Scholars of representation have long realized that the rules and procedures embedded in legislative and electoral institutions have important effects on the ways in which citizens and office holders interact with each other. Institutions influence the degree to which citizens' preferences are reflected in public policy. In Representation and Institutional Design, Rebekah L. Herrick tests a great number of hypotheses related to the impact of institutions on the behavior and attitudes of elected officials in U.S. state legislatures. Utilizing an original survey of state legislators, she undertakes an ambitious examination of the effects of differing institutional arrangements on representation.

Keenly aware of the difficulties involved in attempting to pin down the concept of “representation” with a single definition, Herrick approaches the subject from a number of angles. Factors affecting the following aspects of representation are explored: the incumbency advantage, legislative ambition, legislators' role orientations, symbolic representation, the accuracy of legislators' perceptions of their constituencies, and substantive representation. The groups of institutional variables examined include measures of legislative professionalization, election laws, term limits, and “dyadic relationships”: characteristics of districts such as district size and district magnitude that might affect how legislators interact with those districts.

Herrick's findings generally support, to varying degrees, the contention that institutional design affects representation. Most of the findings are in agreement with previous research. Given the large number of independent variables that Herrick examines (measures of the level of professionalization in a legislature, for example, include leadership strength, committee strength, the number of legislators, the provision of office budgets, the support of legislative staff, and whether legislating is a full-time job) it is difficult to succinctly summarize the empirical results. A few of the more novel findings, however, merit specific mention.

First, Herrick finds that different components of professionalization affect various aspects of representation in different ways. For example, while greater resources are typically associated with greater levels of representation, legislative capacity is not. Larger budgets and full-time status are associated with higher levels of electoral security and lower levels of policy representation, but also with greater levels of symbolic and service representation. Second, while term limits are generally associated with legislators being less sensitive to constituents, they are also associated with higher levels of policy representation. Third, Herrick finds that electoral features that make campaigning more difficult increase policy representation, but decrease the visibility of legislators in their districts. Fourth, legislators who display static ambition (wanting to maintain their current office) and progressive ambition (wanting to move to a higher office) tend to be more like trustees than delegates, and generally provide more symbolic representation than legislators who plan to leave office. Taken together, these findings contribute a substantial degree of theoretically interesting nuance to the literature on institutions and representation.

Although Herrick typically does an excellent job of describing the causal mechanisms potentially at work between institutions and representation, the measurement of certain key variables is not fully satisfactory. For example, Herrick uses electoral vulnerability as the main measure of electoral accountability. While it may be difficult for representatives to be held accountable if they approach electoral invulnerability, the relationship between the two concepts is not at all straightforward. It could be the case, for example, that legislators who pay particularly close attention to their constituents, and are hence quite accountable in that sense, are more difficult to defeat.

The measurement of legislative ambition serves as an additional example of slippage between the concepts Herrick is trying to get at and the measures employed in the book. The survey question tapping ambition asks legislators what they plan to be doing in five years. It seems likely, however, that this measure conflates ambition and the expectations legislators have regarding their ability to win reelection or election to higher office. Herrick is right to think that legislators' ambitions are likely to change over time, but her discussion could be improved by recognizing that expectations may be shaped by institutions to an even greater degree than ambition.

While Herrick's findings taken as a whole are compelling, the book's statistical modeling is not particularly strong. Herrick ignores the multi-level nature of her data. The models all include individual, chamber, and state level measures, but no statistical adjustment is made for the fact that the observations are not independent, due to individual legislators being clustered within legislative chambers. Short of full-scale hierarchical modeling, current practice in such situations typically involves the use of clustered standard errors. Clustered standard errors are often substantially larger than non-clustered errors, with the key implication being that Herrick may have incorrectly identified some chamber-level institutional features as having statistically significant effects on representation.

Another shortcoming of the modeling is that while Herrick theorizes that factors such as legislative ambition serve as intervening variables between a number of independent variables and representation, the indirect nature of these relationships is not directly modeled. As such, in cases where both a set of independent variables that affect an intervening variable and that intervening variable itself are simultaneously used to predict the value of some measure of representation, the overall importance of the set of causally prior variables may be greatly underestimated.

While institutional variation across states provides a great deal of leverage for examining the effects of institutions on representation, some scholars may question the generalizability of Herrick's findings to legislatures in other national and sub-national contexts. Progressive ambition, for example, is likely to be much more constrained in the U.S. Senate than in state legislatures. Moreover, while the U.S. states provide much institutional variation, they provide little variation in terms of electoral systems broadly conceived—e.g., majoritarian vs. proportional representation systems—and representational differences between such systems are the focus of much contemporary research. On the whole, however, the causal linkages identified by Herrick are likely to be relevant in a wide variety of legislative settings.

Scholars of representation are likely to find the chapter on substantive representation somewhat unsatisfying. The measures of policy representation employed deal largely with only one policy issue area (gay rights), and while there is solid logic behind the decision to focus on bill co-sponsorship, the use of additional measures of policy representation, particularly ones related to legislators' general ideological positioning in relation to their constituents and the relationship between state-wide policy preferences and the adoption of policy at the state level, would seem to be warranted. Party membership is also conspicuously absent from the discussion. Given the importance accorded to substantive representation by most representation scholars, its relatively brief treatment is likely to disappoint some readers.

Despite its faults, Representation and Institutional Design is a commendable contribution to the literature on the effects of institutions on representation. Of particular value is the way that Herrick breaks down institutional features such as legislative professionalism into component pieces, each of which may have differing effects on representation, a concept which Herrick similarly—and wisely—treats as having many important facets. The level of detail is impressive, as is the manner in which Herrick pulls together many important strands of a large and varied literature.