International Journal of Sales & Marketing Management Research and Development (IJSMMRD) ISSN (P): 2249–6939; ISSN (E): 2249–8044 Vol. 10, Issue 2, Dec 2020, 17-30 © TJPRC Pvt. Ltd.

DETERMINANTS OF ENSET PRODUCERS MARKET PARTICIPATION DECISION AND INTENSITY OF PARTICIPATION IN ENSET PRODUCT: THE CASE OF DISTRICT, SOUTH WEST SHOA ZONE, NATIONAL REGIONAL STATE,

SHELEMEREFERAJEBESA Socio economics Research Team, Bako Agricultural Engineering Research Center, Bako, Ethiopia ABSTRACT

In Ethiopia, enset is one of the indigenous root crops widely cultivated in the south and south western parts. Particularly in Wonchi district it is a major crop used as source of food and cash income for majority of smallholder farmers but there has been limited performance of producers in the enset product marketing. To this end, this study was intended to assess determinants of enset product (kocho) market participation decision and intensity of participation. Both primary and secondary data were used for the study. Primary data were collected from randomly selected 184 sample enset producers through stratified two stage sampling technique. The data were analyzed using both descriptive and econometric methods. A double hurdle model was used to identify determinants of enset producers’ market participation decision and intensity Article Original of participation. The first stage of double hurdle shows that participation decision in kocho market is positively and significantly related with availability of economically active labor in the family, land allocated for enset crop, decision making on kocho income utilization, ownership of transport facility and access to market information. In contrast, ownership of livestock and distance to the nearest market limit sampled respondents participation decision in kocho market. A second hurdle regression model result indicate that educational level of the household head, enset farming

experience, land allocated for enset crop and decision making onkocho income utilization determine intensity of participation in kocho market positively and significantly. The study recommends spreading market information in chain actors, encouraging farmers to learn adult and formal education and rising experience producers through experience sharing on the enset production.

KEYWORDS: Determinants, Kocho, Market Participation, Participation Intensity & Wonchi

Received: Jul 09, 2020; Accepted: Jul 29, 2020; Published: Aug 25, 2020; Paper Id.: IJSMMRDDEC20202

1. INTRODUCTION 1.1. Background of the Study

In Ethiopia, root and tuber crops are the second largest crops, after cereal crops in terms of quantity of production (CSA, 2017). Those crops contribute a major share in traditional food system of many people. They play a vital role in food security especially in south and south western part of Ethiopia. Enset, anchote, potato, onion, carrot, yams, taro and cassava are the major root and tube crops grown in the country (Yeshitila and Temesgen, 2016).

Enset (Ensetventricosum) is one of the indigenous root crops widely cultivated in the south and south western parts of Ethiopia.Itis a multipurpose crop with all plant parts being used byproducers as human food, animal fed, medical and ornamental uses. It has high significant value in day to day activity of farmers who cultivate this crop and it is a major source of food for them. The enset products;kocho, bulla and amicho are eaten by human.

www.tjprc.org [email protected] 18 Shelemereferajebesa

During the dry season the domestic livestocks are fedenset parts, which are not normally eaten by human. Moderately dried itsleaf sheaths and midribs are used for local house in enset growing areas. The fiber of enset is used in the weaving of shopping bags, hand bags, suitcases, sieves, pouches and mats. It is the most important raw material and vital fencing, wrapping and packing of every material and, keeping animals in and around the house.

According to CSA (2013) the total area under ensetcrop in Ethiopia estimated to 312.17 thousand hectares, whereas the total area under this crop in Southern Nations Nationalities and Peoples Regional State is 217 thousand, Oromia 94 thousand and Gambella region 0.382 thousand hectares. During Meher season1from Ethiopian private peasant, total of 127.3 million ensetcrops were harvested and produced 34.8million quintals of kocho, 1.1million quintals of bulla and 29.4million quintals of amicho.In the same period, 45.7 million ensetplants were harvested and produced 11.9million quintals of kocho2, 680.6 thousand quintals of bulla and 10.1million quintals of amicho3 in Oromia region (CSA, 2017).

From Oromia region Jimma, Borana, Guji, West Arsi, South West Shoa and West Shoa Zones are the enset potential Zones (CSA, 2015). The South West Shoa Zone is one of the outstanding enset crop accommodated area. From 95 thousand hectares of total farming 18.3 thousand hectares were underensetfarming and it cultivated by73153 households in the zone(SWSZ ANRO, 2018). Thecrop is produced in all districts of zone except Sodo Daci and Ilu districts. Wonchi district is one of the potential and well known by enset crop production, about 5428 hectares of the districts’ cultivated land is covered by this crop and all kebeles in the district produce the crop. Mesfinet al. (2018) listedkocho, bulla and amicho as the major food products obtained from the enset. From the three food products of the enset, kochoand bulla were supplied to different markets from production site. Due to its perishable nature, amichois not delivered to markets. According to Abebeet al. (2015) the largest proportion of kocho and bulla were supplied to the market rather than consumption.

