The impact of gender on loan repayment behavior
in group lending
An empirical analysis of the impact of women’s loan repayment behaviour in group lending. Case study:Tanzania
Name: Zahor, Talib
Student number: s1752375
UNIVERSITY OF GRONINGEN
Faculty of Economics and Business MSc Finance (Corporate) Supervisor: Prof. R.Lensink
Abstract
This study measures the impact of gender on loans repayment behavior in the Tanzanian context. The study also measures the influence of marriage on loan repayment behavior. Through data collected using administered questionnaires at the study area, results of the study show that women behave better than men on loan repayment behavior.
Furthermore, the results show that marriage has influence on loan repayment behavior.
Preface
This study was conducted as part of the fulfillment of the Masters Degree in Finance.
The study measures the impact of gender on loan repayment behavior; whether women perform better than men on loan repayment behavior.
Further the study explores the influence of marriage on loan repayment behavior in group lending.
During the process of undertaking my research, several people helped me to accomplish the study.
First, I would like to thank NUFFIC for funding my studies.
Secondly, I thank my mentor, Prof. Robert Lensink for his tireless advice on the subject and various comments.
Also, I would like to thank Dr. Abeid Gasper of the Institute of Finance Management (Tanzania) who gave me courage throughout my studies.
Furthermore, I would like to request Mr. Nassor Hussein and Mr.John Euseby of The institute of Finance Management to accept my appreciation for their comments and advice to the accomplishment of this study
I also thank my lovely wife Subira who provided me with the real life advice, courage and comments that ensured the thesis is accomplished on time and for accepting my absence for the whole period of my studies.
Section one
1.1 Introduction
Microfinance is the provision of small scale financial services to lower income individuals (Hishigsuren, 2007). Individuals who lack credit from banks and other financial institutions use the microfinance as a saviour.
Microfinance has been used as a women’s empowerment tool (Khan, 1999) and an instrument for povery alleviation (Hishigsuren, 2007). Developing countiries use microfinance to fight poverty because of low collateral rules upon which microfinance is operated (Van Eijkel 2006).
Although microfinance is associated with poverty alleviation, there are very few careful impact studies on microfinance (Morduch 2005). However, most of such studies on microfinance show high compliance rate for women than for men. (Morduch, ( 2005);Kinsey, (2002-2005) and Brett, (2006)). Besides, studies conducted in the field of microfinance, show that very little is known on contribution of marriage/relationship to loan repayment behaviour on group lending programs. Instead, most research show large numbers of women participating in group lending programs. Some of such researches associate such participation with the lack of access to finance in large, formal financial sectors such as banks.
Problem statement
The group lending programs, besides being the main financial facility for poor individuals who lack access to formal financial institutions, have women being the main customers. It is said that in group lending programs women have high compliance rate on repayment of the loans (Morduch 2005). This behaviour is influenced by personal character of women, who are considered to be more risk averse than men, cooperative, and easy to monitor because they are less mobile than men.
The question then remains whether such tendency of high repayment rate among women is a result of group lending policy or the personal behaviour of women.
Furthermore, very little is known about the influence of marriage on loan repayment behaviour. While some studies show the great opportunity possessed by men to access finance in formal financial institutions (Coate, 1995), other studies reveal the need for microfinace in women empowerment (Kivane and Bruce, 2001).
The impact of gender on loan repayment behaviour is measured by concentrating on the compliance rate on repayment of loans between men and women in group lending programs compared with those outside group lending.
The sample of respondents outside group lending programs considered from women and men who are not members of any group lending programs but have other local way of finance. Such local way of finance is an agreed arrangement between groups of people to contribute certain amounts of money within agreed periods. Every member of the group is required to submit the monthly or weekly contribution to a group leader and the group leader will then give such (monthly/weekly) collected amount to one member of such group. The monthly/weekly contribution of members will continue (including those who had received the money) until all members of the group had had their turn (to receive the sum of monthly/weekly contribution).
The influence of marriage on loan repayment behaviour was measured by observing whether respondent has ever received financial support from his/her spouse in case of business loss or any other damage that might result failure on loan repayment.
Objectives of the study
The objectives of this study is to assess the gender impact on loan repayment behaviour in group lending programs as concluded by other studies. The study also explores the influence of marriage on loan repayment behaviour.
The rest of the study is arranged as follows: in section 2 the study gives the description of the financial system in Tanzania; section 3 provides literature review, where the background of the problem and empirical studies on group lending on women’s behaviour and other impact are reviewed; section 4 describes data collection procedure and methodology used in this study; section five (5) represents results and discussion of the finding, also provides the conclusion and the recommendations for further studies.
Section two
2.
Descpition of the financial system in Tanzania.
21. Economic reforms in Tanzania
Prior to 1986, Tanzania’s economy was centrally planned; the government was the central player and holder of all means of production. During 1986, Tanzania initiated economic reforms under which the private sector was allowed to participate.
