Determinants of Rural Women Access to Credit in Cheha District, Gurage Zone, Sothern Ethiopia

Access of credit for rural women can play a significant role, especially in rural development activities. This study conducted with the aim of analyzing factors that affect access to credit of rural women in cheha woreda. Primary data collected through structured questionnaire from 100 sample women of chukara and Gasore kebeles selected randomly. The different sources of credit (formal & informal credit source) from which the women use credit to each source. Regarding the result from descriptive statistics showed that, women using the informal credit institution are greater than formal credit institution. The estimation results of the logit model show that marital status ,family size, extension advise, distance to nearest credit institution are the important and significant factors that enhance access to credit. In general as these research indicates most samples rural women are on the problem of low land holding size , high family size , inadequate extension advise service and their livelihood is not sufficient for their family so they are on great demand of credit with more amounts in order to enhance their livelihood. Finally I recommended that women use their times and resources wisely and effectively and they may be choose savings than credit and governments give the advise for the women use their time and resource wisely and effectively. Appreciate women who participate in saving than credit.


Objectives Of The Study 1.2.1. General objectives
The general objective of this study analysed the determinants of women access to credit in the study area.

The specific objectives
 To identify the sources of credit to cheha woredawomen  To identify the factors affecting women access to credit service

Researhch Methods 2.1. Description Of The Study Area
ChehaWoredawas one of the woreda in Guragazone of (SNNPRS) in southern Ethiopia.It had common boundarieswithabeshgeWoreda at north, Geta and EnemorWereda at south, ezha and GumerWereda at east and yemspecialWereda and Oromiya region at west. It located 180 km, south west of Addis Ababa and 22 km south west of Wolkite town. Population of the area male 67509 and female 70156, which total is 137665, from this rural population male 62079 and female 64865 which total is 12944. Economic activity of the area was concerned with agricultural activity, the main animal are 110280, equine 2427, goats and sheep 12672, then total animal population is 125379.topgraph of the area, attitude of this woreda range from 1710-2800 Mater above sea level. From the total area of woreda distribution of land use in hectare cultivated land 25792(13106 annual crop and 12686 perennial crop), grazing land 1465, forest land 5877, miscellaneous land 5163, potential cull tilted land 3171 and uncultivable land (degraded land ) 2504 hectare, which total land of woreda in hectare was 43972.
The main crops of the woredawere divided in two categories: one was perennial crop (ensset, coffee, mango etc.) and other annual crop (teff, wheat, maize, etc.) Agro climate of the woredaisdega 20% and wenedega 80%.The temperature of this maximum 27 C o and minimum 18 c o chehaworeda in there were different informal and formal credit institution. for instance iqub ,idir,lending from friends were informal credit services. commercialbabk of ethiopiawas formal credit service institutions .

Types Of Data And Data Collection Method
Both primary and secondary sources used. The primary data collected by directly interviewing the sampled rural women. Secondary data obtained from published and unpublished available sources. Qualitative data collected through focused group discussions, key informants interview (some local formal and informal leaders), and personal observations. To collect quantitative data, this study used semi-structured interview schedules.

Sampling Technique And Sampling Size
According to the basic principle, the availability of prior information about the target population in the study area and the overall objective of a given study determine the decision of choosing a specific sampling technique.For the achievement of the objective of this research, sampling techniques used to cheha woreda purpusively. A stratified sampling procedure used to select sample households.Atthe first stage, two kebeles cukuara,Gassore randomly selected from 41 kebeles using purposivoly.
From these two kebeles, a total of households randomly selected using the probability proportion to size. From the total sample households, 54 and 46 are users of credit and non-users of credit respectively using every household in the selected kebeles given equal chance of being selected. Simple random sampling techniques used to collect the necessary information from the households. The random sampling was used as an appropriate technique because it avoid bias of representative and all people in the population had an equal chance of being selected.the solvin'ssampling formula with 90 percent confedence level used to determine sample respondant . Solving formula: n=

=100
The above formula shows that the actual sample size for this study is 100.Proportional sample size based on house hold was essential to determine the number of respondants from two kebles. 2260=100 1230=n1 n1=

.Descriptive analysis
Descriptive statistics is, one of the techniques used to summarize the data collected from a Sample representing a given population. By applying descriptive statistics such as percentage, frequency and others, one can compare and contrast different categories ofSample units (in this case women households) with respect to the desired characters so as to draw some important implications about the source of credit for the rural women in the area.

