Authoritarian leaders frequently appeal to social conservative values, sometimes also described as illiberal or traditionalist, but little is known about how or even whether such strategies actually pay political dividends. It is widely presumed that nondemocratic leaders think they are benefiting somehow, and some prior studies do find strong correlations between social conservatism and autocrat support. But correlation is not cause, and existing studies have yet to demonstrate convincingly that the appeal to such values actually puts the autocrat in a stronger position than they would have been by sticking with all the other kinds of appeals they typically make.Reference Abromeit 1
To deepen our understanding of values-based appeals in autocracies, this study exploits the fact that Russia’s leadership embedded President Vladimir Putin’s 2020 attempt to escape constitutional term limits in a plebiscite that was presented to the public as being more about enshrining one dimension of social conservatism (traditional morality) in Russia’s basic law than about extending his rule. Drawing on an original experiment-bearing survey of the Russian population after the plebiscite, it finds evidence that the Kremlin’s appeals to heteronormativity and anti-secularism generated substantial (potentially decisive) added numbers of yes votes for his reform package, primarily by drawing from the large pool of morally conservative voters who did not support him politically (a pool constituting more than half of Putin nonsupporters). These findings expand our understanding of the role of appeals to core political values made by authoritarian leaders, including these values’ potential to be leveraged for the legitimation of institutional and other changes that have no intrinsic connection to the values in question and that might otherwise be wanting in support.
Traditional Moral Values and Authoritarian Support
Studies of authoritarian regime stability commonly focus on either regime mechanics or sources of legitimacy. By mechanics, I mean formal or informal institutional mechanisms or practices that suppress regime-threatening activity, such as mass protests, elite defection, and coups d’état (Brownlee Reference Brownlee2007; Gandhi and Przeworski Reference Gandhi and Przeworski2007; Levitsky and Way Reference Levitsky and Way2022; Pepinsky Reference Pepinsky2014; Svolik Reference Svolik2012). By legitimacy, I refer broadly to public support for the regime relative to perceived alternatives or at least acquiescence to it. Studies have identified a wide variety of factors that sometimes improve authoritarian legitimacy, including performance, especially in the economy (Brancati Reference Brancati2014; Magaloni Reference Magaloni2006; Reuter and Gandhi Reference Reuter and Gandhi2011; Rosenfeld Reference Rosenfeld2018); identity connections (Sharafutdinova Reference Sharafutdinova2020; Zheng Reference Zheng1993); gendered personalistic appeal (Matovski Reference Matovski2021; Sperling Reference Sperling2014); rallying around the flag (Greene and Robertson Reference Greene and Robertson2019); state employment (Rosenfeld Reference Rosenfeld2021); and the strategic manipulation of information (Guriev and Treisman Reference Guriev and Treisman2019; Rozenas and Stukal Reference Rozenas and Stukal2019).
Here we focus on another way autocrats often seek legitimacy: by appealing to values. By “values,” I refer to what political psychologists have called core political values, or “overarching normative principles and belief assumptions about government, citizenship, and … society” that “facilitate position taking in more concrete domains” (McCann Reference McCann1997, 565). When individuals structure such values coherently along a single dimension like “conservatism”–“liberalism,” a process many psychologists believe arises from high cognitive drives for uncertainty and threat reduction, values can constitute elements of ideology (Jost et al. Reference Jost, Glaser, Kruglanski and Sulloway2003). Some complex and broadly applicable values by themselves can “for all intents and purposes” be considered “ideological” (Jost, Federico, and Napier, Reference Jost, Federico and Napier2009). In short, all ideology involves values but not all values are ideological. The possibility that values-based appeals could generate support for autocrats was almost taken for granted during the Cold War era, when regimes frequently invoked leftist or anti-leftist ideology in ways researchers widely considered effective (Friedrich and Brzezinski Reference Friedrich and Brzezinski1956; Linz Reference Linz2000; Malia Reference Malia1995). But with the collapse of the USSR and China’s de facto conversion to a form of economic capitalism, scholarly interest faded in the autocracy-legitimizing potential not only of leftism but also of right-wing values that had been positioned against communism. To the extent values have remained a focus for studies of authoritarian legitimacy, they have concentrated primarily on nationalist values like patriotism and ethnocentrism (Mylonas and Tudor Reference Mylonas and Tudor2023).
Of central interest here are authoritarian appeals to one particular core political value, widely referred to as “traditional morality,” a term capturing the idea that “society should protect traditional religious, moral, and family values” (Schwartz et al. Reference Schwartz, Caprara, Vecchione, Bain, Bianchi, Caprara, Cieciuch, Kirmanoglu, Baslevent and Lönnqvist2014, 903).Reference Akturk 2 As values that legitimize the social status quo, they form part of classic conservative ideology as psychologists have defined it and, more specifically, are a form of social (as opposed to economic) conservatism (Hoffarth, Liaquat, and Jost Reference Hoffarth, Liaquat and Jost2023; Jost, Federico, and Napier Reference Jost, Federico and Napier2009). Because ideology is increasingly understood by psychologists as a product of individually varying cognitive drives in interaction with social context, and because we can therefore expect variation in the extent to which individuals integrate components of ideology into more coherent cognitive structures (Jost Reference Jost2021), we cannot presume that an ideological value like “traditional morality” is a single latent construct shared and experienced identically across individuals. Instead, we expect different elements of traditional morality to be unevenly spread across the population but also to be quite coherently present among some people. Thus, when autocrats appeal to different elements of traditional morality, we would expect not that all people would respond in the same way, but that some people would respond more strongly to some elements than to others. At the same time, people manifesting the greatest coherence among traditional moral values should be especially pronounced in their response to such appeals. Although one must be cognizant of the complex differentiation involved, it is the latter, more general pattern that is of primary interest here.
