ArticlePDF Available

Public Support for Democracy in the United States Has Declined Generationally

Authors:

Abstract and Figures

Support for democracy in the United States, once thought to be solid, has now been shown to be somewhat shaky. One of the most concerning aspects of this declining attachment to democracy is a marked age gap, with younger Americans less supportive of democracy than their older compatriots. Using age-period-cohort analysis of 12 national surveys collected between 1995 and 2019, we show that this age gap is largely a function of a long-term generational decline in support for democracy, with little evidence of an independent life-cycle effect apparent. The combination of generational decline without a positive and counterbalancing life-cycle effect offers a sober prognosis of how support for democracy in the United States might look in the future.
Content may be subject to copyright.
Public Support for Democracy in the United States Has
Declined Generationally
Christopher Claassen
1,
* , Pedro C. Magalh~
aes
2
1
Senior Lecturer, School of Social and Political Sciences, University of Glasgow, Glasgow,
UK
2
Research Professor, Institute of Social Sciences (ICS), University of Lisbon, Lisbon, Portugal
Abstract Support for democracy in the United States, once thought to
be solid, has now been shown to be somewhat shaky. One of the most
concerning aspects of this declining attachment to democracy is a
marked age gap, with younger Americans less supportive of democracy
than their older compatriots. Using age-period-cohort analysis of 12 na-
tional surveys collected between 1995 and 2019, we show that this age
gap is largely a function of a long-term generational decline in support
for democracy, with little evidence of an independent life-cycle effect
apparent. The combination of generational decline without a positive
and counterbalancing life-cycle effect offers a sober prognosis of how
support for democracy in the United States might look in the future.
The commitment of the people of the United States to a democratic system,
long taken for granted,
1
is now in doubt. A growing body of research has
demonstrated their shaky support for democracy when this is understood as
support for concrete democratic norms or institutions or a preference for pro-
versus antidemocratic political candidates (e.g., Bartels 2020;Graham and
Svolik 2020;Gibson 2021;Simonovits, McCoy, and Littvay 2022). There is,
however, considerable evidence now that Americans’ commitment to democ-
racy even in the abstract is also in decline. As Voeten puts it, “the United
States is an example of a country where support for democracy has gone
down while alternatives have become more acceptable” (Voeten 2017,p.3;
c.f. Drutman, Goldman, and Diamond 2020).
*Corresponding author: Christopher Claassen, Politics, University of Glasgow, Adam Smith
Building, Glasgow G12 8RT, UK; email: christopher.claassen@glasgow.ac.uk.
1. Graham and Svolik (2020, p. 392) make this point vividly by quoting Robert Dahl: “It is
nearly impossible to find an American who says that he is opposed to democracy or favors some
alternative” (Dahl 1966, p. 40).
V
CThe Author(s) 2023. Published by Oxford University Press on behalf of American Association for Public Opinion Research.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (https://creativecommon-
s.org/licenses/by/4.0/), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is
properly cited.
https://doi.org/10.1093/poq/nfad039
Public Opinion Quarterly (2023) Vol 00 No 0, 1–14
Downloaded from https://academic.oup.com/poq/advance-article/doi/10.1093/poq/nfad039/7275177 by guest on 18 September 2023
These trends can be seen in figure 1, which reports longitudinal estimates
of support for democracy in the United States. The composite measure
(main panel) is obtained from Claassen’s (2020) dynamic Bayesian estimates
of democratic mood across multiple countries and years. Rather than using
survey questions about respondents’ “satisfaction with democracy,” closer to
Easton’s conceptualization of specific support (Easton 1975, p. 437),
“democratic mood” is based on widely used questions gauging the desirabil-
ity of democracy, its comparison to undemocratic alternatives, or evaluations
of the latter (see, e.g., Norris 2011). In this way, it captures diffuse support
for democracy as a political regime. Although the credible intervals overlap,
Claassen’s estimate of democratic mood in the United States fell from well
above the global average (zero) of 141 countries in the world in 1995 to
close to that global average in 2020. In contrast, there was no commensurate
Figure 1. Declining public support for democracy in the United States, 1995 to
2020. The figure on the left shows Bayesian estimates of latent public support
for democracy in the United States and the other G7 nations, based on Claassen
(2020). Data are standardized such that the mean level of support for democracy
across 141 countries and 33years equals zero and the standard deviation equals
one. The US estimates rely on survey data from the World Values Survey,
AmericasBarometer, Pew Research, and the Comparative Study of Electoral
Systems. Years in which survey data were collected are shown by the rug plot
on the x-axis. The figures on the right show the proportions of US residents of-
fering support for democracy in response to two particular questions routinely
fielded by the World Values Survey (top) and AmericasBarometer (bottom) sur-
veys. For the purposes of this figure, rejecting a strong leader is defined as
selecting the “fairly bad” or “very bad” response options; supporting democracy
is defined as selecting response options 5 through 7. See table 1 for question
wordings and response options. See Supplementary Material figure S1 for more
longitudinal estimates of specific items.
2 C. Claassen and P.C. Magalh~
aes
Downloaded from https://academic.oup.com/poq/advance-article/doi/10.1093/poq/nfad039/7275177 by guest on 18 September 2023
decline in a group of comparator nations—the other six members of the G7
group of high-income democracies.