2.4.2.Econometric models
Regression which involves yes or no is a dummy dependent variable regression model. Which are applicable in a wide variety of fields and are used extensively in survey or census-type of data (Gujarati, 1995).The dependent variable in this study was dummy variable, which assumes a value of zero or one depending on whether or not the borrowers are default. When one or more of the explanatory variables in a regression model are binary, we can represent them as dummy variables and proceed to analysis. The loan repayment performance is a dependent variable, which is dichotomous taking on two values, one if the borrower is a non-defaulter and zero otherwise. Estimation of this type of relationship requires the use of qualitative response models.
In this regard, the non-linear probability models, Logit and Probit are the possible alternatives. The ordinary least squire regression, when the dependent variable is binary, produces parameter estimates that are inefficient. Consequently, hypothesis testing and construction of confidence interval become inaccurate and misleading. To alleviate these problems and produce relevant empirical outcomes, the most widely used qualitative response models are the Logit models credit access is a dependent variable, while different socio-economic and lender related factors considered as independent variables. In this case the value of this dependent variable is 0 and 1, which stands for 1 if the borrower is user and 0 If the borrower is non user. Therefore, credit access treated as dichotomous dependent variable. credit access is, therefore, a non continuous dependent variable that does not satisfy the key assumptions in the linear regression analysis. When the dependent variable to be modeled is limited in its range, using ordinary least squares (OLS) may result in biased and inconsistent.

Specification of the logit model
This study was intend to analyze which and how much the hypothesize repressor can relate to the loan repayment performance of urban women. As already noted, the dependent variable is a dummy variable, which will took a value zero or one depending on whether or not a borrower defaulted. However, the independent variables were of both types, that is, continuous or categorical.
Probit and logit models are similar and yield essentially identical results. Aldrich and Nelson (1984) indicated that in practice these models yield estimated choice probabilities that differ by less than 0.02 and which can be distinguished, in the sense of statistical significance, only with very large samples. The choice between them therefore, revolves around practical concerns such as the availability and flexibility of computer programs, personal preference, experience and other facilities.
The logit models is commonly used in studies involving qualitative choices. The probit probability model is associated with the cumulative normal probability function, whereas, the logit model assumes cumulative logistic probability distribution. The advantage of these models over the Linear Probability Model is that the probabilities are bound between 0 and 1. Moreover, they fit best the non-linear relationship between the probabilities of the dependent variable and the explanatory variables, that is one which approaches zero at slower and slower rates as an explanatory variable (Xi) gets smaller and smaller and approaches one at slower and slower rates as Xi gets larger and larger. Gujarati (1988), Feder et al., (1985, Aldrich and Nelson (1984) and Maddala (1981) have recommended probit model for functional forms with limited dependent variables that are continuous between 0 and 1, and logit models for discrete dependent variables.Hence, the logistic model is selected for this study. Therefore, the cumulative logistic probability model is econometrically specified as follows: To identifying factor affecting access to credit service at the individual household level, Binary logit model used. This method chosed because it was a standard method of analysis when the outcome variable was dichotomous (Hosmer and Lemeshow, 2000), measured as had a value of 1 or 0, where 1 = participant on credit and 0 = non participant on credit . Generally, the Binary logit model written as: Therefore, the cumulative logistic probability model is econometrically specified as follows: Pi=F (zi) =F (α+∑ % & ) = ' ()* = ….1 Where, P i is the probability that an individual will participate in formal credit or does not participate given X i ; e denotes the base of natural logarithms, which was approximately equal to 2.718; X i represents the i th explanatory variables; and α and β i are parameters to be estimated Logit model written in terms of the odds and log of odds, which enables one to understand the interpretation of the coefficients. The coefficient of the logit model therefore represents the change in the log of the odds associated with a change in the explanatory variables. The odds ratio implies the ratio of the probability (P i ) that an individual choose an alternative to the probability (1-P i ) that he/she not choose it.
+ui……………………………………………………….6 2.5. Hypothesis Dependant variable woman in cheha woreda either participant on credit or not participant on credit.so,this study approched the dependant variable in" user" or "nonuser" responses which was dummy variable. credit participation was defined in this study as participation of credit .comprizing all those credit activities. woman participation in credit the dependant variable for logit model. Independent variable Marital status:this variable was adummy variable which take avalue of "1"if the respondent women is married and "0"single .this independent variable hypothe sized to affect access to credit positively .assumed that married womens can handle and manage there ovreallivelihood (social duties and farm activities) better than households who income. Therefore, married women households can get access of credit repay than single households. Age of the household head: It was defined as the period from the respondent"s birth to the time of the interview measured in years. It was a continuous variable. Those farmers having a higher age due to life experience had much better association with cooperatives and other formal credit institutions, and it will be hypothesized that older farmers with higher age may have more access to use credit from the formal sources and increase its income (Samuel,2010 ). Education of thewoman household head (heduc): This variable was measured using formal schooling of the household head and hypothesized to affect access to credit positively. It had taken dummy values 1 if the woman household attended any formal education of any level and 0 otherwise. Education increases woman " ability to get and use information. Educated womans may had the ability to analyze costs and benefits and thereby improve their livelihood. According to Samuel (2010) those womans who have better level of schooling has high chance of being participant. It was hypothesized that educated womans had more access to credit compared to others. Family size: It was the number of people in the household. The larger the family members, the more labor force available for the production purpose. Based on this, families with sufficient labor force were expected to participate in credit program and increase household income. On the other hand, large family size may imply selfinsufficiency in terms of food consumption because large households consume more than do small households. Households who had more number of family members were less likely to participate in the project than households with less family members (Samuel, 2010). Therefore, the effect of family size on credit access and increasing income may be indeterminate a prior. Distance from source of credit institution : It was a continuous variable and measured in killo meter which producers walk or travel to reach the nearest district Micro finance institution. The closer the household was located to the micro finance institution, the lesser would be the transportation cost, loss due to spoilage, better access to market information , and less time spent. Therefore, distance was hypothesized to affect smallholder farmers" participation in credit finance negatively. Total land size in hectare (landsiz):-This was a continuous variable referring the total land owned by households in hectare. It consists of the sum of owned cultivated land, rented-in land and land secured through sharecropping arrangements) by the household. On the other hand, households owning large farms had a lower probability of attaining credit from formal financial institutions. This variable was hypothesized that, the farmer who had larger size of land can utilize more capital and access for credit and therefore he/she more participate in the formal sources. Access to extension service; this is a catagorical variable .it also positively affect credit because extension services give awareness about advantage of rural credit to women, as result, they are interested to take credit.

