Despite their long-held reputation as controlled affairs, autocratic elections continue to surprise. Recent years have seen turnovers in power or unexpected opposition gains in such far-flung places as Guinea-Bissau, Venezuela, Malaysia, and Bhutan. Of course, in many other dictatorships, ruling parties resoundingly win their elections and only increase their aura of dominance. Yet even when turnover is unlikely, elections can loom over autocratic politics like a squall line on the horizon. For instance, many observers claim that Putin’s planned reelection in 2024 makes him less willing to accept defeat in Ukraine. How do we make sense of the inner workings and surprising outcomes of these elections?

The three books under review here tackle this question with a nuance and attention to context we need to see more of in the autocracy literature. To briefly summarize a quarter-century or so of research on autocratic institutions, we can distinguish a handful of phases. The first phase, including work by Barbara Geddes and Jennifer Gandhi, drew attention to the existence of institutions like legislatures and elections in autocracies and claimed there were real politics occurring within them. A second phase, largely functionalist in spirit, argued that these institutions bolstered regime power to explain why autocrats adopted them. Elections, for instance, served to project regime invincibility, to provide a regular occasion for clientelism, and to gather information on citizen discontents. A third phase countered that these institutions often presented significant challenges to autocrats, with scholars like Staffan Lindberg and Jason Brownlee outlining a process of democratization through autocratic elections.

So which is it? Do autocratic elections bolster regime survival or undermine it? The lesson one can draw from these three books is, well, it depends. As such, they are emblematic of a new current of research arguing that autocratic institutions like elections are neither generally stabilizing nor generally challenging. Rather, their effects vary based on the political environment and subtle elements like citizen perceptions. In essence, we can see these books as foundations of a fourth phase that accepts this complexity and tries to make sense of it.

Elvin Ong’s Opposing Power: Building Opposition Alliances in Electoral Autocracies confronts a seemingly straightforward but overlooked question: when do opposition parties in autocracies set aside their differences and form electoral coalitions? It has long been known that opposition coalitions are highly effective at toppling dictators at the ballot box. Inspired by dramatic successes in elections in Senegal in 2001 and Kenya in 2002, scholars like Marc Howard and Philip Roessler (“Liberalizing Electoral Outcomes in Competitive Authoritarian Regimes,” American Journal of Political Science, 50(2), 2006), Daniela Donno (“Elections and Democratization in Authoritarian Regimes,” American Journal of Political Science, 57(3), 2013), and Michael Wahman (“Opposition Coalitions and Democratization by Election,” Government & Opposition, 48(1), 2013) found that opposition coalitions predict incumbent defeat and democratization. The real puzzle that Ong confronts is why opposition parties do not do this in every election as they are generally free to do so and it represents the opposition’s best chance of victory.

Ong provides a thorough explanation of coalition formation, with a close attention to the motives, calculations, and sometimes the hubris of party leaders. Showing the importance of context, it matters critically in Ong’s account who these leaders are, what the parties represent, and whether party supporters will embrace these often nebulous coalitions. The evidence eschews large-N testing to focus on extended case studies of the Philippines, South Korea, Malaysia, and Singapore.

The first step in Ong’s explanation is to note that there are two distinct types of opposition coordination. The first keeps the parties separate but divvies up the districts in which each party will compete, with the intention of avoiding splitting votes across the opposition. The second, and what most scholars mean by “electoral coalitions,” involves opposition parties forming an explicit pact and usually presenting a unified label on the ballot. However, the coalitions typically do not involve the parties entirely surrendering their identities or organization. Rather, the agreements often stipulate a division of power (such as cabinet seats) across the parties conditional on victory, as well as some general principles of reform.

To explain these unified coalitions, Ong presents a persuasive theory: opposition parties coalesce when signs of regime weakness convince them that a coalition can win and no party thinks they can do it on their own (see fig. 1.1 on p. 24 for a summary). A supplementary requirement is that parties need to believe their supporters will cooperate. If the parties do not believe the regime is vulnerable, the parties can instead coordinate on who competes in which electoral district. However, this depends on the electoral system (much easier if there are single-seat districts) and sufficient knowledge of the parties’ relative strengths to allow for a mutually agreeable division of seats.

