Bangladesh is widely regarded as a ‘social laboratory’ where hundreds of non-governmental organisations- microfinance institutions (NGO-MFIs) are working in rural areas on poverty reduction and social development through innovations in entrepreneurship development using microfinance and several other programmes. Among other areas, microfinance has been found to have made notable contribution to poverty alleviation; increase income, savings and investment; and contribute to women empowerment. Microfinance has direct and indirect repercussion on the overall economy through integration of the rural financial sector and stimulating the economy through credit and saving programmes. Given this backdrop, this study is an attempt to assess the socio-economic effects of financial inclusion in Bangladesh to provide a fresh evidence the socio-economic effects of financial inclusion through MFIs. It aims to provide useful for transmitting policy messages in favour of further deepening financial integration through MFIs in Bangladesh.

An in-depth impact assessment is conducted to generate both quantitative and qualitative information relating to the changes in the socio-economic conditions of the beneficiaries of NGO-MFIs. A nation-wide sample survey has been conducted with structured questionnaires to generate quantitative information at household level. In addition, qualitative information is collected which include interesting cases from among the beneficiaries about the impacts of microfinance. The data of both treatment and control borrowers were taken from the database of the respective NGO-MFIs. We conduct survey on 638 Treatment (long-term or old borrowers), 641 Control 1 (non-borrowers) and 634 Control 2 (new borrower) households. As we take double controls, our total sample size is 1,913, which gives more than 90 per cent power. Structured questionnaires are used to conduct survey among the treatment (borrower) and control (non- borrower) households on the broad change variables.

Major Findings

Income, Expenditure, Food Consumption and Poverty

The empirical results reveal that most of the responding households are multiple borrowers. They need to earn higher income from different sources to maintain regular household expenditure and repay the borrowing. They usually invest in the businesses that generate return in short period of time and in regular interval. Their non-barrowing counterparts do not appear to be enthusiastic to invest aggressively.

The study finds that mean income of the non-participating households has increased from Tk.13,640 to Tk.19,158 (40.5 per cent) from 2010 to 2017, while it increased from Tk.15,891 to Tk.25,518 (60.6 per cent) for new borrowers, and it increased from Tk.19,821 to Tk.30,034 (51.5 per cent) for old borrowers during the same time period. Average monthly income of new borrowers is 33 per cent higher than that of non-participating households, while average monthly household income of old borrowers is 17.7 per cent higher than new borrowers and 56.8 per cent higher than the non-borrowers (p.19-22). The positive and significantly different average income of the participating households at 1 per cent level of significance indicate that participation in microfinance is related to higher household income than non-participation.

Average monthly expenditure of new borrower households is 9.25 per cent higher than that of non- participating households, while average monthly household spending of old borrowers is 7.97 per cent higher than new borrowers and 17.96 per cent higher than the non-borrowers. Average monthly expenditure of new borrower households is 5.33 per cent higher than that of non-participating households, while average monthly household spending of old borrowers is 3.73 per cent higher than new borrowers and 9.26 per cent higher than the non-borrowers (p.23-26). It means that participation in microfinance in the long run deepens financial inclusion, which influences to change the spending behaviour of the old borrowers.

Microfinance has been found to influence increasing per capita non-food spending (e.g., on education, health care, clothing, entertainment, etc.), which is observed from the higher per capita spending of the borrowing households compared to that of the non-borrowing ones. It is an indication of higher quality of life due to financial inclusion through microfinance. The ratio of food and non-food spending at national level is 47.7 and 52.3 per cent, respectively, while it is 50.5 and 49.5 per cent (p.19-26), respectively at rural areas as per Household Income and Expenditure Survey (HIES) 2016. Thus, food spending is still dominating the total spending among the survey households, but the share of non-food spending is gradually covering towards the national average.