This study’s choice to focus on traditional moral values is driven by substantive interest: autocrats have been observed increasingly invoking them. This is least surprising in cases like Iran, Saudi Arabia, or Afghanistan where a religious social conservatism was central to the very emergence of their contemporary autocratic systems (Ayoob and Kosebalaban Reference Ayoob and Kosebalaban2008; Bakhash Reference Bakhash1986; Barfield Reference Barfield2022). More puzzling are cases of authoritarian leaders who initially rose to power through other forms of appeal (such as personalism or nationalism) but who later (having already being established politically) added traditional morality to their legitimation strategies. Instances receiving widespread attention include Russia’s Vladimir Putin, Hungary’s Viktor Orban, and Uganda’s Yoweri Museveni (Bluhm and Varga Reference Bluhm and Varga2020; Demidova and Nikolaev Reference Demidova2016; Sanders Reference Sanders2016).Reference Angrist and Pischke 3
We know a lot about traditional moral appeals as deployed by would-be authoritarians in democracies or hybrid regimes, including their ideational foundations (Abromeit Reference Abromeit2017; Brubaker Reference Brubaker2017; Rensmann Reference Rensmann2017; Wodak and Krzyżanowski Reference Wodak and Krzyżanowski2017), the elites and networks who make and spread them (Buzogány and Varga Reference Buzogány and Varga2023; Dimitrova Reference Dimitrova2018; Laruelle and Limonier Reference Laruelle and Limonier2021; Mudde Reference Mudde2007), and why they resonate across many countries’ populations and subpopulations (Berman Reference Berman2021; Gidron and Hall Reference Gidron and Hall2020; Milner Reference Milner2021). Research on democracies and quasi-democracies also indicates that such appeals can be quite powerful, partly or even largely accounting (in combination with appeals to other values) for the electoral rise of illiberal “populist” parties (Bonikowski Reference Bonikowski2017; Noury and Roland Reference Noury and Roland2020) and posing serious threats to democracy itself (Grzymala-Busse Reference Grzymala-Busse2019; Inglehart and Norris Reference Inglehart and Norris2017; Krastev and Holmes Reference Krastev and Holmes2020; Vachudova Reference Vachudova2020).
Far less is known about autocratic contexts, where prior research indicates that connections linking conservative ideology, its cognitive drivers, and political attitudes may differ from what is found in democracies (Hsu and Wang Reference Hsu and Wang2024; Jost, Federico, and Napier Reference Jost, Federico and Napier2009; Pan and Xu Reference Pan and Xu2018). In particular, we lack a firm understanding of whether (or what) autocrats might be gaining politically when they “convert” to traditional morality. These leaders’ initial resort to such appeals, after having risen to power through other means, is commonly explained with reference to some combination of domestic political expediency (e.g., a political crisis requiring a shift in legitimation strategy or the simple perception of a political opportunity) and the influence of transnational conservative networks (Buzogány and Varga Reference Buzogány and Varga2023; Laruelle Reference Laruelle, Kolstø and Blakkisrud2016; Sanders Reference Sanders2016). Accordingly, it is sometimes noted that large portions of the public share traditional moral values in cases where autocrats adopt them (Dimitrova Reference Dimitrova2018; Melville Reference Melville2017; Sanders Reference Sanders2016). What we do not know, however, is the extent (if any) to which articulating this form of social conservatism actually pays off for autocrats when it is deployed and, if so, how.
There are reasons to question how much benefit these autocrats are actually obtaining from traditional moral appeals. Russia, as a substantively important case whose regime has invoked traditional morality even to justify a full-scale invasion of neighboring Ukraine, is useful for identifying the research challenge involved. For one thing, Kolstø and Blakkisrud (Reference Kolstø and Blakkisrud2021) and Laruelle (Reference Laruelle2023) find that Russia does feature the common cross-national pattern of having a grassroots social conservatism that fits with the authoritarian regime’s morally traditionalist rhetoric, but two other things are also true. First, traditional morality has been far from achieving a social consensus, meaning that politicizing these issues could be divisive rather than unifying. Second, there are significant gaps between the traditional moral values many people claim and their actual behavior—for example, on issues of divorce or sexual promiscuity—raising the question whether they would respond positively to politicians who would restrict their personal options. Third, any correlation between traditional morality and support for Russia’s autocrat might be explained by third factors. The Kremlin first mobilized these values only in 2012 as part of a “conservative turn” (Shcherbak Reference Shcherbak2023). This turn arrived long after Putin had already built up a substantial base of popular support through other appeals ranging from personality to perceived competence (Colton and Hale Reference Colton and Hale2009; McAllister and White Reference McAllister and White2008; Rose, Mishler, and Munro Reference Rose, Mishler and Munro2006; Treisman Reference Treisman2011). The turn also coincided with a 2012 presidential campaign to rebrand Putin as “father of the nation,” an effort that relied more on forging identity connections with the public than on traditional morality (Sharafutdinova Reference Sharafutdinova2020). If people who responded to these other appeals also happened to hold traditional moral values for some other reason, this could account for an association between traditional moral values and Putin’s support, without Putin’s invocation of these values having any additional effect on his support at all (Hale Reference Hale, Blakkisrud and Kolstø2025a). Finally, examination of Putin’s rhetoric shows that his talk has not been equally conservative across all aspects of traditional morality. Thus, although he has clearly advocated heteronormativity, he has been found to hold more “Soviet” than reactionary views on women’s proper role in society (Johnson et al. Reference Johnson, Novitskaya, Sperling and Sundstrom2021). The question thus remains open: Are appeals to traditional moral values helping the regime in any way we can actually identify?
