Heterogeneous seed access and information exposure: implications for the adoption of drought-tolerant maize varieties in Uganda

Agricultural and Food Economics

Table 5 ATE-corrected marginal effects of the determinants of adoption of DTMVs under heterogeneous seed access and information exposure

	Model 1 (ATE probit with exposure unconstrained)	Model 2 (ATE probit with exposure–access–unconstrained)	Model 3: (ATE probit with exposure–access–affordability–unconstrained)
Gender of hh head (1 = m, 0 = f)	− 0.076	− 0.082	− 0.047
Years of education	0.002	0.001	− 0.013
Age of head of household	− 0.079	− 0.117	− 0.197
Household size	0.007	0.009	0.007
Membership in association (1 = yes, 0 = otherwise)	− 0.009	− 0.037	− 0.006
Farm size	0.000	0.01	0.008
Distance to market	0.027	0.036	0.039
Information sources (reference group: no information received)
Government	0.148	0.137	0.222
Field days	0.084	0.08	0.233
Radio	− 0.024	− 0.063	0.036
Agro-dealers	0.068	0.131	0.259
Other farmer	− 0.085*	− 0.102	− 0.054
Income status (reference group; insufficient income need borrowing)
Allows to build savings	0.371**	0.420**	0.213
Allows to save	0.213*	0.260*	0.175
Income equal expenses	0.102	0.094	− 0.004
Draws from saving	0.286*	0.331*	0.359*
Frequency of drought	− 0.005	− 0.003	0.007
Received Subsidy (1 = yes, 0 = no)	0.619***	0.606***	0.467***
Livestock units	0.011	0.016	0.007
Eastern region	0.291**	0.342**	0.442**
Western region	0.379*	0.399*	0.375**
Northern region	0.082	0.147	0.24
No. of observations	489	345	214
Log likelihood	− 221.67	− 182.85	− 123.43
LR chi²	86.0	73.10	48.48
Df	29	29	29
Pseudo R²	0.78	0.17	0.16

We only present results of the ATE-corrected adoption models for exposure, seed availability and access to seed at affordable prices
*p < 0.10, **p < 0.5, ***p < 0.01