George Edwards has filled the breech created by the recent death of Lawrence Longley as a powerful voice within political science opposing the electoral college. In the foreword, Longley's longtime associate, Neal Peirce, notes that one of the great puzzles of the 2000 presidential election was that it failed to prompt a massive public outcry against the system that elevated George W. Bush to the presidency, despite Al Gore's 539,893 positive margin in popular votes. Another great puzzle is the lack of any consensus among political scientists on the issue, as described in Choosing a President (2002), edited by Burdett Loomis and myself.

In Why the Electoral College Is Bad for America, Edwards does not address either of these puzzles. Instead, he deepens them. Drawing on massive historical and contemporary evidence and applying clearheaded logic, he provides a compelling prosecutor's case against the college. The evidence is fairly familiar, and the logic is not complex. Perhaps the public can be forgiven for not understanding the issue and its importance. But how can political scientists—whose professional status cannot help but be tarnished by an inability to provide consensual advice on so basic a question as this—be ambivalent and even widely supportive of this deeply flawed method for determining the most powerful political leader in the world?

Edwards begins his case by describing the college with all its complexities, showing how they provide ample opportunities for enormous mischief and malfunction. He then proceeds to a meticulous discussion of how the college violates the most fundamental democratic norm: that all votes count equally. He reminds us that the winner-take-all feature effectively disenfranchises most voters (everyone not living in competitive states and those in the minority within competitive states), that voters in small states are advantaged by the allocation of electoral votes among the states on the basis of their representation in the House and the Senate, that voters have more clout if they live in states with lower voter turnout, and so forth. He shows why these are not trivial inequalities, as they enable the college to achieve outcomes contrary to the popular vote alternative that counts all votes equally. In the 10 reasonably close elections since 1828, the college elected the loser of the popular votes four times: in 1876, 1888, 1960, and 2000, and Edwards describes each of these cases in rich detail, such as why Richard Nixon probably received more popular votes than John Kennedy in 1960. Edwards also considers how the failure of any candidate to get a majority of electors would necessitate a House contingent election, and how such a process would further jeopardize democratic norms. Harrowing accounts are provided of how the country was spared such contingent elections in the “near misses” of 1948, 1960, 1968, and 1976.

Returning to the origins of the electoral college, Edwards shows that the founders were animated neither by animosity toward citizens nor by any coherent theoretical design. Instead, the existing system was patched together out of desperation and expediency. He then presents his new contributions to this age-old debate: how the arguments that have been concocted to defend this relic are either lacking credible evidence or are based on slipshod logic.

While advocates of the electoral college claim, for example, that it protects the interests of states with small populations and the interests of racial majorities, Edwards concludes that the college provides opportunities for any cohesive special interest to have disproportionate power, but that there is no evidence that it has protected either small states or racial minorities. He finds equally flawed the argument that the college protects against fraud and the need for massive recounts throughout the country after a very close election, as a direct popular election provides better inoculation against these problems. He argues that the college does not help winners achieve mandates and is irrelevant to the capacity of presidents to govern successfully. He also concludes that the college is neither an important factor in maintaining the two-party system nor protection against the fragmentation that can arise from third parties.

The evidence that Edwards provides in support of these claims is impressive, as he demonstrates command of both the historical record and contemporary political research pertaining to these claims. His logic is also impressive, as he repeatedly reveals the underlying premises of supporters' arguments and how they are factually inaccurate (e.g., that small states have cohesive interests that are different from those of large states), are conceptually misguided (e.g., that federal principles are implicated in the college), or contain normative judgments that few would defend (e.g., that we should be more concerned with protecting special interests than responding to broader national concerns when evaluating procedures for selecting the president).

So is there something about making prescriptions about how to choose a president that leads political scientists to be wary of the prosecutor's case that Edwards has so eloquently laid out? Two possibilities come immediately to mind. First, the internal logic and evidence may not be as ironclad as Edwards asserts. For example, he finds little possibility that a popular plurality alternative would facilitate the election of extremist candidates, because “polarized parties reflect rather than cause public opinion” (p. 142). Many political scientists would likely dispute this important but undemonstrated assertion. Second, considerations beyond those invoked by Edwards may remain important. The participants in our Choosing a President project returned time and again to the concept of “unanticipated consequences” when thinking about how previous reforms, like the democratizing of the Democratic Party after 1968, had deleterious effects that escaped the evidence and logic that led to the reform. And reform options beyond those considered by Edwards—such as “the instant runoff”—may also seem better than the popular plurality system he favors. Of course, such considerations lead political scientists into the realm of speculation that Edwards abhors, but such matters seem inevitably to color professional judgments about how to choose a president.