These trends remain visible when one examines specific survey questions:
whereas 75 percent of respondents in the United States rejected a system of
government with a “strong leader who does not have to bother with Congress
and elections in 1995, only 62 percent did so in 2017 (figure 1; upper-right
plot). And, while 94 percent of Americans agreed that democracy “is better
than any other form of government” in 2006, only 71 percent continued to do
so in 2019 (figure 1; lower-right plot). In sum, while majorities remain sup-
portive of democracy (Drutman, Goldman, and Diamond 2020), it is clear that
support has declined substantially over the past few decades. To the extent
that dwindling democratic support encourages antidemocratic strategies by po-
litical elites (Seligson and Booth 2009) or increases the electoral prospects of
authoritarian leaders or antisystem parties (Mattes 2018;Cohen et al. 2022;
Lewandowsky and Jankowski 2022), such a trend is troubling news for de-
mocracy in the United States.
One of the most concerning aspects of this ebbing attachment to democ-
racy is a marked age gap, with younger Americans more hesitant about de-
mocracy and more supportive of nondemocratic alternatives (Foa and Mounk
2016;Norris 2017;Malka et al. 2020;Gibson 2021). However, it remains
unclear how we should interpret this age gap. Is it a life-cycle effect,
whereby younger citizens have always been more skeptical about democracy
than their older fellow citizens (e.g., Norris 2017)? Or is it an indicator of
generational change, in which the younger generations have lost support
for democracy (Foa and Mounk 2016)? The consequences could not be
more starkly different: if a life-cycle effect, youthful detachment will
naturally transform into system support with the passage of time; if a
cohort effect, it is political culture in the United States that could be
transformed, as older generations who are more supportive of democracy are
replaced by younger generations who are more open to authoritarian
governance.
To separate the effects of aging (i.e., life-cycle effects) from the effects of
generational change (i.e., cohort effects), we turn, in this paper, to age-
period-cohort analyses. While such analyses of European support for democ-
racy have been conducted (where little to no cohort effects are uncovered;
Wuttke et al. 2020), the US case has been neglected, despite accumulating
evidence of diminishing American support for democracy (e.g., figure 1).
Using Bayesian Generalized Additive age-period-cohort models, and data
from 12 national surveys collected between 1995 and 2019, we find little ev-
idence of a life-cycle effect. Instead, we show that support for democracy
and rejection of authoritarian rule has decreased generationally in the United
States. This suggests that the decline in support for democracy is not easily
reversible.
Support for Democracy in the United States 3
Downloaded from https://academic.oup.com/poq/advance-article/doi/10.1093/poq/nfad039/7275177 by guest on 18 September 2023
Data and Methods
We collect survey measures of support for democracy from two public opinion
projects: the World Values Survey (WVS) and the AmericasBarometer (AB).
Each has polled nationally representative samples of Americans multiple times
across a decade or more. Between 1995 and 2017, in five separate surveys, the
WVS fielded five questions asking US respondents about their support for de-
mocracy or rejection of undemocratic rule. And between 2006 and 2019, in
seven surveys, the AB included a question asking US respondents about their
support for democracy. These survey data are described in table 1.
We analyze each dataset separately, testing if our results remain robust across
the two. Since there are five “support for democracy” items included in the
WVS, we first estimate a scale using a graded response model, a form of item-
response theoretic (IRT) model.
2
The single support for democracy item included
in the AB is treated as ordinal due to the use of a seven-point response set.
It is well known that linear age, period, and cohort effects are not sepa-
rately identifiable even when one has access to data that includes different
age groups and is gathered over a period of many years (see Fosse and
Winship 2019 for a review). For example, when it comes to survey data, a
respondent’s age is nothing more than their year of birth (i.e., their cohort)
and the date when they completed the survey (i.e., the period). Researchers
instead estimate linear combinations of these effects (e.g., the overall trend,
period þcohort effects) or treat one or more of these effects as nonlinear.
Two specific models have been favored for this task by researchers in politi-
cal science: hierarchical age-period-cohort models (HAPCMs; e.g., Schwadel
and Garneau 2014;Smets and Neundorf 2014) and generalized additive
models (GAMs; e.g., Grasso 2014;Jiang and Carriere 2014;Wuttke et al.
2020). The essence of each approach is to separate the effects of interest
(e.g., age, cohorts) into both a linear and a nonlinear component. The latter
can be identified for age groups, periods, and cohorts even though only two
of their linear effects are identifiable. The HAPCM accomplishes this by
modeling cohorts (and often also age groups and periods) as random effects.
The GAM approach models cohorts (or year of birth) using cubic spline
functions. We prefer the latter, as it avoids having to rely on arbitrary cohort
or generational definitions (Jiang and Carriere 2014). Our HAPCM results
are similar, however, and are included in the Supplementary Material.
Specifically, we model respondents’ support for democracy as a function of
their cohort (using cubic splines for year of birth), age (using fixed effects for
2. Creating a scale out of all available items reduces measurement error and is therefore prefera-
ble to examining each item separately (see Berinsky 2017 for a discussion). The additive scale of
the five WVS items shows adequate interitem reliability (Cronbach’s alpha ¼0.67, mean correla-
tion ¼0.28).
4 C. Claassen and P.C. Magalh~
aes
Downloaded from https://academic.oup.com/poq/advance-article/doi/10.1093/poq/nfad039/7275177 by guest on 18 September 2023
age groups 18–29, 30–44, 45–59, and over 60), and the period of the survey (us-
ing fixed effects for survey year). We run these models both with and without
additional control variables. On the one hand, any generational patterns in sup-
port for democracy are of descriptive interest: they are noteworthy regardless of
whether generation remains significant when controlling for other demographic
factors. On the other hand, it is also worth investigating whether any genera-
tional effects are robust to including controls, which helps establish whether
Table 1. Survey measures of US support for democracy.
World Values Survey (WVS)
Question wording Response set Years fielded
I’m going to describe various types
of political systems and ask what you
think about each as a way of govern-
ing this country. For each one, would
you say it is a very good, fairly good,
fairly bad or very bad way of govern-
ing this country?
