All inclusive microfinance: A study of the demand for Islamic microfinance in Malawi.

UPPSALA UNIVERSITY

Department of Economic History

Bachelor of Science Thesis Autumn Semester 2010

All inclusive microfinance

- A study of the demand for Islamic microfinance in Malawi

Lars Eriksson

Mentor: Tom Peterson

[KEYWORDS: MICROFINANCE, RELIGION, ISLAM, MALAWI, DISCRIMINATION, CREDIT CONSTRAINTS]

i

A BSTRACT

Microfinance is the number one buzz word in the development sphere nowadays. The basic

idea of microfinance is to make financial services available for those excluded from the

conventional banking system. By charging market price interest rates on the loans granted

the business is meant to become sustainable and independent of fluctuations in cash flow

from donor funding. The microfinance sector in Malawi is relatively young and still in the

development phase. Since the majority of the charity organizations running microfinance

projects in Malawi are originating from the Western world, the services these institutions

offer are inherently affected by Western (Christian) banking culture. This paper investigates

if this set up results in the exclusion of Malawi’s Muslim population due to the fact that

Islamic law prohibits Muslims from charging or paying interest on loans. The conclusions of

this thesis are that the Muslim population uses the microfinance services to the same extent

as their Christian brethren. However, a large proportion of the Muslim clientele feel that

they are morally prohibited from using the microfinance services because of the interest rate

charged. They only make use of the interest-based loans because they have no other option,

and would thereby prefer services compatible with Islamic law. My recommendation is for

the microfinance institutions to embrace this knowledge and further investigate the need of

Islamic microfinance, and the possibility to implement it, before the consequences becomes

more than a moral issue.

ii

P REFACE

Writing a student thesis as a part of a field study is an adventure enjoyed by a few selected. I had the opportunity to be one of those few. Without bragging I must say that I acquired the possibility to do so all by myself (not counting the Swedish International Development Cooperation Agency which accounted for most of the expenses, which I would like to thank for). But having a possibility to do something is far from actually accomplishing it. Writing this I haven’t actually finished my work, but at the risk of prematurely celebrating the victory I would nevertheless like to thank everyone that has helped me to (hopefully) accomplish this mostly amazing, but sometimes also Sisyphean work. I met a lot of wonderful characters in Malawi who gave me plentiful of guidance and moral boosting. It would take too long to mention you all, but you know who you are and what you did. I would of course like to thank all my respondents for making this work possible. I can only hope that any of my findings eventually will be rewarding to you as well.

Special thanks to my mentor Tom Petersson who’s fierce comments has been like a pitchfork forcing me in the right direction. Who wouldn’t become motivated from feedback like “What did you think of here?” and “read a book and rewrite”? Seriously Tom, you have done a great job and I wish you all the best. Special thanks also to Kyle Scott and Stephen Fox. In you I found not only a safe haven when work was too cumbersome, but hopefully also a lifetime of friendship (when I come to think of it and for the credibility of this thesis, let’s just call it a haven). Thanks mom for letting me go without you having a nervous breakdown, and thanks dad for preventing mom from having a nervous breakdown. And last but not least, thank you Annika for all your support. I wouldn’t have managed without you.

Uppsala, September 2010

Lars Eriksson

iii

C ONTENTS

Chapter 1: Introduction to the Thesis ... 1

1.1 Why Malawi? ... 2

Chapter 2: Thesis Setup ... 4

2.1 Purpose ... 4

2.2 Research Question ... 4

2.3 Delimitation ... 4

2.4 Glossary ... 5

Chapter 3: Research Position... 6

3.1 Microfinance ... 6

3.2 Islamic Microfinance ... 8

3.3 Microfinance in Malawi ... 12

Chapter 4: Thesis Framework ... 14

4.1 In Depth Interviews ... 14

4.2 The Survey ... 15

4.3 Model for Determining Credit Constrains ... 17

Chapter 5: Fieldwork and Results ... 20

5.1 Results of the In Depth Interviews ... 21

5.2 Survey Results ... 22

5.3 Validation of the Survey Data ... 29

Chapter 6: Discussions ... 31

Chapter 7: Conclusions ... 35

References ... 36

Appendix I: Calculations ... 39

Appendix II: Survey Codebook ... 42

Appendix III: Survey Data ... 48

1

C HAPTER 1: I NTRODUCTION TO THE T HESIS

Probably more or less everyone is of the opinion that it stands to reason that rich and

developed countries assist the third world with economic and developmental assistance. But here coherence ends. Recently, traditional principles of aid have received scathing criticism saying it’s more than often is given out of self-promoting and political motives rather than of altruistic concern for the beneficiary. Furthermore, it’s alleged that traditional aid might even be counterproductive for the host countries’ long-term development.

Simplified the argument is that traditional aid deprives people of poor countries the will and ability to provide for themselves. In addition, a good portion of the provided funds tend to end up in corrupt politicians’ pockets instead of where it was originally intended.

Microfinance is an alternative to traditional aid that’s received much acclaim in recent times.

The United Nations appointed 2005 to be the International Year of Microcredit and the following year the Grameen Bank and its founder, Muhammad Yunus, were awarded the Nobel Peace Prize for their pioneering efforts to provide microfinance for the impoverished in Bangladesh. Briefly, the main idea of microfinance is to provide banking services to the poor to enable them to invest themselves in a better future. The purpose of this is to get people involved in their own development process rather than imposing on them a complete plan of action.

With this recent attention, there’s a risk of developing an over-reliance in the powers of microfinance. In the eagerness to use this tool in more and more poverty-stricken areas, I see a possibility that insufficient investigations are being made to ensure that the model is adapted enough to function under the premises that exist in the specific area. Take the Malawi Mudzi Fund as an example. This project was launched is Malawi in 1990 and was the first major attempt to replicate the successful Grameen Bank model in a new location.

Initially the project worked better than expected, but as time went on repayment rates fell steadily and in 1994 the project was shut down. An explanation presented to the repayment- failing levels was that, because of earlier relief efforts the perception amongst the

1 Asante, S.K.B., International Assistance and International Capitalism: Supportive or Counterproductive?, 1985, p. 249-265.

2

population was that the loans didn’t have to be repaid.

A difference in attitude towards loan repayment among the citizens in the different areas might therefore have been the sole factor why the concept worked in Bangladesh but not in Malawi.

