A big part of social science's fascination with political parties is their position in the gray zone between state and civil society. It may be that the advent of public subsidies for their operation, especially in Europe, has brought parties closer to the state. Yet they remain largely autonomous entities. Above all, parties' gatekeeping role in the supply of candidates for elected public office remains almost unchallenged and only lightly regulated.

The ways in which parties undertake this gatekeeping role ought thus to be of central concern to political science, and Reuven Y. Hazan and Gideon Rahat's book opens with a couple of plausible stories about how this filter function can make a real difference to political outcomes. Yet Democracy within Parties joins only a fairly small number of books that focus squarely on how parties choose their candidates for public elections.

What this new work does not do is to offer anything in the way of original empirical analysis. The material that the authors refer to has already been published, much of it by themselves. (Indeed, their command of the relevant literature is impressive. Most of their real-world examples come from Israel, Europe, and North America, but a point is often illustrated with reference to a party from, say, Africa or Latin America.) Instead, Hazan and Rahat make two distinct contributions. The first part of their book is about concepts and measurement, a “framework for analysis.” The second part is mainly about sifting through and assessing the conclusions of previous research. Assumptions about the importance of political institutions underpin the discussion.

Part I is the more demanding of the reader. Four aspects of candidate selection—candidacy rules, the “selectorate,” decentralization, and the distinctions between appointment and voting systems—are each pulled apart in a single chapter. Two propositions from the authors stand out.

One is a 25-point scale on which to classify the “inclusiveness” or “exclusiveness” of the selectorate—that is, those with the right to determine which would-be candidates get to claim the party's endorsement when running for public office (pp. 48–52). At one end of the continuum are the parties in which all registered electors, party sympathizers or not, can jointly make the decision. At the other end are parties in which a single leader does so. Between those extremes are another three categories: selection by party members, by party delegates, and by a smaller elite. The six points that separate each of these five clear-cut categories are to take account of the complication that arises if more than one selectorate is involved (as in the “assorted,” “multistage,” and “weighted-candidate” methods), plus the balance between those multiple selectorates.

The other proposition, which (as Hazan and Rahat show) has been acknowledged by some scholars but neglected by others, is that there is a distinction between, on one hand, the degree of decentralization in a selection system and, on the other, its inclusiveness; and that these variables do not necessarily covary. The book shows (pp. 59–63) that it is possible for a system to be both centralized and inclusive (if, for example, a party holds a binding, nationwide primary in which any electors could take part) and decentralized and exclusive (if the selection is made autonomously by a single person in a small, subnational electoral district).

Part II then addresses “the political consequences of candidate selection methods”. Four chapters address, in turn, participation, representation, competition, and responsiveness, and how previous studies suggest that these properties are promoted or inhibited by the various methods of candidate selection.

For instance, the chapter on competition between aspiring candidates contains what amounts to an intriguing hypothesis. Generated by a study of Israeli parties, the argument is that the most and the least inclusive selection methods are most favorable to incumbents, and that moderately inclusive ones are least favorable. In other words, the relationship between exclusiveness and competition is said to be nonlinear. Gender equality also features consistently in these latter chapters, not least because that reflects the interest shown by existing scholarly work, which in turn has much to do with the relative ease with which the outcome can be observed.

This is all fine as far as it goes—and, given the authors' knowledge of the field, it goes a pretty long way. Considering, too, the dearth of robust cross-national data, Hazan and Rahat can be forgiven for barely applying their own framework to new empirical material. That is the next step for interested scholars. Some readers might wonder just a bit, though, about the increasingly normative tone of the book.

Using clearly defined concepts and logical reasoning to expose flaws in commonly received wisdom is certainly part of social science's mission. It is excellent that the authors point out, for example, that increased inclusiveness in candidate selection is not the same as increased participation, and that the ostensible “democratization” of the selectorate through the spread of intraparty primaries can “result in instant, opportunistic, and corrupt membership” (p. 97ff). They even point out how such “pathologies” might be mitigated. Indeed, the well-known debate about the link between democracy within parties and democracy across parties, a link that some eminent thinkers have doubted, is given due attention.

Especially in light of this debate, however, the authors arguably take normative reasoning a step too far with their proposal, in the final chapter, of an ideal selection method that balances optimally the political goals—“expressing norms and producing democratic outputs, the diffusion of political power, and the health of the party organization” (pp. 173–74)—that they consider to be paramount. I, for one, am not yet persuaded that the intertwining of parties with the state has gone so far that the health of any particular party organization, or the diffusion of power within it, is a public rather partisan concern. An optimal method of candidate selection will depend on what a party really wants to achieve (as many votes as possible, perhaps, or cabinet seats at all costs); and fundamental goal priorities will inevitably vary greatly among parties and over time. Personally, I have no problem with an individual party cheerfully submitting to the iron law of oligarchy and disregarding any sort of internal democracy, even if I may not vote for it.

If I have one final quibble, it is that too many references are a little loose. A rather precise argument or finding is often associated only with a book, without particular pages or sections being identified. Perhaps it is unfair to single out Hazan and Rahat when this sort of imprecision is so widespread in published political science, but it might be especially niggling when the subject matter involves such intricate rules and procedures.

Anyway, this minor complaint should not put anyone off. Hazan and Rahat's writing is clear and fluent, making even the more complex early chapters a thoroughly enjoyable and stimulating read. Above all, the effort to make concepts measurable, and thus to make diverse cases comparable, does a great service to the study of intraparty life.