1: very good
2: fairly good
3: fairly bad
4: very bad
1995, 1999, 2006,
2011 & 2017
- Having a democratic political
system
- Having the army rule
- Having a strong leader who does
not have to bother with Congress and
elections
- Having experts, not government,
make decisions according to what
they think is best for the country
Democracy may have problems but
it’s better than any other form of
government
1: agree strongly
2: agree
3: disagree
4: strongly disagree
1995 & 1999
AmericasBarometer
Question wording Response set Years fielded
Changing the subject again,
democracy may have problems,
but it is better than any other form
of government. To what extent do
you agree or disagree with this
statement?
1: strongly disagree
2-6: unlabelled
7: strongly agree
2006, 2008, 2010,
2012, 2014,
2017 & 2019
Note: Pooled number of respondents available: 8,819 (WVS); 9,609 (AmericasBarometer).
Support for Democracy in the United States 5
Downloaded from https://academic.oup.com/poq/advance-article/doi/10.1093/poq/nfad039/7275177 by guest on 18 September 2023
declining support for democracy in the United States is best characterized as a
generational problem, or, for example, a class one. As control variables, we in-
clude: Republican and Democratic identity (vs. independent), having a college
degree, identifying as female, identifying as white, living in the South, and self-
reported income (standardized within each survey wave). The single support for
democracy item in the AmericasBarometer data is modeled as an ordinal vari-
able using ordered logit GAMs. The IRT measure of support for democracy
obtained from the WVS data is treated as a continuous variable and modeled us-
ing linear GAMs. All analyses are weighted using the weights included in the
WVS and AmericasBarometer datasets.
3
Respondents with missing values are
dropped via listwise deletion.
We estimate all models using Bayesian Markov Chain Monte Carlo
(MCMC) methods, which allows more complex models to be fit and more
accurate variance estimates to be obtained than the corresponding restricted
maximum likelihood methods. We use the brms package, which calls the
Stan modeling platform from R. Weakly informative priors are used. All
models show convergence, as indicated by trace plots R-hat statistics of close
to 1 (specifically, less than 1.02; see Supplementary Material).
Results
One conjecture about the decline of democratic support in the United States is
that American millennials (by convention, those born from the early
1980s until the mid-1990s) have particularly low levels of democratic
support. The main rationale is that this cohort represents the first group
of Americans whose crucial formative years were spent in the post–Cold War
world, a context where threats to democracy became less plausible and vivid
than for previous generations (Foa and Mounk 2016 and 2019). Furthermore,
this millennial generation is also argued to have grown into adulthood under
deteriorating economic conditions (Denemark et al. 2016, p. 184), including
growing income inequality and stagnating incomes for the lower and middle
classes (Foa and Mounk 2019, p. 1021). These conditions may have rendered
this cohort particularly open to contemplating alternatives to the democratic
status quo, in contrast with previous generations.
In figure 2, we examine this conjecture by presenting the smoothed GAM
estimates of support for democracy across year of birth (see also the tables of
parameter estimates—tables 2 and 3). The AmericasBarometer results (left)
show that the more recent the year of birth, the lower the agreement with the
notion that “democracy may have problems, but it is better than any other
3. It is not possible, using the information made available in the public WVS and
AmericasBarometer datasets, to account for design effects. Such effects are not included in the
estimates of uncertainty we present in this paper.
6 C. Claassen and P.C. Magalh~
aes
Downloaded from https://academic.oup.com/poq/advance-article/doi/10.1093/poq/nfad039/7275177 by guest on 18 September 2023
form of government.” Respondents born before and during the Second
World War (i.e., the “Silent Generation”) are more than 80 percent likely to
express support for democracy in this way, while respondents born in the
1970s and 1980s are less than 70 percent likely to support democracy. The
generational effects are even more pronounced when demographic variables
are omitted from the GAM (top left figure).
The WVS results (figure 2, on the right) show, on their face, a slightly different
result. Support for democracy remains stable for the 1910s to 1940s birth cohorts,
and subsequently reaches its highest level for respondents born around the time of
the Second World War. However, support then follows the same precipitous de-
cline as seen in the AmericasBarometer data. As we reach those born since the
1980s, support for democracy is more than half a standard deviation lower than
its Second World War peak—net the effects of age, period, and demography.
4
Figure 2. Generational effects. AmericasBarometer estimates (left) are from a
single item; ordered logit GAM used. “Agreeing democracy is best” is defined
for the purposes of this figure as selecting response category 5 or higher; the
underlying model treats the outcome as ordinal, however. WVS estimates
(right) use a five-item scale; linear GAM used. Demographic controls includ-
ing for models in bottom row and excluded in top row.
4. See the Supplementary Material for additional analyses, including HAPC models
(Supplementary Material figure S3) as well as GAM models employing more restricted three- and
Support for Democracy in the United States 7
Downloaded from https://academic.oup.com/poq/advance-article/doi/10.1093/poq/nfad039/7275177 by guest on 18 September 2023
To be sure, for both the AB and the WVS data, estimates for the most ex-
treme cohorts must be taken with caution, given potential biases caused by
lack of common support (i.e., no data for youthful respondents in the earliest
Table 2. Parameter estimates, AmericasBarometer GAM.