As part of the economic reform program, implementation of the financial sector reforms started in 1991 and aimed to develop a sustainable, efficient and effective financial system. Specifically, the reforms have included liberalization of interest rates, elimination of administrative credit allocation, strengthening of the Bank of Tanzania’s regulatory and supervisory role, restructuring state-owned financial institutions and allowing entry of privately owned financial institutions, (BOT)
The reforms brought about efficiency and competition in the banking sector but unfortunately, they also resulted in a further widening of the institutional gap in the provision of financial services to the lower income segment. Access to basic financial services by the majority of Tanzanians has, therefore, not increased proportionately. This situation prompted the Government to initiate a process for mainstreaming microfinance services. The process was aimed at ensuring the development of a broader based financial system comprising a variety of sustainable institutions, with wide outreach and diverse financial products. (Rubambey, 2005)
2.2 Description of financial system in Tanzania
Tanzania’s financial system is diverse, however it concentrates much on commercial banking, a small section of economy (IMF 2003). It comprises 21 banks, 9 non-bank financial institutions, 2 pension funds that invest in financial assets, 14 insurance companies, 63 exchange bureaus, about 650 functioning saving and credit cooperatives (SACCOS) and a stock exchange. (Table 1).
The use of formal banking system remains low. According to the 2000-01 household budget survey (HBS) only in 6.4% of all households was there a member with a saving or current account. Such percentage was a decrease as it was 18% a decade earlier. The decline reflects in part the restructuring of banking system where many banks were closed (IMF 2003). The non-bank financial system in Tanzania is weak in terms of volume of activity, range and quality of its products and services and degree of market penetration. (IMF 2003). About 250 SACCOs are urban area based as it is the common place of employment. The remaining 400 SACCOs are mostly rural community–based and are small in size.
2.3 Microfinance in Tanzania
In Tanzania, microfinance existed before the introduction of the economic reforms (Minisrty of Finance, 2000). However, it remained weak due to government intervention on the interest rate, lack of internal capacity and weak coordination (Ministry of Finance, 2000). Such government intervention restricted the microfinance institutions to set their own interest rates, hence some of the institutions operated at high costs and ended up in losses.
The financial sector of Tanzania is still new and shallow, consisting of only 21 licenced banks and 11 non banks financial institutions (Randhawa, 2003). Most of these banks are centralized in Dar es Salaam leaving a few to provide coutrywide microfinance network. Due to this factor, the government of Tanzania established the nine-year programe (from 2002) for the promotion of rural financial services. The programe aims at the provision of sustainable increase in incomes, assets and food security for poor rural households by mobilizing savings and other activities that generate income (the Prime Minister’s Office).
2.4 Key players in microfinance in Tanzania
Section three
3.1 Literature review
Group lending and repayment behaviour
A number of studies have been conducted to investigate the contribution of group lending programmes in reducing the moral hazard hence increase the performance and profitability to the lender. Through the joint liability in group, the group lending programmes are considered to be useful in reducing the monitoring cost since every member in the group is held liable in the loan repayment.
The fact that joint liability gives every member a stake in each other’s project, the costless agent reduces the lender’s cost of monitoring every group member while he (the lender) can punish every member upon non payment by any of the members. Thus, this feature induces the group members to behave cooperatively (Aniket 2003). Joint liability is considered as the ideal solution for reducing the auditing cost and verification cost because every member of the group has the incentive to audit the other member due to the liability in the event of default (Ghatak 1999). On the other hand, the idea of joint liability has been considered differently in the study of Hisaki (2006) in which it was observed that the joint liability causes the free rider problem since the member who has been helped (in repayment) in a previous round by other members of the group is more likely to default in the current round.
Moreover, the sequential financing in group lending schemes helps in reducing the monitoring cost (Prabal, 2003)1. If the scheme (sequential financing) applied to finance the projects of two groups ( borrowers) there is the advantage of reducing the scope of collusion since the financing wholly depends on the performance of the first project. But such scheme may cause the great loss if the projects to be undertaken are identical projects because the expected outcome will be lower compared to the static scheme (Aniket, 2003).
Participants in microfinance (group lending).
There is a tendency of high participation of women in group lending programs. Some recent studies manage to draw attention to the large number of women participating in group lending programs and the performance of women in loan repayment in the groups. Khandker (2005) associates the number of women in group lending programs to the scarce resource and illiteracy rate among women which has a negative effect on female education and borrowing. Morduch (2005) links the large number of women in group lending to the lack of opportunity in formal sectors (as men do get favour from formal sector and commercial banks). Kivane and Bruce (2001) also link the great number of women participants in group lending to the women empowerment campaign.
