I offer the following challenges to Professors Lowenstein and McGinnis: Would either of these very capable scholars advise a country attempting to draft a new constitution to emulate the American institution of the Electoral College, either as originally drafted in 1787 or as amended in 1803 with the Twelfth Amendment? Or, if time travel were possible, would either of them, if given the opportunity, remain happily silent at the Philadelphia Convention when the Electoral College was proposed and accepted, almost as an afterthought, as a means of resolving the debate over how to elect the president? Quite frankly, I cannot imagine that either would answer in the affirmative.

The only conceivable argument “for” retaining the Electoral College as a constituent aspect of the American political process is a basically Burkean one. It seems to rely on some mixture of the fact that it is indeed our unique method of choosing a chief executive/head of state; the highly debatable assertion that it has not disserved the country too badly and may even, on occasion, have served us well; and, finally, that it would be either futile, because of the barriers set out by Article V, to try to eliminate the College through constitutional amendment or too dangerous to accept my own proposal, in Our Undemocratic Constitution: Where the Constitution Goes Wrong (and How We the People Can Correct It), of calling a new convention, as is legitimate under Article V itself, charged with examining the many grave deficiencies of our present Constitution.

As should be obvious, I disagree with all of the proposed defenses and believe that the College is a continuing menace to the American polity. Complacent acceptance of its “inevitable” role in electing our presidents is equivalent to an equal complacency about driving a car with slick tires and bad brakes, after having had three drinks, on the ground that one had earlier successfully navigated the route home. Even if true, this is ultimately an adolescent way of thinking. We should recognize that there is also a significant chance that such a car will take us over a cliff and try to guard against such an unhappy future by buying new tires, installing new brakes, and resolving not to drive while drunk.

So what are the primary deficiencies of the Electoral College? I begin with the most obvious one: It regularly sends to the White House persons who did not receive a majority of the popular vote. Since World War II alone, this has included Truman, Kennedy, Richard Nixon (1968), Bill Clinton (1992 and 1996), and George W. Bush (2000). (Gerald Ford’s unelected presidency cannot truly be blamed on the Electoral College, though any spirit of fundamental reflection about the current Constitution might well ask about the necessity or advisability of having a Vice President at all.) More distant beneficiaries include Abraham Lincoln and Woodrow Wilson (1912), either of whom might have been defeated if the United States had adopted such a sensible election process as the Alternative Transferable Vote (ATV) or even run-offs between the top two candidates. One might, of course, applaud both of those presidencies; if that is so, then perhaps we should really be debating if we really believe in “majority rule” at all and why losers in such a process should feel obligated to accept what they believe to be fundamentally wrong decisions on the part of minority presidents.

Of course, there is always the possibility that the “winner” will not even have received a plurality of the popular vote. Everyone knows of the 1824, 1888, and 2000 elections; fewer people are aware that it appears increasingly likely that Richard Nixon in fact received more votes than did John F. Kennedy in 1960. In any event, it’s hard—I believe impossible—to offer a vigorous affirmative defense of such a process. And any complacent assurance that it really doesn’t matter that much who becomes president, perhaps because deep structural imperatives are more important than the mere individuals who inhabit the White House, founders on the presidencies in particular of Abraham Lincoln and George W. Bush.

One explanation for the variance between electoral and popular vote outcome is simply the insupportable boon given small states in the process. That it might have been necessary in 1787 to attain support for the Constitution at all is only a political explanation for its presence in the Constitution; it does not count as an argument in favor of our feeling bound by it in 2007, any more than it constituted a winning argument to say that the necessary compromise with slaveowners in 1787 necessitated perpetual support for the “peculiar institution” thereafter. George W. Bush owes his election not only to the machinations of Kathleen Harris and the United States Supreme Court, but also, and more basically, to the fact that he gained nine electoral votes by winning the two Dakotas and Wyoming at the same time that Al Gore won only five votes by winning New Mexico, even though New Mexico has a slightly larger total population than the other three states added together. Unless one favors affirmative action for the residents of small states, there is no defense for this.

But perhaps it is considered too partisan to suggest that the “election” of George W. Bush is a sufficient trainwreck in itself to justify opposition to the Electoral College. So consider the possibility that Ralph Nader had in fact gained the electoral votes of a single state, say New Hampshire, and therefore threw the choice of the president into the House of Representatives, as required by the Constitution. It should be obvious, incidentally, that any defender of the Electoral College must justify every feature of that dreadful institution. If my adversaries concede that certain features are well worth changing, even at the cost of a potential constitutional amendment, then the obvious question is why not junk the whole thing.

