Unemployment and the European Union, 2000–2017: structural exploration of distant past economic experience and future prosperity

Journal of Economic Structures

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

Table 4 Granger-causality test results

Panel A (dependent: unemployment)
Inflation	Real GDP	Trade	Capital	Labor
22.230 [0.023]**	38.164 [0.000]***	34.603 [0.000]***	17.564 [0.092]*	21.845 [0.026]**
Panel B (dependent: inflation)
Unemployment	Real GDP	Trade	Capital	Labor
15.842 [0.147]	19.277 [0.056]*	10.311 [0.503]	13.136 [0.285]	22.593 [0.020]**
Panel C (dependent: real GDP)
Unemployment	Inflation	Trade	Capital	Labor
24.276 [0.012]**	24.342 [0.011]**	60.274 [0.000]***	12.714 [0.312]	21.058 [0.033]**
Panel D (dependent: trade)
Unemployment	Inflation	Real GDP	Capital	Labor
5.248 [0.919]	14.485 [0.207]	26.484 [0.006]***	28.153 [0.003]***	16.802 [0.114]
Panel E (dependent: capital)
Unemployment	Inflation	Real GDP	Trade	Labor
18.099 [0.079]*	39.735 [0.000]***	92.607 [0.000]***	14.027 [0.232]	21.486 [0.029]**
Panel F (dependent: labor)
Unemployment	Inflation	Real GDP	Trade	Capital
11.796 [0.379]	9.428 [0.583]	9.635 [0.563]	10.925 [0.450]	8.329 [0.684]

The unit of measurement per variable is according to Table 1. The VAR(11) model is estimated with cross-sectional fixed effects (two-way fixed effects are not statistically feasible) and trend as exogenous variables. The lag-length specification of 11 years is based on this paper’s theoretical proposition. AIC estimated from VAR(2) indicates 11 years of lag length to be needed, although SIC indicates 2 years. Autocorrelation LM test rejects serially correlated residuals from the fifth lags. ***, **, and * denote significance at the 1%, 5%, and 10% levels, respectively. Values in square brackets are probability values. All results are calculated using EViews (version 8) VAR and Granger causality