David Primo has written an ambitious book analyzing the ways in which rules alter the size and efficiency of budgets at both the state and federal levels. Even readers who do not buy into the central premise that public spending is driven by the uncontrolled bidding of distributive politics (which we might now call the McCain hypothesis in light of the 2008 election) will find many valuable insights in this study. The comparison of state to federal budgetary processes is by itself an important contribution to the political science literature.

The book begins with the observation that public spending has increased at all levels in the postwar period, and that while federal reforms have largely failed, some states have more successfully lowered their total spending and have more responsibly aligned it with per capita revenue changes. The solution to fiscal restraint, Primo maintains, is in the rules that govern the budgeting process. Properly designed and effectively enforced rules can constrain the inevitable political pressures to spend inefficiently. However, all too often, he laments, the design of effective rules is undermined by political considerations.

Using a game-theoretic model with many typical simplifying assumptions, Primo demonstrates that the allocation of distributive goods will often be inefficient because common pool problems (i.e., the temptation to overproduce concentrated benefits when costs are shared) and the powers that agenda setters have in putting together a support coalition are considerable. Unless these processes can be cabined by effective and enforceable externally or internally imposed rules, spending will increase even when everyone wants to cap it. Moreover, in flush times, legislators will commit to recurring expenses that are unaffordable in the long term.

The author also usefully demonstrates how under certain conditions institutional rules, such as spending limits, supermajority rules, and executive vetoes, can limit inefficient and excessive spending, but not always. Supermajority votes, for instance, can sometimes lead to extra, even if somewhat more efficient, spending as the proposer seeks to increase the coalition to match the higher threshold. In the end, Primo concludes that a no-carryover spending limit enforced by an elected judiciary is the most effective means for restraining budgets.

One of his empirical equations in Chapter 5 shows that public expenditures are indeed lower in states that have no carryover spending limits and elected judiciaries, controlling for other factors. A second, and perhaps more interesting, equation indicates that states with spending limits react to revenue increases more conservatively than do those without, which is consistent with other studies that have shown that states with spending limits have less volatile fiscal patterns and are more likely to put surplus revenue into rainy-day funds.

While there is much of value in this book, there are also a few leaps of faith (and perhaps ideology) and important evidentiary gaps in it as well. Even though Primo acknowledges that distributive goods (especially as he defines them) are only a small fraction of federal and state budgets, he focuses on them as the central objects of restraint. This is indeed analogous to John McCain's attempt to convince us in the last election that earmarks are the core economic problem in America. Quite aside from a point that Primo himself acknowledges—that distributive goods can grease the wheels of budgetary agreement—the main driving forces for budgetary expansion are not distributive goods but, rather, entitlements in the federal budget and nondiscretionary spending at the state level. The empirical model that allegedly proves his case only shows that spending is less in states with spending limits, not that spending on distributive goods is less. But as Primo discusses in Chapter 2, there is a tendency to underprovide for public goods due to free-rider incentives, and there is no way to tell from his empirical models whether spending limits are lowering distributive or public goods.

Another puzzling claim is that elected courts are crucial to budgetary restraint. The evidence for this is thin, hinging on the contention that the combined statistical effect of no-carryover rules and an elected judiciary is greater than their separate effects. Even so, this reader at least would like to have had some specific instances in which the courts acted to enforce spending limits, and some data that demonstrated that elected courts were more likely to take up these issues than nonelected ones. It is not obvious that elected judges are anymore likely to take on a legislature that controls their budget or wade into an issue that is so political than are appointed ones. An unstated assumption of Primo's view is that the courts are defending the public interest in restraining spending, but if the public's opinion is more divided than that, elected judges might see danger in fiscal decisions.

While I do not buy all of the premises or conclusions, there are important takeaway points in this book. The design of budget rules can shape fiscal outcomes in significant ways. External constraints will more effectively limit spending than internal mechanisms such, as Gramm-Rudman-Hollings, but they introduce rigidities that make it harder to deal with crises. And most importantly, since short-term political considerations often undermine long-term fiscal goals, there is a need for well-designed rules to help legislators avoid irresponsible choices.