Given modern partisan gerrymandering, one might have reservations about the House’s ability to make a fair and disinterested choice, but one can also imagine that the ravages of partisan gerrymandering could be limited in the future, whether by judicial decisions or by decisions by states who have become disgusted with their own past practices (which have occurred under both Democratic and Republican legislative control). That would do nothing to resolve the fact that the choice in the House is made on a one-state/one-vote basis, which means, obviously, that the single representatives of Vermont, Wyoming, Montana, Alaska, and North and South Dakota have voting power equal to the combined delegations of California, Texas, Florida, New York, Illinois, and Pennsylvania. Given my own politics, I rather like the idea that Vermont’s Democratic representative can counter-balance the predominantly Republican Texas delegation, and Professor McGinnis might be equally delighted that the Wyoming Republican can balance the entire 53-member California delegation, but I think there is no conceivable defense of this allocation of voting power under any 21^st century theory of what American democracy has come to mean. One person/one vote indeed!

But one ought to be under no illusion that we can necessarily predict the way that a single state would vote even if we know the party affiliation of its representatives. After all, a Democrat or Republican who represents a district that voted reasonably heavily for the other party’s candidate might believe that what David Mayhew famously termed “the electoral connection” justifies voting the constituents’ preferences rather than that of the national party. Or one might imagine more nefarious possibilities, as many thought occurred in 1824, where intense wheeling and dealing could take place to induce a given representative, whose vote might determine the state’s vote in Congress, to vote for a given candidate. Ironically, one of the principal defenses of the Court’s egregious decision in Bush v. Gore, made most convincingly by Judge Richard Posner, is that recourse to the procedures set out in the Constitution and relevant statutes passed by earlier Congresses was itself too risky. A coup d’main by the Supreme Court was, from Posner’s perspective, far preferable to reliance on the political process. Anyone who takes this argument seriously should certainly have reservations about the process set out in the Constitution for breaking electoral college deadlocks. Nor should one think that the prospect of a deadlock is very small. The shift of a relatively small percentage of votes in both 1948 and 1968, when Strom Thurmond and George C. Wallace ran strong regional candidacies that received 39 and 46 electoral votes, respectively, would have required a choice by the House, with quite likely political chaos ensuing especially in 1968.

Professor Robert Bennett has recently subjected the Electoral College to his own critique in Taming the Electoral College. Though he is not so unrelentingly hostile to the institution as I am, he nonetheless focuses on yet additional problems, especially the so-called “faithless elector” who might, by voting “independently” rather than complying with the presumptive wishes of the electorate, generate an arguably illegitimate result. Or consider the question of what electors, who are only occasionally what might be termed widely regarded and leading members of the polity, should do if their candidate(s) were to die between Election Day and the meetings in December. As it happens, Professor Bennett believes that statutory solutions might be possible for some of these problems, thus avoiding the futility of attempting reform via constitutional amendment. But this does not in any way reduce the vulnerability to criticism of the Electoral College; it simply suggests that at least some reforms might be easier to attain than one might otherwise think.

I have focused on what might be termed “formal” problems of the Electoral College derived from its legal structure. But there are also pernicious “informal” consequences as well. Professor McGinnis, in a prior encounter, has challenged the common emphasis on the propensity of the Electoral College to elect candidates who did not win a majority of the popular vote by making the entirely valid point that presidential campaigns would not be dramatically different were elections conducted on a nationwide popular vote basis. George W. Bush may have lagged behind Al Gore in 2000, but one cannot confidently infer that the same outcome would have happened had Bush and Gore been contending in a national popular-vote election. He is surely correct, but that point cuts against the Electoral College itself, for it underscores the extent to which contemporary presidential campaigns have become perversely structured around the reality of the College and its generation of so-called “battleground states” that have become the obsessive focus of modern campaigns. In 2004, for example, a full 99% of all advertising expenditures by the two major-party candidates were concentrated in only seventeen of the states. Florida and Ohio alone accounted for more than 45% ($111 million) of the $235 million spent in all of these states. Wisconsin, another “battleground,” received a total of thirty-one candidate visits, as compared with two visits to California. New York received only one such visit! Other ignored states included Texas and Illinois. This means, among other things, that neither candidate was ever required to prepare serious speeches addressing the needs of the largest states in the Union. As a resident of Texas, I can certainly testify to the fact that it is significantly different from, say, Florida, one of the principal “battlegrounds” in both 2000 and 2004. This undercuts the argument that the Electoral College and the purported benefit given to large states by their ability give the winner of a given state all of the state’s electoral votes (and thus deprive the losing minorities of any representation at all in the College) undercuts, at least to some extent, the small-state bonus. Only some large states are “battlegrounds,” and there is no reason at all to believe that the lucky few are necessarily proxies for their ignored sister states.