Model 1 Model 2
Parameter
estimate
Standard
error
Parameter
estimate
Standard
error
Birth year spline variance parameter 0.64 0.05 0.52 0.36
Birth year fixed effect 3.11 1.19 2.71 1.33
Year: 2008 1.03 0.10 1.02 0.15
Year: 2010 1.10 0.10 1.04 0.14
Year: 2012 1.15 0.11 1.09 0.14
Year: 2014 1.43 0.11 1.33 0.15
Year: 2017 1.23 0.11 1.15 0.15
Year: 2019 1.13 0.12 1.09 0.15
Age: 3044 0.13 0.09 0.18 0.09
Age: 4559 0.19 0.15 0.28 0.15
Age: 60þ0.11 0.19 0.22 0.20
Republican 0.71 0.05
Democrat 0.41 0.04
Has degree 0.29 0.05
Female 0.35 0.04
White 0.08 0.04
Income 0.20 0.02
South 0.04 0.04
Cutpoint1 4.96 0.14 4.77 0.18
Cutpoint2 4.37 0.13 4.19 0.17
Cutpoint3 3.64 0.13 3.45 0.17
Cutpoint4 2.50 0.13 2.26 0.17
Cutpoint5 1.71 0.13 1.43 0.17
Cutpoint6 0.78 0.13 0.45 0.17
N9,584 8,902
Note: Results for World Values Survey Generalized Additive Model estimated using
Bayesian MCMC methods, as implemented in the brms library for R. Three chains were run in
parallel for 2,000 iterations, with the first 1,000 of these being dedicated to warmup of the
MCMC algorithm. Convergence is diagnosed by examination of R-hat statistics and posterior
predictive plots. “Parameter estimates” are the mean of the posterior distributions for each param-
eter across the 3,000 post-warmup iterations (i.e., 1,000 3 chains); “standard errors” are the
standard deviation of these parameter posterior distributions.
four-item scales from the WVS data (Supplementary Material figure S6). Results are fundamen-
tally similar.
8 C. Claassen and P.C. Magalh~
aes
Downloaded from https://academic.oup.com/poq/advance-article/doi/10.1093/poq/nfad039/7275177 by guest on 18 September 2023
cohorts or for elderly respondents in the latest cohorts). Still, the message
common to both sets of results is that there has been a pattern of cohort de-
cline in democratic support taking place at least since those born in the
1940s. Furthermore, both analyses also raise doubts about whether there
might be something unique about the “millennial generation” in terms of its
support for democracy, measured either using a single survey question (as in
the AmericasBarometer data) or with a more nuanced measure encompassing
rejection of autocratic forms of government (as in the WVS data). Instead,
each cohort born after the 1940s has been less supportive of democracy
than the one before. From this point of view, we have no evidence that
Table 3. Parameter estimates, World Values Survey GAM.
Model 1 Model 2
Parameter
estimate
Standard
error
Parameter
estimate
Standard
error
Birth year spline variance parameter 0.73 0.35 0.64 0.33
Birth year fixed effect 0.43 0.72 0.20 0.71
Year: 1999 0.30 0.04 0.33 0.04
Year: 2006 0.32 0.04 0.32 0.05
Year: 2011 0.34 0.05 0.36 0.05
Year: 2017 0.34 0.06 0.47 0.06
Age: 3044 0.02 0.04 0.04 0.05
Age: 4559 0.05 0.07 0.01 0.08
Age: 60þ0.02 0.11 0.14 0.11
Republican 0.08 0.03
Democrat 0.18 0.03
Has degree 0.43 0.02
Female 0.15 0.02
White 0.25 0.03
South 0.10 0.02
Income 0.01 0.01
Intercept 0.25 0.04 0.07 0.05
Residual standard deviation 0.97 0.01 0.94 0.01
N8,797 7,474
Note: Results for World Values Survey Generalized Additive Model estimated using
Bayesian MCMC methods, as implemented in the brms library for R. Three chains were run in
parallel for 2,000 iterations, with the first 1,000 of these being dedicated to warmup of the
MCMC algorithm. Convergence is diagnosed by examination of R-hat statistics and posterior
predictive plots. “Parameter estimates” are the mean of the posterior distributions for each param-
eter across the 3,000 post-warmup iterations (i.e., 1,000 3 chains); “standard errors” are the
standard deviation of these parameter posterior distributions.
Support for Democracy in the United States 9
Downloaded from https://academic.oup.com/poq/advance-article/doi/10.1093/poq/nfad039/7275177 by guest on 18 September 2023
millennials are anything other than at the tail end of a long-term trend of gen-
erational decline in democratic support.
Importantly, these results are obtained while simultaneously adjusting for
(nonlinear) age effects. That younger Americans are less supportive of de-
mocracy than their parents could just be a function of a life-cycle effect
(Alexander and Welzel 2017;Norris 2017), linked to the greater political in-
difference and inattention of younger citizens (Nem
cok and Wass 2021),
their lower experience with the functioning of a democratic regime (Sapiro
2004), or even the systematic descriptive and substantive underrepresentation
of the young in most democracies (Curry and Haydon 2018;Sundstro¨m and
Stockemer 2021). However, as we can see in figure 3, our age-period-cohort
analysis shows there is not much of a life-cycle effect in democratic
support in the United States accordingly, in contrast to the marked
cohort effects we found before. There are no significant differences between
the four age groups once cohort and period effects are considered
(see also Supplementary Material figure S4 for similar results from HAPC
models).
Figure 3. Age effects. Predicted effects of support for democracy by age
group, with other variables set at means or modes. AmericasBarometer esti-
mates (left) are based on an ordered logit GAM. WVS estimates (right) are
based on a linear GAM.
10 C. Claassen and P.C. Magalh~
aes
Downloaded from https://academic.oup.com/poq/advance-article/doi/10.1093/poq/nfad039/7275177 by guest on 18 September 2023
Finally, we also consider period effects, which are shown in figure 4.
Period effects reflect the time-specific factors that have shaped national lev-
els of support for democracy, including short-run events and long-run secular
changes, but also survey-specific methodological factors. As can be seen in
figure 4, we find higher support for democracy in the first wave of
AmericasBarometer and WVS surveys, net the effects of age and generation
(which may well be a methods effect relating to those first-wave surveys).
