In his book, Candidates and Voters, Walter J. Stone offers an optimistic take on the quality of American congressional elections. He argues that once we accurately account for the choices with which voters are confronted, we find that voters do a more than adequate job of selecting the candidates who better represent them on the policy and personal attributes that they value. Further, Stone discusses how these factors aggregate and affect the overall level of representation that citizens receive from their members of Congress. While Candidates and Voters will be of particular interest to researchers focused on spatial models of electoral competition, it will also appeal to scholars of American politics because of the explicit ties to, and implications for, the partisan polarization that has defined the past several decades.

In this thorough yet concise treatment, Stone takes a layered approach to understanding the connection between the policy positions and personal attributes of candidates and the decisions voters make. Specifically, he defines the proximity and valence rules and employs them to consider whether citizens vote correctly. The proximity rule states the well-established assumption that voters will choose the candidate who is closest to the citizens’ ideal point in the policy space. All else being equal, voters will choose the candidate with the higher level of personal attributes that are important to being an elected official, such as leadership, trustworthiness, experience, and integrity (which are collectively referred to as valence).

While formal models of spatial model competition often explicitly consider the policy and valence attributes of both candidates (typically an incumbent versus a challenger), empirical work often falls short. Often, researchers do not possess measures of the challengers’ policy positions. Further, high-quality measures of candidates’ valence are not available. Stone sets out to more effectively assess the importance of policy and valence by considering these attributes for both candidates.

One of the strengths of the analysis presented in the book is the unique data set compiled to measure proximity and valence, which is discussed in Chapter 2 and the Appendix. Given the dearth of data on candidates’ valence and challenger ideology, Stone has utilized a survey of expert informants (i.e., politically involved partisans who are knowledgeable about the district), which he has leveraged in other studies with various coauthors. The survey of expert informants is conducted well in advance of the election, which minimizes the potential problem that evaluations of candidates are shaped by prospects of electoral success (as opposed to the other way around). In all, Stone does a thorough and convincing job of validating the measures of valence and ideology obtained from the survey.

It is also important to note that the expert informant survey was conducted in advance of the 2010 midterm elections, which highlights a second significant strength of the book. Stone does an exceptional job of considering alternative explanations that could account for the empirical link between candidates’ ideology and valence on election outcomes. The most likely alternative explanation rests on the strong links between partisanship and citizens’ attitudes and behaviors. In short, citizens’ identification with one of the two major political parties determines their evaluations of the candidates’ ideology and valence attributes, as opposed to ideology and valence exerting their own independent influence on voting and elections. And this is why the use of the 2010 election is so significant. Due to the strong national partisan tides in favor of Republicans, the attempt to uncover the independent effects of candidates’ ideology and valence in the 2010 House midterm election presents a very difficult test case. Given that Stone does uncover independent effects while controlling for partisanship during that election, the strength of the findings and the importance of ideology and valence are thus highlighted.

A third strength of the book is the layered approach to considering how the proximity and valence rules matter. In Chapters 4 through 9, Stone demonstrates that the proximity and valence rules are significantly related to individual voting decisions, as well as to aggregate election results. Further, he establishes the independent effect of proximity and valence by controlling for other significant alternative factors that may affect voting, including partisanship, candidates’ campaign resources, and the strategic entry of quality challengers.

In addition to establishing the explanatory power of candidates’ proximity and valence, Stone also analyzes whether citizens make correct votes, meaning that citizens choose the candidate who is more proximate and/or the candidate who has higher valence. And more often than not, voters favor the candidate who is superior on proximity and valence, thus leading to the author’s more optimistic tone about the quality of American elections. He also observes that there is a significant number of cross-pressured voters, who must choose a candidate who is more ideologically proximate but has a lower valence, or vice versa. Many previous empirical analyses do not consider this complex choice, which may lead some authors to be more negative on the competency of American voters than is warranted.

This all points to one more additional strength in the manuscript. Stone is able to give a more thorough conception of representation in the American political system. By considering the valence and proximity of both candidates, he can more accurately assess the quality of representation provided by elected officials. Given that citizens are often forced to choose between two options, he demonstrates that individual voters and district electorates do a good job of choosing the better option (even when we account for the cross-pressured citizens and electorates).

One area that could benefit from further consideration is the measurement of policy and ideology. Stone conceptualizes candidates’ policy preferences as being accurately summarized along a single left–right, liberal–conservative policy dimension. He does note the possibility of a multidimensional policy space but ultimately decides that a single dimension is sufficient. However, there is an emerging set of scholars who have demonstrated that citizens’ policy preferences are multidimensional, and there may be a significant number of voters who are cross-pressured across policies. (See D. Baldassarri and A. Goldberg, “Neither Ideologues nor Agnostics: Alternative Voters’ Belief System in an Age of Partisan Politics,” American Journal of Sociology, 120[1], 2014; S. Klar, “A Multidimensional Study of Ideological Preferences and Priorities among the American Public,” Public Opinion Quarterly, 78[S1], 2014; E. G. Carmines M. J. Ensley, and M. W. Wagner, “Political Ideology in American Politics: One, Two, or None?” The Forum, 10[3], 2012; and E. G. Carmines, M. J. Ensley, and M. W. Wagner, “Who Fits the Left–Right Divide? Partisan Polarization in the American Electorate,” American Behavioral Scientist, 56[12], 2012).

A related issue concerns the meaning of liberal–conservative ideology. Christopher Ellis and James Stimson (Ideology in America, 2012) have demonstrated that ideology means different things to different citizens (for some it is a symbolic and social identification; for others it may be more strongly rooted in explicit policy positions). Thus, there are strong reasons to expand the conception of proximity beyond the standard liberal–conservative conceptualization.

Despite these potential criticisms, this book firmly establishes the importance of candidates’ ideological positions and leadership qualities for understanding congressional elections and representation. Further, these criticisms suggest that the problem is one of measurement error, which would imply that Stone’s findings that link candidate proximity and valences are attenuated. Thus, if we did measure policy proximity better, the implication is that empirical links between voting and candidates’ attributes would be even stronger.

Candidates and Voters is essential reading for scholars focused on spatial models of electoral competition, but there is also a broader audience that should heed the findings. Stone establishes that policy and valence have a significant and independent effect on American elections. Thus, scholars who are focused on trying to explain the causes and consequences of partisan polarization in U.S. politics cannot ignore these factors and focus only on national partisan politics. Scholars instead must consider the policy and valence attributes of candidates from both parties if we are to understand the current political divide between the Democratic and Republican Parties in the United States.