From a political scientist’s perspective, the Electoral College is an outstanding example of the importance (and costs) of “path dependence,” whereby decisions made at time X become entrenched and shape future politics, for good and, in this case, decidedly for ill. Perhaps it is correct that there is nothing we can do about the College. But that is completely different from a normative defense of the constitutional iron cage built for us by Framers who were necessarily completely ignorant of the shape the country would take in the ensuing two centuries. I look forward to reading the attempts by two esteemed colleagues to defend the indefensible.

I must decline the first invitation of Professor Levinson’s nicely executed attack on the Electoral College. I readily concede that it might well be possible to devise a somewhat better system for electing the president than the Electoral College. Our Constitution is surely not perfect in this as in other respects. But the salient question in politics is always one of alternatives and thus the issue before us is to compare the Electoral College not with the best law that could be enacted but with a range of possible laws that might be enacted. If one believes, as I do, that the Electoral College works pretty well and is risk averse about changing basic political institutions, one will be loath to innovate, given the vagaries of politics and unpredictability of the content of amendments and their interpretations.

In my view, the Electoral College fulfills the three essential criteria that any election for a president must meet in a democracy. First, it measures what can be measured of the popular will sufficiently well to make sure that the president has a popular basis of support. Second, its results tend to give the president substantial political legitimacy. Finally, it enables democracy to fulfill its core but modest function—making it more likely that leaders will govern in the public interest rather than in their own interest or in the interest of narrow parochial factions.

First, all any presidential election system can assure is that the winner is supported by a substantial portion of the electorate. It is impossible in close elections to make sure that any political figure is supported by a stable majority of the electorate. Thus, Professor Levinson’s main complaint—that the Electoral College and popular vote sometimes diverge—is not very powerful, because in those cases the margin of error is greater than any measure of stable popular support. If the election were held a few days later or even if the weather patterns were different, we might well get a different result regardless of the electoral method used. For instance, in addition to the examples of close elections provided by Professor Levinson, many observers think that Gerry Ford would have beaten Jimmy Carter in both the popular and electoral vote if the election had been held a week later. In a world of fickle voters where many make their decisions at the last moment based on evanescent matters, the errors in measuring real support are far greater than the divergences between the Electoral College and other reasonable election systems.

Our current electoral system thus accomplishes what an electoral system can deliver about as well as a popular vote system. It eliminates candidates without a substantial basis of support, like Pat Buchanan or Ralph Nader. Over the course of time it has also measured changes in popular sentiment, allowing shifts to be represented by the policies advocated by those elected president. If one is dispassionate, it is hard to say that the course of American history would have been better with the popular vote, particularly recognizing, as Professor Levinson now candidly does, that the winner of the popular vote might have been different if constitutional rules had made the popular vote decisive. For every person who dislikes the Electoral College winner who may have lost under the popular vote (like George W. Bush), there is another person who reveres the electoral vote winner who might have lost under a popular vote (like Lincoln).

Professor Levinson’s other complaints about the Electoral College also do not show that it fails in any essential function. He argues that that the Electoral College redounds to the advantage of small states. But the states actually most advantaged by the current system are large states. It is true that small states get an advantage due to the “senatorial bonus” (the two electoral votes awarded any state regardless of population). But so long as states vote under rules awarding the winner all the state’s electoral votes, this advantage is overwhelmed by the greater likelihood that the large state will prove decisive. Mathematicians better than I have calculated, in fact, that the voter in California counts more than twice as much as the voter in Wyoming in a presidential election. Oddly enough, the Electoral College compensates in some measure for the gross disadvantages large state voters suffer from the malapportionment of the Senate. I know Professor Levinson objects to the structure of Senate as well (with greater reason than to the presidential election system), but until that structure is changed, he may want to reconsider at least this aspect of his opposition to the College. This point is a small example of a larger truth. Changing complex rules in a reticulated system may have secondary effects not apparent on the surface.

Small states do enjoy a more modest advantage over some middle size states because of the senatorial bonus, but there is no reason to believe that this is likely to be decisive unless the states with one or two members of Congress differ systematically in their preferences from the rest of the Union. And such small differences are likely to be swamped by the inherent difficulty of measuring stable popular sentiment in a democracy.