Leveraging Traditional Morality through Reform Bundling
This study investigates one specific way in which autocrats have tried to leverage traditional moral values to their political advantage: linking values-based change to pro-authoritarian institutional change so as to enhance the legitimacy of the latter. Although some studies have found that dictators do win support explicitly because of their reputations as “strongmen” (Matovski Reference Matovski2021), blatant efforts to extend or augment their own power can be unpopular (Guriev and Treisman Reference Guriev and Treisman2020; Reuter and Szakonyi Reference Reuter and Szakonyi2021). Autocrats thus have an incentive to “package” self-serving constitutional and other institutional reforms in ways that make them more palatable to the public, often by bundling them with other reforms they believe will have broader support (Baturo and Elgie Reference Baturo and Elgie2018; Dawson and Young Reference Dawson and Young2020; Maboudi, Nadi, and Eisenstadt Reference Maboudi, Nadi and Eisenstadt2023). Such bundling can either lead the citizenry to overlook the power grab (their attention being on the reform bundled with it) or to calculate that the desirable reform outweighs the undesirable one that comes with it. The specific question addressed here, then, is whether traditional moral values are capable of playing such a role (that of the “desirable policy” bundled with the power grab).
By bundling autocratic institutional change with traditional morality, the regime effectively links its legitimation strategy to the mechanics of authoritarianism, with the former invoked to strengthen the latter. This can be done by adding explicitly conservative policy content to the proposed pro-authoritarian institutional reform. Importantly, this tactic does not require that the public believes either (a) that the incumbent sincerely holds traditional moral values or (b) that stronger autocratic rule is needed to defend such values. Instead, the idea is to put the combination of such reforms up as a package, hoping values voters will find the values component intrinsically attractive enough for them to support the package, regardless of their position on (a) or (b). In fact, when values-bundling works as hoped by the autocrat, people can be expected to support autocracy-strengthening institutional reform packages even when they do not support the autocrat personally. In this way, bundling holds out the possibility that authoritarian rulers can expand the base of people mobilized for the purpose of strengthening their regimes beyond their own pool of supporters.
This tactic will only benefit autocrats to the extent that liberals among their preexisting supporters do not react too negatively. More precisely, the gains among morally conservative nonsupporters of the regime must outweigh any losses among nonconservative supporters. The potential for backlash cannot be neglected because the authoritarian leaders addressed here came to power on the strength of something other than a traditional morality platform, meaning they are likely to have both moral conservatives and moral nonconservatives in their coalitions. The values-bundling tactic essentially involves a bet, then, that either nonconservatives in the autocrat’s coalition will be influenced more by their attachment to the autocrat than by any values-based stands of their own or that nonconservatives in the autocrat’s coalition who care deeply about these moral issues will be far fewer in number than conservatives who might be drawn to support the reform package.
Research Design
This study tests for such values-bundling effects by examining one important form of autocratizing institutional change that is often bundled when “sold” to the public: the contravention of constitutional limits on the number of terms a sitting president may serve at a given point in time. Escaping such term limits is widely seen as a step toward increasing the personal power of an incumbent or an aspiring autocrat at the expense of potential constraints on their rule, formally enabling a transition to higher levels of authoritarianism (Baturo Reference Baturo2014; Ginsburg, Melton, and Elkins Reference Ginsburg, Melton and Elkins2011; McKie Reference McKie2019). Term-limit contravention is widely seen as a power grab, and indeed research shows that leaders who accomplish it can greatly extend their tenure in office (Maboudi, Nadi, and Eisenstadt Reference Maboudi, Nadi and Eisenstadt2023). Because autocrats’ authority can be threatened when they show weakness, a bid to contravene term limits is a high-stakes affair, and we can expect authoritarians who undertake such a gambit to invest considerable effort into ensuring its success (Hale, Reference Hale2015). Such leaders have thus frequently buttressed their escape efforts by bundling them with other kinds of reforms, such as additional institutional changes that appear to be technocratic or democratizing and are thus expected to enjoy broader support (Baturo and Elgie Reference Baturo and Elgie2018; Dawson and Young Reference Dawson and Young2020).
Values-based appeals have not frequently been considered in studies on how autocrats attempt to escape term limits. In perhaps the lone exception, Corrales (Reference Corrales2016, 16–17) reports some limited but intriguing evidence that related strategies could be important for defeating term limits cross-nationally, at least for leaders who came to power more or less democratically. In his econometric analysis of data from Latin American presidential attempts to contravene term limits from 1988 to 2016, Corrales finds a correlation between success and “ideological extremism.” He notes, however, that “we cannot know whether the extremism or moderation observed is the cause of success in changing the constitution or its result” (16). The inconclusiveness of such previous work on the role of values in strengthening autocracy helps motivate the present study.
The Case of Putin’s 2020 Soft Contravention of Term Limits
The Russian case offers researchers an important opportunity to explore the utility of appeals to traditional morality for autocrats because these values were explicitly bundled with Putin’s 2020 bid to escape term limits. After winning reelection to a second consecutive term as president in 2018, the Russian leader faced a constitutional term limit that would have barred him from running for the same office again in the next presidential election in 2024 (Chaisty Reference Chaisty, Baturo and Elgie2019). His solution was to change the constitution, with the change resetting his count toward the term limit, a move made previously in, for example, Kyrgyzstan in 1998 and Peru in 2000 (Baturo Reference Baturo, Baturo and Elgie2019, 82). That is, his terms-served count would be set back to “zero” under the logic that the “new” political system brought by the constitutional change should come with a “new” term limit that should not be applied retroactively to include the prior constitutional era. This move thus constituted what Maltz (Reference Maltz2007, 128) calls a “soft contravention” of presidential term limits, involving an extension rather than outright abolition of the limit. Putin sought to legitimize this constitutional change by orchestrating a plebiscite on it.
With more than a majority still backing Putin but facing a gradual decline in support in the lead-up to 2020, and with polls showing substantial public support for term limits, the Russian leader had reason to believe that simply asking voters to approve an extension to his term could be unpopular and hence risky (Chaisty and Whitefield Reference Chaisty and Whitefield2019). Putin’s initial January 15, 2020, announcement that a constitutional plebiscite would be held thus made no mention of resetting his term-limit count. Instead, he framed it as being about the need to make certain, largely technocratic, correctives to Russian basic law, which had been nearly unchanged since the constitution’s adoption in 1993.