In the present study the diversity and quantity of food intake have increased significantly among short-term borrowers compared to non-borrowers and long-term borrowers compared to short-term borrowers. Growth of fish and meat consumption in old borrower households was higher in old borrowers than non-borrowing households, while the result is mixed in new borrowers vs. non-borrowing households. The survey results demonstrate that weekly fruit consumption was lower among borrowing households compared to their non- borrowing counterparts in 2010, which reversed in the following years. It was 2.8 days among non- participating households in 2017, while it is 3 days for new borrowers and 3.2 days for old borrowers in the same year (p.28-30).

The empirical results demonstrate that total savings, investment for business expansion and total investment influence positively on income, while savings in MFIs demonstrates negative impact on income of the surveyed households. Old borrowers had higher total household income, which is evident from the coefficient of old borrower. It implies that as the membership of MFI gets older and the households repeatedly utilise many loans for long periods for investment, it results in significantly higher gross return of the borrowing households compared to new borrower and non-borrowing households. Greater amount of savings are the results of higher total and non-food expenditure (significant at 1 per cent level) as the survey data show that higher amount of total savings is observed in the household which had higher total and food spending. Similarly, greater total investment is observed in the households which earned higher income. The current amount loan has positive and statistically significant effect on total and non-food spending (at 1 per cent level), which implies that the households spend more from the returns of investment due to greater amount of loan among the other factors.

In the present survey the incidence of both upper and lower poverty was lower than that of the national level among the surveyed households in 2010. Poverty was significantly lower among the borrowing households in 2010 compared to that of national level. However, in 2017 the incidence of poverty was consistent among the old borrowing households, while it was higher than that of the national level in 2017 for new borrowers and non-borrowers. The incidence of both moderate and extreme poverty was less among the borrowing households than their non-borrowing counterparts. Significantly higher rate of the reduction of extreme poverty was observed among the borrowing households than the non-borrowing ones from the empirical results (p.33-34).

Savings, Asset Building and Investment

The study finds that average monthly savings of new borrower households is 84.4 per cent higher than that of non-borrowing households, while average monthly household savings of old borrowers is 139.8 per cent higher than the non-borrowers. Similarly, per capita monthly savings has increased by 41.8 per cent among non-borrowing households, while it was 94 per cent for new and 68.5 per cent for old borrower households. The surveyed households overwhelmingly mentioned that they intended to use their savings in coping different type of crises including natural disasters. In 2017 about 74 per cent of the old and78 per cent of the new borrowers intended to use their savings in crisis coping, while about 42 per cent of the non- borrowing households wanted to use it for the same purpose. crisis coping is becoming the most important

priority of the non-borrowers and new borrowers. The next important priority of the intended used of their savings for higher education of their offspring. In 2017 about 20 per cent of the old borrowers,16 per cent of the new borrowers, and 9 per cent of non-borrowers intended to use their savings for this purpose (p.36- 39).

The empirical results reveal that microfinance has brought significant impact on asset holding on the borrowing households. It finds that 56.2 per cent of the non-borrowing households had at least one television set in 2017, while it was 62.3 per cent for new borrowers and 67 per cent for old borrowers possessed at least one television set in the same year. Greater proportion of households also possessed at least one tractor/power tiller and one rickshaw/CNG-run vehicle.

The ownership of tin-shed building has increased at higher rate among borrowing households (+10.4 and +13.6 percentage points among new and old borrowers, respectively) than their non-borrowing counterparts (+7.8 percentage points). We also found that higher proportion of households among the borrowers had at least one mobile phone, at least one electric fan and electricity connection in the list of their assets and utility compared to their non-borrowing counterparts, and the results were statistically significant.

Interesting results have also been found in the dynamics of possession of agricultural land by the surveyed households. The proportion of households having agricultural land was the highest among non-borrowers in any of the reference period compared to the borrowing households. However, the proportion of household which possessed that land has been found to be decreasing among the non-borrowing households from 2010 to 2017, while the borrowing households are increasingly purchasing agricultural land during the same period (p.41-44).

The average household investment for business expansion was significantly higher for borrowers than their non-borrowing counterparts (at 1 per cent level), which implies that the earlier households are investing more due to microcredit. The growth of investment in this purpose was magnificent among new borrowers during 2010-2017 period, which means that entrepreneurship developed most rapidly among this group of respondents (Figure 4.11).