Professor Levinson also observes that the Electoral College makes candidates pay attention to swing states rather than the nation as whole. This tendency does seem to be a defect. But a system of popular vote will also cause candidates to pay more attention to some voters rather than others, most importantly the voters who cost the least per voter to turnout. Thus, candidates are likely to campaign more in cities than in rural areas and more in affluent areas than poor ones. This defect of the popular system may be even more substantial than the current focus on swing states, because the voters in swing states are likely to be more heterogeneous than voters chosen by low turnout cost. Both the popular vote and the electoral vote will not actually treat all voters equally in a practical sense, but no system can be devised that will have equal effects, given that voters are only equal as matter of law, but otherwise differently situated.

The Electoral College also performs well, and perhaps better than the popular vote, in the second important goal of an electoral system: bestowing legitimacy on the president. It has been noted that the Electoral College tends to magnify the winner’s margin of victory. Particularly among the general public as opposed to the political cognoscenti, this greater margin bestows a greater legitimacy. The sense of legitimacy is especially important for the president because in our system the president is not only the head of government, but the head of state, unifying the nation in times of crisis.

Legitimacy is aided simply by the venerable nature of the system. A large majority understands the basic rules of the game and knows that these rules were not invented with any current election in mind. Given its longevity, even the divergence between the popular vote and the electoral votes does not detract much from the legitimacy of the president. For instance, after the election of 2000, President George W. Bush was seen as legitimate in all sectors of society with the possible exception of university professors. Transitioning to a new system, however, would raise questions about whether its details were structured to aid one or the other of the two great parties. Again, I do not wish to argue that this cost is huge or that in time it would not dissipate, but it is a significant cost not faced by our current system.

Moreover, as Professor Walter Dellinger, an advisor to Democratic presidents, has written, the devil is in the details of a popular vote system. According to Professor Dellinger, third party participation in a popular vote may give third parties greater leverage to distort the results. On the other hand, a runoff election is likely to depress voter turnout in the second round. Professor Levinson appears to advocate a single transferable vote system, but that system presents a voter with at least slight complications and even slight complications confuse many voters, who are inattentive, ignorant, or, as it were, cognitively challenged. Any popular vote system also has other potential problems. If there is a close popular election, we face a national recount, which is potentially more destabilizing than the local unpleasantness of 2000. Perhaps these problems can be sorted out, but I fear it as likely that a national popular vote will replace one set of defects with another. Moreover, there can be no certainty that the optimal reform will in fact be passed.

I address briefly subsidiary complaints. Levinson notes the danger of the faithless elector but himself observes that my colleague Robert Bennett has shown that these problems can be handled by statute. He also worries that the election could be thrown into the House of Representatives where the decision is made by state delegations, each with a single vote. Once again, elections decided by the House are likely to be so close that we cannot tell which candidate would have stable majority but either candidate is likely to emerge with sufficient support to govern. Moreover, even if House determinations were a problem, a relatively straightforward constitutional amendment could provide that a candidate be elected with a substantial plurality rather than a majority of the electoral vote, thus making it highly unlikely that elections would ever be thrown into the House.

In closing, let me suggest that the debate about the Electoral College today is as much about the purpose of democracy as its mechanisms. One reason that I am not bothered by the Electoral College is that I have what David Held has called the “protective idea of democracy.” Under this conception, democracy is largely valued as a mechanism to assure that government will not be used to advantage a distinctive class of rulers. Protective democracy diffuses power through society so that politics will not interfere with, but simply sustain the framework for, the sources of real happiness—exchanges within the family and the market. The current method for electing the president in conjunction with our legislative elections accomplishes this modest goal.

But if one believes in what Professor Held describes as participatory or social democracy, the Electoral College may seem very unsatisfactory, particularly in its symbolism. Oversimplifying a bit, under the model of participatory or social democracy, the legitimacy of all social institutions ultimately depends on popular approval and should be subject to a continuous process of social reform. Thus, even if it is impossible to capture a stable majority with the election of a president, it is important to appear to do so if majority will is the source of legitimacy. Accordingly, it is not surprising that as the United States flirted with social democratic ideas, the Electoral College, which reflects the Framers’ more circumscribed view of democracy’s purpose, has come under attack. In a recent book, Professor Levinson himself combines his attack on the Electoral College with an assault on other features of the Constitution, like the presidential veto, which are in fact designed to insulate the order of civil society from rapid social reform by majority will.