Thus, much of the theory rests on the perceptions of party leaders of the regime’s weaknesses, their own electoral strengths, and their supporters. Theoretically dealing with perceptions can be tricky due to individual biases, tendencies toward leader overconfidence, and the like, especially in autocratic environments with limitations on free media and reliable voter surveys. However, Ong does an admirable job showing how these perceptions are often grounded in observable events. Specifically, Ong explains how belief in regime weakness is reinforced by a close succession of “alarming, public, regime-debilitating events” (p. 45) like economic crises, internal splits, corruption scandals, and large protest movements. Beliefs in the relative strength of parties are often based on past election outcomes and the size of rallies, although he catalogues many examples of parties overestimating their appeal.

A significant feature of Ong’s theory is filling out the costs of forming coalitions, which has tended to be passed over in prior work. Most obviously, forming a coalition means sacrificing some control of your party’s policy goals, branding, and identity. If other members of the coalition are especially objectionable to your supporters, you risk losing their loyalty in the future. For instance, Ong discusses how the collaboration of Malaysian parties with the Islamist party PAS damaged the trust of their supporters and led to their long-term decline. These costs explain why parties choose not to coalesce unless victory is within reach, and present an intriguing dynamic where party fates are shaped by past coalition choices.

For evidence, the book relies on paired comparisons of the Philippines versus South Korea and Malaysia versus Singapore. As Ong explains, the opposition in the Philippines in 1986 coalesced around an ultimately successful coalition (UNIDO) supporting the presidential candidacy of Corazon Aquino, who abandoned her assassinated husband’s party label (PDP-Laban). The battle to form this electoral coalition was fraught and decided only at the last minute, with the key point of contention being whether Aquino or UNIDO’s leader Salvador Laurel would run for president. In stark contrast, South Korea in 1987 involved a previously unified protest movement that split between its two most prominent leaders, Kim Dae-jung and Kim Young-sam. Both ended up running for the presidency, dividing the opposition and letting the regime candidate win with barely a third of the vote. As Ong persuasively argues, the difference hinged on how key actors anticipated their strength in coming elections. In the Philippines, Laurel accepted his dependence on Aquino for victory, whereas each of the Kims incorrectly believed they could win without opposition coordination.

The second paired comparison examines a longer span of history than these two pivotal elections. In Malaysia, opposition parties repeatedly agreed on unified-label elections, continually failing to make much of an advance until 2018. In that year, a broad coalition succeeded in ousting the ruling party coalition after about 60 years, although only under the banner of a previous dictator, Mahathir Mohamad. In contrast, Singapore has never seen a comprehensive opposition coalition, with opposition parties instead dividing up electoral districts. Again, Ong’s explanation is persuasive and is grounded in the differing perceptions of regime vulnerability. Malaysia’s regime has faced numerous crises, most significantly in the run-up to 2018 with the 1MDB corruption scandal and a major wave of protests. Singapore has instead seen overwhelming dominance by its ruling party, almost entirely avoiding economic crises, corruption scandals, and major splits within the regime. With little perceived chance of success, Singapore’s opposition has chosen to focus on building individual party brands, with each competing in a small number of electoral districts.

Overall, I found the case studies to be convincing and compelling. The discussion is noteworthy for taking these opposition parties’ calculations and agency seriously, correcting for an overwhelming focus in past work on ruling parties. The cases also weigh against a natural counterargument to Ong’s theory, namely that coalitions fail to form based on the incompatibility in aims among them. This could have been filled out a bit more as an alternative in the theory section, but we see strong counterevidence within the cases. We find coalitions with very high polarization (between the PAS and non-Malay parties in Malaysia) and a failure to coalesce where policy differences were slim (such as in South Korea and Singapore).

For some, a potential limitation of Ong’s book will be the lack of a large-N, cross-country analysis testing the main claims. A reasonable defense for this choice is that the fundamental causes are about perceptions, which are difficult to measure across hundreds of cases. However, future work could test reasonable proxies for these perceptions. As Ong argues, belief in regime weakness often follows measurable events like economic crises, protest waves, and corruption scandals. Moreover, parties’ beliefs in possible victory and their dependence on one another could be proxied by past election results. These proxies could be tested against the role of polarization among the most prominent parties.

