Skip to main content
Journal of Economic Structures

The Official Journal of the Pan-Pacific Association of Input-Output Studies (PAPAIOS)

Table 4 Estimates of elasticity of substitution (Hokkaido)

From: Networks in Japanese regional agro-food economies: an empirical exploration of the network linkage model

id

LS FE

IV FE

\(\theta\)

\(\theta -1\) s.e

\(\theta\)

\(\theta -1\) s.e

1st F

Overidentification

Endogeneity

hok

1

0.903

0.075

0.032

0.419

5

1.8 (0.408)

3 (0.077)

 

2

1.159

0.074

1.709

0.197

200

2.8 (0.244)

15 (0.000)

 

3

1.161

0.101

2.587

0.668

2

5.4 (0.069)

13 (0.000)

 

4

0.908

0.021

0.878

0.025

589

38.1 (0.000)

4 (0.037)

 

5

1.044

0.036

0.495

0.214

18

7.2 (0.007)

10 (0.001)

 

6

1.039

0.024

1.110

0.067

142

2.4 (0.305)

2 (0.205)

 

7

0.990

0.016

1.132

0.054

325

3.6 (0.060)

8 (0.004)

 

8

0.996

0.038

1.398

0.190

35

1.7 (0.187)

8 (0.005)

 

9

1.030

0.023

1.256

0.092

29

6.2 (0.044)

8 (0.006)

 

10

1.015

0.021

1.057

0.032

179

0.9 (0.646)

3 (0.076)

 

11

1.119

0.058

1.578

0.203

90

8.8 (0.012)

6 (0.013)

 

12

1.102

0.033

1.118

0.045

1176

0.8 (0.678)

0 (0.550)

 

13

0.909

0.026

1.122

0.075

21

1.0 (0.593)

8 (0.006)

 

14

0.954

0.019

0.919

0.018

229

3.7 (0.157)

4 (0.047)

 

15

1.373

0.028

3.339

0.111

185

42.9 (0.000)

1514 (0.000)

 

16

0.943

0.026

1.543

0.484

1

18.3 (0.000)

3 (0.074)

 

17

1.058

0.060

0.779

0.309

4

2.8 (0.245)

0 (0.546)

 

18

1.055

0.034

0.842

0.165

3

2.9 (0.089)

1 (0.227)

 

19

0.945

0.052

0.215

0.203

21

32.1 (0.000)

9 (0.003)

 

20

1.025

0.034

1.311

0.258

9

2.3 (0.322)

2 (0.185)

 

21

1.031

0.037

1.573

0.471

3

0.1 (0.741)

2 (0.144)

  1. (1) For sector classification (id), see Tables 2 and 3. Final demand sector (id = 0) is omitted
  2. (2) All variables are taken first difference of log-transformed values in the period 1970–2015 (9 observations every 5 years after 1975 for Okinawa, 10 observations every 5 years after 1970 for the rest)
  3. (3) LS FE is a least square fixed effect estimator with panels of input sectors. Standard error is cluster-robust assuming heteroskedasticity over input sectors
  4. (4) FE IV is a fixed effect estimator with the instrumental variables: macro-TFP in output region, CGPI of output sector and input sectors. Standard error is cluster-robust assuming heteroskedasticity over input sectors. 1st F is the result of F test under null hypothesis that coefficients of instrumental variables equal zero (explanatory variables are weakly correlated with instruments). Overidentification is the result of Sargan–Hansen test under null hypothesis that all instruments are correlated with residuals. Endogeneity is the result of Davidson–MacKinnon’s J test under null hypothesis that an ordinary least squares estimator of the same equation (i.e., LS FE) yields consistent estimates
Back to article page