Financial inclusion and development: What do we know?
There has been an increased focus on the importance of financial inclusion, and it is generally believed that financial inclusion can play an important role in helping people, especially the poor, in improving their livelihoods and can spur economic activity. However, what does the empirical evidence have to say about this positive impact of financial inclusion on development? How does this impact vary across different dimensions of financial inclusion? What are the channels through which these relationships hold? This section tries to answer these questions and provides an overview of the literature that has critically studied the positive and negative impacts of different dimensions of financial inclusion on business outcomes, socioeconomic outcomes, and economic growth, at a microeconomic level as well as the macroeconomic level.
Access to credit is considered particularly important, especially for low-income households, for a number of reasons. Households may want to avail credit to invest in their business or education, buy home or livestock, start a new business, organize weddings, manage unexpected emergencies, or simply to tackle various ups and downs in life. According to the World Bank’s Findex database 2017, a little less than half (47%) of the world’s adult population reported borrowing money in the past year. In the case of developing countries, friends and family were the most common source of borrowing, whereas, in the case of developed countries such as the OECD countries, the most common source of borrowing was a formal financial institution. Globally, about 11% of the adults (one quarter of borrowers) borrowed money from a formal financial institution. However, formal borrowing was quite low in some regions such as South Asia (6.6% of adults) and Sub-Saharan Africa (7% of adults) as highlighted in Figure 0.2. Countries in these regions mostly rely on informal sources of borrowing. A map depicting the concentration of formal borrowing around the world is provided in Appendix 0.2.
due to various factors. First, some of the informal sources of borrowing such as moneylenders can charge exorbitantly high interest rates and prey on poor households who are in dire need of credit. The literature has generally referred to these moneylenders as “loan sharks”. A study of moneylenders conducted in rural Pakistan by Aleem (1990) showed that the average annual interest rate charged by moneylenders was about 78.5% with a standard deviation of 38.3%. This translates to the fact that interest rates ranging from 2% to as much as 150% lie in the 95% confidence interval, indicating very high variability of interest rates amongst moneylenders. Second, borrowing from a formal source might come with better credit terms as compared to informal sources. Third, when people borrow from friends and family or through other informal sources, they are restricted to funds within the community. These funds might not be sufficient to fulfil their credit needs especially if they belong to a poor community, and this can contribute towards increased inequalities and poverty traps (Demirguc-Kunt et al., 2017).
Microcredit, as a credit source, started more than three decades ago with the aim to provide credit to low income households which were previously excluded from the formal financial sector. It is generally believed that the idea of microcredit was first introduced by the Bangladeshi economist Dr. Muhammad Yunus. In 1976, Muhammad Yunus visited the village Jobra in Bangladesh and provided a loan worth $27 from his own pocket to 42 women to help them start a business (Yunus, 1998). He believed that these microloans could help in kickstarting a dynamic development process by helping micro entrepreneurs grow, generate income, hire more people, and reduce poverty. With the help of the Bangladesh central bank and international donors, Muhammad Yunus established the Grameen bank in 1983 and started providing microcredit to low-income households at a much larger scale. Over the next two decades, microfinance started to grow in developing countries all across the world. United Nations named 2005 as the “International Year of Microcredit”. Moreover, in 2006, Muhammad Yunus and the Grameen Bank won the Nobel Peace Prize “for their efforts through microcredit to create economic and social development from below”. Since then, microfinance has expanded rapidly, and the total microfinance clientele has grown by more than 16 times from 8 million in 1997 to about 139 million in 20173. Currently, there are over 10,000 microfinance institutions all around the world providing financial services to low-income households.
There is a large body of the literature that has studied the impact of access to credit on a variety of outcomes and has found mixed results. Most of the studies which were conducted in the early 1990s and 2000s showed a significant impact of microcredit on different socio-economic outcomes, but they were mostly fueled by anecdotal evidence, descriptive statistics, or impact evaluations conducted by the microfinance institutions themselves (Hannig and Jansen, 2010). For example, a study conducted by the Grameen Bank claimed that 65% of their clients had crossed the poverty line (Grameen Bank, 2007). The recent empirical evidence conducted by independent academics has produced relatively modest results. The literature does not generally find a significant effect of microcredit on socio-economic outcomes such as poverty, health or education outcomes. However, the evidence is relatively more positive for some business-related outcomes.