Sarah Zukerman Daly’s Violent Victors: Why Bloodstained Parties Win Postwar Elections shifts more of the focus to incumbent victors in war rather than opponents, but the context of imperfect elections remains. To be clear, the book mostly concerns postwar democratic elections, but for most of the cases democracy is marginal and the elections represent a transition point from autocracy or state collapse. Thus, the bulk of the electoral campaigning occurs under fairly autocratic circumstances.

Daly’s book articulates a fascinating and novel puzzle: why do parties associated with the victors of civil war usually win postwar elections despite committing widespread atrocities? Further, why do communities that were most seriously victimized by this violence nevertheless give substantial support to their oppressors?

Daly’s answer is deceptively simple: voters in postwar environments desire security above all else. They want peace to hold, crime and further violence not to run rampant, and the economy to rebound. By achieving military victory, incumbents can take credit for ending the war and present the most credible claim to maintaining stability. In other words, they have an immense “valence” advantage on the most critical issue for voters, leaving them in a dominant position in postwar elections.

Left at that, Daly would have a testable prediction, but the book goes much further in filling out the implications for campaign strategies of war victors, losers, and other parties. The victors, parties associated with the incumbent government or military that won the war, need to credibly signal that they can maintain peace and yet will not abuse their power. This is a delicate balancing act, representing the ability to coerce future violent opponents and an unwillingness to use that coercion against nonviolent citizens. As a result, these parties are often fronted by civilians unrelated to the worst atrocities and work to position themselves as “restrained Leviathans” (pp. 78–88). Further, their first-mover status lets them claim the ideological middle ground, relying on security valence to win. War losers, who are typically rebel groups allowed to transform into political parties, are instead blamed for the war’s violence and forced to extreme positions to maintain their following. Bystander parties who did not participate in the war can position themselves as promoting security through the rule of law, but Daly argues that this is less credible and compelling than the victors’ appeals.

One of the greatest strengths of the book is a thorough consideration of alternative explanations. Among the alternatives, victory might stem from a general incumbent advantage, use of coercion in the election, or control of state institutions. Again and again when discussing evidence, Daly returns to these alternatives and weighs them against her theory. Personally, control of institutions strikes me as the trickiest alternative to distinguish, but one could argue that this is part of the security advantage. I was also left unclear if voters’ security preference for victors was about seeing the party as more capable or because they believed the party would resort to violence if defeated. After all, dominant military victors should have the capacity to overturn election losses.

Daly’s evidence is the most varied of the three books. A series of survey experiments in Colombia shows that hypothetical candidates are preferred more when they served in the victorious army and had success maintaining security, even when they were explicitly associated with killings. The three case studies cover distinct war outcomes across Central America: military stalemate in El Salvador, government victory in Guatemala, and rebel victory in Nicaragua. The case analysis really shines. Daly covers the varying parties involved down to their campaign slogans, party manifestos, and internal meetings. Consistently, she finds victorious parties emphasizing security. For instance, Guatemala’s FRG, the party founded by the victorious military strongman Efraín Ríos Montt, campaigned with a logo of a hand holding up three fingers representing “Security, Welfare, Justice” (p. 167).

Finally, Daly conducts large-N testing of postwar elections across the globe from 1970 to 2015. Overwhelmingly, voters tended to support the militarily victorious side, with essentially no role played by the level or source of wartime atrocities (pp. 219–25). Further, an analysis of survey evidence from the World Values Survey confirms that voters who value security as the most salient issue tend to vote for military victors in even greater numbers (pp. 229–31). The analysis is well conducted and complements the qualitative work. However, the sample is no longer limited to regimes that are even debatably democratic, meaning that many of these elections were completely controlled by the incumbent victor. Results are shown to hold for a subset of countries with “cleaner elections” (p. 224), but this will still include some autocracies.