But if one believes, as I do, that social democracy is seriously wrong-headed and that majority rule in political decision making is far from sacrosanct, one will welcome the symbolism of the Electoral College. Its structure on its face rejects the notion of unmediated popular sovereignty—a notion that has made it harder to recognize that society is legitimated not by majority will, but by principles of natural justice, like the right to liberty and property. By giving states a role, the Electoral College includes within its structure elements of subsidiarity—a principle of governance that facilitates the exercise of these rights. Thus, I count it as a virtue, not a defect, that the symbolism of the Electoral College reminds us that simple majority will is not the legitimating feature of society, but that instead popular consent is merely an instrument to protect the deep and enduring principles that make us a free people.

Rebuttal by Daniel H. Lowenstein — Five Inconceivable Reasons To Support the Electoral College

Professor Levinson says the “only conceivable argument” for retaining the Electoral College is a “basically Burkean one,” namely that the Electoral College is “our unique method” (Professor Levinson’s emphasis). Professor Levinson is unfair to Edmund Burke, who was one of the great reformers of his day and would have scoffed at the suggestion that an obnoxious practice should be preserved because it is our way of doing things.

But there is a different aspect of Burke’s thought that is basic to Professor Levinson’s and my different views on the Electoral College. Burke saw government as a practical art, to be approached with circumspection or boldness as the occasion may require, but always informed by experience and prudence. He had little tolerance for the “Sophisters” of his time who sought to design governments on theoretical and mathematical bases.

Professor Levinson’s main objections to the Electoral College as an instrument of government are distinctly of the theoretical and mathematical kind. Thus, he worries at length that many presidents have been elected with less than a majority of the popular vote and a few have been elected with less than a plurality of the popular vote. He also agonizes that votes cast in small states are given more weight than votes in large states. As Professor McGinnis shows, it is doubtful whether Professor Levinson can hold his own in that part of the field. Even if he can, these seem to me to be minor concerns.

Admittedly, “rule of the majority” is a popular slogan, giving the opponents of the Electoral College a leg up in popular debate. So far as that goes, I believe in majority rule at least as much as the next person—unless the next person happens to be one of Burke’s Sophisters. But the “republican principle” pointed to by Madison in Federalist 10 was not abstract majoritarianism, as Madison made abundantly clear in that essay and throughout his contributions to The Federalist. The popular principle at the heart of Locke’s Second Treatise, repeated in the Declaration of Independence, is consent, not majoritarianism. As can be seen from Professor McGinnis’s essay, it is risible to suggest that the Electoral College as it operates in the United States is a departure from the principle of consent or from the republican principle.

Before I give reasons for supporting the Electoral College, let me respond to Professor Levinson’s “challenge”: Would I advise a new democracy to adopt the Electoral College? No. I am not smart enough to tell a new democracy how to structure itself. Were I transported by a time machine to the Philadelphia Convention of 1787, would I sit quietly while the Convention incorporated the Electoral College into the proposed Constitution? Yes. I think that upon emerging from the time machine, even I would have enough sense not to lecture the likes of Madison, Franklin, and Washington on the principles of government. True, even those wise men were not wise enough to foresee how the Electoral College would operate, but I would not imagine that I could do better than they.

As I turn to the affirmative case, I am confronted by Professor Levinson’s dictum that there are no conceivable arguments for the Electoral College other than the faux-Burkean argument I have eschewed. Therefore, I offer five inconceivable reasons to preserve the Electoral College.

1. The Electoral College turns the many winners who fail to win a majority of the popular vote into majority winners. It also magnifies small majorities in the popular vote into large majorities. These effects of the Electoral College enhance Americans’ confidence in the outcome of the election and thereby enhance the new president’s ability to lead. Professor McGinnis addresses this point effectively, so I shall not elaborate further.

2. The Electoral College causes presidential elections to be significantly oriented around states. Perhaps because I spent eight years in the state government of California before I became an academic, I share many of the popular prejudices about the inside-the-Beltway mentality and the arrogance of the federal government. Against all the pressures of nationalization, it is important to maintain the states as strong and vital elements of our system, both in practice and in public understanding. Unlike some conservative jurisprudes, I do not believe constitutional limits on the powers of the federal government are a promising way to accomplish this. In practice, the Electoral College is by no means the most important institution we have for strengthening the states, but neither is it by any means the least important.

