Sectoral and spatio-temporal differentiation in technical efficiency: A meta-regression

Figures & data

Table 1. Summary statistics and heterogeneity in MTE estimates

			Mean technical efficiency
Variables	Count	Frequency (%)	Minimum	Maximum	Average	Std. deviation
MTE	223	-	0.073	0.990	0.676	0.188
SSIZE	223	-	3.000	2507.000	210.928	267.455
NONPERR	120	0.541	0.165	0.950	0.672	0.180
PERR	13	0.058	0.470	0.910	0.716	0.159
OTHERCROPS	2	0.009	0.481	0.518	0.500	0.026
ANIM	5	0.022	0.725	0.856	0.808	0.052
MIXED FARMING	15	0.067	0.192	0.928	0.646	0.248
FISH	4	0.018	0.417	0.600	0.532	0.087
AQUACULTURE	5	0.022	0.680	0.840	0.771	0.063
FOODMANU	4	0.018	0.400	0.924	0.659	0.214
PHARMANU	3	0.013	0.500	0.657	0.574	0.079
TEXTMANU	1	0.004	0.665	0.665	0.665	-
WOOD	1	0.004	0.618	0.618	0.618	-
MANNEC	4	0.018	0.151	0.520	0.310	0.173
TRANSPT	1	0.004	0.630	0.630	0.630	-
INTERMED	28	0.126	0.073	0.938	0.718	0.200
INSURANCE	4	0.018	0.650	0.750	0.691	0.049
PUBADM	2	0.009	0.477	0.769	0.623	0.207
EDUPRIM	1	0.004	0.869	0.869	0.869	-
EDUSUP	1	0.004	0.825	0.825	0.825	-
EDUCOTHER	4	0.018	0.760	0.916	0.843	0.066
HOSP	10	0.045	0.385	0.990	0.765	0.185
HEALTHCEN	7	0.031	0.510	0.990	0.807	0.192
COASTL	30	0.135	0.192	0.990	0.677	0.182
MID	38	0.170	0.165	0.965	0.734	0.212
NORTH	69	0.309	0.170	0.916	0.660	0.178
GHANA	38	0.377	0.151	0.990	0.688	0.206
DYEAR	223	-	2000	2019	1998	171.467
DEA	58	0.260	0.073	0.990	0.696	0.220
SFA	165	0.740	0.151	0.950	0.669	0.176
XSECTION	195	0.874	0.151	0.990	0.670	0.186
TSERIES	2	0.009	0.790	0.820	0.805	0.021
PANEL	27	0.121	0.073	0.965	0.709	0.206
JOURNAL	154	0.691	0.073	0.938	0.684	0.183
CONF	16	0.072	0.170	0.950	0.627	0.235
WP	15	0.067	0.380	0.887	0.615	0.138
THESIS	38	0.170	0.151	0.990	0.688	0.206

Table 2. Likelihood ratio test for heteroscedasticity using probit model

Likelihood functions	Likelihood ratio	Critical value (10%)	Decision
−138.44796	0.015	2.706	Accept
−138.45550

Table 3. Hypothesis tests for selection of functional form of the unweighted fractional regression model

	Logit	Probit	Log-log	Cloglog	Cauchit
Misspecification test
RESET	0.763 (3)	0.951 (4)	0.425(2)	1.351(5)	0.004(1)
Generalised goodness-of-functional form test
GGOFF	2.531(4)	2.599(5)	0.532(1)	1.299(2)	2.229(3)
Rank sum	(7)	(9)	(3)	(7)	(4)
P test
HALogit	-	0.247	1.359	0.315	1.049
HAProbit	0.538	-	1.399	0.635	1.423
HALoglog	0.489	0.235	-	0.095	0.177
HACloglog	1.781	1.795	2.379	-	2.763
HACauchit	0.487	0.560	0.662	0.640	-

Numbers in brackets are ranks, 1 is highest (most preferred) and 5 lowest (least preferred).

Table 4. Egger’s test and selected unweighted cauchit fractional regression model

	Egger’s test	unweighted cauchit fractional regression estimates
MTE	Coefficients	Coefficients	Robust Std. Err.	dy/dx	Delta-method Std. Err.
CONSTANT	0.7031***	1.4348**	0.6951	-	-
SSIZE	−0.0004***	−0.0002	0.0002	−0.0000	0.0000
NONPERR	-	−0.3428	0.2121	−0.0769	0.0474
PERR	-	−0.0678	0.3325	−0.0152	0.0745
CROPS	-	−0.6671**	0.2882	−0.1496**	0.0646
ANIM	-	0.5351*	0.3232	0.1200	0.0747
FISH	-	−0.8087**	0.3575	−0.1813**	0.0791
FOODMANU	-	−0.2864	0.4757	−0.0642	0.1068
PHARMANU	-	−0.4756*	0.2747	−0.1067*	0.0606
TEXTMANU	-	0.0145	0.4432	0.0032	0.0994
WOOD	-	0.0466	0.1121	0.0105	0.0253
TRANSPT	-	−0.4595	0.3579	−0.1030	0.0795
INTERMED	-	0.2273	0.2713	0.0510	0.0607
PUBADM	-	−0.0539	0.4599	−0.0121	0.1031
EDUCPRIM	-	1.2795***	0.3515	0.2869***	0.0851
EDUCSUP	-	1.2019***	0.3471	0.2695***	0.0804
HOSP	-	0.3876	0.4236	0.0869	0.0949
COASTL	-	0.3270	0.2985	0.0733	0.0667
MID	-	0.4733*	0.2501	0.1061*	0.0546
NORTH	-	0.2495*	0.1315	0.0560*	0.0296
DYEAR	-	−0.0004	0.0003	−0.0001	0.0001
DEA	-	−0.1167	0.1869	−0.0262	0.0418
XSECTION	-	−0.0124	0.2659	−0.0028	0.0596
TSERIES	-	1.0401***	0.2968	0.2333***	0.0700
JOURNAL	-	0.1733	0.2098	0.0389	0.0470
CONF	-	0.0354	0.2850	0.0079	0.0639
WP	-	0.0346	0.2232	0.0078	0.0501
Model properties
N	223	223
Deviance	36.6806	33.1567		-	-
(1/df) Deviance	0.1660	0.1660		-	-
Pearson	35.4373	32.5672		-	-
(1/df) Pearson	0.1604	0.1604		-	-
AIC	0.9169	1.0805		-	-
BIC	−1158.3040	−1053.6850		-	-
Log pseudolikelihood	−100.2363	98.4743		-	-

***, **, * are 1%, 5% and 10% levels of significance, respectively. Std. Err.: standard error.