Crucially, in making his initial announcement, Putin also called on the broader public to advance proposals for what amendments should be added to the package to be voted on, creating a working group to collect these and help him decide which ones to include. This is the formal process through which both the term-limit contravention and the traditional morality amendments made it into the constitutional reform bill, which ultimately included some 206 constitutional changes. For example, within less than a week, reports emerged that the working group would consider defining marriage exclusively as a union between a man and a woman (Meduza 2020). Public figures echoed this idea, ranging from parliamentary deputy speaker Petr Tolstoy (Viatchanin Reference Viatchanin2020) to Russian Orthodox Church officials (Novosti Reference Novosti2020). Putin himself weighed in on February 13, making clear he wanted an amendment in the package declaring that there would be no “parent number one and parent number two” in Russia, only “papa and mama” (Kommersant 2020b). This amendment was thus included in the bill that Putin personally endorsed and submitted to the parliament on March 2 (Izvestiia 2020), alongside a variety of additional traditional morality declarations, socioeconomic commitments, appeals to nationalism, and other items (Hutcheson and McAllister Reference Hutcheson and McAllister2021). The term-limit plank, however, was not among these others. It appeared only on March 10, when octagenarian parliamentary deputy Valentina Tereshkova, a national hero as the first woman in space, requested the floor during a session on the bill and cited popular demand in making the bombshell proposal (Gamov Reference Gamov2020). Although we cannot assume that legislatures boil down to rubber stamps in autocracies like Russia (Brown et al. Reference Brown, Schaaf, Anabtawi and Waller2024; Noble Reference Noble2020), it is clear here that (at the very least) the Kremlin managed this process in a way that embedded an important pro-authoritarian institutional change in a much broader constitutional amendment package that included strong appeals to traditional morality and that was portrayed as responding to broad public opinion (Hutcheson and McAllister Reference Hutcheson and McAllister2021).
As plebiscite day approached, the regime went to considerable lengths to deemphasize the term-limit reset component and call attention to other elements, including the provisions on traditional morality. For example, mention of the term-limit amendment was initially left out of the informational segment of the Kremlin’s website for electronic voting. Russian journalists noticed this omission, and when asked directly why regime advertising more generally was ignoring the term-limit plank, Putin’s spokesman replied, “There are many amendments there. And it is hardly possible to single out any particular one. They are important in their totality. And you know that in different sources, totally different amendments are mentioned. But there are far too many of them to mention in the setting of an advertisement” (Kommersant 2020a). In other words, the regime chose to emphasize what it wanted to call attention to within the larger whole. And as The Independent reported, “Most official literature concerning the vote has played down the main amendment—the resetting of term limits—and has concentrated instead on populist changes about faith in God and marriage as the union of man and woman” (Carroll Reference Carroll2020). This was widely understood as a Kremlin effort to boost the share of “yes” votes for Putin’s escape from term limits by bundling it with traditional moral (and other) propositions that people could be expected to support in a single up-or-down package (Hutcheson and McAllister, Reference Hutcheson and McAllister2021). The Kremlin ultimately held the plebiscite over a week culminating on July 1, 2020.
By examining the possible effect of conservative values-based appeals on this vote, this study exposes the potential for autocrats to mobilize such values to support institutional changes that can extend or stabilize their rule. Doing so in the Russian context has some additional advantages. For one thing, it is substantively important to understand Russia’s authoritarianism, which is arguably prototypical and influential: not only did its regime produce one of the bloodiest and most internationally destabilizing wars of the post–Cold-War era in 2022 and justify this in part through reference to traditional morality but also, under considerations of duress or expediency elites in other countries have adopted many of the Russian regime’s features (Fish Reference Fish2017; Hall Reference Hall2023; Waller Reference Waller2021). Russia is also frequently identified as an exporter of traditional moral values (though not necessarily exclusively those), adding to the importance of understanding these values’ role in its own political system (Akturk Reference Akturk2019; Stoeckl and Uzlaner Reference Stoeckl and Uzlaner2022).
The Survey
To assess whether this bundling tactic actually worked for the Kremlin, and more generally to establish whether values-based appeals may be facilitating authoritarianism-enabling institutional reforms, this study draws on results from a survey in Russia.Reference Bakhash 5 The data collection effort, conducted by the Russian agency VTsIOM, included telephone interviews of a randomly selected nationally representative sample of 1,500 adult residents of the country, with two additional sets of 600 respondents selected as random representative samples of the population in two Russian regions, during the period April 27–May 6, 2021, less than a year after the plebiscite.Reference Barfield 6 The survey included a large battery of questions on different aspects of traditional morality and related topics.Reference Baturo 7
Method
The study proceeds empirically in two main steps. First, it examines basic relationships among the relevant survey responses. If appeals to traditional morality did indeed yield political dividends in the form of more yes votes in the plebiscite, we should observe a strong correlation between the degree to which an individual shares these values and the propensity to self-report having voted yes. Recall of how one actually voted will, of course, be imperfect nearly a year after the plebiscite took place, but this imperfection should not diminish confidence in the conclusions reached. If recall failure is randomly distributed, this should simply create noise in the data, making it harder rather than easier to detect the predicted patterns and thereby making this a stronger rather than a weaker test. Research on vote recall, however, establishes that error tends to lean in the direction of current preferences (Elsas, Miltenburg, and van der Meer Reference Elsas, Miltenburg and van der Meer2016; Himmelweit, Biberian, and Stockdale Reference Himmelweit, Biberian and Stockdale1978); if this is the case when it comes to the plebiscite, the analysis here would be picking up current propensities to vote for the referendum package. It would thus still constitute a meaningful test of the fundamental proposition at hand, which is that we should see a positive individual-level correlation between support for traditional morality and support for the package of constitutional reforms that bundled values with term-limit contravention.