Despite of enset potential, there has limited performance of producers in the enset product marketing in the district. Determinants of enset producers market participation and intensity of participation in enset product is not documented. In this study enset product is refers to kocho; other enset products suchas bulla, fiber and Amichowere not considered.

1.2. Statements of the Problem

Ethiopian agriculture is playing a key role in generating surplus to speed up the country’s overall socio-economic development (Hassen, 2006). Sustainable success in agricultural growth depends not only on achieving agricultural productivity and household food consumption but also critically depends on increasing better market access and expansion of market opportunities. Market access has been identified as one of the critical factors influencing the performance of smallholders’ agriculture in developing countries (Barrett, 2008).

Enset crop is crucial in the Ethiopian context and specifically in the study district. It has a significant contribution to the livelihood of producers as income sources as well as ensuring of food security. Currently, about 19860households of Wonchi district are engaged in the production of this crop and leading their life based on enset farming.

The activity is mainly meant for additional income generating activity on top of other crop production like maize,

1Meher is a main season between September and February 2Kochois the bulk of the fermented starch obtained from a mixture of the decorticated leaf sheaths and pulverized corm of enset. 3Amichois the fleshy inner portion of the ensetcorm, which is eaten as a root and tuber crop after being boiled.

Impact Factor (JCC): 8.6084 NAAS Rating: 3.37 Determinants of enset Producers Market Participation Decision and Intensity 19 of Participation in Ensetproduct: The Case of Wonchi District, South West Shoa Zone, Oromia National Regional State, Ethiopia teff, wheat and barley and livestock rearing. In fact, enset has been the main food and cash crop in the district (WDANRO, 2018).

Despite the valuable contributions of enset to food security, income and alleviating poverty is substantial, limited attentions has been given to enset research particularly in market aspect. There are several problems that hinder the performance of enset production and productivity. Shortage of enset improved variety, its products’ price fluctuation and market seasonality, poor infrastructure, enset diseases, poor information flow, poor transportation system, using perishable packaging, lack of cooperation between actors and poor policies concerning the enset product market are the major constraints of enset production (Ashenafiet al.,2017; WDANRO,2018).Although, production up to selling the enset product is tedious work producers faced market problems as product price is low and inadequate enset product market information in the study area. It was not clear why all enset producers are not enset product market participants and determinants of the participation and intensity of participation was also unknown in the district.

1.3. Objectives of the Study

The general objective of the study was to analyze determinants of enset product market participation and intensity of intensity of participation in Wonchi District, South West Shoa Zone, and Oromia National Regional State, Ethiopia.

The specific objectives of the study were:

 To identify the determinants ensetproducers participation decision inenset product marketing; and

 To identify the determinants of intensity participation in enset product marketing in study area

RESEARCH METHODOLOGY

2.1 Description of the Study Area

Wonchi district is located in Oromia regional state of South West Shewa Zone, Ethiopia. It is one of the eleven districts in south west shoa zone and about 9kilometer and 123kilometer from town and Addis abeba respectively.

The district has two agro-ecologies; highland (40%) and midland (60%).The mixed farming system of both crops and livestock are common economic activity in the district. The important crops grown in the district are maize, teff, wheat, barley, enset, and onion. According to WDTMDO (2018)in 2017/2018 production year 835352 quintals of cereal grain, 8078quintals of pulse grain, 281723quintals of horticultural crops, 31764 of live-animals, 5253tones of hide and skin were supplied to the market.

Moreover, 39936 quintals of ensetproducts (kocho and bulla) were supplied to the market. In general, Wonchi district is the major producer of enset from south west shewa zone and enset production is considerable sources of cash in the district (WDANRO, 2018).

2.2 Data Types, Sources and Methods of Data Collection

This study used household survey data collected from Wonchi District. In order to generate sufficient information both quantitative and qualitative data from primary and secondary data sources were used. Primary data were collected from randomly selected enset producers and kochotraders.T o collect primary data semi- structured questionnaire were prepared and pre tested on sampled kebeles was made to evaluate the appropriateness of the design, clarity and amended based on feedback. Accordingly Secondary data relevant for this study were also gathered from District Agriculture and Natural www.tjprc.org [email protected] 20 Shelemereferajebesa

Resource Offices and from Office of Trade and Marketing Development. Moreover, journals and websites were visited to generate relevant secondary information for the study.

2.3 Sampling Techniques and Sample Size Determination

A stratified and two stages random sampling method was used to select the sample household heads. The study area was divided into two stratums namely midland and highland based on agro-ecology because of considerable variation in enset production among agro-ecologies in the district. In the first stage out of twenty three kebeles, five kebeles were selected randomly from the district, three kebeles from midland and two kebeles from highland based proportion of kebeles found in each agro-ecology.

There are many approaches to determining the sample size. These include using a census for small populations, imitating a sample size of similar studies, using published tables, applying formulas and also rule of thumb (Glenn 1992; Knapp and Campbell-Heider, 1989). In this study rule of thumb (Knapp and Campbell-Heider, 1989) was used to determine sample size for this study. Accordingly the sample size was determined by equation (1) below;

n  30 10k (1)

Where: n = sample size and k = number of independent variables.