There is otherwise little evidence of temporal variation, which suggests that
the trend of declining support for democracy in the United States shown in
figure 1 is largely a product of more supportive older generations being
replaced by less supportive younger ones.
Conclusions
Support for democracy in the United States is not as solid as it once seemed.
While a growing number of authors have demonstrated weak public support
for specific democratic institutions and norms or candidates committed to
Figure 4. Period effects. Predicted effects of support for democracy by survey
year, with other variables set at means or modes. AmericasBarometer esti-
mates (left) are based on an ordered logit GAM. WVS estimates (right) are
based on a linear GAM.
Support for Democracy in the United States 11
Downloaded from https://academic.oup.com/poq/advance-article/doi/10.1093/poq/nfad039/7275177 by guest on 18 September 2023
democracy, we show that diffuse support for a democratic regime in the ab-
stract is weaker than it once was.
Existing research has found a marked effect of age on support for democ-
racy in the United States, with younger citizens demonstrating significantly
lower levels of democratic commitment. Some have attributed this age gap
to a cohort effect, for which the generation of those born in the 1980s and
1990s is argued to be particularly responsible. Others have attributed this to
a life-cycle effect, in which support for democracy is weaker for younger
individuals but is later learned as one matures and becomes socialized into
the political system.
Using age-period-cohort analysis, we find no evidence for such a life-cycle
effect. Furthermore, we demonstrate that generational decline in support for
democracy did not start with the cohort of those who spent their
crucial formative years in a post–Cold War context. Instead, American support
for democracy has been weakening in one cohort after the next at least since
the Second World War. This trend echoes the findings of age-period-cohort
analyses of political tolerance (Schwadel and Garneau 2014) and civic partici-
pation (Caren, Ghoshal, and Ribas 2011) in the United States. But it diverges
from the findings of a previous age-period-cohort analysis of support for de-
mocracy in Europe, where “generational disparities are narrow in most cases
..., and members of this [most recent] cohort remain committed to democracy
as a viable system of government” (Wuttke et al. 2020,p.10).
The combination of generational decline and the lack of a positive life-cycle
effect offers a sobering prognosis of how support for democracy in the United
States might look in the future. Younger generations already have substantially
lower democratic support than the older generations whom they will replace.
Without a positive life-cycle effect to counter this intergenerational replace-
ment, the United States faces a continued decline in public commitment to a
democratic system of government. Such decline justifies concerns with the fu-
ture endurance of democratic institutions in the face of potential political, so-
cial, and economic crises and with the availability of the American public to
shut out leaders and movements set on undermining liberal democracy.
Supplementary Material
Supplementary Material may be found in the online version of this article:
https://doi.org/10.1093/poq/nfad039.
Funding
Pedro C. Magalh~
aes was supported by the Fundac¸~
aoparaaCieˆncia e
Tecnologia, grants UIDB/50013/2020, UIDP/50013/2020 and LA/P/0051/2020.
12 C. Claassen and P.C. Magalh~
aes
Downloaded from https://academic.oup.com/poq/advance-article/doi/10.1093/poq/nfad039/7275177 by guest on 18 September 2023
Acknowledgements
This paper benefited from feedback provided by James L. Gibson, Anja
Neundorf, Lisanne de Blok, and Alex Wuttke.
Data Availability
Replication data and documentation are available at https://doi.org/10.7910/
DVN/HVYTNM.
References
Alexander, A. C., and C. Welzel. 2017. “The Myth of Deconsolidation: Rising Liberalism and
the Populist Reaction (No. 10).” ILE Working Paper Series. Hamburg: University of
Hamburg, Institute of Law and Economics (ILE). https://www.econstor.eu/bitstream/10419/
170694/1/ile-wp-2017-10.pdf.
Bartels, L. M. 2020. “Ethnic Antagonism Erodes Republicans’ Commitment to Democracy.”
Proceedings of the National Academy of Sciences of the United States of America 117:
22752–59.
Berinsky, A. 2017. “Measuring Public Opinion with Surveys.” Annual Review of Political
Science 20:309–29.
Caren, N., R. A. Ghoshal, and V. Ribas. 2011. “A Social Movement Generation: Cohort and
Period Trends in Protest Attendance and Petition Signing.” American Sociological Review 76:
125–51.
Claassen, C. 2020. “In the Mood for Democracy? Democratic Support as Thermostatic Opinion.”
American Political Science Review 114:36–53.
Cohen, M. J., A. E. Smith, M. W. Moseley, and M. L. Layton. 2022. “Winners’ Consent?
Citizen Commitment to Democracy When Illiberal Candidates Win Elections.” American
Journal of Political Science 67:261–76.
Curry, J. M., and M. R. Haydon. 2018. “Lawmaker Age, Issue Salience, and Senior
Representation in Congress.” American Politics Research 46:567–95.
Dahl, R. A. 1966. Political Oppositions in Western Democracies. New Haven, CT: Yale
University Press.
Denemark, D., T. Donovan, and R. G. Niemi. 2016. The Advanced Democracies: The Erosion of
Traditional Democratic Citizenship. In Growing up Democratic: Does It Make a Difference?,
edited by D. Denemark, R. B. Mattes, and R. G. Niemi, 181–206. Boulder: Lynne Rienner
Publishers.
Drutman, L., J. Goldman, and L. Diamond. 2020. Democracy Maybe: Attitudes on
Authoritarianism in America.” Voter Study Group. https://www.voterstudygroup.org/publica
tion/democracy-maybe.
Easton, D. 1975. “A Re-Assessment of the Concept of Political Support.” British Journal of
Political Science 5:435–57.
Foa, R. S., and Y. Mounk. 2016. “The Danger of Deconsolidation: The Democratic Disconnect.”