There’s a considerable risk of the world’s poor being treated as a homogeneous group of people. But even though they have the dubious pleasure of being part of the not so exclusive group of the impoverished, this doesn’t necessarily imply that they all have identical needs for financial services. As to further complicate the situation, many of the world’s poor areas are also inhabited by a heterogeneous population with different cultural expressions and needs. This entails a risk of excluding certain groups of society from the microfinance market unless taken account for, or as stated in the World Bank Publication The Microfinance

Handbook:

“There are cases in which a certain group within a community cannot or will not take part in a financial services project due to a religious, ethnic, or other social influence. It is important to understand these restrictions when identifying a target market, so that products and services can be developed that takes into account the limitations of some groups”.

1.1 Why Malawi?

First and foremost, Malawi is one of the world’s poorest countries.

Over 80 percent of the population lives in rural areas and a full 21 percent of the residents live below the

established poverty line set at an income of less than one dollar a day. But at the same time the political situation in the country is more or less stable and the state apparatus is

functioning relatively well.

The conditions for microfinance institutions (MFIs) to operate in Malawi are thus at least theoretically good. And indeed, many MFIs have already established themselves on the Malawian microfinance market, but business is still in an early stage of development. In 2007, the United Nations Development Programme (UNDP), together with Malawi’s Ministry of Finance launched the Financial Inclusion in Malawi (FIMA) as a way to

2 Pal, M. S., Building African Institutions: Learning from South Asia, 1999, p. 145-147.

3 Ledgerwood, J., Microfinance Handbook: An Institutional and Financial Perspective, 1999, p. 41.

4 Ranking 160 out of 182 in the Human Development Index 2009, available at http://hdr.undp.org/en/statistics/

(2010-06-14).

5 Background Note: Malawi, Bureau of African Affairs. U.S. Department of State, 2009, available at http://www.state.gov/r/pa/ei/bgn/7231.htm (2010-06-14).

3

promote microfinance business in the country. The project will continue until 2011 and aims to establish a special microfinance unit within the Ministry of Finance with responsibility for developing and implementing a national strategy for a financial system that through

microfinance includes the entire population.

FIMAs developmental work includes trying to identify causes why individuals, or entire segments of the population, might be excluded from the microfinance system, either by choice or by being discriminated against by the MFIs. Reasons for possible exclusion include size of business, type of business, turnover and the like or more individual-based factors such as sex, language, religion or ethnicity. The Malawian population consists of many different ethnicities and religions. This cultural melting pot has given rise to numerous variations, both between and within different regions of the country.

Perhaps as a reflection of the prevailing tendencies in Western politics, when designing microfinance, focus has been on gender issues and how the instrument can be used to strengthen women’s rights. However, an issue that hasn’t been taken into account

sufficiently is how religion affects participants in a micro-financial context. Even though we live in a secular society in the West, religion still has a significant effect on everything from daily life to national policy in many of the countries where microfinance is practiced.

About 20 percent of the Malawian citizens are Muslims, making Islam the country’s second largest religion after Christianity.

Muslims are under Islamic religious law (Sharia) prohibited from paying or accepting interest on loans or savings. Still, none of Malawi’s leading MFIs offer financial services customized to Sharia in addition to their interest-based loan and savings services. If Malawi’s Muslims obey the ban of interest, one fifth of Malawi’s

population would be prohibited from using microfinance services. If so, this would primarily be devastating for those directly affected. But the economic development of the nation as a whole could be kept back or even fail to come as a consequence of that a segment of the population systematically is excluded from one of society’s primary functions.

6 Financial Inclusion in Malawi – Micro Finance, UNDP in Malawi, 2009, available at

http://www.undp.org.mw/index.php?option=com_programme&view=Poverty%20Reduction&task=project&ar ea=project%20area&id=10 (2010-06-14).

7 Background Note: Malawi, Bureau of African Affairs. U.S. Department of State, 2009, available at http://www.state.gov/r/pa/ei/bgn/7231.htm (2010-06-14).

4

C HAPTER 2: T HESIS S ETUP

As much as I would like to skip ahead to the interesting revelations of this thesis, for the sake of academic procedure I’ll better start off with the formalities. Also, it cannot be denied that academic procedures comes to hand for the reader who wants to know what the author intend to achieve and for the author to know if he actually did it.

2.1 Purpose

The purpose of the government and private institutions engaged in microfinance in Malawi is undeniably good. But the eagerness to help the vast majority of the population has overshadowed the possible risk of alienating one of Malawi’s minority groups even more.

The charity work done in Malawi is mostly funded and conducted by Western relief organizations. The purpose of this thesis is to highlight the problems that can arise when only offering banking services of western custom in a partly Islamic society. My intention is not so much to criticize the work done by the MFIs in Malawi, but rather to contribute to increased knowledge of the socio-economic premises in which the microfinance business operates. My hopes are that this thesis will be helpful to all parties involved. An extended integration of the Muslim population into the microfinance business not only helps these new clients, but also helps expanding the microfinance circle of customers. Ultimately, such an expansion would also be passed on to the already integrated communities as increased capital in the business leads to increased capacity to meet the financial needs of the entire population.

2.2 Research Question

The purpose of this thesis can be transformed into the following research question:

- Is there a need for an Islamic alternative to the interest-based services offered by the microfinance institutions in Malawi, and if so, what are the possibilities for such an implementation?

2.3 Delimitation

This thesis is the result of a field study conducted in Malawi in the spring of 2010. I gathered

my empirical material in the four regions of Lilongwe, Nkhotakota, Salima and Mangochi. In

these regions the population consists of a relatively high proportion of Muslims. The results

5

may be assumed to be applicable to other regions with a similar demographic profile. For practical reasons I have not studied the regions with a lower proportion of Muslims, which brings a possibility that the situation in these areas isn’t consistent with my conclusions.

The subject of microfinance has a wide range of possible fields of study. I have chosen to concentrate on the impact of Islamic religious law on the access to microfinance services for the Muslim minority in Malawi. Microfinance institutions have a diverse range of services. I have chosen to study only the interest-based loan services because it is a service that all MFIs offer, and is primarily associated with microfinance, but also because it is a service that Sharia explicitly prohibits.

2.4 Glossary

Field Officer is the denomination of an official who works at a local MFI office managing the loan activity. Field Officers are often recruited locally.

Financial Inclusion in Malawi (FIMA) is a development project launched by UNDP with the objective to develop and implement a national strategy how through microfinance expand the country’s financial sector and give all citizens access to financial services.

Halal is Arabic for permissible. The term is used for goods or services that are authorized under Islamic religious law. The term is often associated with the rules of eligibility for food, but actually covers all human action.

Haram is Arabic for forbidden. The term is used for goods or services which are banned under Islamic religious law. It Includes, amongst others, the prohibition to eat pork or drink alcohol, but also to charge or pay interest on a loan.

Microfinance Institution (MFI) is a commercial or non-profit organization that offers micro- financial services.