Then the study has k=13, n  30+10*13= n  160. But in this study the sample size was inflated to184, to check non-respondents. In second stage, from list of enset producer households in the sampled kebeles, 184householdswere selected randomly. The total number of households taken from each kebeles was based on, Pandey and Verma (2008) proportional sample allocation formula and given by equation (2) below;

nN n  i (2) i N

th Where: ni = Sampled households from i kebele

n = Sample size

th Ni = The total households in i kebele

N = Total households in selected kebele (sum total of households in five kebele)

Table1: Sample Distribution of enset producer Households in Selected Kebeles No Kebeles Agro- Ecology Total Households in the Kebele Sampled Households 1 Harowonchi Highland 1068 46 2 Weldotalfem Highland 1010 44 3 Worabu masse Midland 516 25 4 Harobasaka Midland 646 30 5 Sonkolekake Midland 904 39 Total 4144 184 Source: WDANRO, 2018 and own computation result.

2.4 Methods of Data Analysis

Two types of data analysis, descriptive statistics and econometric analysis were used to meet objectives of the study.

Impact Factor (JCC): 8.6084 NAAS Rating: 3.37 Determinants of enset Producers Market Participation Decision and Intensity 21 of Participation in Ensetproduct: The Case of Wonchi District, South West Shoa Zone, Oromia National Regional State, Ethiopia 2.4.1 Descriptive Analysis

To describe the demographic, socio economic, institutional charactetics of sampledenset producer in the study area the descriptive statistics were used. The inferential statistics (test statistics) such as t-test and chi-square testwere used to compare kochomarket participants and non- participants as well as different market outlet choicesby the sampled kocho producer households.

Econometric Analysis

This sub-section deal with procedures followed to choose the appropriate model to achieve the research objectives. Accordingly, various tests were run to make sure the proposed methodologies best fit the data.

2.4.2 Model Specification for Determinants of Participation Decision and Intensity of Participation in kocho Market

To select the model that best fits the data set fordeterminants of thekocho market participation decision and intensity of participation, a cavernous model specification tests were conducted.

Table 2: Model Specification Tests Test Type Candidate Model Hypothesis Test Statistics Decision ( Γ = -2[ln LT − (ln LP + ln LTR)] ∼χ2 ) 1.Likelihood ratio Tobit Vs Double Ho: Tobit best fits data -2(-556.47 – (-23.02+( Reject test Hurdle Model Ha: Double Hurdle best fits data -402.64) Ho 2 261 > 12 = 26.22*** 2. Akaike's Tobit Vs Double Ho: Tobit best fits data Ha: Double AIC= tobit= 1140.945 Reject Information Hurdle Model Hurdle best fits data AIC=DHM= 903.3399 Ho Criterion Voung test statics= Ho: No difference between Heckman 5.696 Double Hurdle Vs Reject 3. Vuong test model and DHM ( test statics is smaller  Heckman model Z =2.575 at ( Ho than Zcrit) 2 =1%) Source: Own computation from survey result, 2018.

Following the justifications provided in model specification tests above table; Double hurdle model was used to identify determinants of producer market participations decision and intensity of kochosales in the study area. Cragg (1971) Double-hurdle model evaluate two separate decisions regarding market participation and intensity of selling. In the first stage, probit model was used to identify determinants of kocho market participation decision by producers. The probit model that assesses the producer’s participation decision in kochomarket described by equation (3);

* D i =  X i + i i ~ N (0, 1) (3)

= 1, if >0

0, if = 0

Where, is latent variables takes the value 1 if the producers participate in kocho market as sellers and 0 otherwise, α is a vector of parameter to be estimated, X i is independent variables hypothesized to affect households

www.tjprc.org [email protected] 22 Shelemereferajebesa

decision tokocho market participation and i error term.

In the second stage, the truncated regression model was used to analyze determinants of the intensity of kochosales. Therefore, the second hurdle represents the intensity of kochomarketed and modeled as equation (4);

* 2 Yi =  Zi  i  i ~N (0, ) (4)

* * Yi  Y* if 00

0 if =0

2.5 Definition of Variables and Working Hypotheses

Table 2: Summary of Variables used in Double Hurdle Model Dependent Variables Type Measurement Market participation Dummy Yes=1, No=0 Expected Sign Intensity of participation Continuous Proportion Independent variables Participation Intensity Sex of the household head Dummy Female=1, Male=0 + + Educational level Continuous Year of schooling + + Availability of economically active labor Continuous Adult Equilevent + + Ensetfarming experience Continuous Year + + Land allocated for ensetcrop Continuous Hectare + + Participation in off/non farm activity Dummy Yes=1, No=0 - - Means of own transportation Dummy Yes=1, No=0 + + Livestock owned Continuous TLU + + Decision making on kocho income utilize Dummy Female=1 + + Access to market information Dummy Yes=1, No=0 + Frequency of extension service given Continuous Number of days + + Distance to the nearest markets Continuous Kilometer - - Source: Own completion based on literature(2018). RESULTS AND DISCUSSIONS

Descriptive Results

From total of 184 sampled enset producers 118 (64.13%) were participants in kochomarket while the remaining 66 (35.87%) were non- participants during 2017/18 production year. From total sample producers 105 (57.07 %) were female headed including female spouse in male headed and the remaining 79 (42.93%) were male headed households. Bear in mind here, household is one who made decision aboutkocho marketing in context of this study. Among kochomarket participants, female and male headed households constitute 70(59.32%) and 48(40.68%) respectively. Out of non- participants, 35(53.03%) were female headed while the remaining 31(46.97%) were male headed households.

