Public debate in the media matters: evidence from the European refugee crisis

EPJ Data Science

Table 3 Granger-causality test results. Granger-causality test of interactions between asylum acceptance rates and public debate sentiment in the media. The 10 asylum seekers’ countries of origin with the most decisions made (both accepted and rejected) are included in the regressions. We report Granger-causal testing for 1-month (1-lag) and 2-month (2-lag) lagged VAR models. Locations with NA result from monthly data not being available from the EU database. The \(D_{2}\) statistic is computed as in [71] and is the result of combining Granger tests from multiply imputed datasets

Country	p-value corresponds to Wald test
	Public debate sentiment → Acc. rate				Acc. rate → Public debate sentiment
	\(D_{2}\)statistic		p-value		\(D_{2}\) statistic		p-value
	(1-lag)	(2-lag)	(1-lag)	(2-lag)	(1-lag)	(2-lag)	(1-lag)	(2-lag)
Austria	12.10	8.35	<0.001^∗∗∗	<0.001^∗∗∗	9.72	1.66	<0.001^∗∗∗	0.20
Belgium	26.77	10.45	<0.001^∗∗∗	<0.001^∗∗∗	2.49	1.66	0.12	0.20
Bulgaria	12.47	16.12	<0.001^∗∗∗	<0.001^∗∗∗	4.08	0.55	0.04^∗	0.48
Czech Rep.	12.81	11.68	<0.001^∗∗∗	<0.001^∗∗∗	2.73	0.49	0.10	0.49
Denmark	13.92	11.41	<0.001^∗∗∗	<0.001^∗∗∗	4.71	1.10	0.03^∗	0.31
Finland	10.03	9.22	<0.001^∗∗∗	<0.001^∗∗∗	5.47	1.25	0.02^∗	0.26
France	16.44	6.03	<0.001^∗∗∗	0.01^∗	16.85	5.70	<0.001^∗∗∗	0.02^∗
Germany	9.04	3.31	<0.001^∗∗∗	0.07	8.75	3.82	<0.001^∗∗∗	0.05
Greece	9.92	7.14	<0.001^∗∗∗	0.01^∗	0.10	3.61	0.75	0.06
Hungary	5.77	4.44	0.02^∗	0.05	6.92	6.72	0.01^∗	0.01^∗
Italy	0.40	3.83	0.55	0.08	0.04	2.22	0.85	0.14
Netherlands	5.63	3.56	0.02^∗	0.06	4.76	2.14	0.03^∗	0.14
Norway	30.80	11.82	<0.001^∗∗∗	<0.001^∗∗∗	3.77	1.73	0.05	0.19
Poland	8.18	3.88	<0.001^∗∗∗	0.06	1.70	1.13	0.19	0.29
Portugal	0.42	1.02	0.52	0.34	NA	0.69	NA	0.41
Slovakia	8.81	0.45	0.42	0.50	1.79	0.60	0.18	0.44
Spain	6.45	8.39	0.01	<0.001^∗∗∗	4.32	3.54	0.04^∗	0.06
Sweden	9.25	11.44	<0.001^∗∗∗	<0.001^∗∗∗	5.86	2.45	0.02^∗	0.12
UK	21.53	12.91	<0.001^∗∗∗	<0.001^∗∗∗	29.04	9.72	<0.001^∗∗∗	<0.001^∗∗∗