Print Reference This Reddit This

Microfinance As A Tool To Combat Poverty


The term “Microfinance” pertains to the lending of extremely small amounts of capital to poor entrepreneurs in order to alleviate poverty. This particular form of lending was formalized by Mohammed Yunus in Bangladesh during the 1970’s. Yunus won the Nobel Peace prize in 2006 for this effort.

Historically, microfinance has existed among the poor in various shapes and forms. The most common example cited is that of a rotary club, wherein people pool their savings into a certain fund every month, and then are randomly picked to receive the entire fund every year. This method required the member to belong to the same society in order to be trusted. This “pooled savings” method has been popular across the globe.

Today, microfinance institutions allow for both the “pooled savings” model as well as the small lending. The concept is simple, loan small funds to the poor (usually under 10 USD) for a small fixed period of time, and thus the individual is able to access further lending at points of repayment or thereafter. This would empower entrepreneurs to take up a trade and allow them to start earning. For the purpose of this analysis, the latter model of microfinance is assumed.

Women are actually favored by this form of lending, since it is seen that the repayment rates of women higher than men. This research proposal attempts to generalize the finding to include the poor, and attempts to make no distinction in the demographics of the microfinance recipients. Furthermore, this proposal focuses solely on developing countries. This is done to ensure the data is collected on the individuals most in need of micro financing opportunities.

The purpose of this research proposal is to identify and explain if, in the past three decades wherein microfinance has taken place, there are any significant impacts on poverty and inequality in the applicable regions. Furthermore, as the growth in aggregate microfinance loans have increased, is there a direct relationship with growth in entrepreneurial activities? In other words, has the advent of microfinance enabled entrepreneurs to increase their activities?

Literature Review

Microfinance is one of the tools developed to combat poverty. This proposal focuses on five studies to survey a sample. Concerns are more regarding the sustainability of the microfinance enterprises. Loans are constantly being made to high-risk low income individuals, with unique and innovative methods being utilized. Thus, the most significant concern at the moment is whether the formal microfinance institutions are actually impacting poverty in a significant manner. It is with this idea in mind that the literature was selected.

II.1 Research Purpose and Hypotheses

The goal of microfinance is to reduce poverty, however the authors focus on how poverty is reduced. Amin et al (2003) focuses on serving people slightly above or below the poverty line, however the really poor and destitute are excluded. By contrast, Copestake et al (2001) analyzes the impacts of microfinance on firms and individual wellbeing. Copestake focuses on business performance and household income to link the availability of microfinance and wellbeing of the poor.

Evans et al (1999) approach seeks to explain nonparticipation in the microfinance revolution. Kabeer (2001) provides a meta-analysis and shows why various studies conflict in their conclusions as to the impact of microfinance on women empowerment.

Finally, Park (2001) evaluates the actual microfinance programs in China using 3 key measures (targeting, sustainability and overall impact). Of the literature surveyed, analysis of Park (2001) is the most relevant as far as this proposal is concerned since they chose to focus on the overall impacts of microfinance, as opposed to the minute aspects of the financing model. However, the approach taken by Amin et al (2003) is quite strong as it figures out, if the model is making more than marginal impact on poverty.

The research hypotheses are as follows:

1. Amin et al (2003). How effective is the availability of microfinance for the poor and


2.Copestake et al (2001). Estimate the impacts of microfinance programs on business

performance and individual wellbeing.

3. Evans et al (1999). Analyze barriers to entry in microfinance.

4. Kabeer (2001). Conduct a meta-analysis to explore the reasons for conflicting outcomes in the impact of microfinance and women empowerment.

5. Park (2001). Evaluate the effect of microfinance on poverty via targeting, sustainability and impact.

II.2 Research Design Types

Based on the surveyed literature, the most popular design for analyzing microfinance is the Longitudinal Design, specifically a Cohort Design that is prospective in nature. All the authors utilize survey instruments in conducting their research. Copestake et al (2001), Evans et al (1999) and Park (2001) each formulate a Cohort design, though they differ in the shared characteristics of the cohorts. In addition, Evans et al (1999) utilize a “cascading set of qualitative focus group discussions and key informant interviews” in order to lend a qualitative approach to their research. They collect their sample from two villages in Bangladesh and follow the respondents over two years. Kabeer (2001) conducts a meta-analysis on a variety of research where the most common method utilized is the Cohort Design as well.

II.3 Unit of Analysis and Sampling Method

Amin et al (2003), Evans (1999) and Park (2001) focus on the household level as their main unit of analysis. This is pertinent because this allows measurement of the impacts on microfinance on the grassroots level. Copestake (2001) takes a different approach by surveying individuals rather than households, in order to accurately capture the impacts of microfinance on businesses.