Regarding agro ecologies where respondents are found, 94 (51.08%) found in midland agro ecology and about 90(48.92) found in highland agro- ecology. Group comparisons of the kochomarket participants and non-participants were computed using chi2-testfor dummy variables and t-test for continuous variables and the results are presented in table4 and 5 respectively.

Chi2- test statistics indicated that variation in agro ecologies where respondents are found between kocho market participants and non- participants was statistically significant 1%. This implies that producers found in the highland were more participated in the market than midland. The reason might be the basic livelihood of producers found in the highland

Impact Factor (JCC): 8.6084 NAAS Rating: 3.37 Determinants of enset Producers Market Participation Decision and Intensity 23 of Participation in Ensetproduct: The Case of Wonchi District, South West Shoa Zone, Oromia National Regional State, Ethiopia isensetcrop and also the crop is more productive in highland than midland. The other reason is topography of the farm land, the topography of highland is not suitable for traction as compared to midland and they allocated it to enset crop based on human power as this crop is the major field crops in highland in the district.

Table 4: Chi-square Test for Samplekochomarket Participants and Non-Participants Non- Participants Total Sample Participants Variables (N=118) (n=184) 푥2-Value (N=66) N % N % N % Sex of household Female 70 59.32 35 53.03 105 57.06 0.684 Male 48 40.67 31 46.97 79 42.93 Agro ecology Midland 45 38.13 49 74.24 94 51.08 22.083*** Highland 73 61.87 17 25.76 90 49.92 Off/non farm activities participation Yes 67 56.77 37 56.06 104 56.52 0.009 No 51 43.33 29 43.94 80 43.48 Access to kochomarket information Yes 86 72.88 15 22.73 101 54.89 42.998*** No 32 27.12 51 77.27 83 45.11 Own means of transportation facility Yes 109 92.37 36 54.45 145 78.80 36.259*** No 9 7.63 30 45.56 39 21.20 *** indicate significance at 1%. Source: Own survey result (2018).

From the total sample respondents 101(54.89%) of them have accesses to market information. From those about85.15% were participants inkochomarket and the remaining 14.85% were non-participants. Usually the information on the prices of kochoin the market is retrieved from different sources such as friends or neighbors, kochotraders and development agents.On the other hand, chi2- test statistics indicate that variation in access to market information between kocho market participants and non- participants was statistically significant 1% (table 4).

Out of total sample households 39(21.20%) of them have not own means transportation and the remaining 78.8% haveown means transportation. About 5(2.7%),32(17.4%) and 108(58.7%) of sampled respondents have animal cart only, animal cart and pack animal, and pack animal only like horse, mule and donkey as means of transportation respectively. The variation in own means of transportation among kochomarket participant and non-participant is statistically significant at 1% in favor of the former (table 4).

The average age of sample respondents was 46.49 years. This implies that most of respondents were within productive age. The average age of participants was 49.51 years, while for non-participants it was 41.09 years, the t-test statistics revealed that age difference between the groups was statistically significant at 1% probability level. The average level of education of sample respondents is 4.94 years. This average year of schooling is greater than the average year of schooling 2.7 years in Ethiopia (HDII, 2018).The mean years of schooling of kochomarket participants and non- participants is 4.85 and 5.10 years respectively.

The overall mean family size in adult equivalent ratio per households was about 3.85 which are lower than the www.tjprc.org [email protected] 24 Shelemereferajebesa national average 4.6 per households (CSA, 2014). The mean family size of kochomarket participants and non-participants were found to be 3.92 and 3.73 family size respectively.

Labor availability (healthy active labor force) is a prominent input for enset production as well as marketing. The overall mean economical active labor in family in adult equivalent ratio was 3.32 in family members per household. The mean economical active labor in family of kochomarket participants and non-participants were 3.43 and 3.12 respectively. The t-test statistic showed that there was statistically significant mean difference economical active labor in family between market participants and non-participants at 10%significance level. This implies that households with more healthy active labor force participate in kocho market more than households with less healthy active labor force.

The averageenset farming experience of sample respondents was 24.13 years. The mean enset farming experience of kochomarket participants and non-participants were 28.62years and 16.09 years respectively. The mean difference in enset farming experience among market participant and non-participant is statistically significant at 1%significance level (table 5).

