The Impact of Baitul Maal wa Tamwil on Household Welfare: Empirical Evidence from Indonesia

Despite the growing studies of Baitul Maal wa Tamwil (BMT) in Indonesia, few have empirically analyzed how BMT increases household welfare, primarily through credit for productive activities. This study aimed to analyze the impact of BMT availability in the community on household welfare. Considering that credit is targeted for productive activities, the researchers limited the respondents to farmer households and households with non-farming business activities, amounting to 4642 and 2250, respectively. Utilizing longitudinal data from the Indonesian Family Life Survey (IFLS), the analysis was conducted employing the difference-in-difference (DiD) method. The results showed that BMT's existence in the community improved farmer households' mean welfare by 1.65%. After controlling for household and community characteristics, consistent results were obtained with a small decrease in the coefficient of 1.58%. Meanwhile, households with non-farming business activities were not affected by the existence of BMT in their community. Nevertheless, a positive coefficient of diff-in-diff indicated that BMT and welfare had a positive relationship even though it was not strong enough to be a contributor to the outcome.


Introduction
Conventional microfinance products have been sustainably growing in Muslim majority countries, notably the Grameen Bank experiences in Bangladesh (Abdul Rahman & Dean, 2013). Although approximately 44 percent of conventional microfinance customers reside in Muslim countries, it does not fulfill the needs of Muslims optimally (Karim, Tarazi, & Reille, 2008) as they assume the option of conventional financial products is incompatible with the principles outlined in sharia (Komi & Croson, 2005). Furthermore, Karim et al. (2008) argued that although there is a market for poor clients who only want to be engaged in Islamic transactions, clients who engage in conventional practices will prefer Islamic once it is available. Abdul Rahman and Dean (2013) asserted that microfinance practitioners in Muslim-majority countries such as Afghanistan, Indonesia, Syria, and Yemen stated that some conventional micro-borrowers tended to switch over to Islamic products once they became available. Dusuki (2008) affirmed that Islamic microfinance should take the initiative to meet the increasing needs of Muslim clients to alleviate poverty and help the poor people in their distress. Islamic finance has boomed in recent years, but what has hit the headlines is that big money has moved around, following Islamic law principles (Seibel, 2008). Islamic microfinance is a sub-system of the Islamic financial system, and there is a bigger demand for sharia-based financial products and services with the speedy development of the Islamic finance industry. The role of microfinance in fulfilling development and social necessities is indispensable (Hassan & Saleem, 2017). According to Ahmad and Ahmad (2009), microfinance is constituted by a range of financial services for potential customers who are considered non-bankable, mainly because they lack guarantees to protect a financial institution against a loss risk. The appearance of inclusive microfinance and the delivery of small loans to the poor who usually fall outside the mainstream banking institution has focused on growing attention from various parties, including governments, bankers, practitioners, academics, and researchers (Dusuki, 2006).
Islamic microfinance provides financial benefits for the community, especially the unbankable group, and rejects usury because it is oriented to Islamic sharia principles. The prohibition of interest (riba) in Islam is a way to avoid exploitation and uphold justice (Hadisumarto & Abdul, 2010). According to a study conducted by Siddiqi (1981), the main reason why Islam erases interest is that, in essence, interest is a form of oppression. Furthermore, Mannan (1986) revealed that interest deprives humanity's foundations, mutual help, and sympathy. According to Ramzan Akhtar (1997), the reason for the prohibition of usury is since interest guarantees a permanent return to the lender, while the borrower is left to bear all the business risk. In this case, Islam wants to promote justice and equality. Justice requires that investors take the risk with borrowers if they want to make a profit. Thus, it is obligatory for Muslims to eliminate the usury element from all kinds of transactions.
