Skip to main content

Research, Policy and Practice

Table 4 Two-step Heckman model results for fodder market participation

From: Determinants of pastoralists’ participation in commercial fodder markets for livelihood resilience in drylands of northern Kenya: Case of Isiolo

Variable

Decision to buy fodder equation

Quantity of fodder bought equation

Coef.

dy/dx

Coef.

Std. error

Access to market information (1 = yes; 0 = no)

1.329***

0.474

Access to weather information (1 = yes; 0 = no)

0.526**

0.188

− 9.927

7.412

Access to extension services (1 = yes; 0 = no)

0.212

0.076

3.442

5.991

Access to credit (1 = yes; 0 = no)

0.732***

0.261

Land tenure (1 = private; 0 = communal)

− 1.190***

− 0.425

− 20.43**

9.543

Number of shocks experienced (in the last 10 years)

0.470***

0.168

5.801**

2.423

Log age of respondent (years)

0.834*

0.298

13.512

9.838

Sex of household head (1 = female; 0 = otherwise)

− 0.454*

− 0.162

5.233

5.779

Log household size (numbers)

0.125

0.044

− 8.345

6.222

Log land size owned (ha)

0.154

0.055

− 1.301

3.611

Tropical livestock units (TLUs)

0.0042

0.001

1.506***

0.181

Proportion of off-farm income

− 0.914**

− 0.326

3.326

11.002

Ward (town as base category)

0.520**

0.185

7.709

5.448

Constant

− 6.312***

− 86.39**

42.054

  1. Sample (n) = 201 (censored observations = 85, uncensored observations = 116). Rho = 0.618, sigma = 26.626, lambda = 16.451*, Wald chi-square (11) = 91.930, prob > chi-square = 0.000. Source: survey data (2018)
  2. ***Significance levels at 1%; **significance levels at 5%; *significance levels at 10%