) as their major source of income while 138 of the respondents (23%) are engaged in non-agricultural income-generating activities. The average monthly income and assets value for the sampled household is 37,830 PKR (Pakistani rupee) and 2,26,872 PKR (Pakistani rupee), respectively (see Table 2).
Few of the respondents own farmland. The overwhelming majority (83.8%) of the respondents contract their farming land from villages, while 16.2% farm on leased land. In terms of farm size, up to three-quarters of the respondents work on farms no larger than 5 acres, in addition, the proportion of the borrowing households who work on large farms (a size larger than 5 acres) is 25%, which is higher than that of the non-borrowing households (18.3%). This implies that households with larger farm sizes are more likely to become the ZTBL’s microcredit borrowers. The proportion of the borrowing households who live within 5 kilometers of the ZTBL branches is higher than that of the non-borrowing households (61.8% versus 50.6%) and the share of the borrowers living more than 10 kilometers from the ZTBL branches is lower compared to the non-borrowers (7.2% versus 15%). This suggests that households who live physically closer to the ZTBL branches are more likely to access the ZTBL’s microcredit.
Less than half of the respondents have saving accounts in ZTBL branches. Compared to the borrowers, the non-borrowers appear to be more inclined to deposit money with ZTBL’s (43.6% versus 58.4%). In addition, a majority (77%) of the respondents do not hold ZTBL’s shares with a relatively higher proportion of shareholding observed in the non-borrower group.
To conclude, the frequency distributions of ATTITUD and ALTER in Table 2 show that non-borrower respondents are generally more averse to having debt and abler to access alternative credit sources when they need to borrow, compared to the borrower respondents.
Table 3. Logit estimates for households’ accessibility to microcredit.
Number of observations     600
Log likelihood     -240.66796
LR chi(16)        306.15
Prob > chi2        0.0000
Pseudo R2          0.3888
Correctly Predicted   75.3%

Independent                    Estimated    Std. Err.  z     P> |z|                        Marginal 
variables(a)                  coefficients                                           Effect(b)

GEND       0.5987576    0.2636455  2.27      0.023            0.0775907
AGE        0.0104543*    0.0128062  0.82      0.075          0.0013547
HHSZ       -0.2114619    0.0846987  -2.50      0.013         -0.0274025
EDU        0.0120002***    0.0577462  0.21      0.835          0.0015551
HHINCOME (in 1,000 PKR(c))  0.0001818    0.0000258  7.04      0.000           0.0000236
HHASSET (in 1,000 PKR)   -0.000021     3.19e-06  -6.57      0.000         -2.72e-06
OFFICIAL       0.5259286***    0.3405868  1.54      0.123          0.0681531
SAVING    -0.7007796    0.247454  -2.83      0.005         -0.0908113
ATTITUD      -0.9801542    0.2623261  -3.74      0.000         -0.1270144
ALTER       -1.85559     0.3353125  -5.53      0.000          -0.2404588
SHAREHLD    -0.6211151    0.2734035  -2.27      0.023         -0.0804879
FARMSZ       0.2578116    0.0630121  4.09      0.000          0.0334088
SELFEMPL       0.7137147    0.2783922  2.56      0.010            0.0924875
Dummy variables(d)
DIST2       -0.3909111***    0.2721304  -1.44      0.151         -0.0506567
DIST3     -0.5984942***    0.3702695  -1.62      0.106         -0.0775566
_cons            -1.179988***    1.088678  -1.08      0.278

a Dependent variable = 1 if the household has accessed microcredit and zero otherwise. 
b The Marginal effect is at the mean value. For a binary variable, the marginal effect is P|1 _ P|0.
c 1 US$ = 102 PKR (Pakistani rupee).
d to avoid the multi collinearity problem, a dummy variable is dropped from the group.
* 10% significant level.
** 5% significant level.
*** 1% significant level.

4.2. Determinants of rural household accessibility to microcredit in Pakistan
The Logit model (Eq. (3)) was conducted to examine household-level factors that influence households’ accessibility to microcredit and estimated via maximum likelihood estimation technique. Table 3 presents the estimated results of the logit model.