3. The Electoral College produces good presidents. True, we’ve had a few lemons and a larger number of unmemorables. But we’ve had a remarkable number of remarkable leaders. The Electoral College has produced Washington, Jefferson, Jackson, Lincoln, Cleveland, Theodore Roosevelt, Wilson, Franklin Roosevelt, Truman, Eisenhower, and Reagan. None of us may approve of all of these without qualification, but taken together, it is a pretty big group of distinguished chief executives. Probably the only country in the world that could point to an arguably comparable set of chief executives is the United Kingdom, and they, like us, elect their executives indirectly.

Professor Levinson might respond to this history in a manner akin to W.S. Gilbert’s encomium to the House of Lords, which included the following:

When Wellington thrashed Bonaparte, as any child can tell,
The House of Peers throughout the war,
Did nothing in particular,
And did it very well.

In other words, Professor Levinson might argue that the Electoral College is not the cause. These or equally distinguished presidents might have been selected by a popular vote. So they might have. But “did” is better than “might have.” A lot better, in the eyes of us anti-Sophisters.

4. As I mentioned above, the kind of Burkeanism I subscribe to is the kind that places a high premium on learning from experience. The experience of the last several years has made two additional benefits of the Electoral College evident.

The first is the lesson from Florida 2000, that the Electoral College has the great merit of confining such election conflicts to one state (or a few, as in 1876). It is true that there is some weight on both sides of the scale on this question. A squeaky close race in a single state is more likely than a squeaky close vote in a national popular election, because of the smaller number of total votes. The greater likelihood of a contest at the state level no doubt holds, even when one adds that it needs to be a pivotal state in the Electoral College. Point for Professor Levinson.

But the greater point is that both types of problem are very unlikely, though both are also possible. So the tiny absolute difference in the probability is outweighed by the far more catastrophic effects that would occur if we were confronted by a Florida-type conflict in which the whole country was up for grabs. Game for the Electoral College.

5. As recently as a year or so ago, I was saying that although I support the Electoral College, if I had the power I’d reform it to avoid the “faithless elector” problem, which both Professors Levinson and McGinnis address. I was wrong. If we solved the faithless elector problem by making the Electoral College an automatic process rather than one depending on humans, we would lose what might some day turn out to be the Electoral College’s greatest benefit. (There may be other ways to solve the faithless elector problem without the implication I discuss here, but I’ll let that pass for now.)

I should have learned this lesson from the controversy in the 2002 New Jersey Senate election, when Senator Robert Torricelli withdrew his candidacy for reelection after having been nominated in the Democratic primary and after the apparent deadline under New Jersey law for the naming of a replacement candidate. (As most readers probably know, the New Jersey Supreme Court bent the New Jersey statutes and its own precedents into a pretzel for the good reason of allowing the Democrats to name former Senator Frank Lautenberg as a replacement candidate. Lautenberg was elected.)

It took two comparable incidents in 2006 for the point to penetrate my skull.I refer to the Republicans’ unsuccessful efforts to replace House nominees on the ballot who stepped down in Texas and Florida because of scandals. The point is that nominees may die, become disabled, or for other reasons become manifestly unsuitable for the office. Serious though the problems were in the three cases to which I have referred, they were trivial compared to the problems we would face if a presidential nominee—or, even more of a problem, a presidential winner after the election—died, became disabled, or became manifestly unsuitable for any of a variety of possible reasons.

True, one might try to legislate on this problem directly (and probably we should, at least to cover the period after the electoral votes are cast). But it would be tough. There would be three problems: defining the circumstances in which the legislation would go into action; establishing a process for applying the definition to the actual circumstances; and deciding how a replacement should be chosen. These would all be hard problems to solve in legislation, where some neutral or at least objective standards would be needed. True, death would be easy enough to determine and disability would not be too difficult, but manifest unsuitability would be a tough enough nut, I think, to withstand any statutory nutcracker.

What is needed for such problems is a political solution. And the Electoral College is ideal for the purpose. The decision would be made by people in each state selected for their loyalty to the presidential winner. Therefore, abuse of the system to pull off a coup d’etat would be pretty much out of the question. But in a situation in which the death, disability or manifest unsuitability plainly existed, the group would be amenable to a party decision, which seems to me the best solution.

To be sure, it is not a particularly likely situation. But three major incidents in a space of four years make it clear that it is not out of the question. The Electoral College greatly shortens the period during which we would have no handy means of coping with the problem. Even that period could be eliminated if we allowed a very high percentage of the electors who voted for the winner—say, 90 percent—to petition to reconvene the Electoral College any time up to the inauguration. Giving up this and the other benefits I have identified in exchange for Professor Levinson’s mathematical niceties strikes me as a very bad bargain.