Nearly all the borrowing households have been found to invest in 2017, while less than two-fifth (about 37 per cent) respondents of the non-borrowers invested in the same year. However, growth of investing households was the highest among new borrowers, which demonstrates an exponential trend (from about 17 to 89 per cent from 2010 to 2017) (Figure 4.14). which implies that the households among the MFI borrowers are investing more on average than others due to microcredit (p.46-53).

Among the areas of investment, the surveyed households invested more in livestock, crop/fishery, plantation, new business or opening shop, and purchasing vehicle for economic activities. The empirical results reveal that both income and current loan have positive and statistically significant impact on investment, which indicates the household with higher income and higher current loan from MFIs invest more than non-borrowing and lower income households.

The empirical result indicates that in 2017 around 40 per cent of population of old borrower households and 42 per cent population from new borrowers were multidimensionally poor. While the proportion poor population were close among the borrowing households, about 88 per cent population (about half) of the non-borrowing households were multidimensionally poor (p.55-56).

Women’s Advancement

The study finds mixed impact of microfinance on women’s empowerment and advancement. It had mostly positive impact on non-financial empowerment indicators, while the impact was mostly insignificant on

women’s financial empowerment. The study finds that a meagre proportion of women (only 8.4 to 14.5 per cent respondents) can decide to take and use the loan on their own, used the loan for themselves, and decide to use income from loan. about one-fifth of women borrowers (both old and new ones) have been found to keep accounting of income and spending from credit, while around two-fifth of husbands maintained those accounts individually or jointly (p.55-60).

We observe that from 30.8 to 48 per cent respondents reported that significant change occurred in major indicators of women’s empowerment, such as respect in family and society, relationship with husband, sense of self-esteem, health care including that of girl child in the household and freedom of movement from home in the households where women had access to microcredit both for short term and term compared to the time when they did not take loan. The study also finds that the physical/mental torture on the women borrowers at home has not changed significantly as only about 11 per cent of new and 16 per cent of old borrowers (p.61-62).

Way Forward

• Since financial inclusion through MFIs has been found to have made significant impacts in terms of higher levels of savings, investment and consumptions leading to higher pace of reduction of extreme poverty, it will be wise for the financial institutions to continue this linkage programme with the MFIs for their own interests.
• Women borrowers should be trained in raising awareness on their rights and dues in the family and society. MFIs can organise leadership training for them vis-à-vis other regular training and awareness programmes.
• Some of the large and medium MFIs also do not have independent and well-equipped Monitoring and Evaluation (M&E) department.
• MFIs should conduct research on business and economic potential of the localities where their branches are located. If they introduce financial products as per the potential and train the borrowers accordingly, then both borrower and MFIs would benefit.
• Expanding microcredit operations in geographically remote areas, such as char, haor and coastal areas as well as higher poverty-prone areas should be encouraged by government’s incentives because operation of MFIs in these areas are sparse due to higher risks and operational costs. MFIs should provide credit to those areas in more flexible conditions and deliver products especially designed for clients in those areas, e.g., crop loans, and loans for floating farms and hydroponics.
• Agro-based households suffer significantly whenever their seasonal crops are damaged because of any natural disaster. Hence, crop insurance can be of great support to these households, which can be experimentally introduced by MFIs with support from the government, e.g., Ministry of Agriculture.
• Microfinance packages would also include technical support, information sharing and marketing linkage of the products.
• The government could take the lead in this regard through interest subsidies and incentives to MFIs and semi-formal institutions in providing financial services in remote areas through MRA and PKSF. This may in turn induce the private sector to extend its services in such areas.
• Loan policy and packages of the MFIs should be developed keeping in mind the socio-economic profile of an area, culture and ethnicity of the inhabitants, investment pattern, intensity of disasters, etc.
• Long-term borrowers should be provided loans with softer terms. In addition, successful borrowers should be provided SME loan like the ones provided by the banks.
• MFIs should target the unemployed youth to derive more benefit from demographic dividend in Bangladesh through devising suitable products for them to utilise their potential in contributing to economic and social sectors.