Journal of Democracy 27:5–17.
—–—. 2019. “Youth and the Populist Wave.” Philosophy & Social Criticism 45:1013–24.
Fosse, E., and C. Winship. 2019. “Analyzing Age-Period-Cohort Data: A Review and Critique.”
Annual Review of Sociology 45:467–92.
Gibson, J. L. 2021. “After the Storm: Racial Differences in Allegiance to America’s
Democracy.” SSRN.https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3854107.
Support for Democracy in the United States 13
Downloaded from https://academic.oup.com/poq/advance-article/doi/10.1093/poq/nfad039/7275177 by guest on 18 September 2023
Graham, M. H., and M. W. Svolik. 2020. “Democracy in America? Partisanship, Polarization,
and the Robustness of Support for Democracy in the United States.” American Political
Science Review 114:392–409.
Grasso, M. T. 2014. “Age, Period and Cohort Analysis in a Comparative Context: Political
Generations and Political Participation Repertoires in Western Europe.” Electoral Studies 33:
63–76.
Jiang, B., and K. C. Carriere. 2014. “Age-Period-Cohort Models Using Smoothing Splines: A
Generalized Additive Model Approach.” Statistics in Medicine 33:595–606.
Lewandowsky, M., and M. Jankowski. 2022. “Sympathy for the Devil? Voter Support for
Illiberal Politicians.” European Political Science Review 15:39–56. https://doi.org/10.1017/
S175577392200042X.
Malka, A., Y. Lelkes, B. N. Bakker, and E. Spivack. 2020. “Who Is Open to Authoritarian
Governance within Western Democracies?” Perspectives on Politics 20:808–27. https://doi.
org/10.1017/S1537592720002091.
Mattes, R. 2018. “Support for Democracy.” In Oxford Research Encyclopedia in Politics, edited
by W. Thompson. New York: Oxford University Press. https://doi.org/10.1093/acrefore/
9780190228637.013.622.
Nem
cok, M., and H. Wass. 2021. “Generations and Political Engagement.” In Oxford Research
Encyclopedia of Politics.https://doi.org/10.1093/acrefore/9780190228637.013.1756.
Norris, P. 2011. Democratic Deficit: Critical Citizens Revisited. New York: Cambridge
University Press.
—–—. 2017. “Is Western Democracy Backsliding? Diagnosing the Risks.” HKS Faculty
Research Working Paper Series, No. RWP17-012. Boston, MA: Harvard Kennedy School,
Harvard University. https://www.hks.harvard.edu/publications/western-democracy-backslid
ing-diagnosing-risks.
Sapiro, V. 2004. “Not Your Parents’ Political Socialization: Introduction for a New Generation.”
Annual Review of Political Science 7:1–23.
Schwadel, P., and C. R. Garneau. 2014. “An Age–Period–Cohort Analysis of Political Tolerance
in the United States.” Sociological Quarterly 55:421–52.
Seligson, M. A., and J. A. Booth. 2009. “Predicting Coups? Democratic Vulnerabilities, the
AmericasBarometer and the 2009 Honduran Crisis.” Perspectivas desde el Baro´metro de las
Ame´ricas.https://www.vanderbilt.edu/lapop/insights/I0821en.pdf.
Simonovits, G., J. McCoy, and L. Littvay. 2022. “Democratic Hypocrisy and Out-Group Threat:
Explaining Citizen Support for Democratic Erosion.” Journal of Politics 84:1806–11. https://
www.journals.uchicago.edu/doi/abs/10.1086/719009.
Smets, K., and A. Neundorf. 2014. “The Hierarchies of Age-Period-Cohort Research: Political
Context and the Development of Generational Turnout Patterns.” Electoral Studies 33:41–51.
Sundstro¨m, A., and D. Stockemer. 2021. “Conceptualizing, Measuring, and Explaining Youths’
Relative Absence in Legislatures.” PS: Political Science & Politics 54:195–201.
Voeten, E. 2017. “Are People Really Turning Away from Democracy?” Journal of Democracy
Web Exchange.https://www.journalofdemocracy.org/wp-content/uploads/2018/12/Journal-of-
Democracy-Web-Exchange-Voeten_0.pdf.
Wuttke, A., K. Gavras, and H. Schoen. 2020. “Have Europeans Grown Tired of Democracy?
New Evidence from Eighteen Consolidated Democracies, 1981–2018.” British Journal of
Political Science 52:1–13.
14 C. Claassen and P.C. Magalh~
aes
Downloaded from https://academic.oup.com/poq/advance-article/doi/10.1093/poq/nfad039/7275177 by guest on 18 September 2023
... Yet, scholars are disputed on the meaning of these trends among democracies' youngest members. Some argue that young people have grown tired of democracy for performative reasons, but that this tiredness has not affected their democratic aspirations (Foa and Mounk 2016;Norris 2011;Claassen and Magalhães 2023). More recently, however, some scholars have raised concerns regarding younger citizens' commitment to democracy, calling into question the commonly-held premise that a growing democratic disconnect is not necessarily fueling a shift towards undemocratic and illiberal alternatives (Foa and Mounk 2019;Frederiksen 2024;Huttunen and Saikkonen 2023). ...
... In their seminal article on youth and democracy, Foa and Mounk (2019) contend that younger people are particularly vulnerable to the rising tide of illiberalism, suggesting that youth have become increasingly disenchanted with democracy as a political system. In a similar vein, Claassen and Magalhães (2023) demonstrate that support for democracy has decreased generationally in the United States, with little evidence of life-cycle decreases to counter-balance these cohort-level shifts. ...