Indonesia is one of the countries with most of the population is Moslem, which develops microfinance institutions with sharia principle in the provision of capital for microenterprises. One of the rapidly growing microfinance institutions is Baitul Maal wa Tamwil (BMT). BMT is a financial institution with the sharia concept that combines the concept of maal and tamwil in one agency activity. The maal concept emerges and becomes part of the Muslim community's life to collect and channel funds for zakat, infak and shadaqah (ZIS) productively. Meanwhile, the tamwil concept is set up for productive business activities of middle and lower society (micro) for the benefit (Mulyaningrum, 2009). According to Soemitra (2009), BMT is an Islamic financial institution formed to play a social role and seek profit in accordance with Islamic law. Generally, BMT institutions are led by leading Islamic organizations such as Pusat Inkubasi Bisnis Usaha Kecil (PINBUK) or Centre for Micro Enterprise Incubation and Dompet Dhuafa foundation, Sharia banks, Islamic boarding schools (pesantren) and Muslims patrons who have strong socioeconomics and religious influences in the community (Dewanti, 2013). BMT was established as an Islamic financial intermediary, especially for small and micro enterprises (SMEs) (Wardiwiyono, 2012). In addition, in the detailed profile of SMEs, Hatmaka (2013) finds that, the benefit rate of MFI's product such as credit utilization and the rate of saving utilization by SMEs is able to support the growth of SMEs. Rapid development has made BMT's role in Indonesia be reckoned with (Masyita & Ahmed, 2013). Several reasons support the rapid development of BMT in Indonesia. First, as the majority of the Indonesians are Moslems and some Islamic institutes/organizations such Majelis Ulama Indonesia (Indonesian Ulama Council), Muhammadiyah, and Nahdlatul Ulama have affirmed that bank interest is riba (usury), which is unlawful (haram). BMT is the alternative choice of financial services provider (Wardiwiyono, 2012). Second, its extensive network covers all Indonesian regions, making it a strategic intermediary for financial services. The method of "picking up balls" or coming to customers directly on-site is often used by BMT to expand the area's scope while facilitating customer's financial access. Easy administrative prerequisites and extensive network make BMT a promising financial aid choice (Rahajeng, 2014). BMT in fact developed into one of important microfinance institutions (MFI) in Indonesia, whether viewed from the financial performance and the number of people who could serve. Third, as one of MFI, BMT seems more resilient to external shocks due to Indonesia's economy, as evidenced in the era of economic crisis of 1997/1998. Aziz (2004) asserted that the economic crisis that hit Indonesia in 1997 did not significantly impact the BMT stability. Moreover, it has been proven that BMT could survive when the global financial crisis hit the Indonesian economy's stability in 2008. During the global financial crisis of 2008, there was an increase in the percentage of financing provided by BMT (Wardiwiyono, 2012).
In September 2010, PINBUK supervise 3,068 BMTs and financing total 1,67 billion IDR. The total number of BMTs in this year is more than 3,300 BMTs. They attracted to Indonesia civils because BMTs are generally unregistered under the normal practice rules of Bank Indonesia (the Central Bank) and they developed significantly in quantity and spread in wide areas, especially in rural and played significant role in bridging the access of financial services to the poor and SME (Dewanti, 2013). The presence of BMT accommodates the Muslim community's aspirations amid the anxiety towards economic activities with the usury principle and supports funding to develop small and medium enterprise empowerment activities. On the one hand, BMT runs the mission of the sharia economy and carries out the duty of the populist economy by improving the micro-economy. Micro-Enterprises (MEs) were the biggest industrial component in Indonesia, with 98.79% of the whole industry (1.11% small industry, 0.09% medium industry, and only 0.01% large industry) (Statistics Indonesia, 2014). However, with the industry's complexity and scope still limited, micro industries tend to experience obstacles in their development. One of the factors of the non-development of MEs is the difficulty in the capital aspect. Most MEs do not recognize banks or financial institutions and have difficulty accessing them because of their complex procedures. This limitation of banking access causes the SMEs sector to become un-bankable. This condition causes the micro industries, which are generally owned by the poor, unable to grow optimally so that the poverty rate will increase, and the social gap will widen.
Based on the facts, BMT's presence is expected to provide solutions for existing problems, especially faced by small entrepreneurs, to play a maximum role in supporting Indonesia's building and economic fundamentals. Naqvi (1981) stated that the development of MEs (for poverty alleviation) requires the provision of equitable capital. Islamic law not only determines the choice of individuals or groups but also builds the integration of both.
