In opposition to the recent trend in democratization studies that sees this process as if the glass is half empty, this study of Nicaraguan democratization argues that the glass is half full. Recent history in Nicaragua includes momentous events, such as the fall of the Somoza dictatorship in 1979, the triumph of the Sandinista revolution, the subsequent Sandinista electoral victory in the presidential contest of 1984, and their later defeat in 1990, when voters rejected Daniel Ortega in favor of the mild-mannered, conservative Violeta Chamorro. Noting the voters' continuous embrace of what the authors call “democratic conservatism,” they conclude that “Nicaragua appears to be evolving into an inclusive, constitutional democracy” (p. 280). How was this possible, especially in a country that exhibited so few of the prerequisites traditionally associated with the successful establishment of democracy? And why did voters, the majority of them poor, reject the Sandinistas in 1990 and later elections, favoring candidates that came to undo most of what the Sandinistas had achieved in terms of social and economic equality? These are the most important questions Leslie Anderson and Lawrence Dodd seek to elucidate in this suggestive and informative book.

In Part I, the authors try to demonstrate that, conventional wisdom notwithstanding, Nicaragua exhibited three key procedural foundations that favored democracy, political space, social class, and party. The authors argue that since colonial times, as well as during the long Somoza years, Nicaragua developed a political space (i.e., instances of protest and contestation), conscious class differences, and strong political parties that allowed it to move quickly into democratization once the Somoza regime fell. They contend that “Sandinismo did not initiate the democratic foundations of space, class, and party that already existed in Nicaragua prior to the revolution. Instead, the revolution provided specific popular experiences that enabled mass citizen participation, enhanced citizen capacity for reasoned choice, and, through these, fostered democratization” (pp. 52–53). My problem with this view is that these three foundations are pitched at such level of generality as to render them unusable to predict democratic prospects in other places. One can easily identify similar foundations in many of the other Latin American countries, despite their wide variability in democratic achievement.

Part II comprises the heart of Learning Democracy. It studies the 1990 election that led to the defeat of the Sandinistas and the advent of democratic conservatism. Why do the authors assign so much importance to the 1990 election? They believe that it produced a regime change from “radical socialism” to “democratic conservatism” (p. 209). Although I do not mean to downplay the importance of this election, I do question the claim that it marked such a momentous regime change. The 1990 contest changed a government, not a regime: Nicaragua in 1990 was already democratic inasmuch as the Ortega administration was the result of free and fair elections conducted in 1984. The authors advance a “theory of reflective voting” to explain the 1990 election. Why do we need another theory of voting? They argue that current theories assume the existence of stable and well-established democracies, and thus the theories ignore that voters “may face more complex considerations when they attempt to reason through retrospection or prospection” (p. 117). Moreover, current accounts of voting do not usually give much attention to election contexts. To remedy these shortcomings, Anderson and Dodd develop a theory that is both dynamic and contextual. It is dynamic in the sense that it views voters as oscillating between prospective and retrospective concerns as the electoral campaign evolves. It is contextual because it takes into account the meager electoral history of developing countries, which makes the traditional choice between retrospection and prospection more complex.

For Anderson and Dodd, neither retrospective nor prospective evaluations taken separately drive the vote choice. Citizens start the electoral campaign with both prospection and retrospection in mind. They look at the incumbent and the opposition and make a “first cut decision” (p. 128). They use retrospective considerations to evaluate the incumbent but “reflective” (their term for prospection and retrospection combined) concerns to assess the opposition. If both the incumbent and the opposition pass this initial cut, voters then move to the second stage of the campaign, the moment of “attitudinal crystallization and differentiation” (p. 128). At this stage, the incumbent is assessed through a reflective assessment, while the challenger—given his or her inexperience—is assessed primarily though prospection. Moreover, this “reflective voting” also occurs across different elections. In the early stages of democratization, voters engage in reflective choice. As democratization progresses, the electorate moves toward retrospective considerations. Thus, as politics normalize and crises subside, voters in developing nations come to resemble their counterparts in the industrialized world. That is, less-educated voters vote along retrospective lines, while the more educated use prospective considerations (p. 132). They test this theory in Part III of the book.

It is in the application of this theory to the 1990 election that the book demonstrates its greatest weaknesses. The operationalization of retrospective and prospective consideration is not standard. For instance, the authors consider candidates' images and perceptions of candidate competence as indicators of retrospective and prospective attitudes (p. 179). This is highly debatable since it assumes that the voter's image of the incumbent is an indicator of retrospection, whereas the voter's image of the challenger is an indicator of prospection. The implausibility of this assumption is made clear when one considers that the images were measured in response to questions such as “is [the candidate] experienced/ a leader/ prepared/ close to the people/ brave/ patriotic/ respected/ honest?” (p. 141). In addition, the authors interpret the fact that more variables are statistically significant in surveys conducted closer to election day than they were in polls conducted at the beginning of the campaign as an indication that “voters broadened the factors they considered” (p. 180) in deciding their vote. This appears consistent with their theory of reflective voting. However, an alternative and more likely explanation is possible for this pattern: sample size effects. Early surveys had a smaller number of respondents than later ones (the earliest survey had 404 respondents, while the latest had 859). Not surprisingly, it was more difficult for some variables to achieve statistical significance in the earlier surveys than it was in the later ones. Differences in sample size may account for the inconsistency in the performance of retrospective and prospective considerations (emerging as significant in some surveys but not in others), rather than the more elaborate and complex process of “reflective” reasoning that the authors attribute to Nicaraguan voters.