To establish whether the predicted correlation obtains, the study begins by creating a theory-driven set of indices measuring the extent to which individuals express traditional moral values based on their answers to 13 questions in the nationwide survey. Each index is the individual’s average response to relevant questions on a scale from 0 to 1, where all variables are rescaled such that 0 captures the least conservative position possible on the relevant question and 1 the most conservative. The first index includes all 13 measures and captures the general core political value of traditional morality. Reference Baturo, Baturo and Elgie 8 Because theory expects that the coherence of positions on these questions will vary across individuals, indices are also created for five distinct aspects of traditional morality captured in these 13 questionsReference Baturo and Elgie 9: sexism (a belief that men are better leaders and deserve employment priority), marriage traditionalism (defining marriage as a man plus a woman and thinking families should strictly include a cisgender husband and a cisgender wife), gender intolerance (rejecting nonheterosexuals as neighbors and considering gender minorities to be alien to Russia’s culture), personal conservatism (reserving sex for marriage and never allowing for the possibility of abortion), and anti-secularism (here, support for close relations between the Russian Orthodox Church [ROC] and the state, in particular considering the ROC a fundamental part of Russian culture, wanting schools to teach religion, considering it God’s plan that Russia be successful, and thinking that the ROC should back the state).Reference Baturo, Elgie, Baturo and Elgie 10 The precise wording of the questions used to create these indices and which questions go into which index, can be found in online appendix A.
Because the nationwide survey also included a question capturing whether people voted for the 2020 constitutional reform package discussed earlier, we can use the indices of individual-level traditional moral values in two ways: (1) to assess whether traditional morality generally correlates with the propensity for people to have voted for Putin’s constitutional reform package and (2) to explore which particular dimensions of traditional morality may have mattered most. To perform this analysis, I constructed a binary variable that is coded 1 for people who reported having voted yes in the plebiscite and 0 for all others (see appendix A for details). Weighting to known population parameters, the survey puts the yes vote at 39% of the population and 66% of voters, which is lower than the official results of 53% and 78%, respectively. This is consistent with independent analysis indicating that some fraud was present, but that actually cast yes votes did outnumber no votes (Hutcheson and McAllister Reference Hutcheson and McAllister2021, 374).Reference Berman 11
The study’s second main step addresses the thorny issue of causality, exploring whether adding morally conservative amendments tends to increase the likelihood that people (including nonsupporters) would vote for the constitutional reform package, which included the term-limit contravention. It does so through a preregistered survey experiment added to the two regional surveys that are used in this study. The two regions are Arkhangelsk and St. Petersburg. Because this experiment was embedded in the larger LegitRuss survey project, which had other goals beyond this particular study, the selection of these specific regions for the analysis reported here was essentially arbitrary. Yet, because these northwestern Russian provinces stand out historically for relative liberalism (Orttung Reference Orttung1995), they present a challenging test for the proposition tested here—that appeals to conservatism yield political benefits that are not outweighed by backlash from liberals. Thus, although I cannot claim that the experiment’s results are nationally representative, any deviation from the national norm would likely be in the direction of underestimating rather than overestimating average effects nationwide. A total of 900 respondents from these two regional surveys were randomly selected for participation in the experiment, providing for plenty of statistical power.Reference Bjarnegård and Donno 12 These two regional surveys included almost all the same questions as the national survey, but the experiment in the two regional surveys replaced the national survey’s question on how people voted in the constitutional plebiscite.Reference Blakkisrud and Kolstø 13
The experiment was designed to maximally capture the real-world situation that the regime was putting its citizens in with its traditional morality-bundling strategy, which crucially involved the pronounced downplaying of the term-limit-reset component of the package. The core of the experiment was to remind all respondents about the 2020 plebiscite and then ask them how they would vote today if it were to be held again. The experiment’s 900-respondent sample was randomly divided into three roughly equal parts. The control group received only this core question, asking people how they would vote if the plebiscite were held anew. The treatment group of primary interest was given a mention of the package’s traditional marriage plank as an example of what they voted on. To accurately reflect the regime’s downplaying of the power grab, no specific mention was made of the term-limit-reset plank in this treatment. Instead, the design allowed respondents in this treatment arm to bring to the experiment their real-world level of awareness that the package included term-limit contravention, avoiding the artificial emphasis that an explicit mention would have created. The nationwide survey indicates that it was fairly common knowledge at the time of the experiment that term-limit contravention had been a part of the constitutional reform package: about two-thirds (64%) correctly identified it as included when presented with a short list and asked to say which items were and were not part of the bundle.Reference Bluhm and Varga 14
The difference between this treatment group and the control group, then, should be a reasonable measure of traditional moral values’ potential to add support for a real-world bundled term-limit contravention when values are emphasized and the power grab is downplayed. More specifically, if traditional morality bundling was working for Putin, we would expect the group getting the mention of the conservative amendment to be substantially more likely to report intent to vote yes were the plebiscite to happen again than the control group (not getting a specific example of content at all).
To probe deeper, a third arm of the experiment included a second treatment item intended to establish the impact of explicitly mentioning that the term-limit reset was in the constitutional reform package. In the wording of the analysis plan preregistered with the American Economics Association’s RCT Registry (see online appendix H), strong confirmation would reflect that the traditional moral values treatment “produces a statistically significant increase in the propensity to vote yes on the whole package of agreements compared with the control group…[and] compared with the group primed on the term limit amendment.”