Overall the logit model successfully predicts the possibility of households’ microcredit access (75.3%). The likelihood ratio test with the chi-square statistic is equal to 306.15 with 16 degrees of freedom rejects the null hypothesis that the parameter estimates for the model are equal to zero, at the one percent level of significance. It can be concluded that the explanatory power of the logit model is satisfactory and the model can be used to explain the probability of accessing ZTBL microcredit by the rural households.
Based on the estimated results, 11 variables are found to have significant influence on households’ accessibility to ZTBL’s microcredit: DIST3 (-), HHSZ (-), EDU (+), INCOME (+), SELFEMPL (+), ASSET (-), SAV (-), ATTITUD (-), ALTER (-), OFFICIAL (+) and SHAREHLD (-).
The significant positive signs on INCOME, SELFEMPL and OFFICIAL variables can be explained from the perspective of capital requirement. High-income households tend to have more investment opportunities, leading to the stronger potential need for credit support. High-income households may also be more confident in repaying loans if they borrow. Therefore, they are more inclined to access microcredit. Similarly, the probability of accessing microcredit can be significantly improved when households get involved in self-business apart from agriculture production because of the higher capital requirement for investing in self-enterprises. By the same token, households with members working as village or township officials have greater need of credit for off-farm investment and thus have a higher probability of accessing microcredit.
Households with members working as local officials may also access ZTBL microcredit more easily due to their presumed good relationship with local financial institutions such as ZTBL. On the contrary, the significant negative signs on ASSET, SAV and SHAREHLD variables imply that households that are less budget constrained or have surplus funds under their own control would be less likely to access microcredit. For example, assets correspond to household initial capital. The households with higher assets values may be less budget constrained and therefore less likely to borrow from microfinance institutions (MFIs) such as ZTBLs. Similarly, households that deposit money with ZTBLs are able to access their savings in ZTBLs when they need financial support, which in turn weakens the likelihood of taking out micro loans from a ZTBL. It is also expected that the households that bought shares in ZTBLs are likely to have more surplus money under their control, which reduces their intentions to borrow.
A significant positive sign of the continuous variable EDU indicates that households who have acquired secondary education or less have a higher probability of access to microcredit than uneducated households while keeping other factors constant. This relationship is expected because farmers with formal education (for example, secondary or post-secondary school) are likely to have more exposure to the external environment including risks and possess more skills. They, therefore, might require more credit for consumption and/or production, compared to uneducated farmers. In contrast, between HHSZ variables and household access to microcredit, there is a significant but negative relationship, suggesting that larger-size households are less likely to borrow from the ZTBL microcredit program. This is possible because larger-size households tend to have low repayment capacity resulting from the smaller future expected income per capita, which lowers the probability of borrowing. This finding contradicts Maddala (2001) and Ho (2004) findings, who concluded that the probability of accessing formal credit increases with household size.
The estimated coefficients of variables DIST3, ATTITUD, and ALTER are all negative and significantly different from zero at the one percent level. Holding other factors constant, the households residing more than 10 kilometers from ZTBL branches have a significantly lower probability to access ZTBL microcredit compared to those who live within 5 kilometers from ZTBL branches, mainly due to the perceived high borrowing costs arising from travel expenses and time opportunity costs. In addition, an adverse attitude towards having debt could decrease the likelihood of accessing any type of credit by households, including microcredit. Furthermore, the availability of other credit sources (such as informal credit) also tends to reduce the probability of borrowing from a ZTBL microcredit program. This result is consistent with Vaessen (2001), who observes that many poor households are more willing to use informal credit owing to low transaction costs and flexible loan contracts.
The marginal effects are also calculated for the regresses of the logit model to provide a direct economic interpretation on the influence of these variables on households’ accessibility to microcredit. The results are also summarized in Table 3. For example, the marginal effect of HHSZ indicates that an additional member increase in the family would decrease the probability of accessing microcredit by 2.74% on average. In addition, the probability of borrowing from a ZTBL microcredit program would increase by 0.0023% with every 1000 PKR increase in INCOME. By contrast, an additional 1000 PKR increase in ASSET would reduce households’ probability of accessing ZTBL microcredit by 2.72%. This, however, indicates that the marginal effects of both INCOME and ASSET on the probability of accessing microcredit are minimal.