1
Microfinance, gender influence and poverty
Furthermore, some studies measure the impact of microfinance programs on poverty eradication by considering women participation in groups as a great opportunity to reduce poverty due to their direct involvement in family issues (Khadkher, 2005). The proceeds from loan or business made by women go into food, health, and education for their children; hence it has a long term multiplier effect in the economy (Sharda, 1996).
Women continue to be considered as less mobile (easy to monitor), oppressed by their husbands, more risk averse2, cooperative and less likely to default on their loan repayment3. The idea of women being more reliable in loan repayment has also been linked to group formulation, that women are more likely to behave when they form the group by themselves (without gender mix) because of their psychological negative notation toward men’s behaviour (Fiona Graig and Iris Bohnet, 2006). Also, Barr and Kinsey (2002-05) find that men are less responsive than women when it comes the issue of the shame-inducing social sanctions.
On the other hand women are deemed to be less independent in decision making. Hence, this forces them to depend largely on men in their decisions and control of their businesses (Goetz and Gupta, 1996)4. Nearly the same result was found in the paper of Holvoet (2005) in which it is shown that unlike women, the large percentage of men who take loans take their own decisions on the use of such loans. Brett (2006) shows that women fail to generate profit from their loans and therefore they have to use household money to repay their loans.
Hypothesis:
Having reviewed the studies, this study intends to delve into the notion of high compliance rate of women in group lending programs on loan repayment and measures the contribution of marriage on loan repayment behaviour inside and outside group lending programs. Due to the review done, the researcher ended up having two hypotheses. The first hypothesis is that “women behave better than men on microfinance loan repayment behaviour” and the second hypothesis is that “marriage has an influence on loan repayment behaviour of couples”.
Section four
4. Data collection and Methodology
4.1. Data collection procedureData were collected using administered closed-question questionnaires (see appendix 1) for a period of six weeks (from April 2008) at Dar es salaam, Zanzibar and Pemba.
Questionnaires were designed to capture all information regarding the influence of gender on loan repayment behavior of the respondents. Also, it was to capture the influence of marriage on loan repayment behavior in group lending. For instance, respondents were asked to specify their marital status and whether they had ever delayed repayment or not repaid loan. Respondents were also required to state whether they had received any financial support from their spouses on repayment of the loan/contribution.
The selection of area of study was influenced by geographical location of the area, business capacity, cultural influence, financial and time constraints. Dar es Salaam is a business centre of Tanzania comprising people of different races and cultures, highly populated, high literacy rate and high presence of almost all microfinance institutions (group lending programs) and has many formal and large financial institutions such as banks (not less than 20 banks).
The inclusion of Zanzibar and Pemba is due to its political and geographical location (it is the other part of Tanzania) and is made up of people of mixed races, highly influenced by Arab and Indian culture, high illiteracy rate (UNESCO, 2000) has less formal and large financial institutions such as banks (less than 5 for Zanzibar and less than 3 for Pemba). It also has a few microfinance institutions such as group lending ones (less the 5 for Zanzibar and less than 3 for Pemba).
In collecting the data, questionnaires were administered to the study area with the assistance of research assistants who were given initial training. A total of 320 questionnaires was administered with the following distribution: a sum of 80 questionnaires were administered in Zanzibar and Pemba in two equal halves (20 in Zanzibar and 20 in Pemba). Within each distribution, 20 questionnaires went to member of group lending and 20 to those outside group lending programs.
The data collected falls under the distribution shown in Table 2 for sample inside group lending and Table 3 for the respondents from outside group lending:
Table 2: number of respondents (on zones) for group lending
Zones FINCA PRIDE OTHERS TOTAL
Dar es salaam 53 47 20 120
Zanzibar - 20 - 20
Pemba - 20 - 20
Total 53 87 20 160
From the data collected a total of 87 of respondents were female which make 54% of all respondents and 73 were men which represents 46% of all respondents. Also a total of 111 respondents were married which is equal to 69.4 % of all respondents. See table 2b, also the total of 63 women (72% of all women are married.) and 48 men were married (66% of all men)
Table 2a: gender distribution of the respondents inside group lending
Zones PRIDE FINCA OTHERS TOTAL
Male female male female Male female male Female Dar es
salaam 20 27 25 28 8 12 53 67
Zanzibar 10 10 0 0 0 0 10 10
Pemba 10 10 0 0 0 0 10 10
TOTAL 40 47 25 28 8 12 73 87
Table 2 b: marital status of the respondents inside group lending
Dar es Salaam Zanzibar Pemba TOTAL
Male Female male female male Female male Female
Married 29 49 10 8 9 6 48 63
Unmarried 24 18 0 2 1 4 25 24
TOTAL 53 67 10 10 10 10 73 87
For the respondents outside the group lending programs, the data were collected using the following distribution:
Among 160 respondents, 120 respondents were from Dar es Salaam and 40 were from Zanzibar and Pemba (20 each). The gender distribution of the respondents in the control group is: 82 respondents were women which makes 51% of all respondents and 78 were men which is 49% of total respondents.