To test whether the expected relationships exist between dependent and independent variables, this study estimates both logit and OLS models and reports both in the online appendix; which type of model was used made no significant difference for the conclusions. Following recent methodologists’ advice, OLS is presented in the main text because its coefficients are more easily interpreted, and it is as good as or superior to logit for estimating average marginal effects (substantive results of interest) with binary dependent variables (Angrist and Pischke Reference Angrist and Pischke2009, 107). Because all independent variables are scaled from 0 to 1, the average marginal effects here reflect the effects on a dependent variable of moving from the lowest to the highest value on a given independent variable in the dataset while holding all other variables at their average values. The term “full effects” is used to efficiently communicate this notion.Reference Bonci, Cavatorta, Baturo and Elgie 15
Results
Are Moral Conservatives More Likely to Report Having Voted for the Reform?
The answer is yes. An OLS model that controls for standard demographics and for approving of Putin’s job performance estimates that the most morally conservative people in the sample are an impressive 53 percentage points more likely to report having voted yes in Putin’s plebiscite than the least conservative. In fact, as is illustrated in figure 1a, traditional morality is roughly as strongly predictive of a yes vote as is support for Putin, with older age having a similarly positive full effect.
Figure 1 Magnitude of the Full Effects of Moving from the Least to the Most Morally Conservative Values on the Likelihood of Voting for Putin’s Constitutional Reform Package in the 2020 Plebiscite
Note: Controlling for demographic variables and support for Putin (OLS with population weights, nationwide survey of Russia, April-May 2021).
Figure 1b breaks down these results by different dimensions of traditional morality, which have distinct relationships with plebiscite voting: anti-secularism and personal conservatism are found to be the strongest correlates of a self-reported yes vote. Online appendix D presents the results in tabular form with different sets of controls and also shows that findings remain roughly the same whether one codes what are sometimes called “nonresponses” (“hard to say” or “refusal”) as missing or as 0 in the binary dependent variable.
Can Values-Bundling Win Votes for Term-Limit Contravention?
We now turn to results of the survey experiment, probing whether linking morally conservative amendments to Putin’s term-limit contravention increased the likelihood Russians would vote for it. Strong evidence is found for such an effect. Table 1 gives the precise wording of the experiment and the distribution of responses in the control group, which was not given any example of any particular amendment included in the package. As can be seen, the regime’s baseline position, as of spring 2021, was under proverbial water: if the plebiscite were repeated, only 39% would vote yes, and a large plurality of 48% would vote no, with another 7% saying they would not vote at all.
Table 1 Distribution of Responses in the Control Group (N = 302)
Note: Experimental design: “The set of constitutional amendments that were voted on last summer included some hundred changes [treatment text read here for treatment groups]. If a similar vote on the same list of amendments this summer were held today, and if you had the choice, would you likely vote for or against?”
One of the two treatment groups in this study received the text “such as marriage should be between a man and a woman” (N = 299), and the other got “such as letting the current president run for the presidency for another two terms” (N = 299). Figure 2 summarizes the results, reporting the average probability of a yes vote in each group with 95% confidence intervals (see appendix C, table OA4, for more detailed results and different models estimated as robustness checks). As can be seen, the mention of the plebiscite’s morally conservative content produced a percentage-point leap in support for the reform package relative to the control group, turning what would be a 39% failing result into something of a landslide with 65% support among the population. This sum is also far greater than that obtained when the term-limit reset was mentioned instead. This latter mention did raise support for the reform package, though to much smaller effect (7 percentage points) and at a weaker (90%) statistical significance level, which itself disappears in certain model specifications reported in appendix E.
Figure 2 Effects of Mentions of the Values and Term-Limit Content of Russia’s 2020 Constitutional Plebiscite on the Likelihood of a Yes Vote (OLS, 95% CI)
Further exploration reveals that, although the term-limit mention primarily mobilized Putin’s own supporters (figure 3b) as one would expect, the values-based appeal mainly brought in support from people who did not intend to vote for him (figure 3a). Autocrats, this evidence indicates, are able to use values-based appeals to win support for pro-authoritarian institutional reform packages even among people who do not support them politically—and to do so in large number. Moreover, the conservative appeal was found not to cost Putin backing from his own supporters: Putin backers respond to the values mention just as strongly positively as they do to the term-limit mention. He thus avoids backlash while mobilizing new people to advance his rule.
Figure 3 Effect of Mentions of the Values and Term-Limit Content of Russia’s 2020 Constitutional Plebiscite on the Likelihood of a Yes Vote, by Putin Support (OLS, 95% CI)
Note: Results in tabular form can be found in appendix C, table OA5.
There is also evidence that voters’ own traditional moral values are driving these effects, which one would expect if the values-based appeals were working as posited. Because the effects of mentioning the values primarily occur among nonsupporters, figure 4 focuses on who among these nonsupporters are the ones moving their position. The movers, it shows, are overwhelmingly the moral conservatives. As panel 4a reports, the most consistent (ideological) moral conservatives who did not support Putin—after getting a mention that the proposed constitutional change package contained a traditional morality plank—became a remarkable 74 percentage points more likely to say they would vote for it. Nonconservatives who did not support Putin are found not to have responded to the values appeal at all. This finding is important, because it means there was no backlash among more liberal nonsupporters that could have potentially offset the gains among conservatives. Of course, this is likely partly because the most liberal elements of Russian society were already strongly in opposition to Putin and so were inclined to vote no. In any case, all told, the appeal to values appears only to have been a plus for Russia’s top autocrat in his term-limit contravention bid.
Figure 4 Effect of Mentions of the Values Content of Russia’s 2020 Constitutional Plebiscite on the Likelihood of a Yes Vote, by Levels of Conservatism, among Putin Nonsupporters Only (OLS, 95% CI)
Note: Results in tabular form can be found in appendix C, table OA7.
Panels b–f in figure 4 report results of breaking down conservative values into their different dimensions. Here we find the primes’ effects are strongly moderated by all dimensions of traditional morality, except personal conservatism (arguably emphasized less by the regime during its “conservative turn”). The values treatment, therefore, strongly appears to have effectively activated in consciousness the more general traditional morality content of the amendments. Indeed, the steeper slope in figure 4a relative to the other panels indicates that people who are consistently conservative across all dimensions of traditional morality tended to respond most powerfully to the values treatment.