Riba is Arabic for usury. Riba includes charging or paying interest on a loan and is prohibited according to Islamic religious law.

Sharia is Arabic for road or trail. It is also the term used for the Islamic religious law. Sharia is

based on texts from The Quran about Muhammad’s living and learning.

6

C HAPTER 3: R ESEARCH P OSITION

Microfinance has in the recent 10 to 15 years been a hot topic of discussion in the socio- economic development sphere. Therefore, there are a great many publications available on the subject. Furthermore, there’s also plentiful available concerning Islam’s ban of interest and its impact on economic development. Many of the Muslim countries have since the oil discoveries in the Arab world experienced a huge economic boom and a well-functioning banking system has therefore been crucial. Initially Muslim intellectuals urged for the acceptance of the Western (interest-based) banking system in order not to hamper the development. But above all, considerable efforts have been made to develop an alternative financial system that’s compatible with Sharia, but also suitable for banking in a global economy.

For those interested in immersing themselves in the theoretical framework of microfinance and Islamic microfinance there’s a lot of very good and fundamental literature available. But to help the uninitiated reader follow my line of thoughts through this essay, I have still chosen to briefly review the basic theory needed and at the same time resume the research position of the subject.

3.1 Microfinance

The idea of microfinance is to give impoverished people the securities and opportunities which follow with having access to financial services. This allows them to use credit for investments or current consumption, and saving to store an occasional surplus for future needs. These possibilities help to reduce financial risks and at the same time provide the opportunity to invest in small scale businesses to increase the revenue.

Poor people are usually excluded from the conventional financial institutions because they don’t meet the standard requirements of financial security, steady income, material collateral etc. Through alternative ways of collateral (e.g. group loaning where group members are responsible for each other’s repayment), microfinance circumvents the security problem in a simple, yet functioning manner.

8 Robinson, M. S., The Microfinance Revolution: Sustainable Finance for the Poor, 2001, p. 9.

9 Ledgerwood, J., Microfinance Handbook: An Institutional and Financial Perspective, 1999, p. 137-138.

7

Unlike traditional aid, the ambition with microfinance is to make it operate under market conditions. This is stated to be the only way to achieve a sustainable wide outreach, and ultimately phasing out the dependence on external assistance.

To make this possible, microfinance customers are (in conformity with conventional banking practice) supposed to pay market-based interest rates on their loans. This to build sustainable institutions that can promote long-term development regardless of fluctuations in external funding. An additional argument often invoked to the benefit of microfinance, is that it acts directly where it’s most needed, which reduces the risk of embezzlement and corruption.

Microfinance can with a little imagination be likened to an economic revolution as development is expected to come from the lower strata of society and make its way up in the hierarchy instead of vice versa.

Assistance in the form of microfinance delegates much of the development responsibilities to local population. As a result of this, it is only fruitful to engage in microfinance where it is possible for people to invest the loan in a way that can provide enough profit.

It is

therefore essential that there is a degree of economic-political stability in the region to make profitable long time investments possible.

Microfinance have because of these reasons received some criticism saying it’s not able to help the poorest of the poor, who usually find themselves in a situation where they don’t have the opportunity to invest in any business sector activity.

At worst, there’s a risk that microfinance clients get caught in a debt trap because of their investment not generating enough income to cover the repayment with interest.

Some critics even claim that a suicide epidemic amongst farmers in India was a result of them not being able to repay their microfinance loans.

Because of this, it’s crucial to use microfinance responsibly to achieve a satisfying result.

10 Robinson, M. S., The Microfinance Revolution: Sustainable Finance for the Poor, 2001, p. 46-49.

11 Ibid. p. 71-74.

12 Ibid. p. 20.

13 Ledgerwood, J., Microfinance Handbook: An Institutional and Financial Perspective, 1999, p. 26.

14 Obaidullah, M., Introduction to Islamic Microfinance, 2008, p. 10.

15 Robinson, M. S., The Microfinance Revolution: Sustainable Finance for the Poor, 2001, p. 20.

16 Sharad, J., The Other side of micro-finance, The Hindu Business Line, 2007 available at

http://www.thehindubusinessline.com/2007/09/05/stories/2007090550080900.htm (2010-09-14).

17 Orissa: Loans driving farmers to suicide, NDTV, 2010, available at: http://www.ndtv.com/news/india/orissa- killing-fields-killing-finance-18040.php (2010-09-14).

8 3.2 Islamic Microfinance

Islamic religious law (Sharia) prohibits the use of interest (riba) when loaning or lending money.

Sharia is derived from Islam’s holy book, The Quran, and the Prophet Muhammad’s living and teachings.

The Quran explicitly prohibits interest (or usury) on several occasions.

Examples of this can be seen in the following verses:

“O you who have believed, do not consume usury, doubled and multiplied, but fear Allah that you may be successful.”

“Allah has permitted trade and has forbidden interest.”

Also the Prophet Muhammad treats the subject in his last sermon:

“God has forbidden you to take riba, therefore all riba obligations shall

henceforth be waived. Your capital, however, is yours to keep. You will neither inflict nor suffer inequity. God has judged that there shall be no riba and that all the riba due to `Abbas ibn `Abd al Muttalib shall henceforth be waived.”

In short Sharia sees interest as a tool of oppression and a means of usurping others’ assets in an equitable manner. By requiring interest on loans, society’s resources accumulate with the already wealthy. This leads to economic inequalities with poverty as a result, which is

incompatible with Islamic equity.

However, this prohibition doesn’t mean it’s impossible to adjust interest-based microfinance to make it compatible with Sharia. On the contrary, The Quran contains segments that mandate for approaches similar to microfinance to curb poverty. Islam promotes charity from the donor’s point of view, but simultaneously wants to minimize the recipient’s dependence on charity. The following story of the Prophet Muhammad’s living exemplifies this well:

18 Obaidullah, M., Islamic Financial Services, 2005, p. 22.

19 Obaidullah, M., Introduction to Islamic Microfinance, 2008, p. 13.

20 The Quran, 3:130.

21 The Quran, 2:175.

22 Last Sermon of Muhammad given on 10 Dul-hajj 10 hijra, mentioned in all book of Hadith.

23 Obaidullah, M., Introduction to Islamic Microfinance, 2008, p. 20.

24 Obaidullah, M., Islamic Financial Services, 2005, p. 22-28.

9

“A man of the Ansar came to the Prophet (peace be upon him) and begged from him.

He (the Prophet) asked: Have you nothing in your house? He replied: Yes, a piece of cloth, a part of which we wear and a part of which we spread (on the ground), and a wooden bowl from which we drink water.