I have enjoyed firing off inconceivable arguments and look forward to what will surely be a stimulating response from Professor Levinson.

I should say at the outset that what most disappoints me about the responses of both Professors McGinnis and Lowenstein, whose work I greatly respect, is their tone of utter complacence. My own view is that the Electoral College is only one aspect of our Constitution that leads me to analogize our present situation to driving a car with slick tires and very bad brakes. Even if this is somewhat hyperbolic, I believe that anyone should recognize the core of truth in the analogy. There is something profoundly troubling about denying past perils and inferring future success from what is truly lucky past performance. Recall, for example, that little more than blind luck saved us from an election thrown into the House of Representatives in both 1948 and 1968, not to mention the fiasco of 2000. I am confident that neither of my respondents would give to their children such a car; what mystifies me is their almost casual acceptance of the risks posed by the Electoral College.

I take comfort in the fact that neither Professor McGinnis nor Professor Lowenstein is really willing to defend the electoral college system in toto. Both, for example, seem to recognize problems attached to the allocation to the House, on a one-state/one-vote basis, of the choice of a president should the Electoral College deadlock. Although it would scarcely satisfy me, I wonder if they would suspend their risk aversion regarding contemplation of constitutional amendment with regard to simply changing the House vote to a one-member/one-vote basis; this would surely enhance considerably the odds that the person chosen could make a plausible claim to majoritarian legitimacy. (We could, incidentally, have an extended debate about the notions of “consent of the governed” suggested by Lowenstein and McGinnis, especially the latter, who is commendable in his candor about his relative uninterest in “majority rule.”)

But, of course, I want to extirpate the College root and branch, so I return to that more central debate. Professor McGinnis writes that the College “mak[es] it more likely that leaders will govern in the public interest rather than in their own interest or in the interest of narrow parochial factions.” I confess that I literally do not understand the connection between the College and such an optimistic conclusion. Indeed, I think the evidence has become absolutely clear that the president's incentive is to pander to the interests of “battleground” states, particularly in the era of what has come to be called the “perpetual campaign.” Consider only the scandalous trade policy adopted by the Bush Administration solely for the purpose of appealing to vulnerable steelworkers in Ohio. Given my own politics, I share the concern for those steelworkers, but there is no conceivable argument, other than the desire to gain Ohio in the 2004 election, justifying the Bush Administration's protectionism. This is one time I found myself agreeing with the editorial page of the Wall Street Journal. Is it possible that Professor McGinnis disagrees? Nor, I think, has the mindless determination, on the part of Democratic and Republican presidents alike, to carry on the dysfunctional boycott of Cuba, served the country well, even if it has served to attract votes of Cuban-Americans in Miami and Jersey City. This is exactly what is meant by governing “in the interest of narrow parochial factions.”

This is probably the appropriate place to note that Professor Lowenstein's encomia to presidential elections that are “significantly oriented around states” is also odd, especially coming from a dedicated Californian like himself. If there is one thing we can be certain of, it is that none of the last three presidential elections has been remotely “oriented” around California, which has been viewed, altogether correctly during this period, as a safe Democratic state. The last elections, by objective measure of advertising expenditures and time spent by candidates, focused on no more than 17 states, not including, recall, California, New York, Illinois, and Texas. This fact also places in a different perspective Professor McGinnis's assertion that “candidates are likely to campaign more in cities than in rural areas . . . .” That may well be so, but surely we can all agree that all cities are not the same. To take several nonrandom examples, Los Angeles, Chicago, Austin, and Philadelphia are all considerably different from, say, Miami, Milwaukee, St. Louis, and Columbus. It is the latter cities that were the object of the candidates' maximum solicitude and not the former, which could all be taken for granted by either Bush or Kerry.

Professor McGinnis's argument about close elections—say, the Carter-Ford election of 1976 or the Bush-Gore election of 2000—has some substance, though it remains unclear why the person coming in first shouldn't prevail. But his argument fails totally when we take into account elections with significant third-party candidates, as occurred, say, in the elections of 1824, 1860, 1912, 1924, 1948, 1968, and 1992. In six of these seven elections, the third-party candidate won both an impressive number of popular votes and at least some electoral votes. Ross Perot was an unusually successful third-party candidate if measured by his percentage of the popular vote, but, of course, the fact that he ran a national campaign rather than a regional one, like George C. Wallace or Strom Thurmond, meant an electoral vote wipeout. In all such situations, I am confident that we would be better off with a national election with voters casting an alternative transferable vote than the present system. Or, frankly, I would settle, as an interim measure, for states adopting the ATV, so that the winner of a state's electoral votes would no longer be “first past the post” but would, instead, be able to claim plausible support by the majority.

