Overestimation of the Level of Democracy Among Citizens in Nondemocracies

• Author: Eddy S. F. Yeung
• DOI: http://doi.org/10.1177/00104140221089647
• Published date: 01 February 2023
• Date: 01 February 2023
• Subject Matter: Articles

Article

Comparative Political Studies

2023, Vol. 56(2) 228–266

© The Author(s) 2022

Article reuse guidelines:

sagepub.com/journals-permissions

DOI: 10.1177/00104140221089647

journals.sagepub.com/home/cps

Overestimation of the

Level of Democracy

Among Citizens in

Nondemocracies

Eddy S. F. Yeung

Abstract

Overestimation of the level of democracy is prevalent among citizens in

nondemocracies. Despite such prevalence, no research to date has sys-

tematically documented this phenomenon and examined its determinants. Yet

given the renewed interest in the role of legitimacy in authoritarian survival,

studying whether and why this phenomenon arises is important to our

understanding of authoritarian resilience. I argue that, even in the absence of

democratic institutions in nondemocracies, autocrats exercise media control

in order to boost their democratic legitimacy. This façade of democracy, in

turn, benef‌its their survival. Combining media freedom data with individual

survey response data that include over 30,000 observations from 22 non-

democracies, I f‌ind that overestimation of the level of democracy is greater in

countries with stronger media control. But highly educated citizens over-

estimate less. These f‌indings shed light on media control as a strategy for

authoritarian survival, and have important implications for modernization

theory.

Keywords

non-democratic regimes, comparative public opinion, media control,

democratic legitimacy, authoritarian resilience

Emory University, Atlanta, GA, USA

Corresponding Author:

Eddy S. F. Yeung, Department of Political Science, Emory University, Tarbutton Hall, 1555 Dickey

Drive, Atlanta, GA 30322-1007, USA.

Email: shing.fung.yeung@emory.edu

“Authoritarian equilibrium rests mainly on lies, fear, or economic

prosperity.”

—Adam Przeworski (1991, p. 58)

Introduction

In recent years, many citizens in democracies have expressed substantial

dissatisfaction with the functioning of democracy in their country, as signaled

by the rising trend of antiestablishment political movements in established

democracies.

1

This contrasts with the curious observation that many citizens

in nondemocracies

2

have displayed considerable satisfaction with the au-

thoritarian governance in their country, as indicated by their widespread

acknowledgment of autocratic leaders (Frye et al., 2017,Guriev & Treisman,

2020b).

This suggests a possible misalignment between the perceived level of

democracy and the measured level of democracy. That is, people may un-

derestimate the level of democracy in democracies, but overestimate it in

nondemocracies.

3

Data from the World Values Survey (WVS) reveal this

interesting pattern (Figure 1). While underestimation of the level of de-

mocracy is prevalent in established democracies such as the U.S. and Japan,

overestimation is evident in nondemocracies such as China and Singapore.

4

Similar phenomena can also be observed in other countries. However, this is

unlikely due to how individuals in democracies and nondemocracies def‌ine

democracy differently, since their conceptions of democracy are, in fact, fairly

similar (Figures S1–S3). This leads to an unsolved puzzle: Why do citizens in

nondemocracies overestimate the level of democracy in their country?

While underestimation of the level of democracy in democracies is closely

linked to the abundant literature on democratic def‌icit,

5

surprisingly no re-

search to date has focused on the overestimation side. Yet, studying over-

estimation is also important: If we study democratic def‌icit due to its

implications for democratic survival, then we should also study why citizens

overestimate regime democraticness given its potential implications for au-

thoritarian resilience.

In nondemocracies, citizens’overestimation of the level of democracy

contributes to authoritarian resilience through regime legitimation. This is

premised on congruence theory in the democratization literature: Autocracies

gain legitimacy when their citizens perceive them to be democratic; such le-

gitimacy, in turn, empowers autocrats to not “supply”democracy as citi-

zens believe that their “demand”for democracy is already satisf‌ied (Inglehart &

Wel ze l, 2 00 5;Qi & Shin, 2011;Rose et al., 1998). Consequently, auto-

crats are incentivized to create a façade of democracy to stabilize their re-

gime (Gerschewski, 2013;2018;Tannenberg et al., 2021). But how is a

nondemocracy able to make this façade of democracy so credible that its

Yeung 229

citizens are willing to believe that their country is already democratic when it

is not?

I argue that media control is an important strategy.

6

By controlling the

media, autocrats can selectively disseminate information that could mislead

citizens into believing that the regime is already democratic. This could

include, for example, the occasional instances where popular policy demands

are aptly responded to by the government (Truex, 2017), as well as the “rule of

law”development within the state (Stockmann & Gallagher, 2011;Whiting,

2017). Crucially, media censorship can cover up common autocratic practices

of the government, including violent repression and electoral fraud. Unable to

obtain information about such regime-revealing practices, ordinary citizens in

nondemocracies may in turn overestimate the democraticness of their country.

Figure 1. Perceived and Measured Levels of Democracy—All Sampled Countries in

the World Values Survey (Wave 6: 2010–16).

Note: Variables are rescaled to range from 0 to 1. Data on perceived levels of democracy are

obtained from the WVS (Wave 6: 2010–16). They are country-average data after dropping

nonrespondents, based on the variable V141 in the WVS. Data on measured levels of democracy

are obtained from V-Dem’s Electoral Democracy Index (corresponding years). Details on

these variables, as well as a discussion of their comparability, are provided in Section 4.

230 Comparative Political Studies 56(2)