Islamic teachings set the spirit of brotherhood as the primary goal of the parties who transact (Billah, 2003). It is an essential foundation for the development of MEs because it is expected to provide mutual benefits. Furthermore, BMT's existence is useful for freeing the people from the usury system and upholding justice, it is also expected to be a bridge for the development of MEs and poverty alleviation. This research gives new contribution for the previous literatures on the effect of BMT presence in the community as microfinance with sharia principles on household welfare, especially the actor of MEs and farming in Indonesia. The remainder of the study was divided into the following sections. Section 2 discusses the literature review, section 3 studies the methodology, and section 4 is the empirical finding. Finally, in section 5, the researchers summarized and concluded the findings.

Research Method
This study used panel data from the Indonesian Family Life Survey (IFLS) wave four (2007) and wave five (2014). IFLS is longitudinal data conducted by RAND Corporation in collaboration with SurveyMETER, representing 83 percent of Indonesia's population. This data contained detailed information about the individual, household, and community characteristics, including information on the existence and accessibility of BMT at the community level. This study focused on the effect of BMT presence in the community (within the village/kelurahan) on household welfare measured by households' net income from farming and non-farming business activities. The researchers chose those groups because BMT's main customer target is business actors, especially small/micro businesses in farming and non-farming activities. This study's outcome was adjusted to the rate of price increase (inflation) in the second period of study, in 2014.
BMT is established in a specific area without random placement. BMT's existence considers several factors that support the BMT's continuity, such as potential market share and customers' availability. Thus, the region's characteristics that could potentially be the reason for BMT's establishment should be controlled. Population density is one of the most substantial reasons for BMT presence in the area. The high density of the population is a good signal to develop economic activity. On the other hand, the economy at the regional level may be a consideration for economic actors to establish a financial institution, such as BMT. However, since microfinance institutions are designed for the middle-lower class, the regional economy level is not very important to be controlled.
In order to control other factors that potentially affected household welfare changes, the researchers involved variables of household and community characteristics, such as household size, urban, female household head, household asset ownership, highest education in the household, and population density in the community, where the household lived. This study's sample consisted of two groups: the farmer household group consisting of 4642 households and the household with non-farming business activities amounting to 2250 households. The following is a statistical description of the farmer household group. In general, household characteristics in Table 1 had no significant differences except the economic level and the population density. Households who lived in communities with BMT had lower net income and asset ownership than non-BMT households. Besides, BMT was also more likely established in communities with relatively low density. From the 4642-household sample, 247 households stated that there was at least one BMT in the community. Meanwhile, for households with non-farming business activities, there were 2250 samples, and 173 of them stated that there were BMTs in their community. The remaining 2077 stated that there were no BMT. The descriptive statistics of the group are presented in Table 2. The pattern of household composition in this group was almost the same as the previous group. However, for non-farm businesses, households with BMT had higher welfare compared to non-BMT households. It could be seen from the mean of business net income and ownership of household assets. Comparing the two groups in Table 1 and Table 2 shows that BMT was established chiefly in communities whose farming production was relatively small and its non-farming enterprises were relatively advanced. The mean asset ownership supported this statement.
The method employed in this study was the difference-in-difference (DiD). This method assumes a time-invariant unobserved heterogeneity. Basically, DiD compares treatment groups and control groups before and after the intervention period. Panel data consists of two periods with setting t=0 before intervention and t=1 after intervention and and are the treatment and control groups' outcomes. The DiD equation can be written as follows.
= ( 1 − 0 | 1 = 1) − ( 1 − 0 | 1 = 0) (1) Where 1 = 1 means treated household in t=1 and 1 = 0 denotes untreated household. In this study, intervention refers to the presence of BMT in the community. This research's primary focus was to know the changes in household welfare before and after BMT's availability in the community. Households who lived in communities with BMT availability in their communities in 2014 were treatment groups, whereas households that did not have BMT in their communities either in 2007 or 2014 were control groups. Involving some control variables, the DiD estimate could be calculated using a regression framework. The estimating equation is specified as follows: To ensure that the regression results were not overestimated, re-regression was done by adding other financial institutions' presence in the model. In this study, larger financial institutions' existence was measured with the existence of Bank Rakyat Indonesia (BRI) in the community. The reason behind choosing BRI was that BRI is a large banking financial institution with equal distribution in Indonesia.