He said: Bring them to me. He then brought these articles to him and he (the Prophet) took them in his hands and asked: Who will buy these? A man said: I shall buy them for one dirham. He said twice or thrice: Who will offer more than one dirham? A man said: I shall buy them for two dirhams.

He gave these to him and took the two dirhams and, giving them to the Ansari, he said: Buy food with one of them and hand it to your family, and buy an axe and bring it to me. He then brought it to him. The Apostle of Allah (peace be upon him) fixed a handle on it with his own hands and said: Go, gather firewood and sell it, and do not let me see you for a fortnight. The man went away and gathered firewood and sold it. When he had earned ten dirhams, he came to him and bought a garment with some of them and food with the others.

The Apostle of Allah (peace be upon him) then said: This is better for you than that begging should come as a spot on your face on the Day of Judgment.

Begging is right only for three people: one who is in grinding poverty, one who is seriously in debt, or one who is responsible for compensation and finds it difficult to pay.”

Muhammad asks the beggar to use his existing assets to build wealth instead of relying on handouts. This mindset is similar to the case of microfinance. But since microfinance is supposed to operate under market economy conditions, the MFIs need to somehow

generate an income of the business. This is usually solved by charging interest on the loans, which runs counter the Sharia prohibition of riba. Fortunately, great efforts have been made to develop an alternative Islamic banking system that among other things enables lending

25 Sunan Abu Dawood, Kitab al-Zakah, Book 9, Number 1637.

10

institutions to charge their clients without violating the Sharia law. These alternative

solutions are not restricted to loans, but also include services such as savings, insurance and payment transactions. But since the scope of my essay is limited to loans this is the only of the Islamic banking services I will discuss further.

Somewhat simplified, Islamic theory regards a loan as a partnership between lender and borrower, because both parties are supposed to benefit from the deal. The loan provider benefits from the deal because he or she may charge the borrower for the service, while the borrower benefits from the deal because of the opportunity to invest the borrowed money in a business to increase its future yield. Both parties are hereby interdependent for making a profit, and according to Islam both parties should thereby share the financial risks.

According to Islamic financial theory, the risks taken by conventional banks when granting an interest-based loan aren’t sufficient for legitimizing its profits.

If some individuals or

corporations charge interest for lending money without sharing enough risk, this

accumulates a majority of the financial means in the pockets of a handful, while a majority of the citizens are forced into a financial deficit. The fact that some live in luxury while many others live in poverty is not compatible with the Islamic vision of justice and thus it is not permissible to take interest on a loan. This egalitarian approach to the lending of money has been important in the shaping of Islamic banking.

One theoretically elegant solution is that the bank decides whether a client’s investment plan is good enough. If so, the bank provides money for the investment in exchange for a share of future returns, if any. In this way, the borrower isn’t forced to repay the loan if the investment goes to nothing. But on the other hand, the repayment grows if the investment turns out to be very successful.

This procedure (called Musharaka or Mudarabah

depending on whether the client also invests own capital or not) is in practice suitable for larger loans because it requires the bank to investigate the client’s investment plan to a greater extent in comparison with a traditional interest-based loan since the repayment of the loan is directly dependent of if the investment plan is solid or not. In addition, the bank

26 Seibel, H. D., Islamic Microfinance in Indonesia, 2005, p.5.

27 Obaidullah, M., Introduction to Islamic Microfinance, 2005, p. 47.

28 Interview with Mr. A.Y. Qadri, imam at the city mosque, Lilongwe, 2010-01-20.

29 Obaidullah, M., Islamic Financial Services, 2005, p. 57-60.

11

must continuously monitor how the investment is progressing to ensure that the right proportion of the profit will reach the bank.

These factors increase the administrative costs of a loan drastically which makes this approach not very suiting for microfinance, which is already stuck with high service costs (interest-based or not) due to a high cost per loan ratio.

The administrative costs of microfinance becomes large because of the simple fact that it takes a lot more time to issue for instance 1000 loans of $ 100 each, compared to a single $ 100 000 loan.

Due to these prevailing circumstances, developers of Islamic microfinance have been forced to look for other ways to enable a commercially viable business. One of the most popular and most useful approaches for Islamic microcredit is the so called Murabaha loan. This loan works in such a way that the client informs the bank that he needs to buy a certain product or service. The bank then buys the product or service from a third party for a price known by the client. Then the bank sells the product or service to the client for a higher price than the bank paid for it, where the difference becomes the bank’s earnings. The client’s benefit is that the payment for the product or service is made at a later time or can be made as installments over an agreed period of time.

There are several other approaches to solve the lending of funds in accordance with Sharia, but Murabaha is the method that’s had the greatest breakthrough in other geographical areas because of its simple design for both lenders and borrowers. This makes it a likely candidate for a possible implementation of Islamic services in Malawi.

Since alternative solutions to interest-based banking already exists, the step towards Islamic microfinance shouldn’t be far. Implementation of non-interest-based services has in fact already been made in countries with a high proportion of Muslim inhabitants, such as Indonesia

and Afghanistan

. But in Hans D. Siebel’s paper Islamic Microfinance in Indonesia from 2006 he clearly illustrates the difficulties and hazards of implementing an Islamic financial system next to the conventional one. According to his findings, only the conventional banks seem to be able to master the art of Islamic banking, but at the same

30 Obaidullah, M., Introduction to Islamic Microfinance, 2005, p. 52-53.

31 Obaidullah, M., Introduction to Islamic Microfinance, 2005, p. 46-47.

32 Seibel, H. D., Islamic Microfinance in Indonesia, 2005.

33 FINCA International - Afghanistan, available at http://www.finca.org/site/c.erKPI2PCIoE/b.2676443 (2010- 06-24).

12

time they lack the experience needed to be able to practice microfinance. The Islamic MFIs on the other hand, have not been able to develop efficient and sustainable businesses due to many of the managers being retirees from conventional banking, with little or no

knowledge of Islamic banking principles.

In addition to this, there seemed to have been a lack of demand for Islamic banking in Indonesia when these services were first introduced.

According to surveys, out of a 97% Muslim population only 11% felt the need for Islamic banking and little has shown that there should have been any change in demand since.

The problem with the implementation of Islamic microfinance in Indonesia makes it obvious that the complexity in such a venture is more than what first meets the eye. This gives me yet another confirmation on the importance of my research, since the affirmation of a need for Islamic microfinance in Malawi is crucial not to repeat the Indonesian mistakes. If Islamic services were to be implemented in Malawi without an actual need for it, this could

potentially have disastrous results for the MFIs and their clients. Offering services without a demand only brings further administrative costs without making any profit from it. At worst, this could make some MFIs going bankrupt, with the loss of the clients’ savings as a result.