The subject I most wanted to hear more about was why these parties allowed free competition in the first place. Why not take advantage of their coercive powers to lock in electoral victory for as long as possible? This would fit with the recent work of Steven Levitsky and Lucan Way (Revolution and Dictatorship: The Violent Origins of Durable Authoritarianism, 2022), who find that successful revolutionary parties produce highly durable autocracies. Two answers seem most plausible. First, democratic elections (and the transformation of losers into political parties) are often a component of peace settlements. Second, Daly’s findings reduce the need to compromise elections as victors should be confident of winning anyway. The evidence in Daly’s case studies certainly suggests highly confident parties.

Masaaki Higashijima’s Dictator’s Dilemma at the Ballot Box: Electoral Manipulation, Economic Maneuvering, and Political Order in Autocracies aims for the most general theory of the three books. The puzzle driving the book is why autocrats pursue radically different approaches to controlling elections. Why do some rely on clientelism and local public goods while others resort to violence and ballot stuffing? Moreover, what are the consequences of these choices?

Higashijima’s theory divides dictators’ attempts to control elections between two broad strategies. First, “electoral manipulation” refers to all the “coercive and institutional measures that favor the manipulator by making election results deviate from the aggregate of citizens’ vote preferences” (p. 47). This encompasses coercion, ballot stuffing, and strategic choices over electoral rules. Second, “economic maneuvering” involves dictators influencing voters through economic incentives like patronage, clientelism, social spending, and local public goods. The core idea is that each strategy has distinct costs and benefits, leading dictators to strategically determine which one they rely on more. Most significantly, electoral manipulation erodes many of the benefits of holding elections by damaging legitimacy and limiting the information gained from the outcomes.

To explain why dictators opt for each strategy, Higashijima argues that dictators need certain resource and organizational advantages to make the economic route work. Specifically, he predicts that dictators with natural resource revenues will have sufficient wealth to buy people off. Further, dictatorships with strong ruling parties and dominant ethnic groups will have the organizational means to economically co-opt sufficient numbers of citizens. In contrast, dictators without these advantages or who face strong oppositions have to resort to electoral manipulation.

Extending the theory further, Higashijima argues that these choices determine postelection stability. When dictators rely too much on electoral manipulation, this provokes outraged citizen protest in response. When dictators under-manipulate and allow public signals of their weakness, they become prey to coups.

For evidence, Higashijima relies the most of the three authors on large-N empirical testing. Chapter 3 confirms that oil wealth predicts less blatant electoral fraud, especially when combined with a ruling party or dominant ethnic group. This uses a measure of fraud that only extends to 2004 and it would have been advantageous to see a more contemporary measure used as a supplement. Chapter 4 finds that oil revenues, especially when combined with ruling parties, predict less reliance on plurality-rule electoral systems and more on proportional representation. Chapter 5 examines economic strategies, demonstrating that regimes holding elections with opposition parties and with less blatant fraud are more likely to run election-year deficits. It was unclear why this chapter did not test the effect of oil revenues, parties, and ethnic groups given that these prior explanatory variables were justified as being most proximate to economic maneuvering.

Chapter 6 then shows that the gap in fraud—the difference between actual fraud and what the model expects—predicts coups and protests according to the theory. This is an interesting approach, but I was not convinced that it was the gap that should matter here. If citizens are outraged by high levels of fraud, it should be the overall level of fraud that predicts postelection protest and not the excess predicted by the model. Similarly, if coups stem from signs of leader unpopularity, then the election outcomes themselves should matter most. The large-N work is supplemented by paired case studies of Kazakhstan, a stable regime reliant on co-optation fueled by natural resources, and Kyrgyzstan, which has experienced heavy electoral manipulation and regime collapse.

Higashijima offers a very ambitious and comprehensive theory here, covering electoral conduct, electoral rules, fiscal behavior, and leader turnover in the aftermath. One issue I have is why some methods of electoral control were bundled together and some were largely absent. The combination of blatant fraud and electoral rule choice as “electoral manipulation” in the theory section seems tenuous, so I was glad to see them tested separately. Although many dictators certainly determine election rules strategically, this is unlikely to be perceived by citizens as similar to ballot stuffing or coercion, nor does it degrade information to the same degree. I also felt an important category of regime dominance over the media, sources of campaign funding, and state institutions was overlooked as this does not necessarily correspond with blatant fraud. Singapore is a good example of an electoral autocracy that leverages its dominance to win elections without outright fraud or vote buying.