... Existing research provides ambiguous evidence regarding young citizens' fit in either of these concepts. On one hand, some scholars argue that young people have grown tired of democracy for performative reasons, but that this tiredness has not affected their democratic aspirations, essentially placing them in the democratic apathy category (Foa and Mounk 2016;Norris 2011;Claassen and Magalhães 2023). More recently, however, some scholars have raised concerns regarding younger citizens' overall commitment to democracy, arguing that a growing democratic disconnect is fueling a shift towards undemocratic and illiberal alternatives (Foa and Mounk 2019;Frederiksen 2024;Huttunen and Saikkonen 2023). ...
Preprint
Full-text available
Recent scholarship has highlighted growing concerns about the erosion of democratic norms and values, particularly among younger citizens. This research agenda remains ambiguous, however, especially regarding the extent to which young people’s democratic disconnect results in a shift towards undemocratic politics. In this research, I rely on a multi-method approach to examine differences in democratic attitudes across age groups. Using observational and experimental data, I provide robust cross-national evidence that younger citizens’ support for democracy is substantially lower than older citizens. Building on these findings, I further demonstrate that youth are far more tolerant of a wide range of undemocratic practices and democratic norm violations. Lastly, I present the results of a conjoint experiment which confirms that younger citizens have significantly lower preferences for democratic societal attributes when compared to their older counterparts. Substantively, these findings contribute to a growing literature on the vulnerability of consolidated democracies to younger peoples’ shifting attitudes towards democratic institutions and norms.
... S'il est en revanche un domaine où les jeunes se distinguent, c'est celui du rapport à la citoyenneté et à la politique, particulièrement distant et défiant (Tiberj, 2017; Lardeux and Tiberj, 2021). Plusieurs contributions récentes vont plus loin et montrent un attachement plus faible des jeunes au régime démocratique Mounk, 2016, 2017;Belchior and Teixeira, 2023;Claassen and Magalhães, 2023;Grassi, Portos and Felicetti, 2023). Cette thèse a suscité une vaste controverse ancrée dans un débat plus large sur la « déconsolidation » démocratique (définie ici comme la remise en question de la légitimité de la gouvernance démocratique) 1 . ...
Article
À partir d’une enquête auprès de 4 072 personnes âgées de 40 ans et moins, cet article analyse les déterminants du soutien à différents modes de gouvernement chez les jeunes Français, pour comprendre de quelle manière les inégalités spatiales, sociales et politiques expliquent ces préférences. Des analyses descriptives confirment d’abord une certaine forme d’érosion du soutien à la démocratie au sein de la jeunesse française, moins susceptible de soutenir le gouvernement démocratique que ses aînés et plus attirée par un gouvernement militaire. Des analyses inférentielles démontrent ensuite le rôle négligeable du lieu d’habitation pour expliquer ces préférences, contrairement à celui du niveau de diplôme, mais aussi la spécificité des 18-24 ans, particulièrement prompts à soutenir un gouvernement militaire.
Article
Full-text available
Existing literature sees attitudinal support for liberal democratic values as a crucial way to prevent the success of illiberal actors and their policies. But recent studies show that democratic support can coexist with lack of punishment for backsliding practices. To reconcile these findings, I argue that social norms generate incentives to declare that one opposes illiberal actors even if one does not—what I call staged democrats. Staged democrats are not a stable safeguard against illiberal politicians, because information shocks can reveal that they do not sincerely oppose illiberalism. Studying Switzerland, where referendums provide information on the sincere preferences of citizens, I collect data on referendum results, public opinion surveys, and party positions. My analyses show that if referendums reveal that the positions of far-right parties are more popular than expected, those parties become more exclusionary. Attitudes only prevent illiberal policy if they are sincerely held.
Book
Full-text available
Este documento de trabajo tiene como finalidad analizar semejanzas y diferencias de cuatro cohortes generacionales sobre su evaluación del contexto nacional, sus posturas políticas, actitudes hacia la democracia y confianza institucional. El estudio incluye un análisis comparativo de las opiniones de las generaciones frente al contexto del país, a mediados de 2019; su postura frente a la política, los partidos políticos y las elecciones en el marco de los comicios presidenciales de 2019; así como las actitudes ciudadanas hacia la democracia y la confianza institucional.
Article
The twenty‐first century has been one of democratic backsliding. This has stimulated wide‐ranging scholarship on the causes of democratic erosion. Yet an overarching framework that identifies actors, behaviors, and decision processes has not been developed. I offer such a structure that includes elites (e.g., elected officials, the judiciary), societal actors (e.g., social movements, interest groups, media), and citizens. I discuss erosive threats stemming from each actor and the concomitant role of psychological mechanisms. The framework highlights the challenge of arriving at a holistic explanation of erosion within a given country during a finite period. It also accentuates why scholars should regularly consider the implications of their specific findings for democratic stability. I conclude by discussing various lessons and suggestions for how to study democratic backsliding.
Article
Full-text available
Many democracies are witnessing the rise and continuing success of parties and politicians who oppose fundamental principles of liberal democracy. Recent research finds that voters support illiberal politicians, because they trade off policy congruence against attitudes toward liberal democracy. Other studies, however, suggest that authoritarian and populist voters might actually have a preference to vote for illiberal candidates. We argue that both factors interact: Authoritarian and populist voters are more willing to trade off policy representation against support for liberal democracy. To test this mechanism, we rely on a survey experiment conducted in Germany. The results clearly demonstrate that voters indeed trade off policy congruence against liberal democracy. Moreover, this effect is particularly strong for populist and authoritarian voters. Overall, the results have important implications for understanding when and which voters support or oppose liberal democracy.
