This is a complex, important, original, and unusual work. It should be studied carefully (not simply read) by all those interested in constitutional studies broadly understood. This includes constitutional lawyers (especially those with an interest in comparative constitutional law, and in constitutional design), political theorists of many different stripes, and most notably scholars in comparative politics interested in institutional questions. If in your research or teaching or consulting you rely on the work of Arend Lijphart, you should definitely read this book. You should also read it if you are interested in institutional design more generally.

On the surface, the work can be described straightforwardly. Working in what we might call the Indiana Workshop tradition (with debts to V. Ostrom, D. Elazar, and C. Hyneman acknowledged in the dedication), Donald Lutz outlines a set of principles of constitutional design, summarized in Chapter 7. These include “match the government to the people,” “seek the best possible political system under the circumstances,” “political power is an unavoidable danger that must be understood and faced, if the design is to succeed,” and “the idea of a constitution is to marry justice with power.” (pp. 218–20) And, as summarized in Chapter 8, he discusses a number of interesting patterns that emerge in constitutional design. Four “interesting curves” he calls them. The first relates separation of powers and popular control, the second relates constitutional amendment difficulty and constitutional amendment rate, the third relates size of legislature and the cube root of population, and the fourth relates the number of written constitutions in the world to the number of real constitutional democracies.

In the past, Lutz's work has been mainly about the constitutionalist tradition in the United States and its historical roots. We can find traces of that past work in this book, but this is most definitely a book with global scope. His basic data set is 75 constitutions from constitutional republics around the world, both large and small (from India to San Marino and Palau). For reasons never explained, constitutions from the past are not included in the main analyses.

Principles of Constitutional Design does not yield its riches easily, however. A straightforward summary, such as I gave, seems to me misleading. And reading the introductory chapter is largely useless if we want help in understanding the book's contribution. We need to approach the book differently.

There are two kinds of important books. The first type is a definitive statement on some topic. The second is a book that opens up large and important intellectual possibilities. Lutz's work belongs in the latter category. It is tentative and sketchy in a number of crucial places and just begs for others to build on what he has started. This is especially true of the fundamental relationship uncovered by Lutz in constitutional design between popular control and separation of powers.

The book does not fit easily into the existing division of intellectual labor (and not just within political science). This makes it both more difficult to write and more difficult to read. It is more difficult to write because Lutz tries simultaneously to give an inclusive outline of a field that does not yet really exist and to contribute both to its methodology and to its substantive claims. He includes a general statement of principles of constitutional design (but that is not where the book's main contribution is, many of those principles being hardly novel). He also presents a strategy for the study of such principles through a search for patterns in design (his “interesting curves”). But on closer inspection, some of those curves turn out to be far more fundamental than others. The book's main contribution is somewhat hidden by all the other things it does. This main contribution is Lutz's completely novel approach to popular sovereignty as the central principle of constitutional design.

The principle of popular sovereignty, the core of this book, is presented first through a discussion of the notions of sovereignty (with special attention to Jean Bodin) and popular sovereignty in the history of political philosophy. It is presented, second, through the development of two indices that measure popular control and separation of powers, respectively, as these appear in constitutional designs (not necessarily in the actual practices of constitutional democracies). Popular sovereignty is achieved by combining popular control and the separation of powers, in a system of limited popular control. The key empirical finding of Lutz's study is that popular control and separation of powers are closely related in constitutional designs: The more popular control, the more separation of powers. This relationship seems to me unexpected, and potentially very important. It certainly should be added to the repertoire of key relationships in the study of democratic constitutions.

The author dismisses the preoccupation in comparative politics with the contrast between presidential and parliamentary system, and proposes the separation of powers variable as a more significant dimension of democratic constitutionalism. To back this up, he identifies a powerful relationship between separation of powers and popular control. Is he right? The evidence presented suggests that he may well be right, but it will take much more than one book to establish such an important claim.

Lutz is surprisingly relaxed in the way he constructs the two crucial indices of popular control and separation of powers. The Index of Popular Control (to take one example) requires something like a theory of democracy to really back it up. It gives some weight to various features of a constitution that contribute to popular control (frequency of elections, who gets directly elected, role of referenda, and many others). The weights given to each feature, and the selection of which features to include, constitutes a tacit theory of democracy (determining what is more and what is less important for popular control). Without such weights there can be no index, and plainly, many alternative reasonable weights can be proposed. The specific decisions need to be defended (he does a little of this, but not nearly enough), or alternative indices need to be constructed to show that the relationship between popular control and separation of powers is not a byproduct of some arbitrary aspect of the choice of weights in constructing the indices.

As I said, Principles opens up a series of important topics for the future, such as the relationship between popular control and separation of powers. It does not by itself establish such a relationship. I hope it will be recognized for the pioneering work that it is, especially in comparative politics. It presents itself more as a work in political theory (it is that, too), and so it is at risk of not reaching that audience in comparative politics that could most effectively built on its main contributions.