Skip to main content

Table 7 Maximum likelihood estimates of determinants of technical inefficiency

From: Correcting for sample selection in stochastic frontier analysis: insights from rice farmers in Northern Ghana

Variable

Irrigation only

Rainfed only

Pooled

Coef.

Std. Err.

Coef.

Std. Err.

Coef.

Std. Err.

Const.

2.542

0.792

− 5.15a

0.981

− 1.552a

0.479

Age

− 0.038b

0.016

0.057

0.016

0.014

0.009

Sex

− 1.005a

0.366

1.916

0.459

0.656a

0.237

HH head

− 0.218

0.315

− 0.734

0.299

− 0.54a

0.194

Education

0.055a

0.021

− 0.018b

0.026

− 0.002

0.015

Commercial

− 1.467a

0.283

− 0.221

0.24

− 0.579a

0.163

Experience

− 0.002

0.021

− 0.078a

0.021

− 0.042a

0.012

Location

−0.722b

0.368

1.886a

0.436

0.631a

0.222

FBO

− 0.565c

0.311

− 0.432

0.287

0.052

0.177

Subsidy

− 0.31

0.375

− 0.869a

0.285

− 0.713a

0.179

Training

− 0.273

0.376

0.731a

0.278

0.112

0.184

Credit

0.553

0.508

0.465

0.332

0.575b

0.246

CC perception

− 0.064

0.378

0.83a

0.264

0.344c

0.187

HH size

0.006b

0.022

0.035

0.021

0.046a

0.015

Log-likelihood

5.77

 

− 160.88

 

− 318.61

 

N

223

315

538

  1. a, b and c represent 1%, 5% and 10% level of significance respectively
  2. Source: Field data, 2017