Most democratic theorists do not spend much time defining, let alone discussing, the concept of constituency. They assume that what is crucial to democratic institutions is the manner in which votes are distributed (equally and justly), not how democratic institutions define and draw electoral constituencies. Andrew Rehfeld's insightful and important book challenges these assumptions. For Rehfeld, democratic theorists need to think about constituency because constituencies are “the quintessential institution of official exclusion”: How democratic institutions define constituencies determines whose interests are catered to and whose interests are ignored.

Rehfeld also takes a provocative position on how constituencies ought to be determined in democracies. He argues that democratic institutions should not necessarily arrange citizens' political lives around where they live. For the author, territory-based constituencies—what he calls “territorial constituencies”—hold significant practical and normative costs for democratic institutions. For instance, he holds that territorial constituencies have contributed to the explosion of public policy “pork.” He does not see “local pork” as an inevitable feature of democratic politics, or even as a political device used to facilitate political compromise. Rather, territorial constituencies “skew” legislative decisions toward merely local interests, and away from the public good.

The main purpose of this book, though, is to initiate a public debate about how constituencies are drawn in contemporary democracies. Democratic citizens should not simply draw their constituencies according to what is presumed “given” or “natural”; rather, they need to consent consciously to the organization of constituencies.

The Concept of Constituency is organized into three parts. The first, and analytical, part provides an extensive discussion of the meaning of constituency and identifies 10 different normative justifications for territorial constituencies. The second part is historical: It explores how and why electoral constituencies in the United States came to be based on where citizens live. The third part is normative and offers three arguments. The first argument contends that territorial constituencies are ill-suited for large nation-states. The second proceeds by sketching “a default position” that should be used in assessing existing ways of arranging constituencies. On this default position, constituencies should be permanent; each constituency should in composition resemble the nation as a whole; and individuals should not decide the composition of constituencies. Electoral constituencies that do not correspond to this default position violate the legitimacy of democratic institutions and therefore require additional justification. The third argument provides a thought experiment as a heuristic device for exploring alternatives to territorial constituencies. In this thought experiment, every citizen is randomly assigned to a single-member electoral constituency for life. In particular, the author argues that each constituency should “look like the nation they collectively represent” and “in no way be based on where a person lived” (p. 210).

This is a rich and provocative book. For instance, it raises the issue of whether the United States should “count” House members' votes in proportion to the size of their constituencies or whether U.S. citizens should be allowed to group themselves by selecting which features of their identities they would like as the basis of their constituencies (voluntary constituencies). Rehfeld raises another interesting issue when, in the course of discussing the need to differentiate authorization and accountability, he considers whether democratic citizens should be given two votes—one to signal their assessment of their representative's past performance and another to signal which representative they endorse for a particular election. These are just some examples of the many interesting ideas explored in the book.

Some of Rehfeld's arguments are, by his own admission, speculative. In particular, his arguments assume, without sufficient evidence, that his policy proposals are tenable, that they would have certain effects, and that those effects would be desirable. This makes the arguments difficult to assess. For instance, it is unclear whether one could in fact design random constituencies that would “resemble the nation as a whole” without taking into account where any particular constituent lived. Rehfeld's argument is also predicated on the assumption that constituencies organized differently, for example, around work or religion, would produce entirely different interests, such as vocational and theological interests, respectively. It is uncertain, though, whether the elimination of territorial constituencies would, in fact, significantly diminish local pork. Nor is it clear that a particular community subjected to illegal dumping of hazardous waste could gain enough political clout to fight these violations effectively, when its voice is “randomly” distributed, and so diluted, across the national constituencies. With territorial constituencies, a representative of such a community who possesses sufficient political clout could trade his or her vote to secure the required enforcement of the relevant environmental regulation. (As far as I can tell, it is an open question whether Rehfeld would, or would have to, consider this as an instance of local pork). In short, it is not clear whether we should, with the author, assume that eliminating all the leverage that territorial constituencies provide would be an unqualified good. Assessing these assumptions would require amassing empirical evidence that we do not currently have.

That said, what is most impressive about the book is that it succeeds in opening up a new and important area of inquiry, one that underlies and informs existing theoretical discussions of political representation. Rehfeld often flags explicitly that his arguments are meant to be suggestive. The suggestive character of this book ought to be kept in mind, especially by those readers who want to keep constituencies territorial. In particular, his arguments are unlikely to persuade those who are skeptical of relying on technologies, like the Internet, to safeguard democratic practices or who favor democratic practices that require citizens to have face-to-face meetings. Indeed, his rejection of territorial constituencies might anger some oppressed groups whose identities are inextricably tied to certain locations, for example, Native Americans whose religious practices are tied to sacred lands. Rehfeld's willingness to accept a House that would be 100 percent Republican or 100 percent Democratic might also strike some as antideliberative and even dangerous. Such objections, though, do not touch the main point of the book—namely, that democratic citizens need to rethink their blind attachment to territorial constituencies and to deliberate about and consent to the design of their constituencies.

In conclusion, The Concept of Constituency is a thought-provoking and insightful book that is a must read for anyone interested in political representation, democratic legitimacy, and institutional design. In particular, Rehfeld's discussion of the concept of constituency challenges the institutional reforms posed by theorists of group representation. For this author, the current emphasis on being present in legislatures is misplaced. Institutional reforms, like gerrymandering electoral districts, gender quotas, and even proportional representation, will fail unless theorists of group representation change how democracies organize constituencies. More importantly, Rehfeld's “default position” provides a normative argument against contemporary redistricting efforts in the United States. Following his thinking, redistricting should not be left to individuals and therefore not to members of state legislatures. The failure of democratic theorists to address how democracies ought to organize their electoral constituencies allows “the quintessential institution of official exclusion” to persist.