The sample is based on three strata, new, established and pipeline (forthcoming) borrowers. Additionally, they utilize focus groups and informant interviews as well as content analysis to provide an insight into microfinance nonparticipation.

The sampling methods used in the analyses follow the random sampling method in order to maintain uniform and external validity to the studies. This is the cost effective method of collecting the individual and household level data. The only exception that one could make on the selected studies is that they usual pertain to a single country. External validity, then, would follow a similar route in that the studies are relevant in being generalized to the populations studied in the select country. It is this area where this research proposal attempts to diverge.

II.4 Operations of Key Indicators

In reference to a measure of female empowerment utilized by various studies analyzed by Kabeer (2001), the measure is defined as the extent to which microfinance allows women to be independent in the paternalistic societies. Examples of operations of the measure include an index of “managerial control” over the loans by women, and surveys conducted to analyze the rate of domestic abuse and child schooling patterns.

II.5 Section Summary

Time is an important factor. The use of a true experimental design is also unfeasible because of the nature of microfinance. Since microfinance is a tool used to alleviate poverty, systematic exclusion of any group from accessing microfinance is unethical.

This implies that we are trying to gain external validity to analyze the effects of microfinance, both on grassroots and national level. Therefore, the best design for this specific research proposal would be quasi-experimental time series design as used by Amin et al (2003).

Most authors opt for stratified random sampling method, which indeed assures representativeness with respect to each strata (or cohort) in the analysis. This method requires the strata to divide on income level, or where the respondents are in relation to the poverty line. Therefore, the best method is similar to the multistage random sample used by Amin et al (2003).

Finally, as far as the operations of the relevant variables are concerned, there are two main methods out of which the best would be using the income level of the respondents in relation to each country’s own poverty line, in order to calculate the level of poverty experienced by the respondent.

III. Methodological Plan

III.1 Unit of Analysis

Perhaps the more obvious solution to study this issue is to study the impacts of microfinance at the country level. The overall figures of developing economies are known to have measurement problems, especially when the macro-level variables are recorded in a manner tailored to the developed economies.

Country-level data helps us identify the direction between microfinance and poverty. The main method of variable measurement is on the country level with household level evidence for accuracy.

The surveyed literature on microfinance uses the household unit of analysis because the principal investigator has greater control in the measurement of the key explanatory variables. The first research design can remain unchanged, but there will be differences in the second design since not all the household characteristics can be described in aggregate terms. Since the sampling method utilized on the household level can allow for generalizability, we can use this data to cross-check against the aggregate variables for the region in question; in order to make sure that the causal direction satisfied. Data would then be collected from the selected households via an annual survey.

In sum, data collected on the household level can be tested against data collected on the country level in order to ensure the external validity of the research design.

III.2 Research Designs

it would be unethical to propose a population to be systematically excluded from the opportunity. However, the study lends itself to a Quasi-Experimental Time Series Design similar to the one employed by Amin et al (2003). Observations are made on the sampled households at the beginning of the analysis and then every subsequent year over a 10year period. Using the observational data, we can calculate the regional income and microfinance participation levels. The changes in income every year would provide an answer to the nature of the relationship. In order to set up a natural experiment with the “control” group containing non-participants and the “treatment” group consisting of microfinance participants. Observing these two groups over time can add further validity to the design.

On the aggregate level, since we are interested in the direction of the relationship between the dependent and independent variables, and we know when the independent variable was introduced into the model coupled with the fact that the aggregate level variables have been measured before the advent of formal microfinance institutions, a simple retrospective directional analysis can be readily available. Furthermore, in order to cross-check the results of the household level design, we can additionally set up a prospective research design by using country level statistics provided by the World.

III.3 Sampling Method

For the country level unit of analysis, no sampling is necessary for the retrospective design, since there would be no reason to exclude any country from the analysis, especially since the countries are studied individually. This assumes, as a matter of course, that the level of micro financial activity for all countries and for past years is available. However, if microfinance lending data cannot be obtained from all countries, a systematic sampling technique can be utilized to obtain data from the poorest countries. For the prospective country level research design, the only data we would require is the one from the regions selected at the household level, and thus the sampling technique below would dictate the countries selected for this proposal.

For the household unit of analysis, a multi-stage random sampling technique would be utilized. The population for this analysis would be all households in developing economies, with the first stage being the random selection of countries, the next stage a random selection of rural districts (emphasis added) and finally, a random selection of households with equal probability within the certain district (regardless of income level). The reason for selecting just rural households is that as a greater number of developing economies are being industrialized.