Table 5: T-test for Sampled kochomarket Participants and Non-Participants Non- Total Participants Participants Sample Variables (N=118) t-value (N=66) (N=184) Mean Mean Mean Age 49.510 41.090 46.490 -6.041*** Education level 4.850 5.100 4.940 0.521 Family size 3.920 3.730 3.850 -0.788 Availability of economically active labor in 3.430 3.120 3.320 -1.355* family Enset farming experience 28.620 16.090 24.130 -9.963*** Total land size owned 1.490 1.670 1.550 2.252** Total land allocated for enset 0.350 0.140 0.270 -10.500*** Total livestock owned 4.760 4.810 4.780 0.151 - Quantity of kocho produced 11.980 5.460 9.640 16.480*** Quantity kocho consumed 5.090 5.460 5.220 2.366*** Frequency of Extension service given 15.270 10.900 13.700 -3.522*** Distance to the nearest market 4.270 5.650 4.770 5.488*** ***, ** and * indicate significance at 1%, 5% and 10%. Source: Own survey result (2018).

3.2. Econometric Results

Before running double hurdle and multivariate probit models necessary tests that verify the hypothesized independent variables and existence of econometric problems such as multicolinearity problem, heteroscedasticity and outliers were performed using appropriate test statistics. According to Gujarati (2003) multicolinearity is a situation where explanatory variables are highly correlated. To detect whether the data at hand exhibits multicolinearity problem or not, Variance Inflating Factor and Contingency Coefficient testswere used.

The VIF was used to test multicolinearity of continuous variables using STAT command ‘Collin’ without regressing OLS in this study. If the VIF of a variable exceeds 10, that variable is said to be highly collinear and can be excluded from the model. Contingency coefficient values which ranges between 0 to 1, measures the degree of correlation between dummy variables based on chi-square measure of association. Contingency coefficient value with 0.75 or more

Impact Factor (JCC): 8.6084 NAAS Rating: 3.37 Determinants of enset Producers Market Participation Decision and Intensity 25 of Participation in Ensetproduct: The Case of Wonchi District, South West Shoa Zone, Oromia National Regional State, Ethiopia shows a strong degree of association between dummy variables. Therefore, the mean VIF of continuous explanatory was 1.27and for all dummy variables included in the model, the correlation coefficient is below 0.75. Indicate that there was no serious problem of multicolinearity and no data lost due to multicolinearity problem. Heteroscedasticity exists when the assumption that the variance of the error term is constant across the observation is violated (Greene, 2008).

Accordingly, heteroscedasticity was tested using Breusch-Pagan/ Cook-Weisberg and White test. Since Breusch- Pagan/ Cook-Weisberg and White tests evaluates the null hypothesis of a constant variance in the data. In both tests constant variance of the error terms across observations was rejected (Breusch-Pagan/ Cook-Weisberg test, Prob> chi2= 0.0522) and white test Prob> chi2 = 0.0488) meaning that heteroscedasticity was a problem, therefore heteroskedasticity- consistent standard errors (robust) method was used to overcame the problem . For all variables outliers were checked using the box plot graph, so that there were no serious problems of outliers and no data lost due to outliers.

Factors Affecting market Participation and Intensity of Participation in Kocho Market

Out of the twelve explanatory variables fitted first hurdle model, seven variables are significantly influence sample respondents decision to participate in kocho market. These are: availability of economically active labor in a family, total land allocated for enset crop, livestock owned, decision making on kochoincome utilization, distance to nearest market, own means of transport facility and access to market information. In same ways eleven variables included in second hurdle four variables significantly affect intensity of participation in kocho market. These are: educational level of household head, experience of ensetfarming, total land allocated for enset crop and decision making onkocho income utilization.

Table6: Double Hurdle Model Estimation Results of Factors Affecting Market Participation and Intensity of Participation in kocho Market 1st Hurdle 2nd Hurdle Variables Robust Robust Coeff AME Coeff Std. Err. Std. Err. Sex of household head 0.716 0.436 0.049 1.650 1.421 Educational level 0.050 0.078 0.003 0.406* 0.237 Availability of economically active labor 0.331* 0.185 0.023 -0.565 0.389 Enset farming experience 0.043 0.029 0.003 0.418*** 0.076 Land allocated for ensetcrop 10.407*** 2.369 0.720 18.977*** 4.805 Participation in off/non farm activity -0.733 0.466 -0.050 1.720 1.525 Means of own transport facility 0.939** 0.422 0.065 - - Livestock owned -0.226** 0.098 -0.015 0.609 0.398 Decision making on kocho income utilization 4.846*** 0.937 0.335 3.402** 1.695 Access to market information 1.970*** 0.422 0.136 0.319 1.704 Frequency of extension contact 0.048 0.044 0.003 -0.026 0.083 Distance to the nearest market -0.252** 0.102 -0.017 0.059 0.498 Constant -4.650*** 1.429 29.825*** 4.206

Sigma 7.339 *** 0.505

***, ** and * indicate significance at 1%, 5% and 10%, Log likelihood = -425.66, N1hurdle = 184, N2hurdle=118, Wald chi2 (12) = 26.65, Prob> chi2 = 0.0087*** Source: Own survey result (2018).

