Power Sharing and Media Freedom in Dictatorships

Figures & data

Figure 1. Media freedom in China, Russia, and Myanmar, 2004–2017.

Note: The index of media freedom is taken from Solis and Waggoner (2020).

Figure 2. Power sharing and media freedom, 1960–2010.

Note: The index of media freedom is taken from Solis and Waggoner (2020). The index of power-sharing is taken from Geddes et al. (2018).

Table 1. Power sharing and media freedom in dictatorships, 1960–2010

	Model 1	Model 2	Model 3	Model 4	Model 5	Model 6
Power Sharing	0.132**	0.058+	0.058+	0.058**	0.013*	0.005
	[0.039]	[0.030]	[0.032]	[0.019]	[0.006]	[0.027]
GDPpc		−1.909	−3.156	−3.156**	−3.279***	−1.512
		[1.758]	[1.988]	[1.148]	[0.854]	[1.059]
Oil and Gas		0.186	0.326*	0.326***	0.009	0.270**
		[0.157]	[0.148]	[0.090]	[0.041]	[0.095]
Population		16.666***	−3.119	−3.119	3.641***	10.053***
		[2.883]	[6.901]	[3.698]	[0.491]	[1.878]
Internal Conflicts		−0.498	−0.549+	−0.549*	−0.032	−0.306
		[0.349]	[0.307]	[0.217]	[0.091]	[0.227]
Polity Score		1.944***	1.710***	1.710***	0.689***	1.823***
		[0.149]	[0.144]	[0.102]	[0.032]	[0.108]
Repression		−0.944	−0.629	−0.629	−1.758***	−0.781
		[1.031]	[1.070]	[0.641]	[0.393]	[0.649]
Cold War						−10.134***
						[1.901]
Power Sharing X						0.067*
Cold War						[0.028]
Constant	18.119***	−218.821***	115.722		9.748***
	[2.645]	[40.130]	[119.152]		[0.337]
Country Dummy	No	Yes	Yes	Yes	Yes	Yes
Year Dummy	No	No	Yes	No	Yes	No
R-squared	0.034	0.584	0.638	0.363	0.172	0.604
No. of Countries	114	98	98	98	98	98
No. Observations	4354	2924	2924	2924	2826	2924

Note: The dependent variable is the index of media freedom constructed by Solis and Waggoner (2020). Country-level clustered standard errors are reported in brackets in Models 1 to 3. Models 4 and 6 estimate heteroskedasicity- and autocorrelation-corrected (HAC) standard errors. Models 5 estimates a fixed-effects model with an AR(1) disturbance. + p < 0.1, * p < 0.05, ** p < 0.01, *** p < 0.001.

Table 2. Robustness checks I

	Model 1	Model 2	Model 3	Model 4	Model 5	Model 6
Power Sharing	0.058*	0.001*	0.006+	0.008	1.181*	0.071***
	[0.023]	[0.000]	[0.003]	[0.006]	[0.578]	[0.020]
GDPpc	−1.023	−0.011	−0.523**	1.527***	−2.493*	−6.433***
	[2.220]	[0.014]	[0.178]	[0.407]	[1.234]	[1.158]
Oil and Gas	0.429***	0.001	0.016*	0.002	0.255**	0.460***
	[0.100]	[0.001]	[0.007]	[0.023]	[0.096]	[0.098]
Population	−2.002	0.005	−0.211	1.415	−3.061	−5.783
	[5.650]	[0.038]	[0.318]	[1.598]	[3.672]	[4.304]
Internal Conflicts	−0.643+	−0.007	0.003	−0.304***	−0.335	−0.733**
	[0.332]	[0.003]	[0.022]	[0.090]	[0.236]	[0.251]
Polity Score	0.783***	0.026***	0.022	0.338***	1.742***
	[0.138]	[0.001]	[0.014]	[0.029]	[0.117]
Repression	−4.665***	−0.019*	−0.178**	0.147	−0.467	−0.489
	[0.937]	[0.008]	[0.068]	[0.194]	[0.643]	[0.702]
Executive Constraint						4.179***
						[0.358]
Country Dummy	Yes	Yes	Yes	Yes	Yes	Yes
Year Dummy	Yes	Yes	Yes	Yes	Yes	Yes
R-squared	0.207	0.385	0.090		0.302	0.263
Log likelihood				−389.449
No. of Countries	78	99	76	53	96	98
No. Observations	1069	2966	981	1617	2645	2823

Note: The dependent variable (DV) in Model 1 is Freedom House’s measure of press freedom (1992–2010). The DVs in Models 2 and 3 are the V-Dem project’s indices of freedom of expression (1960–2010) and government effort on censoring the Internet (1992–2010), respectively. The DV in Model 4 is media environment (1960–2010) constructed by Whitten-Woodring and Van Belle (2017). The DV in Models 5 and 6 is the index of media freedom (1960–2010) constructed by Solis and Waggoner (2020). The data of Power Sharing in Model 5 are taken from the Gandhi and Sumner (2020). Standard errors are reported in brackets. Country-level clustered standard errors are estimated in Model 5. Other models estimate heteroskedasicity- and autocorrelation-corrected (HAC) standard errors. p < 0.1, * p < 0.05, ** p < 0.01, *** p < 0.001.

Table 3. Robustness checks II

	Model 1	Model 2	Model 3	Model 4	Model 5	Model 6
Power Sharing	0.040**	0.043+	0.111***	0.055**	0.042+	0.059**
	[0.015]	[0.022]	[0.023]	[0.019]	[0.024]	[0.019]
GDPpc	−5.812***	−3.175*	−7.143***	−3.267**	−8.736***	−3.704**
	[0.953]	[1.422]	[1.279]	[1.220]	[1.770]	[1.182]
Population	−11.568**	−2.123	−7.358	2.821	−12.494*	−6.497+
	[3.796]	[5.553]	[4.993]	[3.764]	[5.395]	[3.575]
Internal Conflicts	−0.570**	−0.804**	−0.487+	−0.449*	−0.414	−0.477*
	[0.194]	[0.282]	[0.249]	[0.222]	[0.300]	[0.223]
Oil and Gas	0.371***	0.437***	0.477***	0.299**	0.295**	0.348***
	[0.069]	[0.095]	[0.101]	[0.092]	[0.111]	[0.089]
Repression	1.486*	−2.713***	−0.748	−0.631	−0.650	−0.892
	[0.602]	[0.766]	[0.782]	[0.628]	[0.828]	[0.657]
Liberal	71.838***
Component	[3.385]
Multiparty		7.812***
Elections		[0.572]
Military			4.716*
			[1.961]
Monarchy			13.668***
			[3.170]
Party			1.080
			[1.710]
Polity Score				1.647***	1.602***	1.631***
				[0.100]	[0.131]	[0.101]
Foreign Aid				1.083**
				[0.358]
FDI					−0.036
					[0.064]
State Capacity						0.400
						[3.845]
Country Dummy	Yes	Yes	Yes	Yes	Yes	Yes
Year Dummy	Yes	Yes	Yes	Yes	Yes	Yes
R-squared	0.452	0.450	0.123	0.369	0.346	0.362
No. of Countries	99	95	99	95	91	95
No. Observations	2924	656	2926	2765	1582	2835

Note: The dependent variable is the index of media freedom constructed by Solis and Waggoner (2020). Heteroskedasicity- and autocorrelation-corrected (HAC) standard errors are reported in brackets. + p < 0.1, * p < 0.05, ** p < 0.01, *** p < 0.001.

Solis, J. A., & Waggoner, P. D. (2020). Measuring media freedom: An item response theory analysis of existing indicators. British Journal of Political Science, 51(4), 1685-1704.https://doi.org/https://doi.org/10.1017/S0007123420000101

 Google Scholar

Geddes, B., Wright, J., & Frantz, E. (2018). How dictatorships work. Cambridge University Press. https://doi.org/https://doi.org/10.1017/9781316336182

 Google Scholar

Whitten-Woodring, J., & Van Belle, D. A. (2017). The correlates of media freedom: An introduction of the global media freedom dataset. Political Science Research and Methods, 5(1), 179–188. https://doi.org/https://doi.org/10.1017/psrm.2015.68

 Web of Science ®Google Scholar

Gandhi, J., & Sumner, J. L. (2020). Measuring the consolidation of power in nondemocracies. Journal of Politics, 82(4), 1545–1558. https://doi.org/https://doi.org/10.1086/708683

 Web of Science ®Google Scholar

Data Availability Statement

The data described in this article are openly available in the Open Science Framework at https://doi.org/10.7910/DVN/LQMGBG.