Since we plan to sample households regardless of income level, we would need to reach out to areas where the probability of sampling a poor household is higher. Since we are sampling from the rural areas alone, there is a concern regarding the generalizability of the data, especially in comparison with the country level variables.

III.4 Main Independent Variable

The main independent variable for all three hypotheses outlined above is the amount of microfinance lending. This is operationalized for the country level design by calculating the total amount of loans made by the major microfinance houses operating within each sampled country for each fiscal year. The data is proposed to be collected from the tax filings of the lending institutions in operation within each country for every year included in the study. There are nations where the tax filings of the microfinance institutions are not made public. In order to get the data for these areas we would take the data from the household survey as described below.

On the household level, a more relevant figure would be achieved by asking the respondents on the survey to state the amount borrowed from the microfinance lending institution in the prior year. This would also provide a viable alternative in the event that the country level data from the microfinance institutions is not readily available.

III.5 Dependent Variables

Based on the hypotheses proposed above, three main dependent variables will be tested against the main independent variable. Measures for obtaining data for the dependent variables are given in the following order: poverty, income inequality and finally, entrepreneurship activity.

The World Bank defines the concept of “Absolute Poverty” as the people living under 1 USD per day. Keeping this concept in mind, we propose to test for both reductions in absolute poverty as well as poverty defined in relation to the countries’ own poverty line.

Furthermore, the level of poverty (for the household unit of analysis) can be measured by first defining a “poverty line” for the country, and then the difference between the household income level and the poverty line would represent the level of poverty faced by the household.

Since we have a more accurate picture of individual household income levels, the more the household income is below the poverty line, the greater the level of poverty for the household. The actual data on household income levels would be a component of the survey conducted and then included in the calculation outlined above.

Secondly, in regard to income inequality, the aggregate level data can be obtained by the government statistics department of each country. Income inequality is usually defined by the statistical “Lorenz Curve” where the percentage of income held by each quintile of the population is reported. Thus, the income share held by the lowest 20% of income recipients in the population is used. The amount of micro financing is directly proportional to the income share of the lowest 20%. This can also be thought of as a measure of how the entrepreneurship activity in impacting the general economy.

It is important to mention that our sample contains rural households and thus the measure for income inequality will not accurately translate to the aggregate figures. However, we can still compare the directional outcomes for both measures in order to solidify the result of the hypothesis.

Finally, the last dependent variable is the level of entrepreneurship within each of the regions used in our sample. A lot of the entrepreneurship activity within the poor is not recorded as a part of the “formal sector”. Thus, an individual accessing microfinance loans could not be expected to register his business with any government agency. However, a rough estimate for this figure may be calculated by the number of loans made by the operating microfinance institutions. The underlying theory is that since microfinance lending primarily operates to induce entrepreneur activity in the poor, the higher the number of loans made would mean a higher aggregate level of informal sector business activity.

On the household level of analysis, the survey would ask the respondents to state the number of loans received in the current year. Any household is free to take multiple loans in a given year. Assuming each loan facilitates the entrepreneur in conducting his/her business. The survey would actually provide us with a far more accurate measure of the level of entrepreneurial activity in each household

IV. Statement of Limitations

There are still a number of problems with the design as far as internal and external validity are concerned. We have suggested a multiple design approach to combat some of the issues that may arise from a more simplistic methodology. The absence of randomized control groups continues to be a problem, though that is handled by using non-participants as an additional control in the design. Furthermore, history presents a problem in that we plan to sample the households at a specific point in time but circumstances occurring prior to the research may impact the results. We suggest a retrospective country level analysis to control for such factors, however it is unclear whether the data collected over the past will be able to explain some of the variation in the final design, especially at the household level of analysis. Instrumentation can also be a factor and thus the researchers would need to carefully plan out the survey and avoid making any changes after the first one is administered in order to ensure data consistency.

Maturation and attrition would continue to pose problems for this design however. Plus, since the data is analyzed over time, it would be very difficult to replace the respondents accurately. This is why we would need to use a large sample in the beginning so as to ensure that both problems can be accurately handled. Furthermore, a large ample size up-front would also allow us to handle the potential disadvantages of using a multistage random sampling technique.


By doing this research proposal about microfinance industry, we got to know the different research designs, we also made a few hypothesis, key indicators in research and their operation, the methodological plan, sampling methods, dependent and independent variables, limitations of research and many more.

Therefore, once this research is done, it will answer a few questions that arise in minds of different people.

I would sincerely like to thank all my faculties’ friends and parents for their support in helping me do this assignment.

Need help with your literature review?

Our qualified researchers are here to help. Click on the button below to find out more:

Literature Review Service

Related Content

In addition to the example literature review above we also have a range of free study materials to help you with your own dissertation: