Skip to main content

Democracy Erodes from the Top: Public Opinion and Democratic “Backsliding” in Europe

  • Chapter
  • First Online:
Democracy in Times of Crises

Abstract

Observers have described a “crisis of democracy” in contemporary Europe, with “exploding” popular support for right-wing populist parties fueled by anti-immigrant sentiment, political distrust, and dissatisfaction with democracy itself. However, opinion surveys provide remarkably little evidence of such a “crisis” in public opinion. Across Europe, attitudes regarding immigration, European integration, political trust, and satisfaction with democracy have remained largely unchanged over the past two decades. In the two European countries where democracy has eroded significantly, Hungary and Poland, “backsliding” has been engineered by politicians and imposed from above, not in response to any mandate from voters. Public acquiescence in these developments seems more plausibly attributable to substantial improvements in subjective well-being under illiberal governments than to any popular hankering for authoritarianism.

This is a preview of subscription content, log in via an institution to check access.

Access this chapter

Chapter
USD 29.95
Price excludes VAT (USA)
  • Available as PDF
  • Read on any device
  • Instant download
  • Own it forever
eBook
USD 99.00
Price excludes VAT (USA)
  • Available as EPUB and PDF
  • Read on any device
  • Instant download
  • Own it forever
Softcover Book
USD 129.99
Price excludes VAT (USA)
  • Compact, lightweight edition
  • Dispatched in 3 to 5 business days
  • Free shipping worldwide - see info
Hardcover Book
USD 129.99
Price excludes VAT (USA)
  • Durable hardcover edition
  • Dispatched in 3 to 5 business days
  • Free shipping worldwide - see info

Tax calculation will be finalised at checkout

Purchases are for personal use only

Institutional subscriptions

Notes

  1. 1.

    On the “folk theory” of democracy, see Achen and Bartels (2016).

  2. 2.

    See Norris and Inglehart (2019, p. 9). By this tabulation, the average vote share for “populist” parties in 32 Western democracies increased from 10.9% in the 1980s and 9.9% in the 1990s to 11.4% in the 2000s and 12.4% in the 2010s.

  3. 3.

    Of the eight countries with the highest levels of right-wing populist sentiment in 2014–2017, only two (Hungary and France) had right-wing populist parties attracting as much as 10% of the vote. On the other hand, right-wing populist parties flourished in Switzerland, Denmark, and Norway, all of which were among the half-dozen European countries with the lowest levels of right-wing populist sentiment in 2014–2017. As Rovira Kaltwasser (2012, p. 188) observed, “populist radical right parties have shown a great success precisely in those regions of Europe where the structural prerequisites for their rise were hardly existent.”

  4. 4.

    The description of Hungary and Poland as “mildly authoritarian regimes” is borrowed from Levitsky and Ziblatt (2018, p. 188).

  5. 5.

    On the relationship between these attitudes and support for right-wing populist parties, see Bartels (2017).

  6. 6.

    The 354,829 respondents represent 183 country rounds; the country round samples range in size from 985 to 3,142 and average 1,939. Surveys were not conducted in the remaining 24 country rounds (11.6%). My characterizations of European opinion are based on weighting each country round in proportion to its adult population. My substantive conclusions remain essentially unchanged when each country round is weighted equally.

  7. 7.

    Data and documentation are available from the ESS website (http://www.europeansocialsurvey.org/). My analysis generally includes EU countries as of 2006 and those in the Schengen area. It excludes countries admitted to the EU after 2006 (Bulgaria, Croatia, and Romania), some small countries with little or no ESS data (Cyprus, Iceland, Latvia, Luxembourg, and Malta), and several other countries represented sporadically in the ESS dataset (Albania, Israel, Kosovo, Montenegro, Russia, Serbia, Turkey, and Ukraine).

  8. 8.

    Because the set of countries represented in each ESS round varies, I track European averages using a statistical analysis including country fixed effects for the whole period. The countries accounting for most of the missing observations—Italy, Greece, and Lithuania—are in some respects unrepresentative, though only Italy is populous enough to have much impact on the European averages. More elaborate statistical procedures designed to account for sample selection produce little evidence of bias due to correlations between countries’ opinion climates and their participation in specific ESS waves.

  9. 9.

    Three questions asked how willing the respondents’ country should be to allow immigrants (1) “of the same race or ethnic group as most [country]’s people,” (2) “of a different race or ethnic group,” and (3) “from the poorer countries outside Europe.” Responses to these three questions were recoded to range from zero (for “allow none”) to ten (for “allow many”). The other three questions asked (4) whether immigration is good or bad for the country’s economy, (5) whether the country’s cultural life “is generally undermined or enriched” by immigration, and (6) whether immigration makes the country “a worse or a better place to live.” The correlations between responses to the six questions range from .46 to .80, and their loadings on a common factor range from.72 to.85. I included respondents who answered at least five of the six questions, imputing neutral values for the sixth when necessary.

  10. 10.

    The estimated increase is.42 (with a standard error of.17). The corresponding estimated increases for the six separate items range from.05 (for enriching cultural life) to.63 (for allowing more immigrants of the same race or ethnic group). All of my longitudinal cross-national analyses allow for disturbances in individual survey responses to be correlated within country-waves.

  11. 11.

    Dionne (2017), Nolan (2015), Scheppele (2015), Drozdiak (2017), pp. 3–4; Eddy (2017). Data on immigration and asylum-seeking are recorded in the OECD’s International Migration Database (http://stats.oecd.org/).

  12. 12.

    A separate item tapped trust in the European Parliament on a zero-to-ten scale. The average level of trust in 2018–2019, 4.4, was only slightly lower than the average level in 2002–2007 (4.5), and only slightly lower than the average level of trust in the respondents’ national parliaments (4.6).

  13. 13.

    The estimated increase is.18 (with a standard error of.15).

  14. 14.

    From 2002 to 2019, the average annual growth rate of real GDP per capita was 4.0% in Poland and, despite a sustained downturn in the wake of the Great Recession, 3.1% in Ireland. The corresponding growth rate for the 19-country Euro area as a whole was 0.8%.

  15. 15.

    The estimated increase in support for further integration in this cohort from 2004–2005 to 2018–2019 is.14 (with a standard error of.13).

  16. 16.

    Another item tapping trust in political parties did not appear in the first ESS round, but has produced generally similar responses in subsequent rounds. Additional questions regarding trust in the legal system and the police have produced higher ratings and are less strongly correlated with the more specifically political trust ratings.

  17. 17.

    The cumulative decline amounts to just.05 points (with a standard error of.16 points) on the ten-point scale.

  18. 18.

    Denmark exhibited an even higher but generally declining level of trust.

  19. 19.

    Although some of these countries are missing from as many as four or five ESS rounds, the statistical analysis generating the estimated average levels of political trust for Europe as a whole takes them into account by including country fixed effects for the entire period. Tests for selection bias based on relating survey participation to population, economic conditions, protest activity, and immigration rates reveal little evidence of correlation between non-participation and opinion climates within countries over time.

  20. 20.

    Kirk A. Hawkins et al., The Global Populism Database (https://populism.byu.edu/Pages/Data)

  21. 21.

    “We Lied to Win, Says Hungary PM,” BBC News, 18 September 2006

  22. 22.

    See Schöpflin (2007). Of course, by this standard almost every parliamentary democracy is “a semi-democratic system.”

  23. 23.

    Office for Democratic Institutions and Human Rights, Limited Election Observation Mission Final Report, 11 July 2014 (https://www.upr-info.org/sites/default/files/document/hungary/session_25_-_may_2016/osce-odihr_upr25_hun_e_annexe1.pdf)

  24. 24.

    My measure of conservative worldviews is based on responses to ten ESS items tapping the importance of security, tradition, creativity, diversity, and the like. It is only modestly correlated with left-right ideology (R=.15). The scale runs from zero (for people who ascribed maximal importance to liberal values and minimal importance to conservative values) to ten (for those who ascribed maximal importance to conservative values and minimal importance to liberal values). I included respondents who answered at least nine of the ten questions, imputing neutral values for the tenth when necessary. On the political relevance of these attitudes, see Stenner (2005) and Hetherington and Weiler (2018).

  25. 25.

    According to Albertazzi and Mueller (2013, pp. 358–361), the League of Polish Families’ youth wing “was staffed by large numbers of skinheads, quite open about their Nazi sympathies and responsible for attacks against gay and feminist groups, members of ethnic minorities and others,” while Self-Defence had been organized in the early 1990s “as a militia aimed at defending farmers from debt collectors and it had not been a stranger to violence in the past.”

  26. 26.

    Weighting all countries and ESS rounds by population, 4.5% of the respondents said they were “very close” to some party, 27.3% “quite close,” 14.8% “not close” or unspecified, and 1.3% “not at all close”; the remaining 52.1% said they did not feel closer to any particular party.

  27. 27.

    Analyzing the success of populist parties in 27 countries in Europe and the Americas, Castanho Silva (2019, p. 280) concluded that “elite collusion and corrupt governments are the most important factors behind the rise of populists.”

  28. 28.

    Here, too, there is a strong parallel with Bermeo’s (2003, p. 232) conclusion that democratic breakdowns in twentieth-century Europe and Latin America often occurred “where civil society was relatively dense.” On “hollowing out,” see Mair (2013).

  29. 29.

    The standard deviation of left-right placements in Hungary increased from 2.29 in 2009 to 2.55 by 2019. The proportion of respondents placing themselves at 0, 1, or 2 on the zero-to-ten left-right scale increased from 8.7% to 13.4% (an estimated 4.7% increase with a standard error of 1.3), while the proportion placing themselves at 8, 9, or 10 decreased from 21.5% to 20.5% (an estimated 1.0% decrease with a standard error of 1.8).

  30. 30.

    The standard deviation of left-right placements in Poland increased from 2.32 in 2015 to 2.42 in 2016–2017 and 2.47 in 2018–2019. The proportion of respondents placing themselves at 0, 1, or 2 on the zero-to-ten scale increased from 7.9% to 9.3% (an estimated 1.4% increase with a standard error of 1.1), while the proportion placing themselves at 8, 9, or 10 increased from 24.3% to 25.5% (an estimated 1.2% increase with a standard error of 1.7).

  31. 31.

    Kate Connolly and Ian Traynor, “Hungary Receives Rescue Package, With Strings Attached,” The Guardian, 29 October 2008 (https://www.theguardian.com/business/2008/oct/29/hungary-economy-imf-eu-world-bank ). “Hungary Offered A Tripartite Bailout—With Strings,” Forbes, 29 October 2008 (https://www.forbes.com/2008/10/29/hungary-imf-aid-markets-economy-cx_vr_1029markets7.html#4ae90d1c3404)

  32. 32.

    In the V-Dem project’s annual assessments of “Liberal Democracy,” Hungary’s rating fell from .782 in 2009 to .370 in 2019. Poland’s rating fell from.811 in 2014 to.533 in 2019. By way of comparison, France and Germany’s ratings in 2019 were .798 and .838; Bulgaria and Romania’s were.513 and.467.

  33. 33.

    Zaller (1998, p. 186) interpreted President Bill Clinton’s popularity in the wake of a major scandal as demonstrating “just how relentlessly the majority of voters can stay focused on the bottom line,” meaning not just prosperity but, more broadly, “political substance.” He noted that the public might not be “either wise or virtuous. For one thing, its sense of substance seems, in the aggregate, rather amoral—usually more like ‘what have you done for me lately’ than ‘social justice’.”

  34. 34.

    Important early works include Prothro and Grigg (1960) and McClosky (1964). Two recent examples are Bartels (2020) and Graham and Svolik (2020).

  35. 35.

    I am grateful to Vanderbilt’s May Werthan Shayne Chair for financial support, to Nancy Bermeo for advice and encouragement, to Kaitlen Cassell for splendid research assistance, and to Cassell, Benjamin Page, John Sides, and participants in the Vanderbilt Political Science faculty workshop for helpful criticism of a preliminary draft of this report.

References

Download references

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to Larry M. Bartels .

Editor information

Editors and Affiliations

Rights and permissions

Reprints and permissions

Copyright information

© 2022 The Author(s), under exclusive license to Springer Nature Switzerland AG

About this chapter

Check for updates. Verify currency and authenticity via CrossMark

Cite this chapter

Bartels, L.M. (2022). Democracy Erodes from the Top: Public Opinion and Democratic “Backsliding” in Europe. In: Economou, E.M., Kyriazis, N.C., Platias, A. (eds) Democracy in Times of Crises. Springer, Cham. https://doi.org/10.1007/978-3-030-97295-0_3

Download citation

Publish with us

Policies and ethics