A sum of 107 respondents outside group lending was married. The distributions of marital status and gender for the respondents outside group lending programs are recorded in Tables 3a and 3b:
Table 3a: the gender distribution of the respondents from outside group lending
Gender Total
zone Male female Total Dar es Salaam 58 62 120
Zanzibar 10 10 20
Pemba 10 10 20
Total 78 82 160
Table 3b: distribution of marital status for the respondents outside group lending
Dar es Salaam Zanzibar Pemba TOTAL
Male Female male female male Female male Female
Married 33 40 10 8 8 8 51 56
Unmarried 29 18 0 2 2 2 31 22
TOTAL 62 58 10 10 10 10 82 78
Distribution of the respondents
respondents outside group lending, though the actual number and percentage of the women participants in group lending programs and local financing are more than what has been reported because women largely lack opportunities in formal financial institutions like banks (Morduch 2005).
Moreover 72.4% of women participating in the group lending programs are married and 72% of women outside group lending are married while 65% of men in group lending are married and 62% of men outside group lending also married.
The gender and marital status variation has great influence on participation in group lending programs because in most African countries, Tanzania cannot be an exception, men are deemed to be the heads of family and hard worker and that they are expected to find any means to make sure their families survive and stay in proper and reasonable life while women especially uneducated and married are expected to be the housewives or doing the small business at their homes.
4.2 Methodology
4. 2.1 Analytical framework
In this study “misbehave” is the dependent variable for testing the tendency and impact of gender on loan repayment behaviour. Consideration of misbehave as dependent variable is influenced by the conclusion of some studies showing that gender has an impact on compliance (repayment of loan in group lending) (Morduch 2005). Misbehaviour is measured by studying the tendency of the respondent to delay or fail to repay the loan/contribution on time. The respondents were asked to answer the questionnaires whether they had ever been late to repay loans or not paid at all (see appendix 1).
Misbehave is regressed against independent variables like education, gender, monitoring, distance and religion in order to answer the question on whether women behave more than men in group lending.
The logit model is used to test the dependent and independent variables to test the hypothesis that women behave more than men in group lending.
1)
l ogit(ϒ = +a βxi
Where:
Y1 =(dummy 1) the presence of misbehave and 0 if not
In measuring the “misbehave”, respondents were asked to specify if they had ever delayed or not paid loans or submission of contributions for respondents in group lending and outside group lending respectively. Respondendents were provided with options to select in the questionnaires for late or non payment of loans. But for whatever reason provided by respondent, if respondent had ever delayed to repay the loan or not pay at all or failed to submit the contribution on time was considered to be “misbehave”.
Xi1= gender, dummy, 1 if respondent is female and 0 if male
Xi2= monitoring, 1 if the respondent monitors his/her fellow members and 0 if the respondent
does not monitor other members.
Monitoring was measured by asking respondent whether he/she knew the businesses of his/her fellow members. The question was simply set to address whether respondent knew type and places of business of his/felow members.
members and normaly meet before the group meeting, it was deemed to be monitoring of each other.
Xi3 =education, dummy 1 if the respondent possesses the post secondary level (higher level)
or higher education and 0 if posssess below the post secondary level. Xi4= religion, 1 if the respondent is muslim and 0 if otherwise.
Furthermore, the study tests another hypothesis which states that marriage has influence on loan repayment behaviour on group lending. In testing this hypothesis other variables were added to the first model. Variables added were control and support.
The respondents were asked whether they had full control of their businesses. The questionnaires were set to ask respondent if he/she alone controlled the business or he/she together with his/her couple controlled the business.
“Control” was dummy 1 if respondent had full control the business and 0 if the respondent did not have full control the business.
Support was measured by asking the respondent if he/she received any financial support from his/her spouse to repay loans or contributions for group lending and outside group lending respectively. The dummy 1 means respondent received financial support and 0 if respondent has never received the financial support from his/her spouse.
Various statistical analyses are presented in reaching the conclusion of this study. Firstly, correlation matrix of all independent variables is reported to see their relation among themselves (appendix 2). The result of the correlation matrix shows that all variables are independent, hence there is no variable which shows strong relatedness. Thus the problem of multicollinearity does not exist. (Brooks, 2002).
Moreover chi square test is presented to see the influence of independent variables on dependent variables (goodness of fit test, as attached in appendix 3). The likelihood ratio is used to measure the goodness of fit model because likelihood ratio has more desirable properties in binary and multinomial situations ( Mitchell, 1992).
In describing the impact of group lending on gender, the logit model is applied to explain the relation between the dependant variable (misbehave) and various independent variables. The use of such a model is highly influenced by the nature of the data (the binary data) used where the logit /probit came to be favourably compared to the normal linear regression model (John, T. Pohlmann 2003).