3.3 Microfinance in Malawi

Microfinance in Malawi is growing. The first operations were established already in the 1990’s and have by now reached over twenty major stakeholders, both governmental and private. Between March and June 2006, the number of loans issued went from 158 918 to 182 099, an increase of over 20 000 loans in only three months.

But it’s still long before the financial sector matures. Most institutions suffer from limited resources, which in turn limit the number of loans they can issue. Geographical penetration is not much deeper than larger cities and secondary towns whilst over 80 % of Malawi’s population lives in the rural areas.

In addition to the MFIs, there are also some associations supporting microfinance operations in the country, such as previously mentioned FIMA, and the Malawi Microfinance Network (MAMN) which is a joint organization for the microfinance institutions. MAMN had at the

34 Seibel, H. D., Islamic Microfinance in Indonesia, 2005, p. 12-14.

35 Seibel, H. D., Islamic Microfinance in Indonesia, 2005, p. 10.

36 USAID Malawi, Malawi Microfinance Network (MAMN), 2007.

37 Burritt, K., Expanding Access to Financial Services in Malawi, 2006, p. 88.

13

end of 2007 as many as 20 MFIs connected to their network.

None of the information I have found about these MFIs or the support organizations suggests that there is any Islamic microfinance conducted in the country or the need for it being discussed. I have searched the websites of the organizations that have one,

and used the Google search engine with keywords ”Islamic Microfinance Malawi” and “Islamic Banking Malawi” to find information about other MFIs that might offer Islamic services. The only thing that was close to what I was looking for was the above mentioned Malawi Mudzi Fund (which no longer exists), and the charity organization Islamic Relief, which has had some activities in Malawi, but not any microfinance according to information on their website

. However, a number of

organizations haven’t published a website it’s possible that they still provide Islamic

microfinance, but that the information about it isn’t available in the Internet. What’s actually the case was something I simply had to investigate on site in Malawi.

38 The Africa Microfinance Network – MAMN Malawi, available at http://afminetwork.org/members_mamn- malawi_70.html (2010-06-14).

39 Organizations whose websites I have searched for information about Islamic Microfinance: Malawi Savings Bank, Concern Universal Microfinance Operations, FINCA Malawi, Microloan Foundation and Opportunity Bank Malawi.

40 Islamic Relief website, Islamic Relief in Malawi, 2009 – 2010, available at http://www.islamic- relief.com/wherewework/9-MW-malawi.aspx (2010-06-14).

14

C HAPTER 4: T HESIS F RAMEWORK

If (contrary to the picture I got from my initial research) the Muslims of Malawi already had access to Islamic microfinance, investigating if there was a need to introduce it would immediately be proven rather unnecessary. Since no comprehensive information about the various activities of the MFIs was available I had to find it out for myself. Luckily, I had the opportunity to spend a period of ten weeks in Malawi in order to collect the materials needed to answer my research question. The first thing to do on site in Malawi was to find out what the economic situation really was for the country’s Muslim population.

I began my fieldwork with a series of interviews with representatives from FIMA, field officers from some of the larger MFIs and scholars within the Muslim community. I did this partly to ascertain whether or not the MFIs in Malawi actually offered Islamic banking

services, but also to gain a more thorough understanding of the factors in both Islam and the microfinance business that were the basis for the problems. At the risk of being premature (but at the same time to motivate further reading) I choose to reveal that none of the interviewees were aware of any MFI in the country that offered Islamic services. I was thereby able to dismiss the fear of my research question being superfluous.

When I knew that none of the present MFIs offered Islamic services, I still had to find out if there was actually a need for it. Most of the time on site was therefore devoted to a survey conducted in the segment of the population that’s the target group of microfinance, i.e.

poor people with the premise to start some kind of business, or small-scale merchandisers who want to invest in order to increase business yields. The data collected through the survey was then used for a statistical study to examine the need for Islamic microfinance in Malawi. As a basis for the statistical study, I used a model developed to investigate the differences in credit constraints between different ethnic groups on the Zimbabwean debt market. I modified the model in order to make it suitable for testing credit constraints in a microfinancial context depending on religion.

4.1 In Depth Interviews

To get a deeper knowledge of the area studied, I started my fieldwork with a number of

qualitative interviews with key figures of the different stakeholders on the Malawian

microfinance market. Since these interviews partly addressed a topic that I possessed little

15

knowledge of, I felt that it was appropriate to let the interviews follow a semi-structured approach. This not to be restricted to interview questions constructed with insufficient knowledge on the subject. Before each interview session, I prepared a number of wider issues or topics that I wanted to broach. As the interviews progressed they tended to become more of a discussion and I had to improvise my questions depending on where the dialogue took us. Of course there are a number of flaws coming with this kind of

“insufficiently” prepared interviews. But in hindsight I can truly say that the results of these interviews were sufficient for proceeding with my fieldwork, and thus this rationalization gave me more time to spend on the survey.

4.2 The Survey

The Muslim population in Malawi is not evenly distributed across the country, but the majority live along Lake Malawi’s coastline and in the larger cities. In the district with the largest share, as much as 72 percent of the citizens are Muslims.

This is quite descriptive to the uneven distribution that’s prevailing when the country’s overall population consists of about 20 percent Muslims. Since I was interested in exploring the Muslim population’s attitude to, and use of microfinance, I chose three districts with a high percentage of Muslim residents where to conduct my survey. To make my questions relevant to the respondents, it was also essential that MFIs was present in the selected regions.

Using demographic data from the National Statistical Office of Malawi, I sorted out districts with a high proportion of Muslim inhabitants in order to collect the statistical data needed to proceed with the investigation.

Thinning ultimately resulted in the districts of Mangochi in southern Malawi, and Nkhotakota and Salima in the central region. Northern Malawi has almost no Muslim inhabitants and therefore no district was selected here. Nkhotakota is a historically significant place for Malawi’s Muslim population. This is from where the Arabs (who brought the Islamic faith to Malawi) managed the slave trade over Lake Malawi and to the Arab world.

The overall percentage of Muslims in the selected districts was about 52

41 National Statistical Office of Malawi, Population Characteristics, 2008, available at

http://www.nso.malawi.net/data_on_line/demography/census_2008/Main%20Report/Statistical%20tables/Po pulation%20Characteristics.xls (2010-06-14).

42 National Statistical Office of Malawi, Population Characteristics, 2008, available at

http://www.nso.malawi.net/data_on_line/demography/census_2008/Main%20Report/Statistical%20tables/Po pulation%20Characteristics.xls (2010-06-14).