Result and Discussion
Regression analyses were carried out to examine the changes in household's welfare. The regressions' basic set included a year dummy, a BMT dummy, and interaction BMT*year, which gave the difference-in-difference. The second set of regression incorporated some control variables with the same method. As mentioned in the method sub-section, DiD could be calculated with equation (1) and equation (2), which produced the same result. Firstly, the analysis was performed for the sample that contained farmer households. Farmer households who lived in a community with at least one BMT became the treated group, while farmers without BMT in their community became the control group. Table 3 displays the diff-in-diff results on farmer household: A positively significant coefficient in diff-in-diff provided the information that the availability of BMT raised farmer's welfare. Moreover, the significance level remained stable at 1% even some control variables were added to the model. The coefficients were not really different; adding control variables changed the coefficients to 1.577, which signified that BMT's availability improved 1.577% of the net income from farming activities. Despite the relatively small percentage, if the researchers looked deeper into the data composition, BMT was located mainly in the low-productive area. Profit generated by the farmer in the without-BMT area was 14% higher. The negative difference in net income before intervention and more prominent outcome conditions after intervention showed that BMT's availability increased the household's welfare. Table 3 gives information that untreated households could not generate higher profits than treated households in 2007. In the following research period, however, the treated household's profit could exceed the untreated household. As BMT's main target was the middle-lower class whose majority's livelihood as a farmer, BMT's availability positively impacted the community.
To ensure that the analysis results were not overestimated, this study tried to add a variable to the existence of other larger financial institutions in the community. The results proved that there were no significant changes compared to the previous results. The sign of coefficient and significance were almost the same.
Secondly, the analysis was conducted for the sample that contained households with nonfarming business activities. The household that lived in a community with at least one BMT became a treated group, while farmers without BMT in their community became control groups. Table 4 shows the diff-in-diff results on this household: Table 4 points out that there was no significant difference in welfare changes in the household with BMT in his community and household without BMT. Although there was a significant difference between treated and control groups in the after-period, the baseline period was not significant. It produced diff-in-diff results to be insignificant.
Nevertheless, the positive coefficient of diff-in-diff indicated that BMT and household welfare had a positive trend but were not strong enough to explain the increase in household welfare.
Based on the data composition, BMT's presence tended to be located in areas where people had business with higher profit. The mean net income of households from business profit in the BMT area was almost 10% higher than households in the non-BMT area. Besides, asset ownership was also bigger, and the highest education level experienced by household members of BMT household was also better. Therefore, it is not surprising if BMT's existence was not strong enough to explain the community's increasing welfare.
There is a possibility the existence of other financial institutions potentially contributed to the increase in household welfare. Thus, the researchers conducted a similar regression analysis, but the variables related to BMT were changed to Bank Rakyat Indonesia (BRI). As explained earlier, BRI is considered to represent other financial institutions other than BMT because BRI is the most widely distributed financial institution in Indonesia. The same method showed results that were not much different from the previous analysis.
Variable of BRI interaction with the year was not significant in affecting welfare. The argument behind this result is that a more modern business climate is likely to give a more significant contribution than just the existence of financial institutions in the household's community.

Conclusion
Baitul Maal wa Tamwil (BMT) is one of the sharia microfinance institutions in Indonesia with rapid growth. As a driving force for the community's economy, microfinance institutions, in this case, BMT, generally improved welfare through business credit, both farming and non-farming. Using the data of Indonesia Family Life Survey (IFLS) waves 4 and 5, this study analyzed the effect of BMT presence in the community on welfare obtained through farming and non-farm business. Thus, this research's sample was farmer households and households with non-farming business activities.
Employing the difference-in-difference (DiD) method, the results obtained that sBMT's existence in a community improved the farmer household's welfare equal to 1.65%. Analysis involving control variables, including household and community characteristics, gave a similar result with the previous result, which was 1.58%, with the same level of significance. On the other hand, the household with non-farming business activities was not affected by BMT's existence in their community. Nevertheless, the positive coefficient of diff-in-diff indicates that BMT and welfare have been shown to have positive correlation but the existence of BMT does not strong enough to explain the increase in household welfare. There was no significant difference in welfare changes in the household with BMT in his community and household without BMT. Even though there was a significant difference between treated group (farmer households who lived in a community with at least one BMT) and control group (farmers without BMT in their