The second hypothesis “marriage has impact on loan repayment behaviour” influenced other variables to be added at the first model. In this context, the variables like “support” and “control” were added. The added variables are also dummies. “Support” is 1 if the respondent (of either gender) received financial support from his/her spouse, and 0 if not. And “control” is 1 if the respondent had full control of his/her business and 0 if she/he had partial or no control at all.
In these additional variables, support is expected to be negatively related to misbehaves because the tendency of spouses to help each other reduces the repayment burden rather than in cases where the members stood on their own.
Section five
Results and discussion
This section presents the results of gender impact on loan repayment behaviour and the influence of marriage on loan repayment behaviour.
Section 5.1 presents the statistical analysis of the model used to test the impact of gender on loan repayment behaviour and section 5.1.2 presents the statistical analysis of the model used to measure the influence of marriage on loan repayment behaviour.
5.1 Statistical analysis 1
Statistical analysis of group lending and outside group lending
GROUP LENDING OUTSIDE GROUP L
Coefficient P-value coefficient P-value
Variables (standard error) (standard error)
Gender -1.721032 0.0000* -1.429226 0.00000* (0.368576) (0.348207) Religion -0.498223 0.1603 -0.191549 0.5789 (0.354804) (0.345151) Education -1.398647 0.3431 -0.474303 0.7517 (1.475228) (1.499347) Monitor 0.281416 0.4874 0.227643 0.6491 (0.405273) (0.500348) Observations 160 160 prob.chi square 0.36 0.6201 Key
*= statistically significant at 1 % significance level, **= statistically significant at 5% significance level and ***= statistically significant at 10%.
This result can also be expressed and supported by the result of cross tabulation analysis. In cross tabulation, total number of respondents who behave in group lending is 93 which equals 58.1% of all respondents in group lending, among whom 61 respondents make up 66% of all respondents who behave are women (appendix 4a). This finding does not contradict that of (Morduch 2005) in the study conducted to see the differences on loan repayment behaviour between men and women.
When “monitoring” was included in the model, unlike other studies, this study could not find the contribution of monitoring toward member’s loan repayment behaviour on either the group lending or outside group lending.
The monitoring is measured by checking whether group lending participants at least know the businesses of other members or they have the tendency of seeing each other, other than at group meetings. The monitoring is not statistically significant when regressed against misbehave.
The results of monitoring not being significant toward expressing misbehave on loan repayment behaviour contradict with group lending policy expectation. Monitoring was expected to have strong influence on shaping the loan repayment behaviour of borrower on group lending as the group lending policy proposes. The peer monitoring as expressed by (Tirole 2006) could not be supported in this study.
This insignificance is caused by the way of in which people live in Tanzania. From my observation, most of the people do not invest in monitoring each other; instead they just live by trusting one another and that is why 76.5% of members of group lending progrms who do not monitor each other did not misbehave (See the appendix 4a). The same result does apply in the case of the outside group lending where the 88.2% of the respondents do not monitor each other but they behave.
When education was regressed as explanatory variable it was found to be not statistically significant. This is influenced by the nature of the participants of the group lending programs. The statistics shows that 99% of respondents in group lending are uneducated but they behave. The same result is seen in the outside group lending. The uneducated for the purpose of this study means respondent possesses formal education below the post secondary education level (Tanzanian Grades).
5.1.2 Influence of marriage on loan repayment behaviour
Statistical analysis 2
GROUP LENDING OUTSIDE GROUP L
coefficient P- Value coefficient P-Value
(standard error) (standard error)
support -0.819883 0.0894*** -2.573569 0.0002* (0.482665) (0.69397) gender -1.231589 0.0024* -0.766055 0.0574*** (0.405834) (0.403099) Education -1.881481 0.2282 -1.353828 0.3828 (1.561525) (1.55109) control 1.349424 0.0169** 0.136484 0.7718 (0.56481) (0.470529) Monitoring 0.624103 0.1693 0.732606 0.2223 (0.454108) (0.600329) observations 160 160 Chi.sq(prob) 0.4651 0.8830 Key
*= statistically significant at 1 % significance level, **= statistically significant at 5% significance level and ***= statistically significant at 10%
The result of the statistacal analysis 2 above reveals that “support” is negatively related to misbehave and is statistically significant at 10% significance level This result implies that those who get support from their spouses are unlikely to misbehave. Such results are also supported by the results of the bivariate cross tabulation between “gender” vs “support” and contol vs gender. The bivariate cross tabulation of group lending shows that a total of 55 respondents equal to 34% of all respondents in group lending who get the financial support from their spouses among whom 44 respondents equal to 80% are women( see appendix 4b).
