Willingness to Pay for Assisted Reproductive Technologies by Pastoral Herd Owners in Southern Rangelands of Kenya

8 Uptake of Assisted Reproductive Technologies (ARTs) can widen and accelerate access to quality breeding 9 stock under pastoral systems, but user participation without subsidy is uncertain when the herders have to meet 10 the direct cost. Analysing willingness to pay (WTP) for ARTs may provide insights into potential uptake of 11 ARTs when development agencies withdraw subsidies and private entrepreneurs take up the service delivery. 12 This study estimated WTP for ARTs among 130 pastoral herd owners participating in Sahiwal cattle breed 13 upgrading program implemented in Transmara Sub County in the southern rangelands of Kenya. The average 14 WTP was KES 3,643 (USD 33.4), equivalent to 21.4% premium with reference to base price (KES 3,000) (USD 15 27.5). Sixty-seven percent of the pastoral herd owners expressed WTP above the hypothesized market price, 16 twice more than those expressing WTP below the hypothesized market price (33%). The first choice attributes 17 influencing the WTP for ARTs were high milk yield (62%), high value calves (37%), and high growth rates 18 (1%). Gender was the only socioeconomic factor significantly (P<0.005) influencing WTP, with men expressing 19 higher WTP than women (KES 3,870 (USD 35.5) versus KES 3,223 (USD 26.6). The results indicate a high 20 likelihood of pastoral herd owners continuing to access and use ARTs at own cost, which provides opportunities 21 for private sector participation in ARTs delivery and upscaling to increase access to superior Sahiwal genetic 22 resources in pastoral herds. This will need putting in place policy interventions supportive of efficient ARTs 23 delivery mechanisms. 24

strategy has been associated with increasing inbreeding levels and loss of genetic diversity among Sahiwal cattle 39 population in Kenyaand will be tragic, if remains unchecked (Ilatsia et al., 2012;Mbuku et al., 2019).

40
Effective selection efforts for resilient and productive animals are required. Continued bull services could prove 41 expensive as they are associated with risks of disease carriers, injurious to young heifers, increasing inbreeding 42 levels and loss of genetic diversity with intense use (Mbuku et al., 2019). To increase access to high quality previous study of AI utilization in ASALs (Narok county), observed a higher level of awareness about AI 50 services among pastoralists (70%), but those still using bull services as breeding method were predominating 51 (Khainga, 2015). Continued use of bull services was partly attributed to preference for bull service (Janssen-

79
applying the analytical framework of Hanemann et al., (1991). This model uses hypothesized bid levels that 80 reflect the cost of accessing the service as the basis for calculating the mean WTP and given that the herd 81 owners were aware of the ARTs but unable to attach true value to use of ARTs, close-ended questions approach 82 was adopted.

83
The herd owner was initially asked if he/she was willing to pay an amount say 'Bi' for the provision of ARTs 84 services per animal served. The level of the second bid level was contingent upon the response to the first bid; 85 "higher than the initial bid if the response was 'yes' and this assumed that the Bi ≤WTP<∞", or "lower if the 86 response was 'no' and this assumed that the 0≤WTP< Bi". The subsequent bids played an important role in 87 placing an upper and lower bound on the respondents' unobserved true WTP (Alberini and Cooper, 2000).

88
Four possible outcomes can be obtained from the double bounded dichotomous model. These are presented as 89 yy, yn, ny and nn where yy implies that both answers are 'yes' (WTP is higher than the upper bid) and yn implies 90 first answer was 'yes' followed by 'no' (WTP is between initial bid and the upper bid). On the other hand, ny 91 implies a 'no' answer followed by 'yes'(WTP is between lower bid and the initial bid) and nn implies 'no' 92 answer in both (WTP is between zero and the lower bid) (Hanemann et al., 1991). The possible outcomes and 93 inferences from double bounded dichotomous choice questions on WTP are presented in Table 1.

98
The respondent was therefore assumed to make decisions following possible discrete outcomes, that is; where WTP is 0 if the respondent is not willing to pay for the offered bid price and 1, if he/she is willing to pay 101 the bid price for accessing ARTs. The dependent variable (WTP) was therefore hypothesized to be influenced 102 by a set of socio-economic factors attributed to the respondents (Table 2)  analysis. The four (4) excluded were incomplete or inconsistent responses on one or more variables. Table 3 119 presents the summary descriptive characteristics and mean WTP for ARTs based on gender, age, education 120 level, primary livelihood source, production system, grazing area, farming experience, market distance and first

135
In general, 67% and 33% of the herd owners expressed WTP above and below the bid prices respectively.

136
Seventy one percent (71%) and 29% of male and female herd owners expressed WTP value above the base price 137 while an equal proportion (50%), declined the first bid with WTP below the bid price (Table 4). Middle aged 138 herd owners (between 31-60 years) expressed 69% and 64% above and below bid WTP price respectively 139 relative to above 61 years who expressed 2% and 5% above and below bid WTP price respectively. Higher milk 140 yield attribute as first choice for using ARTs by herd owners had 62% above the bid WTP and 57% below the 141 bid WTP; high value calves above 37% and below is 43%; and high growth rate above the bid WTP at 1%.

142
In the regression analysis, only gender was retained as socio-economic factor of significant (P<0.05) influence 143 on the level of WTP for utilizing ARTs by herd owners (Table 5).

170
The average WTP was an equivalent of 21.4% premium on the base price of the ARTs.

188
On the other hand, females considered higher milk production as their initial reason for WTP for ARTs. To