Skip to main content

Table 8 OLS versus SDM estimation results

From: Spatial interaction and economic growth: a case of OECD countries

Variables

OLS

SDM

Direct

Indirect

Total

ln(\({s}_{k}\))

0.184*

(11.81)

0.354***

(15.24)

0.356***

(14.976)

0.041

(1.050)

0.396***

(8.216)

ln(\({s}_{h}\))

2.958***

(22.865)

0.941***

(14.595)

0.955***

(16.074)

0.361***

(3.681)

1.316***

(15.394)

ln(\({s}_{\tau }\))

− 0.004

(− 0.456)

0.058***

(5.141)

0.058***

(5.270)

− 0.013

(− 0.982)

0.044***

(2.835)

ln(\(n+\delta\))

0.936***

(6.853)

− 0.125***

(− 3.665)

− 0.119***

(− 3.451)

0.116*

(1.775)

− 0.003

(− 0.039)

Constant

3.008***

(15.674)

    

W*dep.var

 

0.147***

(2.661)

   

R- square

0.609

0.781

   

log-likelihood

 

1083.259

   
  1. Dependent variable ln(Ypc): Natural log of real GDP divided by working-age population (15–64). The OLS model expresses the results of traditional economic analysis in the second column. The SDM results start in the third column. The direct, indirect, and total effects are given in the following columns. t-stats are given in parenthesis. *, ** and *** indicate significance at the 0.10, 0.05, and 0.01 levels, respectively