Availability of Economically Active Labor in Family

As expected economically active labor availability in a family affects probability of kochomarket participation positively and significantlyat10% significance level. The average marginal effect of this variable on probability of market participation showed that as economically active availability in family increase by one adult equivalent ratio, the probability of the household to participate in kocho market increased by 2.3%. Due toenset production and processing was www.tjprc.org [email protected] 26 Shelemereferajebesa more tedious and it needs more labor. The result is in line with Nuri (2016) household with more number of labor produce more kochoand participates in kochomarket. Also, the findingAtaul and Elias (2015) householdswith a large number of active household labors can reduce their cost of production and produce surplus to be market-oriented.

Land Allocated for ensetcrop

As expected, this variable has a positive and significantly affect decision to participate in kochomarket by producers’ at 1% significance level.

When the land allocated for ensetis large, there is more number of ensetstands ready for harvesting which enables the household to have large volume of kochoproduced. The average marginal effect implies that given other factors are unchanged, when the land allocated for enset crop increase by a hectare the probability of kochomarket participation decision increases by 72%.The result is in line with Tessema (2017) area under ensetcultivation to have a positive relationship with kochomarket participation of ensetproducer households.There a positive and significant relationship existed between intensity of kocho sale and land allocated for enset crop by producers. It affected the intensity of kochomarketed significantly at 1% probability level. The model output indicate that one hectare increase in the land allocation for enset, itwill increases intensity of kochomarket participationby 18.97% holding other variables constant. The result is in line with Addisu (2018) found that size of land under teffproduction was positively and significantly affect the level teffcommercialization.

Means of Own Transport Facility

This variable affects participation in kochomarket positively and significantly at 5%. As compared to households who have no own means transportation, households who have own means transportation, probability of participating in kocho market by 6.5% more. The reason was ownership of transport facility reduces the cost of transporting kocho to the kocho market. This is in line with Efaet.al. (2016) found that ownership of transport equipment is positively and significantly influences teff market participation.

Livestock Owned

There are two opposing arguments made by different empirical studies regarding livestock owned and market participation. The first group argued that ownership of livestock reduces market participation, since these livestock offer alternative sources of cash income and hence the relationship is negative (Tadie and Lemma, 2018). The second group claims that the first group livestock size has positive contribution to market participation decision (Habtamu, 2013). Finding of the present study is consistent with the first argument that concludes the number of livestock owned by household influence kocho market participation negatively and significantly at 5%. Holding other explanatory variables constant, the result shows that when the livestock owned increased by one tropical livestock unit, probability of kochomarket participation decreases by 1.5%. This was due to producers who have more livestock tend to sell them instead of selling kocho produced to cover their yearly expenditure; they may tend to specialize in livestock production as a means of generating cash.

Decision Making on kocho Income Utilization

Consistent with a prior expectation there is positive relationship between decisionsmaking onkocho income utilization and kocho market participation at 1%. As compared to households who decision of income utilization given by male or negation between husband and wife, those households who decision of income utilization given by female, the probability

Impact Factor (JCC): 8.6084 NAAS Rating: 3.37 Determinants of enset Producers Market Participation Decision and Intensity 27 of Participation in Ensetproduct: The Case of Wonchi District, South West Shoa Zone, Oromia National Regional State, Ethiopia of kocho market participation increases by 33.5%. The reason might be more work load is left for female starting enset production to kocho selling in market, when decisionmade by them they more motivated to processkocho and supply to market and also as expectation there is positive relationship between decision on utilizing incomekocho and kocho market participation at 5%. As compared to households who decision of income utilization is given by male or negation between husband and wife, those households who decision of income utilization given by female, the intensity of kocho market participation was increases by 3.4%.

Access to kochomarket Information

Ashypothesized this variable influence kocho market participation positively and significantly at 1%. As compared to household who had not get access to kocho market information, households who had get increases the probality of participating in kocho market by 13.6%. Producers who have kocho market information prior to marketing tend to sell more of their produce than those without. The result is line with Nuri (2016) found that the positive relation exists between market information and kocho market participation.

Distance to the Nearest Market

As hypothesizedthis variable influence kocho market participation negatively and significantly at 10%. When distance to the nearest kocho market increases by one kilometerthe probability of participating in kocho market is reduced by 1.7%. This was due to the far distance the household resides from the nearest kocho market; the less likely the household involved in selling kocho due to long walking time and relatively high marketing costs and less access to market information. Tessema (2017) found that as distance in kilometer increase from the farmers’ residence to kochomarket, the probability of its market participation decision was decreased. This result is also consistent with the findings Tadeleet al. (2017) reported that distance to market of selling wheat in minutes of walk from wheat producers homestead influence the level of market participation negatively and significantly.

Educational Level of Household Head

As hypothesized educational level of household head was found to have positively and significantly affected the intensity of kochoparticipation at 10% significance level. The average marginal effect indicated that as the level of formal education of the household head increases by one grade, intensity of participation is increases by 0.406%. This indicates that attending formal education increases ability to get and use information and to adopt better production practices and produce large volume of kocho.The result is in line with Takeleet al. (2017) education was positively affecting the intensity of participation in mango market.