Discussion
Overall, these results indicate that it was a successful strategy for Putin to have bundled term-limit contravention with traditional moral values as part of a broader values-based legitimation strategy that his regime had pursued for several years. By playing up the values and downplaying the term-limit escape, he was able to build new support for a legislative package including that contravention. Of course, experiments frequently face challenges of external validity because the process of identifying cause creates a controlled environment that does not match reality. This study thus does not claim to have gone back in time and established that Putin’s traditional morality-bundling strategy actually generated this share of votes in the 2020 plebiscite itself. What is established, however, is that values were capable of making a large difference in support for Putin’s institutional reform in spring 2021, reflecting their impact at that particular time. The new support came primarily from people who did not count among Putin’s supporters but who did support these values. The political gains are found to have been substantial: figure 1 indicates that a mere mention of the amendments’ conservative content generated new support from one-fifth to one-quarter of Russia’s adult population. And this, it is found, would have made the difference between success or failure for a repeat plebiscite having a 50% threshold for success.
Such an important effect is plausible, given the landscape of preferences in Russia. The nationwide survey analyzed here finds that the pool of morally conservative nonsupporters of Putin in spring 2021 included 28% of the adult population, or 60% of all non-Putin supporters and 38% of all moral traditionalists.Reference Bonikowski 16 And figure 4a found that the conservative values mention had an over 70% efficiency rate in generating support for the plebiscite among the most morally conservative people in the sample who did not already support Putin.
What about the other end of the traditional morality spectrum, Russia’s most secular-liberal individuals? Figure 1a shows they were far less likely to report having voted for Putin’s constitutional amendment package than were the most conservative people. However, the survey experiment indicates their nonsupport did not hinge on Putin’s values-based legitimation strategy. Putin supporters, be they morally liberal or conservative, mostly tended to support the president’s initiative anyway, without having to be influenced by the values included in the bundle. And Putin’s most morally nonconservative nonsupporters, as shown in figure 4a, were generally not moved at all by the mention of the plebiscite’s values-oriented content. They appear to have been lost to Putin long ago, essentially unwinnable to authoritarian-strengthening changes he might propose. In short, the authoritarian gains we find in institutional reform support reflect not only new backing by traditional moralists but also the absence of any noticeable backlash from nonconservatives.
The major overall values-mention effect differs from what happens when we simply mention the term-limit content of the plebiscite. At the aggregate level, the latter does not clear a 95% statistical significance standard. Even if one were to ignore this significance threshold, any effect appears to have involved the mobilization of only those few Putin supporters who did not otherwise support the constitutional reform package.
Although I cannot rule out that some respondents were not replying sincerely to the survey’s questions in Russia’s authoritarian context, the most likely bias would make the results underestimates, rather than overestimates. If people are systematically overreporting their willingness to support Putin’s amendment package, this leaves less scope for the traditional morality prime to increase support for the amendments. That said, recall that supporters of the amendments in the national survey and the experiment’s control group were a clear minority (only 39%), and that mentioning the extension of Putin’s rule included in the package did not result in a pronounced increase in professed support. This suggests levels of dissembling were not high.
One might also wonder whether respondents, given a mention of the plebiscite’s traditional marriage component, might simply have been reporting their level of support for traditional marriage instead of for the whole reform package. This is unlikely. The question is clear in asking about the reform package as a whole, and as noted earlier, most respondents at baseline were aware of items other than traditional marriage that it contained. If the support they cited pertained only to traditional marriage, then the aggregate levels of this support should be the same as the share of people who support traditional marriage when asked directly. But this is not the case: the share of reform supporters in the experiment is 65%, which is significantly lower than the share backing traditional marriage in a direct question (82%), as we would expect for a package that bundled the traditional marriage plank with a less popular initiative like term-limit contravention.
Conclusion
This study’s most important contribution is to provide heretofore elusive evidence that values-based legitimation strategies can, in at least some instances, have significant effects on autocrats’ efforts to buttress their own regimes. By going beyond correlation to establish at least some element of cause through a survey experiment and associated econometric analysis, the study makes a prima facie case that more research should be devoted to the roles values can or do play in autocratic regimes. It further indicates that the effects can be large, in this case potentially making the difference between clear minority and majority support for an initiative the autocrat considered important. More specifically, the study demonstrates a role for values in autocrats’ efforts to contravene term limits, a phenomenon comparative researchers have more typically explained through either the leader’s popularity or raw political power, including the regime’s manipulative or repressive apparatus (Baturo and Elgie Reference Baturo, Elgie, Baturo and Elgie2019; Cheeseman Reference Cheeseman, Alexander and Robert2019; Corrales Reference Corrales2016). Because the term-limit contravention studied here involved changes in formal law, these findings contribute to a growing literature establishing that modern authoritarian policy making cannot simply be boiled down to a rubber-stamp process (Gandhi, Noble, and Svolik Reference Gandhi, Noble and Svolik2020); it is shown here that values are an important tool that autocrats sometimes leverage to shape and build legitimacy for desired legislation and institutional reforms through a process that can be less predictable than often assumed.
In so doing, the study additionally advances our understanding of interrelationships between the mechanics of authoritarian rule and autocratic legitimation strategies. Rather than being two distinct sources of regime stability that can be studied separately, it is shown here that they can be powerfully linked. In contrast with informational autocracy theory, which emphasizes how regime mechanics can influence legitimation by manipulating the information available to subjects (Guriev and Treisman Reference Guriev and Treisman2019), this study shows that the causal arrow can also point in the other direction: legitimation strategies can facilitate refinement of the mechanics of authoritarian rule. And although the same informational autocracy theory also downplays any need for ideology, focusing instead on autocrats’ appeals to “competence,” this study shows that invoking ideological values can provide critical leverage for autocrats intent on institutionally buttressing their rule—and in a paradigmatic case for informational autocracy theory (Russia). Indeed, arguably unlike demonstrations of competence, values-based appeals are shown capable of helping autocrats adopt desired institutional change without strong constraints from the limits of their personal support base.
