Immigration as a thermostat? Public opinion and immigration policy across Western Europe (1980–2017)

Figures & data

Table 1. Source information, by country.

	Start date	End date	Time points	Item series (questions)		Survey items (administrations)		Explained variance
				N	%	N	%
Austria	1990	2016	27	24	4.63	83	123.37	69.99
Belgium	1981	2017	37	27	5.21	111	4.51	76.08
Denmark	1981	2015	35	49	9.46	226	9.18	56.40
France	1981	2017	37	68	13.13	295	11.99	59.79
Germany	1980	2017	38	82	15.83	354	14.38	69.59
Great Britain	1981	2017	37	64	12.36	280	11.38	51.41
Ireland	1981	2017	37	28	5.41	116	4.71	65.38
Italy	1981	2016	36	11	2.12	45	1.83	90.02
Netherlands	1981	2017	37	49	6.46	354	14.38	43.40
Norway	1982	2017	36	34	6.56	150	6.10	78.88
Portugal	1989	2017	29	27	5.21	100	4.06	80.30
Sweden	1980	2017	38	35	6.76	176	7.15	48.63
Switzerland	1989	2017	29	20	3.86	171	6.95	53.48
Total			453	518		2.461

Figure 1. Restrictive immigration opinions and asylum rates, by country (1980-2017).

Note: Immigration opinions have been standardised within each country. Higher values indicate more restrictive opinions.

Table 2. Descriptive statistics across years, per country.

	Restrictive immigration opinions			Asylum rates
	Years	Mean	Std. Dev.	Years	Mean	Std. Dev.
Austria	27	52.14	7.41	37	2.31	1.76
Belgium	37	52.98	4.84	37	1.45	0.92
Denmark	35	49.85	4.25	37	1.39	1.05
France	37	48.58	3.70	37	0.63	0.24
Germany	37	50.67	6.89	37	1.58	1.77
Great Britain	37	51.40	4.76	37	0.51	0.40
Ireland	37	45.39	7.35	26	0.91	0.90
Italy	36	45.34	5.76	37	0.29	0.41
Netherlands	38	47.77	4.15	37	1.19	0.88
Norway	36	53.01	7.27	37	1.63	1.32
Portugal	29	48.23	6.33	37	0.05	0.05
Sweden	38	42.54	6.33	36	3.21	2.86
Switzerland	29	47.18	8.09	37	2.65	1.47
Total	453	48.80	6.72	580	1.21	1.50

Table 3. Summarised factor results of immigration opinions.

	Factor loading	Uniqueness
Austria	0.08	0.26
Belgium	0.94	0.04
Denmark	0.88	0.03
France	0.93	0.03
Germany	0.60	0.07
Great Britain	−0.92	0.04
Ireland	−0.86	0.09
Italy	−0.80	0.16
Netherlands	0.66	0.10
Norway	0.54	0.12
Portugal	0.61	0.30
Sweden	0.35	0.22
Switzerland	0.65	0.09
EV	6.75

Note: We refer to Section B in the supplementary materials for the full pattern matrix of the factor analysis.

Table 4. Public responsiveness for immigration.

	Δ Restrictive immigration opinionst
	(1)	(2)	(3)	(4)	(5)	(6)
Δ Inflow asylum seekerst-1	0.014	0.015	0.016	0.016	0.027	0.014
	(0.020)	(0.020)	(0.020)	(0.020)	(0.018)	(0.020)
Δ Inflow foreignerst-1	0.011	0.024^	0.030*	0.031*	$0.051*	0.065*
	(0.014)	(0.014)	(0.014)	(0.014)	(0.017)	(0.020)
Δ GDP per capitat-1	--	−0.083*	--	−0.025	−0.049	−0.080*
	--	(0.024)	--	(0.026)	(0.030)	(0.038)
Δ Unemploymentt-1	--	--	0.130*	0.116*	0.110*	0.069^
	--	--	(0.026)	(0.030)	(0.037)	(0.039)
Constant	−0.039	0.001	−0.044^	−0.032	−0.036	−0.054
	(0.024)	(0.029)	(0.025)	(0.029)	(0.025)	(0.035)
Observations	373	373	370	370	264	284
R-squared	0.003	0.024	0.051	0.052	0.068	0.090
Rho (average)	0.081	0.078	0.059	0.060	−0.125	0.020
Countries	13	13	13	13	13	11

Note: Standard errors in parentheses; * p < 0.05, ^ p < 0.1; We refer to section E in the supplementary materials for alternative models. All results remain substantively similar, also when including a dummy for the migration crisis.

Table 5. Policy responsiveness for immigration.

	Δ Immigration policyt
	(1)	(2)	(3)	(4)	(5)
Restrictive immigration opinionst-1	−0.214*	−0.232*	−0.234*	−0.266	−0.349*
	(0.098)	(0.103)	(0.091)	(0.208)	(0.130)
Left-wing governmentt-1	--	0.055	0.067	0.013	0.070
	--	(0.059)	(0.049)	(0.097)	(0.066)
GDP per capitat-1	--	--	0.007	0.026	−0.004
	--	--	(0.013)	(0.024)	(0.012)
Unemploymentt-1	--	--	0.025	0.085	−0.019
	--	--	(0.036)	(0.078)	(0.045)
Asylum crisist-1	−1.967*	−1.961*	−2.004*	−2.217^	−2.065*
	(0.534)	(0.538)	(0.533)	(1.250)	(0.513)
Constant	0.143	−0.006	−0.0435	−1.512	0.245
	(0.130)	(0.176)	(0.654)	(1.348)	(0.728)
Observations	420	419	406	269	282
R-squared	0.117	0.123	0.123	0.134	0.150
Rho (average)	0.152	0.157	0.138	0.255	0.169
Countries	13	13	13	13	11

Note: Standard errors in parentheses; * p < 0.05, ^ p < 0.1; We refer to section F in the supplementary materials for alternative models. All results remain substantively similar.