Enset Farming Experience

As prior expectation there is positive relationship between experience inenset farming and intensity kocho market participation at 1%. Producer with longer production experience had more knowledgeable and good conditions forecasting ability, this improves the kocho production and intensity of kocho sold. That is as experience of ensetfarming increases by a year intensity of kochomarketparticipation increases by 0.418%. The result is in line with Toyibaet al. (2014) found that experience in papaya production had a positive and significant effect on papaya volume marketed.

SUMMARY, CONCLUSIONS AND RECOMMENDATIONS www.tjprc.org [email protected] 28 Shelemereferajebesa

This chapter summarizes the findings of the study and makes conclusions based on the results of the descriptive and econometric model. Specific policy recommendations emanated from the findings are also given.

4.1 Summary

Enset (Ensetventricosum) is one of the indigenous root crops widely cultivated in the south and south western parts of Ethiopia. Its product specifically kocho has a significant contribution to the livelihood of producers as source of food and income as well as ensuring of food security in wonchi district. Despite of enset potential, in the district, there has limited performance of producers in the ensetproduct marketing in the district. Determinants of ensetproducers market participation and intensity of participation in ensetproduct is not documented. Therefore, specific objectives of the study were identifying determinants ofensetproducers’ market participation decision and intensity of participation in kocho market.

To meet the objectives of the study primary data were collected from randomly selected 184 ensetproducer’s households in the five kebele in 2018/2019 cropping season and from a total of using pre tested semi-structured questionnaire. Descriptive statistics and econometric models were used to analyze the collected data. Double hurdle model was used to identify the factors affecting kochomarket participation decision.

The descriptive statistics shows that, out of surveyed sample respondents, 64.13% of sampled respondents were participants in kocho market. The chi-square test reveals that there is significant difference in terms of agro ecology, access to market information and own means of transport facilities between kochomarket participants and non-participants and t- test also revealed that mean difference in age of respondents, land allocated for enset crop, ensetfarming experience, distance to the nearest market, frequency of extension service given and quantity of kochoproduced was statically significant between two groups.

Double hurdle model was used to identify the factors affecting participation decision in kocho market and intensity of participation. The result showed that, availability of economically active labor in family, total land allocated for enset crop, decision making on kocho income utilization, own means of transport facility and access to market information influence kochomarket participation decisions positively and significantly while livestock owned and distance to nearest market negatively and significantly affect market participation decision of sampled respondents. Moreover the result indicated that the intensity of participation was affected by educational level of household, experience of enset farming, total land allocated for enset crop and decision making on kocho income utilizationpositively and significantly.

4.2 Conclusions and Recommendations

Depending up on the findings of the study the following conclusions and recommendations are made.

Econometrics result shows that availability of economically active labor in family determineskochomarket participation decisions positively and significantly. Therefore, improving the capacity of available productive labor of the family through training is important to increase enset production, kochomarket participation. Distance to the nearest market negatively and significantly affects the participation decision in kochomarket Therefore, developing market infrastructure as building market center and repairing roads to reduce walking time and minimize transportation costs which in turn improve kochomarket participation and better offering outlets like wholesalers rather than being exploited by nearby village collectors.

Impact Factor (JCC): 8.6084 NAAS Rating: 3.37 Determinants of enset Producers Market Participation Decision and Intensity 29 of Participation in Ensetproduct: The Case of Wonchi District, South West Shoa Zone, Oromia National Regional State, Ethiopia The result also shows that own means of transport facility and access to market information determinekochomarket participation decisions and wholesalers outlets positively and significantly. Therefore, improving transportation access of the producers is essential. Access to kocho market information is also an important issue for ensetproducers to help them decide on participates in the market. So it is important to spread market information to all the kochochain actors. District trade and marketing office could extend up to date kochomarket information collaboration with agricultural extension agents. Educational level of the producers and enset farming experience of the producers were positively and significantly related with intensity of kocho market participation. Therefore, the government should give emphasis on encouraging farmers to learn adult and formal education and rising experience producers through experience sharing on the enset production to enhance intensity of participation by ensetproducers.

REFERENCES

1. Abebe Uma, PaulMansingh .J,FikaduAbdise and MuluDebela (2015). Market performance of ensetproducts: Kocho and Bulla,the case of Dire Inchiniworeda Oromia Regional State, Ethiopia.

2. AddisuHailu( 2016). Value chain analysis of vegetables: The case of Ejere district, West Shoa Zone, Oromia region, Ethiopia. MSc Thesis, Haramaya University, Haramaya, Ethiopia

3. Ashenafi Chaka, TadesseKenea and GirmaGebresenbet (2017).Analysis of the supply chain and logistics practices of Warqefood products in Ethiopia. International journal on food system dynamics, 7(3): 213-228.

4. AtaulGani and Elias Hossain (2015).Market participation decision of smallholder farmers and its determinants in Bangladesh.A review.