This study in no way downplays the desirability for autocrats of appearing competent, of course. Instead, values-based legitimation is most likely a powerful alternative strategy that autocrats can apply either at the outset of their tenure or later in their rule when they fear informational efforts to bolster competent reputations are not fully reliable. That is, it can constitute a form of ideational legitimation that Gerschewski (Reference Gerschewski2023) posits can sustain an autocracy in the absence of performance legitimation. Thus Putin resorted to it initially during the regime’s legitimacy crisis of 2011–12 and felt the need to invoke it to ensure the 2020 plebiscite’s passage after an unpopular 2018 pension reform had damaged the Kremlin’s public standing.
The study’s limitations point to important avenues for future research. One is that it focuses on only a single core political value, traditional morality, which Putin has increasingly embraced (Blakkisrud and Kolstø Reference Blakkisrud and Kolstø2025). To be sure, other autocrats have embraced very similar themes, including Hungary’s Orban and Uganda’s Museveni, assisted by transnational networks dedicated to the spread of precisely such values (Buzogány and Varga Reference Buzogány and Varga2023; Laruelle Reference Laruelle, Kolstø and Blakkisrud2016; Sanders Reference Sanders2016). That said, other autocrats have at times opted to mobilize certain progressive values, even including women’s rights (Bjarnegård and Donno Reference Bjarnegård and Donno2024; Tripp Reference Tripp2019). It will be important, therefore, for future studies to explore whether other values—including leftist or theocratic ideology—are similarly being leveraged in some cases to secure support for autocratic initiatives. Future work could also productively explore the extent to which autocrats seek to mobilize other elements of conservative ideology beyond traditional morality, such as economic conservatism, deepening a body of research that finds variation in the extent to which different elements of conservatism hang together in influencing behavior (Hsu and Wang, Reference Hsu and Wang2024; Wollast et al. Reference Wollast, Phillips, Yahiiaiev, Malysheva, Klein and Sengupta2025).
The study also examines only one tactic that autocrats might use to link values to institutional change: embedding pro-authoritarian institutional change in a vote that includes many other propositions believed to be popular. Such values-based bundling tactics are not unique to the case examined here. For example, autocrats in Tajikistan in 2016 and Tunisia in 2002 bundled their escapes from term limits with planks enshrining certain principles of secularism and human rights, respectively, in their constitutions to win package approval in referenda (Bonci and Cavatorta Reference Bonci, Cavatorta, Baturo and Elgie2019; DeutscheWelle 2016). But researchers would do well to investigate other ways in which autocrats link values to desired reforms beyond explicit and single-package bundling; for example, by holding multiple referendum questions simultaneously (as in Azerbaijan in 2009) or more simply by justifying desired changes as necessary for promoting or protecting the values in question (as in Egypt in 2019).
Another limitation is the study’s focus on an autocrat who embarked on values legitimation more than a decade after having established strong popular support on other bases. Although such cases are of course important and include influential leaders like Orban and Museveni, it will be interesting to investigate whether traditional morality legitimation might be similarly effective for autocrats who initially came to power through morally traditionalist appeals or for leaders in democracies seeking to undermine democratic institutions. Regarding the first possibility, autocrats long identified closely with conservatism would likely not be positioned to gain additional support by continuing to make the same appeals, although doing so should enable them to retain previous supporters. The case of Donald Trump in the United States, however, is suggestive of the second possibility: Christian evangelicals first joined the illiberal Trump coalition—despite major doubts about his personal adherence to traditional morality—because his appeals to their agenda convinced them it would be advanced with his presidency (through, among other things, appointments to the Supreme Court). Examining other cases would also enable us to explore more deeply what precisely mediates and moderates the relationship between values and support for authoritarian actions, including the possibility suggested by psychological studies of ideology that more unstable political contexts might elevate cognitive drives for uncertainty and threat reduction and hence enhance the power of authoritarians’ moral appeals (Jost et al. Reference Jost, Glaser, Kruglanski and Sulloway2003).
Finally, it is worth considering other purposes for which some form of values-based bundling or association tactics could be deployed. Here, we have shown this tactic’s effectiveness in generating support for only one type of pro-authoritarian change, a term-limit-contravention amendment to a country’s constitution. But surely further research would turn up authoritarian attempts to use traditional moral values to legitimate other actions they might take.
Putin (Reference Putin2022) supplies perhaps the most substantively important recent example, adding the following claim to all the others he made to justify his country’s deadly full-scale invasion of Ukraine on announcing it on February 24, 2022: “The United States and other Western partners… sought to destroy our traditional values and force on us their false values that would erode us, our people from within, the attitudes they have been aggressively imposing on their countries, attitudes that are directly leading to degradation and degeneration, because they are contrary to human nature.”
If autocrats like Putin believe such appeals could help them literally get away with murder, surely we should be studying them intently. Indeed, if the results of this study are indicative of broader patterns, advocates of democracy and even international peace will need to find ways not only to thwart autocrats’ political machinery but also to counter the values-based appeals that many of them seem increasingly inclined to make worldwide.
Supplementary material
To view supplementary material for this article, please visit http://doi.org/10.1017/S1537592725102132.
Acknowledgments
For feedback on earlier drafts of this article, the author is grateful to Pål Kolstø, Peter Rutland, Lisa Sundstrom, and other participants in a workshop co-organized by PONARS and Tbilisi State University; a panel discussion at the 2023 Annual Meeting of the British Association for Slavic and East European Studies (BASEES); and a panel at the 2024 Annual Convention of the Association for the Study of Nationalities (ASN).
Data replication
Data replication sets are available in Harvard Dataverse at: https://doi.org/10.7910/DVN/ROCHZJ.
Open access