43 Briggs, P. and Bartlett, M., Malawi: the Bradt Travel Guide, 2006, p. 143-145.

16

percent, which makes the distribution between Muslims and non Muslims close to ideal for my study.

The next step in the preparatory work for the survey was to determine an approach for selecting respondents from the population. At first I had the intention to use civil registers or similar to randomly select people for the survey to get a simple random sample.

Unfortunately I had to reject this idea due to practical circumstances when I realized that the work required to try to locate each individual would be too time consuming. In addition, the number of responses needed would become very large because many of the respondents would fall outside the framework of criteria that were of interest to me. So, of pragmatic reasons, I had to bend a bit on the laws of random sampling which usually surrounds the art of collecting statistical data.

The micro-enterprises that were of interest for my study were often located in designated market places. Instead of trying to locate specific individuals on a predetermined schedule, I simply conducted my survey among the various companies in these clusters and tried my best to keep an even distribution between trades, location, sex, age etc. In the case of religious affiliation the criteria of random sampling solved itself since in most cases the Muslim respondents didn’t use any physical attributes which, at least I, associate with believers of the Islamic faith.

I can find many valid objections to this selective sample of respondents. And I am the first to admit that this approach contains flaws. It may well be the case that my coming conclusions aren’t representative for the population as a whole. But given my restrictions in time and financial support, I find these deviations from normal procedure to be motivated. To validate the quality of my data material I have performed several statistical tests, showing that there are no significant differences in my data between Christians and Muslims when it comes to factors like educational level or annual profit. This shows that my data material is unbiased and the risk of hidden dependent factors that affect the results are minimal. The

implementation and results of these tests are shown in Chapter 5: Fieldwork and Results.

17

4.3 Model for Determining Credit Constrains

In the article Explaining Ethnic Differentials in Credit Market Outcomes in Zimbabwe from 1999, authors Mayank Raturi and Anand V. Swamy uses a method to investigate how clients’

ethnicity affects their possibility of getting loans.

Since this study was methodologically similar to the study I wanted to conduct, I decided to further develop this model to gain my purposes.

The model defines a formula for calculating the probability that a company owned by an individual with a specific ethnicity is limited in its ability to be granted a loan. Since

microfinance targets individuals and micro-enterprises, I have chosen to use the term person instead of firm that is used in the original model. The following variables are defined: W = person wants a loan, A = person applies for a loan, NA = person does not apply for a loan, D

= person is denied a loan and R = person is credit constrained. The formula is described by the following equation, which states that the probability of a person being credit constrained is the sum of the probability of wanting a loan and not applying for one and the probability of applying for a loan and not getting it:

. (1)

Since the probability of not applying for a loan if you want one is the same as one minus the probability of applying for a loan if you want one, i.e. P(NA|W) = 1 - P(A|W), the equation can be developed as follows:

. (2)

The probability of a person applying for a loan, P(A), can be rewritten as P(W) * P(A|W).

Inserted in equation 2 this result in:

. (3) Solving for P(W) gives us the final statement:

. (4)

44 Raturi, M., Swamy, A. V., Explaining Ethnic Differentials in Credit Market Outcomes in Zimbabwe, 1999, p.

586-589.

18

It becomes obvious that the probability of being credit constrained depends on three

factors. To begin with, P(R) increases if the person belongs to a group that to a higher extent want loans, i.e. if P(W) increases. Further, P(R) increases if the individual’s group has a lower probability of applying for loans even if wanting one, i.e. if P(A|W) decreases. Finally, P(R) increases if the individual belongs to a group that has a higher probability of having a loan application rejected, i.e. if P(D|A) increases.

The dependent variables were constructed from the responses to the following questions, with response options in parentheses:

 Have you ever applied for a loan from an MFI? (Yes; No)

 Have you ever been granted a loan from an MFI? (Yes; No)

People who had never applied for a loan also got the additional question:

 Why have you never applied for a loan? (Inadequate collateral; Didn’t want to incur debt; Didn’t need one; Didn’t think I’d get one, Interest rate too high; Not in

compliance with my religion; Others)

In the response alternatives to the last question I have expanded the model by including the option Not in compliance with my religion in order to adapt the method to the matter of investigation. When the data material has been collected the responses are divided after the respondent’s religious affiliation which results in four possible categories for each religious group. The first two categories are made up by respondents that fit in either of the two categories below:

 Applied and granted.

 Applied and rejected.

The last two categories consist of persons who have never applied for loan and are placed in a specific category depending on what answer they gave to the question why they have not applied, according to the following:

 Never applied, but wants a loan. (Inadequate collateral; Didn’t think I’d get one; Not

in compliance with my religion)

19

 Do not want a loan. (Didn’t want to incur debt; Didn’t need one; Interest rate too high)

With probabilities from the survey for the four categories a value for P(R) is calculated for each group. These values are then compared to see if there’s any statistical difference between different religious affiliations in terms of their access to microcredit.

It may seem arbitrary to distinguish between the responses Inadequate collateral and Interest rate too high since both alternatives reflect the cost of taking a loan. The model designers don’t justify this choice further than by demonstrating that the model is robust to alternative classifications of these categories. I insist on this classification, but would like to expand the grounds of this distinction.

The interest rate on a loan is the same regardless of what religious group you belong to, and is thereby no factor that excludes a particular group. However, there might be differences between various groups concerning how much capital they have available to use as collateral for a loan. This is according to me a valid justification for separating the two responses.

However, the response alternative Other is difficult to place in a specific category since it could possibly hide factors that may be based on religion, but which I haven’t managed to identify. In my analysis I will make alternative calculations with “other” placed in different categories to see if the model is robust to alternative classifications of this option.

In addition to these three questions linked to the credit constraint model, the questionnaire

also contained a variety of other questions to enable me to analyze if there are other factors

that could be responsible for a person’s access to microcredit. These were factors such as an

individual’s language proficiency, ethnicity or educational level, and questions about the

company’s nature of current activity, size, annual profit and geographic location. Statistical

methods are then applied to the data to try to find additional links and patterns, thereby

searching deeper explanations to the situation. To see how the questionnaire was designed

view Appendix II: Survey Codebook.

20

C HAPTER 5: F IELDWORK AND R ESULTS

In the preparations for the fieldwork I managed to establish contacts with representatives from FIMA. Their project description was very close to the intentions of my study, albeit in a more general level. While I focused on one specific factor of exclusion from the microfinance market, their goal was to through microfinance establish an all inclusive financial system in Malawi. With this in mind I began my work in Malawi by visiting the Ministry of Finance to discuss with the FIMA-representatives what might be possible to do and how to do it. I figured that if anyone could get me on track with my fieldwork, these were the ones.