Result And Discussion
In this part the result of the study are discussed including econometric analysis and statistical data analysis.and presented. The result of the study presented by using descriptive and inferencial statistical analysis. As indicated in the table above, education level of rural women affects the access of credit which means as the women are learned they have more awareness to take credit and they are confidential to take and use it. In other words, women who are not learned fear to take credit because of lack of awareness. Out of the total respondents who use credit and illiterate are 44%, who use credit and literate are 56%, From out of respondents who does not use credit and who are illiterate are 52% and who does not use credit and who are litrate are 48%. Since education helps to acquire skills and knowledge. According to the above table, from 50 credit users, 56% can get access of extension contact and the rest 44% have not get access of extension contact. Whereas from 50 credit non users 16% only get access of extension contact and 84% does not get access of extension contact as a result they cannot get credit. In the above table we have summarized the descriptive statistical analysis of continuous variables used in this survey. Mean, Standard deviation, minimum, and maximum of the four continuous variables i.e. age, family size, farm size, and distance. The mean age of the accses tocredit is estimated from the stata table and shows that most of thewomens are in the average years of 40 and womens in this age of the total womens shows that most of the rural womens are at the productive stage to credit access. Mean size of family size is 3.just this shows that family size with 40 years old has at least 1 and at most 6 family members respectively. On other side they can get credit access.7 mean distance shows most of the rural womans have not easy access to credit.