Taking these books together, we can find several significant lessons for observers of autocracies and postconflict democracies. As noted above, a common throughline is that the conduct of elections and the implications for regimes vary dramatically by context. As Ong shows, opposition cohesion (a major predictor of regime defeat) depends on recent regime performance and the interrelations and beliefs of individual party leaders. Daly finds that election outcomes hinge on the balance of military power and citizen perceptions of who can maintain order. Higashijima argues that dictators’ conduct in elections and their chances of surviving hinge on economic and organizational resources.

As a running theme, we can also see the pressures of electoral competition as a central pivot point in these countries’ politics, even where they are fairly well controlled. Competitive pressures yield specific electoral coalitions in Ong, distinct political platforms and party leadership choices in Daly, and sharply varying strategies of electoral control and risks of turnover in Higashijima. This effectively flips a common view that the character of autocratic elections is largely downstream from power.

Another significant contribution of these books is the emphasis on taking the perceptions of citizens and opposition leaders seriously. Ong’s theory largely revolves around perceptions of regime strength and relative power among opposition leaders, as well as how citizens react to their parties forming coalitions with rivals. Daly’s account revolves around citizen preferences for security and their perceptions of who can deliver it. Higashijima’s theory also hinges on citizen and rival elite perceptions of fraud levels, regime popularity, and economic benefits. Besides seeing a role for citizen beliefs, a consistent observation is that controlling these perceptions is not easy. People are not blind receptacles of regime propaganda nor are they laser focused on the more “democratic” outcome at all costs. They want security, jobs, and favorable policies and they calculate, sometimes with great shrewdness, who can provide them.

The books are also commendable for examining the aftermaths of the major events they analyze. For instance, Daly has a short chapter on the policy consequences of electing war-victor parties. Using a regression discontinuity design, she finds that mayors associated with victorious paramilitaries reduce crime rates but have worse outcomes in education and health. Ong provides some interesting analysis of how opposition coalitions often fray after victory, perhaps most dramatically in Malaysia in 2020. Finally, Higashijima predicts the occurrences of coups and protest movements after elections.

It is worth noting, especially for graduate students and young scholars, that all three books showcase the convergence of style and organization seen in a lot of current comparative work at top academic presses. The research is driven by a clear, compelling puzzle. An introduction outlines the argument, the evidence, and the importance of the subject, with no dramatic twists or secret explanations left to future chapters. A theory chapter deepens the analysis, but without getting bogged down in literature review. A large-N, cross-country analysis may be present, but is not sold as the sole (or even primary) source of evidence. Instead, each book heavily relies on carefully considered, paired comparisons of chapter-length cases within the same region. This carries the most weight in Ong’s book, but the emphasis on the virtues of case analysis is clear across all three. Scholars of autocracy can learn a lot from reading these books about relatively overlooked cases (like Higashijima’s account of Kazakhstan and Daly’s account of Guatemala’s 1996 election) and under-studied features of prominent cases (like Ong’s account of coalition formation in Malaysia and Singapore).

Much more remains to be done in scholarship on autocracy. For starters, the insights of each of these books can cross over to inform one another. For instance, Ong’s theory explains how military nonwinners can coalesce to defeat winners, as occurred in Nicaragua in 1984. Daly’s theory on security valence and the organizational advantages of war winners can add other explanatory variables to Higashijima’s explanation of autocratic election strategies.

I will stress again that these books reinforce the need to better understand citizen beliefs and behavior in autocracies and postwar environments. What do they care about when they vote? How do they learn about the ruling party and opposition parties in controlled media environments? How do they make sense of what opposition parties are doing when victory is implausible? We get a great deal of information on these questions from these books, but a comprehensive research program is needed to make further progress.

Lastly, an ever-present question for scholars is how our theories and understandings of these regimes carry to the present day and how they will do so in the future. Autocrats learn, as do their opponents. Technology changes. The three books in this review cover a wide scope of history, although among the main case studies, only Ong’s discussion of Malaysia focuses on events within the last 15 years. How these theories and others will travel to the next generation of autocracy and conflict remains to be seen.