The marked increase in the number of countries that started to hold multiparty elections, at least de jure, in the wake of the end of the Cold War is a well-known fact by now. Many have described the extension of the third wave into the former Soviet Union, Eastern Europe, Africa, and Asia, followed by the temporary Arab Spring. A long and still ongoing debate on the role of multiparty elections as either drivers of democratization or, alternatively, means of authoritarian stabilization has engaged many prominent analysts.

In this context, Thomas Edward Flores and Ifran Nooruddin’s Elections in Hard Times suggests “a new theory of elections that explains why elections advance democratization in some times and places but not others” (p. 13). The gist of their theory is that while winners of clean elections are conferred with some amount of contingent legitimacy, it remains short-lived. What matters is really the inherited stock of performance legitimacy (greater socioeconomic development), and the supply of democratic-institutional legitimacy. The latter refers to the depth of entrenched support for democratic norms and institutions in a society. Leaders who win elections and find themselves with paltry supply of these two types of legitimacy are—ceteris paribus—reaping a much smaller dividend from continued democratic elections than are rulers with ample sources of legitimacy. The theory also predicts that the former are much more likely to engage in activities undermining democracy than the latter. This implies what the authors call “low-legitimacy traps,” since undermining democratic institutions is more appealing in the short run for insecure rulers than building hefty stocks of performance- and/or democratic-institutional legitimacy (p. 47).

The empirical analyses are restricted to elections in “developing democracies” (p. 85) and consist of 144 nations between 1946 and 2010, making a total of 1,755 elections. The analyses take as the point of departure a calculation of the “democratic dividend” of elections, which is the change in Polity2 score from the year of the election to two and five years into the future, respectively. By this method, the only period when elections generated substantive dividends was in the 1980s (p. 11), leading the authors to ask: “Why did elections lose their democratizing power soon after the beginning of the electoral boom, during which they spread globally and improved in quality?” (p. 13). This is where their theory enters, and the rest of the book is dedicated to demonstrating its usefulness by means of extensive testing through time-series, cross-sectional analyses with fixed or random effects.

The authors’ main finding is provided in Chapter 3: The 1990s meant that multiparty elections spread to younger, poorer, and more ethnically heterogeneous countries. Accordingly, the argument goes, these younger countries had less democratic-institutional legitimacy, and were also starved of performance legitimacy and badly placed to build stocks of either in a realistic time frame. Therefore, elected leaders in these countries, found mainly in Africa and the former East bloc, have reverted to patrimonial politics and oppression in order to stay in power.

Clean elections, or high electoral integrity as some would prefer to put it, matters but only in the very short term, according to the analyses in Elections in Hard Times. In particular, the so-called founding elections matter and carry a democratic dividend. In fact, according to the regressions used for results presented in Chapter 4, only founding elections produce positive effects on the level of democracy two or five years later, while subsequent polls have no or even negative effects. In short, the authors agree with one claim in the literature, namely, that elections have become a mode of transition. They find no evidence, however, that holding repeated elections has any long-term positive effect on democratization (p. 94).

Instead, Flores and Nooruddin find that a stock of democratic-institutional legitimacy helps countries hold better elections in the future and “also improve political and civil rights” (p. 97). While they present the concept of democratic-institutional legitimacy as a complex one that includes all norms, rules, practices, and organizations collectively defining democracy at the national level, the operationalization in the book is relatively simple: a natural logarithm of the number of years a country has been democratic since independence according to the Polity2 measure. The authors also include an alternative but similar measure that takes into account a decay rate during nondemocratic spells (p. 98). Regression analyses demonstrate that both measures are correlated at statistically significant levels with holding multiparty elections at a later point in time that, in turn, generates a democratic dividend. Drilling down into the components of the Polity measure, it is particularly the indicator of executive constraints that drives this relationship. The authors interpret these findings to mean that leaders in countries with longer democratic histories find it much harder to oppress the opposition and use other means of subversion than do winners of elections in less experienced nations. It is a trap, but one that can be slowly dissolved, according to the authors, as countries over time gain greater experience with democracy and thus improve their probabilities of democratic dividends.

The second main finding is found in Chapter 6 and speaks to the argument that low performance legitimacy is a key factor explaining why elections have democratic dividends, or not. The analyses in this chapter show that leaders ushered in by multiparty elections who inherit a greater fiscal space, which makes it easier for them to provide public goods, are more likely to oversee an improvement of democracy. Again, the authors find evidence of a trap where governments coming into power with little fiscal space run a great risk of descending into patrimonialism and shenanigans to win the next election and thus stay in power. In Chapter 7, they find a similar logic with regard to violent civil conflict that is found to be inherently bad for the prospects of elections leading to an improvement of democracy (especially low-level conflicts), and this creates another trap for poor, conflict-ridden countries. Finally, analyses in Chapter 8 suggest that democratic aid seems to have some potential to address these challenges, especially in young countries with little previous experience of democracy.

While many of the book’s empirical arguments are sound, however, its overall arguments are not as groundbreaking or novel as the authors claim. For example, the claim that legitimacy of various kinds help both with establishing democracy and with increasing its survival probabilities has been with us at least since Seymour Martin Lipset’s “Some Social Requisites of Democracy: Economic Development and Political Legitimacy (American Political Science Review, 53[1], 1959) and David Easton’s “An Approach to the Analysis of Political Systems” (World Politics, 9[3], 1957), if not even in Max Weber’s elaborations on political legitimacy in Economy and Society: An Outline of Interpretive Sociology (1922). It is a little odd to find that Elections in Hard Times leaves out two of these key references (Easton and Weber) in a work that forwards legitimacy as a key explanation, and only refers to the third (Lipset) in a brief mention of economic explanations to democratization (p. 123).

The authors’ approach to measuring the so-called democratic dividends (improvements in levels of democracy following elections) is also a little puzzling. The starting point for the calculation is “the moment when an election ends and an election winner takes office” (p. 40). But it is well known that elections are one of those phenomena in social science that have effects before they occur. Indeed, there was a whole school of thought shaped by this insight from Guillermo O’Donnell and Philippe C. Schmitter’s Transitions from Authoritarian Rule (1986), a paradigmatic work arguing that political rights and civil liberties tended to improve in anticipation and leading up to (“founding”) elections. Being one of the authors in the so-called democratization-by-elections literature that the authors of Elections partly criticize, I have tried very hard to read their evidence with an open mind to see if the argument I first laid out in 2006 (Democracy and Elections in Africa) would be proven wrong. Yet if it is true, as I have argued, that it is the year leading up to the holding of the election that generates much of the positive change, or democratic dividend, and this is typically followed by a slump between elections, then the findings in the present book are largely an effect of the measurement strategy. The authors simply have picked the highest scoring year (usually the election year) and compared it to levels of democracy two years later and found a lack of democratic dividends. Using that measure in a time-series, cross-sectional design to show that elections rarely lead to improvements in democracy is misleading if the level of democracy steadily goes up but in a wavelike pattern where the election year scores the highest in each election cycle. It therefore would have been desirable for the authors to have shown that their results are robust to alternative (and more standard) measures of the relationship between elections and changes in levels of democracy.

This is even more important since the book’s overarching argument is based on data generated by the measure of democratic dividends. If one is not fully convinced that the indicator of dividends is a valid measure of changes in levels of democracy associated with elections, questions remain about all of the findings that follow. Having said that, Flores and Nooruddin provide food for thought in Elections in Hard Times for all of us who study elections and democracy in both new and old countries.