Appeals to the people: A content analysis of references to the people in traditional media, social media, and parliamentary materials

Figures & data

Figure 1. Percentage of items in news, parliamentary and social media data containing the relevant dependent variables.

Note. Total sample of items that refer to the people: N = 1668.

Figure 2. Percentage of content pitting the people against other actors.

Table 1. Parties’ mean score for left/right alignment and populism according to the 2017 Chapel Hill Expert Survey.

Party	CDA	PvdA	VVD	D66	GL	SGP	SP	CU	PVV	PvdD	50+	DENK	FvD
Left/right alignment	7	3.87	7.87	5.47	2.57	7.93	1.27	5	9.27	2.79	5.23	3.83	9.53
Populism	2.2	4.2	2.87	5.4	5.25	1.47	7.07	2.15	9.67	6.5	7.3	5.5	9.8

Table 2. Main effects of election year, populism, ideological positioning, and communication platform.

	Advocacy	Opposition
Election year	.85 (.21)	.57 (.27)
Populism score	1.14** (.03)	1.2** (.03)
Left/right positioning	.86** (.03)	.91** (.02)
Social media	11.74** (.34)	1.24** (.21)
Newspaper	.68** (.14)	.77 (.15)

Notes. Reported are odds ratios and standard error; **p < .01; N = 1235.

Table 3. Main effects of independent variables on positive references to the people across different communication forms (N = 1223).

	Advocacy
	News	Parliament	Social media
Election year	1.69a (.29)	.42*b (.34)	.79 (.98)
Populism score	.99a (.04)	1.25**b (.04)	2.13 (.48)
Left/right positioning	.93a (.04)	.82**b (.04)	1.38 (.34)

Note. Reported are odds ratios and standard error; *p < .05, **p < .01. Means with differing subscripts within rows differ significantly at the p < .05 level based on post-hoc coefficient comparison Chi-Squared tests.

Table 4. Main effects of independent variables on data containing opposition between the people and other actors across different communication forms (N = 463).

	Opposition
	News	Parliament	Social media
Election year	.77a (.34)	.36*a (.54)	3.72a (1.05)
Populism score	1.16**a (.05)	1.19*a (.04)	1.93**b (.17)
Left/right positioning	.95a (.05)	.95a (.03)	.66**b (.11)

Note. Reported are odds ratios and standard error; *p < .05, **p < .01. Means with differing subscripts within rows differ significantly at the p < .05 level based on post-hoc coefficient comparison Chi-Squared tests.

Figure 3. Percentage of content from left wing/right wing sources pitting the people against other actors.