Initially the cooperation with my contacts at FIMA went very well. They helped me to get more information on the microfinance situation in Malawi. And at the same time they seemed pleased to be offered a helping hand in their cause, because resources were scarce.

Before I left for Malawi I had through e-mail correspondence briefed FIMA about the premises of my study and my intentions to explore how social factors might affect a

community group’s access to microfinance. But when I specified that I wanted to explore the economic conditions of the country’s Muslim minority, they seemed very dubious to gains of such a study. They said that they didn’t think I would find anything of importance through a study of religion as an excluding factor in microfinance, because religious contradictions was not a problem in Malawi. Instead they insisted on a focus on one of the factors that they already knew could lead to exclusion, such as educational level, turnover or geographic location. But since my theoretical investigations indicated a possibility of religion as an excluding factor I chose to continue with this setup. Besides, even if I in line with their predictions were to find that religion really isn’t a factor in the Malawian microfinance business, this would still be a valuable result because it would rule out a possible factor of exclusion.

Unfortunately the people at FIMA didn’t share my stance. Instead they argued that my work

potentially could lead to exposure of yet another group who were restricted in their access

to economic resources. This, they said, could in turn lead to further problems for FIMA to

solve. This argument seemed to me very misguided since the finding of excluded groups is a

necessity to include them, which is the sole purpose of FIMA’s work and the very reason that

21

UNDP and UNCDF invested U.S. $ 4 million in the project.

When I insisted on carrying out my original idea they chose to terminate our cooperation, or as they expressed it “withdraw our moral and intellectual support from your study”.

5.1 Results of the In Depth Interviews

Without FIMA’s backing, I chose to continue on my own. I had at this stage already interviewed Mr. A.Y. Qadri, scholar at Lilongwe City Mosque which is one of the biggest mosques in Malawi. During this interview, I seized a deeper understanding of the Muslim perception of interest-based banking and the problems they believe it causes. I also got further confirmation of the importance of my research since Mr. Qadri (which may be considered of having superior knowledge on Malawi’s Muslim community) felt that the absence of an Islamic option to interest-based microfinance inhibits the Muslim population in their access to financial resources. Mr. Qadri told me that since there are no real options available, it cannot be held against Malawi’s Muslims if they choose to use the MFIs interest- based services, but there is a potential that Muslims refrain from taking loans at all because it contradicts their religious believes. This could (according to Mr. Qadri) inflict a major retain on the financial development of the country’s Muslim citizens.

Mr. Qadri was of the opinion that it could be problematic to simply extend the range of services offered by the established MFIs with Islamic options. This was because the money would be “mixed” behind the counter and then the money from an Islamic loan could later be used for an interest-based loan. This would violate Sharia which prohibits the support of unlawful activity. Separate MFIs for conducting Islamic microfinance was therefore

necessary according to Mr. Qadri.

At my first excursion to Nkhotakota I started by interviewing Sheikh Ahmed Abdulla Othunam-Asisi, imam at the Taba mosque. Also he considered it to be a problem that the country’s MFIs only offer interest-based services. Furthermore, he claimed that lack of access to capital afflicted Muslims worse because they tended to be more enterprising than the population in general, and thus in a greater need of access to financial institutions. Mr.

45 Financial Inclusion in Malawi – Micro Finance, UNDP in Malawi, 2009, available at

http://www.undp.org.mw/index.php?option=com_programme&view=Poverty%20Reduction&task=project&ar ea=project%20area&id=10 (2010-06-14).

46 Interview with Mr. A.Y. Qadri, imam at the city mosque, Lilongwe, 2010-01-20.

47 Interview with Mr. A.Y. Qadri, imam at the city mosque, Lilongwe, 2010-01-20.

22

Abdulla-Asisi however, was not of the opinion that separate Islamic MFIs were needed. It was perfectly fine if the established MFIs were to offer Islamic services. The fact that the different orientations would be accommodated “under one roof” was not a problem according to his interpretation of Sharia.

In Nkhotakota I also interviewed two field officers from two of the local MFIs to get their opinions on the research problem. Both field officers wanted to remain anonymous because they presented their personal point of view which, they said, not necessarily consisted with the official position of their employers. The first field officer (who was a Muslim himself) felt strongly that there was a problem for Malawi’s Muslims that the country’s MFIs only offered interest-based services. He stated that a majority of his clients were Muslims, but they used the interest-based services simply because they had no other option. However, he had noticed that many of his Muslim clients tried to be very discreet with their business with the MFI because they were embarrassed by the fact that this went against their faith. He

therefore believed that most Muslims would prefer non-interest-based alternatives if these were to be offered. Unfortunately local MFI field officers have no mandate to initiate such an expansion of services, but this would have to be initiated by corporation management.

The second field officer I interviewed in Nkhotakota (who was not a Muslim) had not

experienced that religion was a problem when his clients would apply for loans. Many of the clients he managed were indeed Muslims, but the clients’ religious affiliation was nothing they kept record of.

5.2 Survey Results

To ensure whether or not there actually is a need for Islamic microfinance in Malawi, I couldn’t just rely on the opinion of authorities, regardless of if they worked for the Ministry of Finance, the microfinance business or represented the Muslim community. To get an accurate view of the situation I had to let the group concerned be heard, i.e. the Muslims of Malawi who have a need for microfinancial services.

I spent about a week in each of the chosen districts, where about three to four days was devoted to the actual survey while the remaining time consisted of planning the operation.

48 Interview with Sheikh Othunam Ahmed Abdull-Asisi, imam at the Taba Mosque, Nkhotakota, 2010-02-11.

49 Interview with anonymous field officer at Microloan Foundation, Nkhotakota, 2010-02-10.

50 Interview with anonymous field officer at Pride Malawi, Nkhotakota, 2010-02-11.

23

This meant locating the market place and hire an interpreter to help me communicate with the informants when their English (or my Chichewa) wasn’t good enough, which often was the case in the rural areas.

The purpose of my inquiry was, as previously stated, to examine whether or not there is a need for Islamic microfinance services in Malawi. Such a need could be proven if there was a difference in access to microfinance between Christians and Muslims in the country. For the purpose of investigating the existence of such a difference I stipulate a null hypothesis (H

) and an alternative hypothesis (H

) in order to use them in the future statistical analysis. The two hypotheses are defined as follows:

H

= There is no difference in access to microfinancial services between Christians and Muslims in Malawi.

H

= Access to microfinancial services differ between Christians and Muslims in Malawi.

