Political science has invested very heavily in the question of what determines the number of parties that compete in a given polity. The general finding is embodied in the M+1 rule, in which M is the number of seats allocated in a district. This states that strategic voting will reduce contests with more than M+1 candidates or parties to contests in which at most M+1 competitors are seriously in the running for seats (this argument has been made by Gary W. Cox, Making Votes Count: Strategic Coordination in the World’s Electoral Systems, 1997). However, the M+1 rule only operates at the district level. Why the same parties compete in all or almost all districts, or, in other words, how local party systems merge to form a national party system has been barely explored. This is an astonishing omission as cross-district coordination is central to the representative process. Recent research (see Daniele Caramani and Ken Kollman, “Symposium on ‘The Nationalization of Electoral Politics: Frontiers of research,’” Electoral Studies 47 2017) has shown that the geographic basis of a party’s support—nationalization—affects the provision of public goods, the legislative careers of representatives, the politics of coalition making in parliamentary regimes, the consolidation of new democracies, national unity, and the attribution of political responsibilities.

Scott Morgenstern’s Are Politics Local? is the most comprehensive analysis of party nationalization ever published, and it is written by one of biggest names in the field. Morgenstern measures nationalization across two dimensions, termed static and dynamic. The former implies the “degree of homogeneity in a party’s vote across a country at a particular point in time,” while the latter captures the “consistency in the change in a party’s vote in each district across time” (p. 5, emphasis in the original). The author’s premise is that “together the two dimensions provide a window into the relative importance to voters of local versus national concerns” (p. 3), local politics being the inverse of party nationalization.

The book is divided into three parts. The first describes, measures, and compares the two nationalization dimensions. Morgenstern develops a hierarchical statistical model for estimating the two dimensions of nationalization simultaneously. District-level electoral results are separated into three components: variance between districts, measuring static nationalization; variance between elections, measuring volatility; and the residual variance or the unexplained variance, both in districts and across time, capturing dynamic nationalization. This sophisticated measure is used to establish the phenomenon and generate a typology of parties, combining the static and dynamic dimensions of nationalization.

The second part of the book explains cross-national and cross-party variation in static and dynamic nationalization. The analysis is focused on institutional and societal variables. While static nationalization is mainly driven by electoral systems (i.e., the greater the number of districts, the more heterogeneous the distribution of partisan support in a given election), the regime type is related to dynamic nationalization (i.e., voters are more likely to respond to national party appeals in parliamentary regimes than in presidential ones).

Morgenstern ends by focusing on the impacts of nationalization on the behavior of voters and legislators. When party nationalization is high, accountability for national conditions increases, legislators cosponsor legislation, and cross-district partnership is more likely. The book breaks new ground and makes a strong case, but a number of concerns remain.

Substantively, there is no concrete definition of party nationalization in the book but, rather, operationalization based on the geographical and temporal homogeneity of partisan support. As a result, the theoretical attributes included in the empirical definition of party nationalization are not clear. The first issue with this undertheorized definition is that the interpretation of nationalization scores, particularly when they are not very high, is problematic. Imagine a national election focused on a national positional issue such as pensions, income tax, or abortion. If nationalization refers to the extent to which party and voter behavior in mass elections concerns national or local issues, the election is highly nationalized. However, districts are always heterogeneous in demographics such as employment, average age, population density, or elderly population, and these politically salient features are correlated with partisan political sentiment. The electoral performance of the incumbent party should vary substantially across districts and across elections, and, therefore, party nationalization measured in terms of homogeneity of partisan support would lead to the conclusion that party nationalization is not high. There is thus a mismatch between theoretical expectations and empirics.

The second issue is that static and dynamic nationalization measured in terms of the homogeneity of parties’ vote shares captures other political phenomena that are stricto sensu different from nationalization. These include the strategic allocation of resources in election campaigns, the effect of varying district magnitude, or simply random variation.

A further substantive concern is that both static and dynamic nationalization might greatly differ across elections (e.g., Ireland in the 2007 and 2011 elections), but the institutional and societal determinants suggested in the book do not. Political parties have a number of decisions to make when an election occurs, crucially affecting the salience of national and local issues. Should they enter the race or stay out? Should they run candidates nationally or only in selected districts? What issues should they emphasize? The Lega Nord, which increased the focus on national issues (particularly immigration) rather than independence for northern Italy in the 2018 Italian election shows that the salience of national and local issues in elections is not written in stone. Unfortunately, the role of party strategies is not explored in the book.

Methodologically, though valuable and compelling, the hierarchical model that decomposes the variance of the district-level vote is very demanding in terms of data. To estimate precise parameters, parties have to compete in a minimum of three successive elections in a relatively high number of districts under exactly the same electoral rules (e.g., district boundaries must not change). This means that countries using few districts, which is not unusual in proportional representation systems (e.g., in Luxembourg, Iceland, or Slovenia, with four, six, and eight districts, respectively), and new democracies in their earlier stages are excluded. As the number of districts and the age of parties affect nationalization in a negative and positive fashion, respectively, the most highly and poorly nationalized parties or countries in all probability are not part of the analysis.

Are Politics Local? would also benefit from the use of survey data. Given its focus on “the relative importance to voters of local versus national concerns” (p. 3), the analysis of party nationalization cries out for individual-level data (i.e., surveys). To the best of my knowledge, there are no systematic studies with individual data on the salience of national and local issues for voters under different social and institutional conditions (i.e., cross-nationally and over time).

I recommend Are Politics Local? to people who are interested in governance and political representation. Any approach to party nationalization necessarily goes through that presented in this book.