Article
Full-text available
Democracies without democrats are not sustainable. Yet, recent studies have argued that Western citizens are turning their backs on the system of self-governance, thereby eroding the societal foundations of consolidated democracies. This study contributes to discussions about citizen support of democracy by (1) analyzing new cross-national survey data in 18 European countries that facilitate assessments of the temporal and geographical generalizability of previous findings, (2) disentangling age, cohort and period effects, thereby aligning the analytical methods with the theoretical arguments and (3) transparently reporting all evidence derived from pre-registered analyses to avoid cherry-picked findings. The findings show that citizens of consolidated democracies continue to endorse self-governance. Yet in some (but not all) countries, there is evidence of a growing number of ‘democrats in name only’, particularly among the young generation. These findings suggest a second phase in research on democratic fatigue that broadens the analytical scope for the multi-faceted nature of democratic support.
Article
Full-text available
Significance Growing partisan polarization and democratic “backsliding” in various parts of the world have raised concerns about the attachment of ordinary Americans to democratic institutions and procedures. I find that substantial numbers of Republicans endorse statements contemplating violations of key democratic norms, including respect for the law and for the outcomes of elections and eschewing the use of force in pursuit of political ends. The strongest predictor by far of these antidemocratic attitudes is ethnic antagonism—especially concerns about the political power and claims on government resources of immigrants, African-Americans, and Latinos. The strong tendency of ethnocentric Republicans to countenance violence and lawlessness, even prospectively and hypothetically, underlines the significance of ethnic conflict in contemporary US politics.
Article
Full-text available
This article conceptualizes the relative absence of youth in legislatures, a feature we perceive as a democratic deficit with detrimental consequences. It introduces a new operationalization: the Youth Representation Index. Rather than calculating youths’ representation by the percentage of Members of Parliament 35 or 40 years old and younger or legislatures’ median age, we argue that scholars should assess youths’ parliamentary presence relative to their proportion of the voting-age population. We contribute by assessing the magnitude of youths’ underrepresentation across countries, finding that adults 35 years old and younger are generally underrepresented in legislatures by a factor of three and those 40 years old and younger by a factor of two. We illustrate that youths’ presence increases under proportional representation electoral systems and with candidate age requirements set at 18 years. Finally, our results illustrate that countries with a younger population display a stronger discrepancy in youth representation.
Article
Full-text available
If the values of younger citizens and voters are the trend of the future, in what direction do they point? Scholars have long noted a decline in political engagement and knowledge among youth in developed democracies, with the fear that this may undermine the stability of liberal institutions. However, youth electoral behaviour appears inconsistent: in much of continental Western Europe, younger voters support populist parties of both left and right, but in the United States and the United Kingdom, only left-wing populist movements benefit from youth mobilization. We explain this divergence by arguing for a distinction between democratic apathy and democratic antipathy. Democratic apathy is characterized by scepticism regarding the value of democratic institutions, low turnout and lack of interest in politics, whereas democratic antipathy involves the active embrace of illiberal movements hostile to pluralistic institutions. In societies where youth do not face economic and social discrimination, democratic apathy is the more common trend, whereas in parts of continental Europe where youth face systematic social exclusion, apathy has become active antipathy.
Article
Democracy is in decline worldwide, primarily because voters elect candidates harboring antidemocratic aspirations. Scholars argue that elections animate the democratic spirits of winners and deflate those of losers, but what about contests ending in the victory of authoritarian candidates? To answer this question, we consider the dynamics of commitment to democracy in Brazil's 2018 presidential campaign. Drawing on AmericasBarometer data and an original five-wave panel survey, we confirm that Jair Bolsonaro's campaign attracted skeptics of democracy. Although his election and inauguration boosted his supporters’ allegiance to the political system, it also exacerbated their tolerance for institutional ruptures such as executive-led coups. Meanwhile, election losers retained their democratic commitments. As a result, the authoritarian victory narrowed preexisting winner–loser gaps in support for the political system, but widened gaps in tolerance for certain antidemocratic maneuvers. Thus, authoritarian electoral victories can foster short-term satisfaction among democracy's critics while abetting future instability.
Article
Recent events have raised concern about potential threats to democracy within Western countries. If Western citizens who are open to authoritarian governance share a common set of political preferences, then authoritarian elites can attract mass coalitions that are willing to subvert democracy to achieve shared ideological goals. With this in mind, we explored which ideological groups are most open to authoritarian governance within Western general publics using World Values Survey data from fourteen Western democracies and three recent Latin American Public Opinion Project samples from Canada and the United States. Two key findings emerged. First, cultural conservatism was consistently associated with openness to authoritarian governance. Second, within half of the democracies studied, including all of the English-speaking ones, Western citizens holding a protection-based attitude package⸺combining cultural conservatism with left economic attitudes⸺were the most open to authoritarian governance. Within other countries, protection-based and consistently right-wing attitude packages were associated with similarly high levels of openness to authoritarian governance. We discuss implications for radical right populism and the possibility of splitting potentially undemocratic mass coalitions along economic lines.
Article
Is support for democracy in the United States robust enough to deter undemocratic behavior by elected politicians? We develop a model of the public as a democratic check and evaluate it using two empirical strategies: an original, nationally representative candidate-choice experiment in which some politicians take positions that violate key democratic principles, and a natural experiment that occurred during Montana’s 2017 special election for the U.S. House. Our research design allows us to infer Americans’ willingness to trade-off democratic principles for other valid but potentially conflicting considerations such as political ideology, partisan loyalty, and policy preferences. We find the U.S. public’s viability as a democratic check to be strikingly limited: only a small fraction of Americans prioritize democratic principles in their electoral choices, and their tendency to do so is decreasing in several measures of polarization, including the strength of partisanship, policy extremism, and candidate platform divergence. Our findings echo classic arguments about the importance of political moderation and cross-cutting cleavages for democratic stability and highlight the dangers that polarization represents for democracy.