5. Barrett, C.B. (2007). Smallholder market participation: Concepts and evidence from Eastern and Southern Africa. Food and agricultural organization workshop on staple food trade and market policy options for promoting development in Eastern and Southern Africa, Rome, March 1‐2, 2007.

6. Baptiste, Ettien Djecthi Jean, Sorho Fatogoma, and Brahima Kone."Productivity of new yam assessions as affected by mosaic virus in transition forest-savanna zone of Côte d'Ivoire." International Journal of Agricultural Science and Research (IJASR) 4.3 (2014): 137-145.

7. Cragg, J.G. (1971). Some statistical models for limited dependent variables with application to the demand for durable goods. Econometrica, 39(5): 829-844.

8. CSA (Central Statically Agency).(2017). Area and production of major crops on private peasant holdings, Meher Season in Ethiopia. Volume I

9. CSA (Central Statistical Agency) (2014). Agricultural sample survey: Report on area and production of major crops (private peasant holdings, meher season), Statistical Bulletin. , Ethiopia.

10. CSA (Central Statistical Authority) (2013).Agricultural sample survey of Ethiopia. Federal Democratic Republic of Ethiopia. Addis Ababa, Ethiopia

11. CSA( Central Statistical Agency) (2015). Agricultural sample survey of Ethiopia. Report on area and production of major cops. Addis Ababa, Ethiopia.

12. EfaGobena, DegyeGoshu, TinsaeDemisie, and TadesseKenea (2016.) Determinants of market participation and intensity of marketed surplus of teffproducers in Bachoand District, Ethiopia. Journal of Agricultural Economics, 5(2): 020-032.

www.tjprc.org [email protected] 30 Shelemereferajebesa

13. Ejigu, Yazachew Genet, Pangirayi Bernard Tongoona, and Beatrice Elohor Ifie. "General and specific combining ability studies of selected tropical white maize inbred lines for yield and yield related traits." International Journal of Agricultural Science and Research 7.2 (2017): 381-396.

14. Garrity, D., Dixon. J and Boffa.J.(2012). Understanding African farming system science and policy.

15. Gujarati, D.N. (2003). Basics Econometrics,4th Edition. McGraw-Hill, New York.

16. HabtamuYesgat (2013). Transformation in smallholder agriculture through intensification in Ethiopia: Determinants and implications. InternationalJournal of Agricultural Science, Research and Technology, 2(3): 123-128.

17. Hassen, R. (2006). The climate change and African agriculture: Measuring the economic impact of climate change on the Ethiopian agriculture. Discussion papers, Pretoria University

18. HDII (Human Development Induces and Indicators) (2018). Statistical update.

19. Ismael, Yassin, Adam Bekele, and MengistuKetema. "Determinants of level of smallholder farmers participation in Potato sales in Kofele District, Oromia Region, Ethiopia." Journal of Agricultural Science and Research 3.2 (2016): 23-30.

20. Knapp, T. R. and Campbell-Heider, N. (1989.)Numbers of observation and variables in multivariate analyses.Western Journal of Nursing Research,11: 634-641.

21. Nuri Lefebo( 2016). Value chain analysis of Ensetin Hadiya zone, Southern Ethiopia. PhD Dissertation,Harmaya University, Haramaya, Ethiopia.

22. Pandey, R. and Verma, M. R. (2008).Samples allocation in different strata for impact evaluation of developmental programme.Rev. Mat. Estat, São Paulo,26(4):103-112.

23. SWSZANRO (South West Shoa Zone Agriculture and Natural Resource Office) (2018.)

24. TadeleMamo, WudinehGetahun, AgajieTesfaye, Ali Chebil, Tesfaye Solomon, Aden Aw-Hassan, TolessaDebele and Solomon Assefa (2017) Analysis of wheat commercialization in Ethiopia: The case of SARD-SC wheat project innovation platform sites. African Journal of Agricultural Research,12(10): 841-849.

25. TadieMirie and Lemma Zemedu(2018).Determinants of market participation and intensity ofmarketed surplus among teff producers in Dera Districtof South Gondar Zone, Ethiopia.Journal of Development and Agricultural Economics, 10(10): 359- 366.

26. TakeleHonja, EndriasGeta and Amsalu Mitiku (2017).Determinants of intensity of market participation of smallholder mango producers: The Case of Boloso Bombe Woreda, Wolaita Zone, Southern Ethiopia. Journal of Marketing and Consumer Research,32: 56-63.

27. TessemaErchafo (2017). Market chain analysis of enset products the case of Doyogena district, southern Ethiopia.M.Sc.Thesis, Haramaya University, Haramaya, Ethiopia.

28. Toyibashafi, Lemma Zemedu and EndriasGeta( 2014). Market chain analysis of papaya (Carica Papaya): The case of Dugda district, Eastern Shewa Zone, Oromia National Regional State of Ethiopia. Academe Research Journals, 3(8): 120-130.

Impact Factor (JCC): 8.6084 NAAS Rating: 3.37