Pitt and Khandker (1998) conducted one of the earliest impact evaluations of microcredit in Bangladesh by looking at microcredit offered by three microfinance institutions including the Grameen Bank. They found a positive impact of microcredit on household consumption, labor supply, assets, and school attendance rates of children. They also highlighted that the impact was more significant for female borrowers as compared to male borrowers. Moreover, a similar study conducted by Pitt, Khandker and Cartwright (2006) showed that microcredit had a significant impact on women’s empowerment in Bangladesh. They showed that microcredit program improved women’s decision-making power in the household, women had greater bargaining power, and also greater freedom of mobility. However, Morduch (2000) and Roodman and Morduch (2014) questioned the identification strategy of their empirical analysis and doubted the validity of their results, which led to a dynamic discussion between the authors on the validity of the early positive results on microcredit4. Similarly, some other studies have also shown that having access to credit has a positive effect on monthly income (Honohan and King, 2012), household income and consumption levels (Mahjabeen, 2008), household welfare (Khandker and Faruqee, 2003), business creation (Bruhn and Love, 2014), and investment in human capital (Amendola et al., 2016).
In 2005, Banerjee et al. (2015) carried out the first randomized control trial of expanding access to credit in an urban market in Hyderabad, India. They partnered with one of the fastest growing microfinance organization in India named Spandana and selected 104 neighborhoods where the microfinance organization would have been interested in opening up a branch. They randomly selected 52 of these neighborhoods as treatment group where Spandana opened up a branch. The rest of the neighborhoods became the control group. The main findings of their study showed that there were no changes in health outcomes, education, or women’s empowerment even after three years of intervention. However, they did find that small business investments and profits of pre-existing businesses increased. Moreover, they found that expenditure on durable goods for businesses and households increased while the expenditure on temptation goods (tea, alcohol, tobacco, gambling) went down.
Similarly, in 2006, Crépon et al. (2015) conducted one of the first randomized control trials on microcredit in a rural setting in Morocco. Their findings suggested that the take up of microfinance was surprisingly quite low (around 13%) which pointed towards a relatively low demand for microcredit. Secondly, they reported that households who had access to credit expanded their self-employment activity and their profits increased. Lastly, they also did not find any impact of having access to credit on education, health or consumption related outcomes. However, Bédécarrats et al. (2019) replicated this study using the same data and raised questions about the external validity as well as some concerns over the internal validity of their main findings.
Similar studies were conducted in Mexico (Angelucci, Karlan and Zinman, 2015), Bosnia and Herzegovina (Augsburg et al., 2015), Ethiopia (Tarozzi et al., 2015) and Mongolia (Attansio et al., 2015) and they reached parallel conclusions. Banerjee et al. (2015) summarized the findings of the six widely cited randomized control trials and concluded that access to microcredit had “a pattern of modestly positive, but not transformative effects”. They concluded that whilst the businesses appeared to benefit from access to microcredit, it did not translate into broader development impacts such as poverty reduction or better education or improved health related outcomes. They also noted heterogeneity in business related outcomes across and within different RCTs, with some
studies finding a notable change taking place at either tails of the distribution of business outcomes such as profits.
Some studies have also looked at the impact of improved access to credit on economic activity and have generally found positive effects. Bruhn and Love (2014) used a natural experiment in Mexico where a microfinance institution named Banco Azteca opened up 800 branches simultaneously. They showed that this increase in access to finance caused a 1.4% increase in overall employment, a 7.6% increase in informal businesses, and a 7% increase in income on average. Similarly, Burgess and Pande (2005) used state level panel data for India and showed that access to finance led to a significant reduction in rural poverty. There is also some evidence which suggests that flexibility in credit product design could result in improved impact (Field et al., 2010).
On the other side, there is some evidence which also alludes to the negative side of credit expansion. This strand of research highlights that microcredit can sometimes do more harm than good and can lead to increased poverty levels, exploitation of women, child labor, increased workload, and the creation of a culture of dependence which has an adverse effect on economic growth (Rooyen et al., 2012; Bateman and Chang, 2009; Copestake, 2002). The reasons cited behind this negative side of credit expansion generally include issues pertaining to over-indebtedness, exploiting nature of lending products, and non-productive use of credit. The crises of over-indebtedness and delinquency in certain microfinance markets around the world (prominently in India, Pakistan and Morocco) are a few examples which also support this view on the negative impact of credit expansion. Bateman (2010) even argues that microcredit is ineffective and it only offers an illusion for entrepreneurship creation and poverty reduction.