In like manner, it is not uncommon to find wealthy women do the same to support their husbands. It is therefore very difficult to separate the individual efforts for loan repayment in the instance of couples.
The results for outside group lending is almost the same with those for group lending. Support is highly significance at 1% significance level. The bivariate cross tabulation of group lending between gender and support shows that 43 respondents, being 27% of all respondents get financial support from their spouses, among whom 37 respondents make up 86% of women.
The difference in percentage of respondents in group lending who get support compared to outside group lending is resulted by the nature of the local finance itself. Most of these local financing is conducted on friendly basis and amounts involved are not that huge. Hence, the participants can delay for a while and submit the money thereafter, so most couples manage to stand on their own.
The results for control differ from that of group lending and outside group lending. The results in group lending show “control” is positively related to misbehave and is statistically significant at 5% significance level. The result means that those who have full control their businesses are likely to misbehave.
More over results of bivariate cross tabulation between gender and control shows that a total of 123 respondents in group lending programs which is equal to 77% of all group lending participants, control their businesses. Of these, 67 respondents (54.5% of all respondents) are men. And those 37 respondents equal to 23% (of all respondents) who don’t have full control of their business, 31 respondents equal to 84% of such respondents who don’t have control are women. It can be said from such results that, marriage has great influence on loan repayment behaviour. This is because a large number of couples share the control of their businesses and this induces them to repay the loans on time. Such results support the findings of Gupta in a study in Bangladesh to assess who control the credit (Gupta, 1996). The result of outside group lending can also be supported by the result of bivariate cross tabulation between “control” and “gender”. Cross tabulation shows that 128 respondents (80% of all respondents) in group lending program control their businesses among which 65 respondents equal to 50.8% of the respondents who control their businesses are women. A total of 32 respondents do not have full control of the business. Of this, 17 representing 53% of the respondents who do not have the full control for their business are women.
also been made by Brett (2006) who realizes that the women eat less and sacrifice so as to pay (Brett 2006).
5.2 Conclusion
From this study, it is realized that women behave better than men in group lending programs, a conclusion which most researchers have established. But this study could not establish a causal relationship for this behaviour of women. It is therefore difficult to say that group lending programs have any impact on women behaviour since the same behaviour of women is observed when they are in group lending programs or outside.
In my opinion, the behaviour of women in general is highly influenced by the social status they possess. They also respond to the perception of the public about them because women in most of societies are deemed to be polite, trustworthy, and more collaborative than men. Therefore responding to such public perception toward them women might behave as a societal norm or convention.
The study also concludes that the marriage plays a crucial role in loan repayment behaviour in group lending. It is difficult to separate the part played by each one in repayment of the loans by couples because a high percentage of couples have shared ownership of the businesses and they financially help one another.
5.2.1 Recommendations for further studies
.Due to the limited area covered by this study and few microfinance institutions involved within the constraints of time, there is a great opportunity for further and detailed study to be conducted in Tanzania and Africa in general which can provide reliable conclusions on the topic.
It has been stated that women behave better than men in group lending and that the marriage has an influence on loan repayment behaviour. However, this cannot be entirely conclusive. This study, for instance, does not provide the degree of such marriage influence on loan repayment behaviour. There is therefore still room to conduct further detailed study which can measure the degree of such influence, though it is challenging, especially in the African context where clear differences in couples’ belongings and expenditure are difficult to establish.
6. Appendix
6. Appendix
Appendix 1 (questionnaire) Appendix 1 (questionnaire) This questionnaire will serve the need of data collection for the purpose of writing the thesis as part of the requirement for the completion of the Master’s degree programme on MSc BA Finance (Corporate) offered at the University of Groningen, the Netherlands. Thank you very much for filling this questionnaire The personal information of the respondent Please choose the correct answer by ticking at the box provided 1. Agea) 18 – 30 years... b) 30 ‐ 44 years ... c) 45 and above...
2. Gender 1) Male... 2 female... 3. marital status 1) Married.... 2) Single ... 3) divorced... 4) Widow... 4. Tribe (please mention) ... 5. Religion 1) Muslim... 2) Christian... 3 others... 6. Educational level
1) Primary education ... 2) secondary education.. 3) higher secondary education ...
7. skills
1)Accountancy course.... 2) business course... 3 computer course ....
4) marketing.... banking ... Others ... ( please specify)...
8. occupation
1) employed ... 2).Self employed 9. Your spouse’s occupation
1)employed ... 2) self employed ... 10. Your estimated montly income
1) 20,000≤ 80,000... 2) 80,000 ≤ 100,000... 3) 100,000 ≤ 150,000....
4) 150,000.. ≤.200,000 5) 200,000≥
11. your spouse’s estimated monthly income
1) 30,000≤ 50,000... 2) 50,000 ≤ 80,000... 3) 80,000 ≤ 150,000....