Below I have in a set of tables compiled the answers to the questions associated with the model for the determination of Credit Constraints described in the Theory and Methods chapter. Each row represents a response option, the different columns shows if the

respondent is a Christian or a Muslim and the numbers indicate how many of each religion has provided a specific answer. Overall I collected responses from 98 respondents

distributed between 59 Christians and 39 Muslims. Table 1 displays the outcome of those respondents who had applied for a loan.

Table 1: Outcome of the respondent’s loan application.

Christians Muslims

Received

21 14

Denied

6 5

Total 27 19

Table 2 shows the reasons given by the remaining respondents to the question why they had not applied for a loan. As we can see in the table, not a single respondent in any of the two groups indicated that they hadn’t applied for loans because it would be against their

religious beliefs. One might come to the conclusion that it shouldn’t be necessary to proceed

24

in analyzing the data material after such an outcome. That none of all the 98 respondents or even any the 39 Muslims indicated Not compatible with religion might be taken as evidence enough to reject the alternative hypothesis. But if we take a look at the distribution between the two groups concerning the other answers we can see that response rates differ

considerably in some cases, mainly the alternatives Didn’t need one, Interest rate too high and Other. This could be a sign of religion being a factor that indirectly affects the two religious groups’ access to microcredit. I take this as a justification to proceed with the statistical analysis of the data.

Table 2: Reasons why the respondents haven’t applied for a loan.

Christians Muslims

Inadequate collateral

3 2

Didn’t want to incur debt

2 2

Didn’t need one

11 2

Didn’t think I’d get one

0 0

Interest rate too high

13 6

Not compatible with religion

0 0

Other

3 8

Total 32 20

To see if I can get a more nuanced view of the situation I will now take further use of the model to determine credit constraints. In Table 3 I have compiled the responses in Table 1 and Table 2 into the four different categorizations presented in detail in the theoretical part.

The bottom row, ∑ Credit constrained, is the sum of rows Applied and rejected and Discouraged borrower.

Table 3: Distribution between Christians and Muslims.

Christians Muslims

Applied and received

21 ^35.6% 14 ^35.9%

Applied and rejected

6 10.2% 5 12.8%

Did not want loan

29 49.2% 18 46.2%

Discouraged borrower

3 5.1% 2 5.1%

∑ Credit constrained 9 15.3% 7 17.9%

25

It can easily be seen that the result is nearly equal between Christians and Muslims according to the model. 15.3% of the Christian respondents can be classified as credit constrained, while the corresponding share is 17.9% for Muslims. To get a picture of the distribution between credit constrained and non-constrained in the two groups, I have in Figure 1 transformed the results into a bar graph.

Figure 1: Bar graph over the credit constraints distribution.

Yet to decide whether the difference is statistically significant or not I performed a chi- square test on the result to see if it’s possible to detect a dependency between religious affiliation and credit constraint. A chi-square test calculates the probability of getting the distribution that the empirical data shows without any dependencies between the variables.

There is handy statistical software available to simplify the procedure, but for the interested I have described the theory and calculations in detail in Appendix I: Calculations. Following chi-square calculations are performed according to the same principles and therefore not accounted for apart from the outcome.

The result of the chi-square test gave us a chi-square value 0.125, which corresponds to a probability value (p-value) of 0.72 which is here interpreted as a 72% probability of that no dependencies exist between religious affiliation and access to credit. It can be mentioned that to be considered as a significant difference in statistical tests, the p-value has to be no more than 0.05, which was in this case by far exceeded. The null hypothesis can thereby not be rejected in favor of the alternative hypothesis at any significant level and thus can no relationship between religious affiliation and credit constraint be demonstrated in this way.

There is nor any significant difference within any the four categories. It is not possible, for

15.3% 17.9%

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Christians Muslims

Credit Constrained Non Constrained

26

instance, to say that any of the religious groups would have a higher probability of being denied loan applications. One additional fact that can be concluded out of these figures is that the Applied and received- and Applied and rejected-rates are approximately the same for Christians and Muslims. This means that the MFIs in no way are being discriminative against any of the groups when coming to granting or rejecting loan applications.

In the previous calculation the answering alternative Other was classified under the Did not want loan category. The number of respondents who indicated the reason Other for not applying for a loan was 3 of the Christian respondents and 8 of the Muslim ones. In percentage, this corresponds to 9.4% respectively 40% of those who had not applied for a loan in each group. Needless to say, it is a big difference between the two groups when it comes to the proportion of those who had not applied for a loan and have a different reason for this than the five predefined answers. This may suggest that the answering alternative Other conceals a factor that is linked to religious affiliation. I have therefore made an alternative calculation in which I instead classify the answer Other under the category Discouraged borrower to see how this affects the outcome. The results of this

reclassification are shown in Table 4 and Figure 2.

Table 4: Recalculation of Table 3 with alternative classification of Other.

Christians Muslims

Applied and received

21 ^35.6% 14 ^35.9%

Applied and rejected

6 ^10.2% 5 ^12.8%

Did not want loan

26 ^44.1% 10 ^25.6%

Discouraged borrower

6 ^10.2% 10 ^25.6%

∑ Credit constrained 12 20.4% 15 38.4%

27

Figure 2: Bar graph over the credit constraints distribution with recalculation for Other.

With this reclassification of the answering alternative Other we can now see an 18%

difference between the groups concerning credit constraint. This is a substantial increase from the 2.6% that was noted with the original classification. The difference can also be assured at a 95% level of significance (chi-square value = 3.863 which corresponds to the p- value = 0.049). This means that the null hypothesis in this case could be rejected in favor of the alternative hypothesis, i.e. access to microfinance loans differ between the two religious groups, to the disadvantage of the Muslims.

But let’s not jump to conclusions. Since I do not know what is the underlying reason to that the groups to such a different extent has indicated the answer Other, I cannot just assume that it’s a factor of credit constraint. 9.4% of the Christian respondents have also given another answer than the predefined to why they have not applied for a loan, which shows that there can exist additional reasons for not applying for a loan which are not linked to religious affiliation and thereby not classified as credit constraining. If these reasons are evenly distributed between Christians and Muslims this means that 9.4% out of the Muslims that answered Other are not to be classified as discouraged borrowers. This gives us yet another scenario where 9.4% of the Muslim Other-answers are transferred to the Did not want loan category and the rest remain in the Discouraged borrower category. This result in a situation where instead 36.5% of the Muslims would be considered as credit constrained compared to the pertaining 20.4% for the Christians. Rounding off to closest integer, this means that 14 of the 39 Muslims are to be classified as credit constrained, compared to 12 of the 59 Christians, which isn’t a significant difference according to the chi-square test (chi- square value = 2.916 which corresponds to the p-value = 0.088).

20.4%

38.4%

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Christians Muslims

Credit Constrained Non Constrained