Endogenous Switching Regression Results for Adoption and Impact of Adoption on Net Returns
| Net Returns | |||
| Variable | Selection | Nonadopters | Adopters |
| CONSTANT | − 0.126(- 0.428) | 1.257* (1.814) | − 1.097 (- 0.292) |
| AGE | − 0.105 (- 1.580) | − 0.138 (- 1.359) | 0.095 (1.019) |
| EDUCATION | 0.476** (2.298) | 5.167** (2.487) | 3.825* (1.932) |
| EXTENSION | 0.246** (2.253) | 2.374* (1.864) | 4.698** (2.095) |
| FSIZE | 0.498* (1.896) | − 0.532** (2.104) | 0.423 (1.546) |
| FGROUP | 0.378** (2.362) | 3.570** (2.246) | 5.563** (2.179) |
| CREDIT | 0.403** (2.465) | 3.697** (2.282) | 6.347** (2.194) |
| MACHINERY | 0.131** (2.178) | 0.077 (1.361) | 1.262** (2.297) |
| NONFARM | − 0.249* (- 1.835) | 0.384* (1.775) | 0.676 (1.543) |
| ILLNESS | − 0.148* (- 1.872) | − 0.157** (- 2.151) | − 2.183** (- 2.091) |
| IMPROVED | 0.314* (1.924) | 4.634** (2.374) | 5.758** (2.360) |
| FERTAPLIC | 0.425 (1.156) | 0.387 (1.573) | 0.168 (1.620) |
| FERTPRICE | 0.158 (1.278) | − 0.541** (- 2.377) | − 0.269** (- 2.613) |
| LABOR | 0.062 (1.405) | 0.049 (1.382) | 0.074 (1.603) |
| WAGERATE | − 0.186 (1.297) | − 1.782** (- 2.282) | − 1.871** (- 2.446) |
| FERTILE | 0.234** (2.316) | 3.618** (2.181) | 5.982** (2.327) |
| PERMEABLE | 0.149** (2.185) | 0.083* (1.724) | 0.206** (2.313) |
| SLOPE | 0.186** (2.021) | − 1.162* (0.761) | − 2.267* (1.884) |
| KULDA | − 0.337* (- 1.861) | − 0.324 (- 1.528) | − 0.415 (- 1.267) |
| ZUWARI | − 0.138* (- 1.689) | 2.867* (1.929)* | 2.649** (2.283)* |
| TAMALE | 0.189 (1.536) | 0.475 (1.019) | 0.376 (0.9691) |
| PERCEPTION | 0.107** (2.386) | ||
| NONFARMRES | 0.034 (0.879) | ||
| EXTENSIONRES | 0.157 (1.598) | ||
| ln σN | 0.454*** (3.293) | ||
| ρNε | − 0.087 (- 1.428) | ||
| ln ρA | 0.562*** (4.615) | ||
| ρAε | − 0.529*** (2.831) | ||
| Log likelihood | − 521.76 | ||
| χ2-Statistic for | 0.859 [0.62] | ||
| overidentification | |||
| Likelihood ratio test | 23.16*** | ||
| of independent | |||
| equations χ2(1) | |||
Note: Absolute t-values in parentheses and p-value in square brackets. NONFARMRES and EXTENSION- RES denote residuals from the first-stage regressions for nonfarm work and extension agents, respectively.
↵* Significant at the 10% level;
↵** significant at the 5% level;
↵*** significant at the 1% level.