4) 150,000.. ≤.200,000 5) above 200,000
Section B the group lending programes
12. Are you member of any group lending programe?
1) YES ... 2) NO... 3) NO but my spouse is ...
1) FINCA.... 2) PRIDE.... 3) others... 13 .How did you select your group?
1) I choose the group... 2) the programe choose the group for me 14 . How many members you are in your group?
1) 5 members... 2) 10 members... 3) others...(sepcify) 15 Have you known your group member before
1) YES... 2) NO... 16 If yes what relation you have with them
1) family relative... 2) friends... 3) neighbours... 17 How often do you meet with your group members?
1) everday... 2) during group meeting.... 3) occasionaly ... 18 What are the business of your group members ( please mention)?
1)...2)...3...4... 19 What is the approximate distance to the meeting place?
1) Walking distance... 2 far... 3) avarage... 20 A)Do you attend the meeting regularly?
1).Yes ... 2) NO...
b) If no, why (please choose among the provided options)
1) A spouse prohibition..., 2)family responsibilty... 3.business affairs... other specify ...
Loan and payment
21 Have you ever got/received loan from the group lending programe? 1) YES... 2) NO... 3) NO but my spourse got... 22 How many times?
23 What was the purpose for your borrowing/loan?
1)Business... 2) familiy problems... 3) buying house hold equipments (TV, Radio etc) ... 4) building house... 5) others...
If the purpose of loan is business...Who control the business?
1) My self... 2) me and my spouse... 3) others... 24 Do you have any outstanding loan?
YES... NO...
25 How did you managed to repay your loan?
1) profit from business... 2) salary (employment)...
3)support from my spouse... 4) support from family/relative... 26 Have you ever delayed to pay your loan
1)YES... 2) NO...
B) if yes,Why..(please choose among the answers given below)
1 )Business loss..., 2) family problems... 3). any unforsean accident (theft, death of close relatives etc)...
27 How you managed to repay thereafter (after delaying)?
1) My fellow group members pay for me..., 2) support from family..., 3)loan from friend/close relatives ... 4) Others...
28 Do you have any other sources of finance/loan 1) YES... NO...
If yes mention ( please choose among options provided below)
1) SACCOS... 2) local saving (upatu)... 3) bank loan... 4) others... please mention...
1) From my business profit... 2) support from my spouse...3) direct deduction from salary... 4) others ... specify please...
30 What are the terms of payment of the loans/saving.
1) Payment on every week... 2) every month... 3) upon profit... 4) others...
31 The expected punishment for late payment of the loan/ saving are.
1) Pay with interest... 2) no more loan granted to you... 3) no punishment... 4) others... please mention...
32 the expected punishment fo non payment of the loan
1) Police ... 2) personal properties will be taken... 3) I dont know... . 4) others...
33 Have you ever being late in payment or non payment at all? 1) Yes ... no...
If no why? Please select
1) It is a sin... 2) to keep my social reputation... 3) always my spouse help me in need.... 4) others please specify...
The end...
Thank you for filling this questionnaires
2a:Correlation matrix of the independent variables in group lending
MARITAL GENDER EDUCATION DISTANCE CONTROL MONITOR SUPPORT RELIGION
MARITAL 1 GENDER 0.13 1 EDUCATION -0.05 -0 1 DISTANCE -0.12 0 -0.1 1 CONTROL -0.3 -0 0.06 -0 1 MONITOR 0.18 -0 0.06 -0.2 -0 1 SUPPORT 0.34 0 -0.1 0.04 -0.6 0.08 1 RELIGION 0.12 0 -0 0.05 -0.1 0.01 0.081 1
2b:Correlation matrix of independent variables outside group lending
CONTROL EDUCATION GENDER MARITAL MONITOR RELIGION SUPPORT
3:The chi square test for control and test group
The likelihood value and probability
gender Support Religion marital
status Control business
purpose monitoring employment education distance
TEST GROUP Value 24.15278 29.47341 2.3589 9.771106 29.3163 0.948219 0.455287 0.335789 0.007927 0.111021 Probability 0.0000 0.00000 0.1246 0.0018 0.000 0.3302 0.4998 0.5623 0.9291 0.739 CONTROL GROUP Value 18.65343 34.6414 0.054209 2.53402 0.02573 NA 1.155859 NA 0.054209 NA Probability 0.000 0.000 0.8159 0.1114 0.8726 NA 0.2823 NA 0.8159 NA
4a:Cross tabulation for group lending
Gender marital Religion monitor education support control business distance
0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 mis behave 0 32 61 25 68 25 43 52 16 92 1 29 39 30 38 5 63 54 14 % 34 65.6 26.9 73.1 36.8 63.24 76.5 23.5 99 1.08 43 57.35 44.1 55.9 7.35 92.7 79.4 20.6 1 46 21 26 41 45 47 66 26 66 1 76 16 7 85 11 81 75 17 % 69 31.3 38.8 61.2 48.9 51.09 71.7 28.3 99 1.49 83 17.39 7.61 92.4 11.96 88 81.5 18.5
4b: Cross tabulation for outside group lending
gender marital religion monitor education support control
Bivariate cross tabulation test group support Control gender 0 1 0 1 0 62 11 6 67 % 59 20 16.2 54.5 1 43 44 31 56 % 41 80 83.8 45.5
References
Bank of Tanzania. (n.d.). bot‐tz.org. Retrieved May 28, 2008, from www.bot‐tz.org: http://www.bot‐tz.org Bikki Randhawa, J. g. (2003). Microfinace regulation in Tanzania: Implication for Development and performance of the industry. Africa region working paper series no 51 . Bohnet, F. G. (2006). why women cooperate with women not with men: evidence form a Slum Nairobi Kenya. Brett, J. A. (2006). "we sacrifice and eat less" The structural complexities of microfinance participation. human organisation Vol 65 no 1 . Brooks, C. (2002). Introductory econometrics for finance. Cambridge: Cambridge university press. Coate, T. B. (1995). Group lending, repayment incentives and social. Journal of Development Economics vol 46 , 1‐18.