4.3. Survey Findings: Other factors affecting households’ accessibility to microcredit
Some qualitative data were also collected in a household survey to examine the factors affecting households’ access to microcredit in rural Pakistan, except for those that were analyzed empirically.
4.3.1. Knowledge of ZTBL microcreditprogram
From the 219 non-borrower respondents, 67 (30.5%) reported that they had no knowledge about the microcredit program operated by the ZTBL. Three main reasons are found for such a lack of knowledge. One of the most cited reasons is the lack of understanding of the concept ‘microcredit’ (n = 102; 46.5%). This is followed by the inadequate promotion of microcredit program by the ZTBL (n = 68; 31%) and the unawareness of the ZTBL branches nearby (n = 49; 22.5%).
4.3.2. Need to borrow
The survey results show that 60.7% (n = 133) of the total non-borrower respondents had no need to borrow money in the past two years. This further confirms that credit demand determines households’ access to microcredit to a large extent. For the other 86 non-borrowers who signaled credit needs, 34 had applied for micro loans from ZTBLs but had been rejected, and 52 had resorted to either formal lender (e.g., the Commercial Banks) or informal lenders (e.g., friends, relatives).
4.3.3. Reasons for loan rejection
The 34 non-borrower respondents whose loan applications were rejected were asked to provide reasons why they were denied loans. The result reported that 25 (73.5%) of the respondents attributed the loan rejections to their poor repayment capacity arising from low household income. Inability to provide loan security (e.g., collateral or co-signer) was also mentioned by 25 (73.5%) of the respondents as an adverse factor in their loan application. Moreover, 6 (17.6%) of the respondents deemed their failure in securing micro loans to be a result of their blemished credit history caused by previous loans defaults. This implies that creditworthiness potentially impacts the households’ access to microcredit. Furthermore, 3 (8.9%) of the respondent’s report that the difficulty in meeting the required documents by the ZTBL loan officers also prevented them from accessing microcredit.
4.3.4. Reasons for not applying for micro loans
All the non-borrower respondents (n = 219) were asked whether they might need to borrow in the future and if so, would they apply for micro loans from ZTBLs. The result reported that 46 (21%) of non-borrowers not interested to get microcredit in the future and 173 (79%) of the non-borrowers signaled borrowing intention in the future, of whom 120 (69.3%) expressed that they would give priority to micro loans if they have credit needs. The remaining 53 (30.7%)respondents who indicated an unwillingness to access ZTBL’s microcredit program were further asked to provide reasons why they would not borrow from ZTBL. Income is found to be a determinant in the households’ future borrowing from a ZTBL microcredit program.
5. Concluding Remarks
This study examines key factors that influence the accessibility of microcredit by rural households in Pakistan. Overall, our results suggest that rural households (especially the poor) and women in Pakistan have limited access to institutional credit, including the microcredit provided by ZTBL. The empirical analysis based on logit model has established eleven household-level factors important in affecting households’ likelihood to access microcredit, including household size, educational level, distance, income, assets value, being self-employed, savings, official status, shareholding status, attitude towards debt, and access to alternative credit sources. The official status, household income and self-employment are these three contributors of the households’ accessibility to credit because of a higher microcredit demand resulting from the higher capital condition (on/off farm), increases the likelihood of accessing credit by households. Conversely, household assets and savings can be used as proxies for household initial capital, and a higher value of either of them can potentially decrease the probability of accessing microcredit by the households. the households with huge family size would be fewer likely to access a credit program due to the perceived low repayment capacity rising from the smaller future expected income per capita. Equally, the probability of accessing microcredit would be substantially reduced if the households are averse to debt or can access alternative credit sources.
Such as documented in this study, the varied nature of rural households is important in accounting for the variance opportunities of accessing microcredit. Consequently, simply expanding microcredit programs in rural areas may be insufficient to increase credit access by rural households. On the demand side, limited credit access can be largely attributed to the low or no credit demand by the rural households for either agricultural production or off-farm activities, where the demand for credit is determined by a number of household-related factors such as those identified in this study. In addition, the poor households have limited access to microcredit because they effectually take themselves out of the credit market for the causes such as the inability to provide collateral and low repayment potential rising from their poor wealth circumstances.
