This variable was found to adversely affect supply of livestock to the market. As hypothesized, increased number of kilo meters to the market center led to the decrease of supplied volume of livestock to the market. Therefore, this variable negatively influenced the supply side of live animal to the market. It further reveal that, farmers who resided far away from lare town treat livestock similar to saving accounts and typically only supplied less number of livestock when certain necessary expenditure arise. This finding is linked with Mebratu (2016) who indicated that distance to market adversely affecting the total sale of Turmeric in Yeki woreda of South Nation and nationalities Region.
4.3.7 Access …show more content…
A unit’s increase in the accesses to market information led to the increases of livestock supplied by 46.7%. Therefore, earlier hypothesis shows that market information plays superior role in determining volume of livestock supplied to the market. This hypothesis hold true when comparing to regression result. Given the emphasis placed on the multi-functionality of livestock, market information can help farmers to gain bargaining power so that they can sell their livestock with the expected price.
4.3.8 Total livestock units
This variable does not necessarily influence total volume of livestock supplied to the market. Given the fact that farmers holding more number of livestock without access to market information and improved education level in which they have strong market bargaining. So any price expectation would likely and highly be inaccurate to them. Therefore, farmers would prepare holding livestock as store of wealth than liquidating it. In this regard total livestock units holds by a given farmers is considered to have negative effect livestock volume supply to the market.
4.3.9 Access to extensive …show more content…
The test result for multicollinearity shows no multicollinearity problem because the results of VIF for variables were less than 10.
Moreover, Breusch-Pagan/Cook-Weisberg test for heteroscedasticity reveals that variables were free from heteroscedasticity problem. Further test for (IIA) independent of irrelevant alternative was run, and assumption was on the basis of likelihood ratio. The model shows the existence of 36% variation of market channel choice among livestock producers in lare woreda.
Table 11. Marginal fixed effect after multinomial logit model
Source; field survey 2017
Note: Primary market is the base outcome among others two outcome (farm gate and secondary channels, dy/dx is marginal effect, N =120, ***, **, * are significant level at 1%,5%, and