Denise Anthony, C. H. (2003). Gender and cooperation: explaining loan repayment in Micro‐ credit groups.
social psychology quartely, Vol 66 No 3 , 293‐302.
Guinnane, G. a. (1999). the economics of lending with joint liability: A review of theory and practice. Journal of
development Economics , 195‐228.
Gupta, A. M. (1996). who take tha credit? Gender, power and control over loan use in rural credit programes.
World development vol 24 no 1 , 45‐63.
Hishigsuren, G. (2007). Evaluating Mission Drift in Microfinance: Lessons for Programs with social mision.
Evaluation Review 31, DOI: 10.1177/0193841X06297886 , 203‐260.
Holvoet, N. (2005). The impact of microfinance in decision‐ making agency: evidence from south India.
Development and change , 75‐102. Horne, D. A. (Sep., 2003). Gender and Cooperation:Explaining Loan Repayment in Micro‐Credit Groups. Social Psychology Quarterly, Vol. 66, No. 3, , 293‐302. IMF. (2003). Tanzania:financial system stability assessment. Washington,DC: IMF. John T. Pohlmann, D. W. (Dec, 2003 ). A comparison of ordinary least squares and logistic regression. The Ohio Journal of Science, . Jonnathan, M. (1999). Microfinance promise. Journal of economic literature, Vol 37 No 37 , 1569‐1614.
Khan, M. R. (1999). Microfinance, Wage Employment and Housework: A Gender Analysis. Development in
Practice, Vol. 9, No. 4. , 424‐436.
Khandker, s. R. (2005). Microfinance and Poverty: evidence using panel data from Bangladesh. World bank
economic review vol 19 no 2 , 263‐286.
Kono, H. (2006). Is group lending a good enforcement scheme for achieving high repayment rate? Evidence from field experiments in Vietnam. Discussion paper, Institute of developing economies, Japan . Kumar, A. (2003). sequencial Group lending with Moral hazard. WOP . Ministry of finance. (2000). National micro‐finance policy. Dar es salaam, Tanzania: Government printer. Mitchell, T. M. (1992). goodness of fit measures for probit and logit. American journal of Political science Vol 36 , 762‐784. Morduch, B. A. (2005). The economics of Microfinance. London, England: MIT Press, Cambridge Massachusett. Naidoo, S. (1996). They might just forget. women and the enviroment, agenda no 29 , 109‐112. Paul E Green, D. T. (2000). Reseach for Marketing Decision. New Delhi: Prentice‐Hall of India.
Prabal, R. (2003). group lending sequencial financing, lender monitoring and joint liability. discussion paper,
Indian Statistical institute, Delhi India .
Pride. (n.d.). Retrieved June 1, 2008, from www.pride‐tz.org: http://www.pride‐ tz.org/pwinner.asp?pcat=aboutus&cat=microfinance&sid=42
Prime minister's office. (n.d.). Retrieved June 1, 2008, from http://www.rfsp.org/00about.htm: http://www.rfsp.org/00about.htm
princeton. (Revised 2007, September). Retrieved May 26, 2008, from data.princeton.edu:
http://www.data.princeton.edu/wws509/notes/c3.pdf
Remco van Eijkel, N. H. (2006). Group Lending and the Role of the Group Leader:. Department of
Finance,Center of International Banking, Insurance and Finance (CIBIF),University of Groningen,The Netherlands .
Rubambey, G. (2005). Policy, Regulatory and supervisory enviroment for microfinance in Tanzania. essays on
regulation and supervision no 15 .
Tirole, J. (2006). The theory of corporate finance. Princeton: Princeton University Press.
