Table 5 Long-run and short-run effects, with asymmetric responses, of the population on per capita GDP.

From: Declining population and GDP growth

Variables

Coefficient

Standard error

p-value

Long-run coefficients

\(POP_{i,t}^ +\)

0.0177**

0.0054

0.001

\(POP_{i,t}^ -\)

0.0085*

0.0039

0.028

Short-run coefficients

ΔGDPt-1

−0.2475

0.1888

0.190

ΔGDPt-2

1.3506**

0.0494

< 0.001

ΔGDPt-3

−0.8493**

0.0544

< 0.001

ΔGDPt-4

0.2018**

0.0188

< 0.001

\({\Delta}POP_{i,t}^ +\)

0.0133**

0.0035

< 0.001

\({\Delta}POP_{i,t - 1}^ +\)

−0.0443*

0.0174

0.011

\({\Delta}POP_{i,t - 2}^ +\)

0.0447

0.0277

0.107

\({\Delta}POP_{i,t - 3}^ +\)

−0.0107

0.0221

0.628

\({\Delta}POP_{i,t - 4}^ +\)

−0.0134

0.0125

0.285

\({\Delta}POP_{i,t - 5}^ +\)

0.0058

0.0036

0.103

\({\Delta}POP_{i,t}^ -\)

0.000024

0.0023

0.992

\({\Delta}POP_{i,t - 1}^ -\)

−0.1517

0.1694

0.370

\({\Delta}POP_{i,t - 2}^ -\)

0.0390

0.0552

0.479

\({\Delta}POP_{i,t - 3}^ -\)

−0.0617

0.0755

0.414

\({\Delta}POP_{i,t - 4}^ -\)

0.0397

0.0491

0.418

\({\Delta}POP_{i,t - 5}^ -\)

−0.0139

0.0161

0.387

Intercept

−74336.44**

28874.13

0.010

Joint Hausman test

1.35

0.508

Error correction coefficient (φ)

−0.2934**

0.0301

< 0.0001

Dynamic Specification: ARDL (4, 5, 5)

Model selection method: Akaike information criterion (AIC)

Maximum likelihood method, dependent variable lags: 4 (automatic selection), Dynamic regressors (5 lags, automatic selection)

Estimator

Pooled Mean Group (PMG) controlling for country-fixed effects

No. of countries

19

  

Period

1990–2021 (32 time periods)

No. of observations

608

  
  1. Source: Authors’ estimates.
  2. ** and * denote statistical significance at the 1% and 5% level, respectively.