In summary, studies discussed in this section have highlighted that the evidence on the impact of access to credit is mixed at best. Studies have not generally found a significant effect on aggregate welfare of the households, although some small businesses seem to benefit from credit. An important point to emphasize here is that the impact of access to financial services on the labor market, especially on entrepreneurship creation, remains an understudied phenomenon.
Financial inclusion is not merely about credit. Another very important dimension of financial inclusion is access to savings accounts. Individuals save money in order to manage future expenses and investments. This includes, but is not limited to, saving for future investments in business or education, healthcare, old age, weddings, funerals, and other potential emergencies. Saving money at a formal financial institution can have several potential benefits as compared to saving at home or saving informally. First, saving at a formal financial institution provides safety against theft and other hazardous events such as fire or flood. Second, having savings at a formal financial institution can allow individuals to earn interest on their money which can offset the impact of increased inflation. Third, having savings at a formal financial institution rather than at home can curb impulsive spending and encourage better cash management (Karlan et al., 2017). Fourth, having money at a formal financial institution provides confidentiality and more control by making it more difficult for friends and family to access this money. This phenomenon is well documented by Anderson and Baland (2002) and Baland et al. (2011). Using case studies from Cameroon and Kenya, they highlight that women sometimes participate in Rotating Savings and Credit Associations (ROSCAs) to escape forced solidarity and to protect their savings from their husbands and family relatives. Similarly, Demirguc-Kunt et al. (2017) highlight that saving at a formal bank account can strengthen women’s economic empowerment by giving them more control over the household funds.
According to World Bank’s Findex database, as of 2017, about half (48.4%) of the world’s adult population reported saving money in the past year. 71% of the adults in high income economies and about 43% of adults in developing countries reported saving. However, only about one quarter of the adult population (about half of the savers) around the world saved at a formal financial institution. Moreover, there are large gaps in formal savings between high income countries and developing countries. About 55% of the adult population (three quarters of savers) in high income countries saved formally, whereas only 21% of the adult population (half of savers) in developing countries saved formally.
In the case of developing countries, many people save either semi-formally or informally. Semiformal saving generally refers to saving in a savings club. One common example of these are the Rotating Savings and Credit Associations (ROSCAs). ROSCAs are generally community-owned and run by local people themselves. They are also known as committees and are quite common in developing countries. People who are part of a ROSCA generally pool their deposits weekly or bi-weekly and disburse the lump-sum amount to a different member each week or two-weeks. In 2017, 11% of the adult population in developing countries reported saving semi-formally in the past year, according to the World Bank’s Findex database. However, one of the most common form of saving in developing countries remains informal savings (Figure 0.3). A large number of people keep their savings at home and ‘under the mattress’, or they save in the form of livestock or jewelry. These informal sources of savings were reported by about one third of the savers in developing countries (World Bank, 2017).
Table of contents :
II. What is financial inclusion?
III. Financial inclusion and development: What do we know?
IV. The Context of Pakistan
V. Outline of the Thesis
Chapter 1: Access to Microfinance and the Economic Ladder
1.2. Details about the Data
1.3. Empirical Methodology
1.4. Results and Discussion
Chapter 2: Financial Inclusion: An illusion for creating women entrepreneurship?
2.2. Data and the Mexican Context
2.3. Empirical Methodology
2.4. Results and Discussion
2.5. Robustness Checks
2.6. Conclusion and Policy Recommendations
Chapter 3: Pushed into the Corner: An Empirical Study on the Drivers of Financial Exclusion
3.2. Theoretical Backdrop
3.3. Measuring Financial Exclusion Indicators
3.4. Drivers of Financial Exclusion
3.5. Empirical Strategy and Data
3.6. Results and Discussion
3.7. Methodological Issues and Limitations
3.8. Conclusion and Policy Recommendations
Chapter 4: Assessing Household Financial Vulnerability using Machine Learning
4.2. Empirical Literature on Measuring Household Financial Vulnerability
4.3. Overview of Machine Learning Methods
4.4. Empirical Strategy
4.5. Results and Discussion
I. Summary and Policy Recommendations
II. Limitations and Prospects for Future Research