One effective way to facilitate household access to microcredit is to encourage households to create investment opportunities when switching farming facilities on/off farm activities. This will give rise to additional capital requirements, which potentially increases households’ demand for credit.
In addition to demand factors, our analysis shows that supply factors, such as interest rates, document requirements, and loan processing time can impair households’ access to microcredit. Therefore, microfinance institutions (MFIs) such as ZTBL should improve their micro lending policies (such as simplifying lending procedures) and re-design their micro loan products to allow for more flexible terms and conditions that better suit the diverse needs of the local rural households. These innovations (especially client-responsive loan products) are thought to be more desirable by the poor whose living conditions are generally associated with uncertainty and vulnerability. This is likely because such flexible services can help them better deal with risks and thus reduce vulnerability. Another observation in this study is that the households’ inadequate access to microcredit can be due to their poor knowledge of ZTBLs’ microcredit programs. Thus, to improve household access to microcredit, it is essential that MFIs intensify the promotion of its microcredit programs among rural households so that households fully understand the characteristics of microcredit. This can be done through village meetings (or social gatherings) and mass media such as radio and newspaper.
A strong relationship between the ability to repay (perceived by households) and access to microcredit indicates that increasing household repayment opportunities helps improve their access to microcredit. Hence, it is important for MFIs to combine micro loans with other services and products that help improve the efficiency of loan use, which in turn helps build the households’ confidence in repaying loans. A useful service is to provide loaned households with an assessment of the profitability of projects supported by credit. Other services may include an agricultural technical extension, off-farm business introductions, and training in cash flow and risk management.
Our findings indicate that informal credit plays an important role in meeting the credit needs of the Pakistani rural households. This includes not only households that fail to obtain financial support through formal channels (such as ZTBL’s microcredit program) but also those who may be able to obtain formal credit but choose to borrow from informal lenders due to their potential merits (for example, flexible lending schemes). This implies that the existence of informal finance may not simply be a result of insufficient supply of formal credit or credit rationing by formal institutions. It is likely that different approaches to lending, taken by formal and informal creditors, force them to serve different groups of borrowers with various problems. This is another main reason for the persistent co-existence of formal and informal finance in many developing countries including Pakistan.
Policymakers in Pakistan should re-evaluate the role of the informal financial sector in rural credit delivery and formulate new policies regarding the development of informal finance. For example, rather than trying to eliminate informal finance, it would be more appropriate to reinforce the linkages between the formal and informal financial sectors in Pakistan. Better linkages between the two sectors would enable one sector to overcome its weaknesses by drawing from the other’s strengths. An example of this would be for banks to make use of the outreach and local knowledge of informal lenders, while informal lenders can benefit from formal lenders’ strong ability to mobilize resources and their access to wide networks across the region. Consequently, strengthening the association between the formal and informal sectors helps expand credit delivery and improve the overall efficiency of the financial system, and hence, accelerates the development of Pakistan’s rural economy.
1. This is not to suggest that the effects of microcredit are universally lauded. Some critics cite evidence that microcredit benefits only the better-off of the poor, which leads to increasing economic inequities. There is also concern about the extent to which women actually control loans disbursed in their names, placing in question its presumed benefit for women’s empowerment.
2. The notion of targeted microcredit grew from a concern that rich rural elites were the major beneficiaries of formal credit institutions (commercial banks, specialized agricultural credit agencies, rural cooperatives, etc.) due to high asset-based collateral requirements.
3. For administrative purposes, Pakistan is divided into four provinces and a Federal Capital. Each province comprises several districts, further divided into tehsils as administrative divisions. As entities of the local government, tehsils exercise certain fiscal and administrative powers over the villages and municipalities within their jurisdiction.
4. 1 US$ = 102 PKR (Pakistani rupee)

* Corresponding author. Tel.: +86
E-mail addresses:

EssayHub’s Community of Professional Tutors & Editors
Tutoring Service, EssayHub
Professional Essay Writers for Hire
Essay Writing Service, EssayPro
Professional Custom
Professional Custom Essay Writing Services
In need of qualified essay help online or professional assistance with your research paper?
Browsing the web for a reliable custom writing service to give you a hand with college assignment?
Out of time and require quick and moreover effective support with your term paper or dissertation?