Summariy of continuious variables by describitive statics
The mean area of the land in (ha) would be 3.2(ha).this shows that the size of land each womans owns is very small to complete woman requirements and this forces individuals to move out of the farm activities to diversify their income sources .As human wants are unlimited but resources are limited, because land is a limited resource. There are different source of credit for rural women. The table above shows that the source of credit to rural women who use cedit are get 62% from informal institutions and 38% from formal institutions.from this there are different types .commercial bank,development bank,nib bank,constraction and business bank,OMO micro finance institution.womens also get credit from informal credit sources for instance from money lenders,money tradesmen,friends and relatives,neighbors,etc. Pseudo R2 = 0.5243 3.2.1. Discussion on significant variable Marital status: influenced negatively the marital status of rular woman in credit access service. The effect was significant at 35 percent. Other things remain equal; result from the marginal effects reveals that the decrease the number of womens parcipate in credit access. but the number of men participate in credit access is increase. The possible reasons are as mens become more likely access to credit service compared withwomens because they affect social norms,cultural norms ,far apart from credit institutions. Family size: other things remain constant when the size of family increase by one member the probability of woman user in credit access decrease by 24.88 percent as the marginal effect shows. this study result disagrees with the finding of (Samuel, 2010).huose holds who have more number of family members are less likely to participate in the project than house holds with less family member. the larger the family members ,the more labor force avaliable for the production purpose.based on this,family with sufficient labor force are expected to participate incredit program and increase house hold income .on the other hand, large family size may be imply self-insufficiency interms of food consumption because large huose holds consum more than do small house holds. Extension service:The marginal effect analysis shows extension advise positively affects rular women access to credit as the effect was significant at 1 percent . womans which get extension advise are expected to have more information that influence farm household's demand for credit access from the micro finance institution. there fore this variables positively influences womens to use credit access . Distance to nearest credit institution:: this is expected variables which were hypothesized to affect rular womens's uses credit access negatively. as the marginal effects analysis shows other things remain constant distance from credit institution is increased by 1 killometer the probability of rular women uses credit access rises by 18.8 percent as rular womens become less users of access to credit. the possible reasons are rular women users in credit access even if there is users of credit access because of the enforced credit nature of informal credit lending systems.most of rular womens choose informal lending systems thoseare local money lenders,friendsand relatives. because they are attracted by the outcomes of informal credit institution like, friendship and mutual assistance among members. Farm size: this was a continuous variable referring the total land owned by households in hectare.this variables also un expected variables which werehypothesized to affect rular womens uses credit access positively.as the marginal effects analysis shows other things remain constant farm size increased by 1 hectar the probability of rular women uses credit access decreased by 17.79 percent. as rular womens become less users of access to credit.

Conclusions and Recommendations 4.1 Conclusions.
 The objective of the study were to know rural women's access to credit in cheha woreda ,to identify major sources of credit and the factors that affecting rural access to credit.  As the descriptive result shows most of womens are participated or users of informal credit sources.  The logistic regression analysis result show that among 7 explanatory variables which were include in the model namely, age, marital status, education status ,family size, extension service ,distance to nearest to credit institution farm size.from those variables 5 variables (marital status,family size,extension advise,distance from credit institution ,farm size)were stastically significant .among those significant variables extension advise distance from credit institution and family size affects positivelly .the rest of other variables affect negatively.

Reccomandetion
 Based on the results obtained from descriptive analysis and econometric estimates and also based on personal observation during the study the following recommendations are recommended.  As the descriptive result shows most of womens are participated or users of in informal credit sourcesSo government and other stakeholders should have give emphasis for those informal financial institution and they have to interfere and encourage and appreciate informal credit source in terms of giving training how to hold their money and on how to maintain books of accounts. As one of the problem in informal credit sourece like iquib ,Idir ,Local money lender ,borrowing from friends and relatives are mostly friendship and their social relation is use as a collateral to be a member so this result in frightening and loss of confidence among members so government should enter to those informal credit source .  Even if womens users of different types of informal credit source some informal financial institution are not going with the interest of households for instance money lenders and borrowers since they are not require collaterals and have high transaction cost they uses but their interest rate is very high since they charge interest rate based on their personal relations this may affect badly especially the poor womens so government should interfere and have to regulate the interest rate to be charged by money lenders.  As the logistic regression results shows extension service is one of the variables which have positively effect on women's users in credit access. in informal credit source that means the presence of formal credit source does not contracdict with and can work together with informal credit source institution.if both sectors work together they may get mutual benefit. but marital status have negative effect on womens users in credit access.this means that womens are married access of credit become decrease.because the married womens choose income gets from working and save there moneys and use wisely and effectivelly. Because they affect interst rate and fail of repayment.but we recommended in the future governments focus rular womens accessto credit giving more advise about credit .family size also have negative effect on womens partcipation of credit source. distance from nearest to credit source institution have positive effects on womens participation of credit source .  we recommended that governments not give attension of distance .it give the attension of how peoples attracted and give informations about credit source effect and purpose,reduce the interst rate of credit.farm size have negative effect on womens access to credit . we recommended that womens use there farm size wisely and effectively and they may be choose savings than credit and governments give the advise for the