I do not have the methodological training to debate whether large states or small states are more benefited by the College. Perhaps all of us can agree that states like my native state of North Carolina and other medium-size states are unequivocally penalized by the College, getting neither the indefensible “senatorial bonus” that small states get nor the advantages of indefensible aggregation that large states receive. As to the latter, moreover, the fact is that the aggregation practice means that my Democratic vote in Texas and Professor McGinnis's presumably Republican vote in Illinois are equally pointless and irrelevant. But this is simply to restate the problem with the “battleground state” phenomenon that has so corrupted our presidential politics in every way. Disaggregation would change things considerably. Maine and Nebraska have led the way. Would Professors McGinnis and Lowenstein support similar disaggregation in all the states?

Professor McGinnis says that we can ignore all such problems because the Electoral College helps to “bestow[] legitimacy on the president” by “magnify[ing] the winner's margin of victory.” Professor Lowenstein agrees. I have no doubt that this was true in the past. But, for better and maybe even for worse, I suspect that most informed American citizens are now all too aware of this conjurer's trick, which can work only if one doesn't know how it is performed. By now, most Americans can see the Wizard-of-Oz aspects of the College. In any event, I think there is little empirical evidence that American presidents enjoy greater legitimacy than, say, British prime ministers or even French presidents because of the distortions produced by our unique system of electing chief executives.

Moreover, it is scarcely the case that at the beginning of his first term George W. Bush was “seen as legitimate in all sectors of society with the possible exception of university professors.” A CNN poll published on April 24, 2001, revealed that 48% of the public believed that he had not won the 2000 election “fair and square” and 23% of the public refused to accept him as a “legitimate” president. That surely goes well beyond university professors like myself. Perhaps the “good news” is that 76% of the public accepted its legitimacy, but I don't think this is enough to vindicate Professor McGinnis's appreciation for the legitimacy-enhancing features of the College. (I'm also confident that if the public at large knew in April, 2001, what kind of president he was going to become, they would have rioted in the streets, but that's another matter.)

Professor Lowenstein forthrightly admits that he would not advise any country today to emulate the College. Nor, I take it, would he advise its adoption were the United States itself to be in the “Constitution adoption” business. So the argument is that we should feel bound to an ancient (or at least 218-year-old) practice because it was good enough for the denizens of the Philadelphia Convention. He invokes the authority of Madison, Franklin, and Washington and says, with what I am tempted to describe as ostentatious modesty, that he “would have enough sense not to lecture” them “on the principles of government.” The obvious question is why not? The central greatness of the Framers—and the lessons they should teach us today—was their openness to the “lesson of experience,” a value that Lowenstein himself acknowledges. These were, after all, revolutionaries, in the case of Washington an armed revolutionary. They were all impatient with similar arguments from authority made by those who counseled continued acceptance of the British Constitution and the supremacy of King George III. Ironically, even Edmund Burke himself supported the Revolution, though one suspects it was largely on prudential grounds. I am far from calling for a revolution. What I would like is something more than arguments from authority based on absolutely outmoded political understandings of 1787.

To be fair, Professor Lowenstein does indeed offer five reasons to preserve the Electoral College. Some I've already addressed. But consider his belief that the College helps to “strengthen[] the states.” I fail to see how. This is fallacious in the same way that it is fallacious to argue that the Senate strengthens the states. The Senate acts as an affirmative action program for the residents of small states, nothing more, nothing less. In no conceivable sense does it strengthen states as such. If one wanted to do such a thing, the German Constitution offers a far more promising way, by drawing the officials of the Bundesrat from state officialdom itself. Or we might repeal the 17^th Amendment and return to the original scheme of selection of senators by state legislators.

Indeed, I do wonder what Professor Lowenstein would have argued in 1912 if a hothead like me had argued for the desirability of getting rid of the original system of selecting senators. Would he have said that what was good enough for Madison, Franklin, and Washington was good enough for him? Indeed, such arguments have been made with regard to almost every single amendment to the Constitution, including the Reconstruction Amendments and the 17^th and 19^th Amendments. I really don't believe for a moment that Professor Lowenstein would have stood with reactionaries against such amendments, but the logic of his more liturgical language regarding the Framers does make one wonder.

Professor Lowenstein raises a genuinely interesting point with regard to the importance of retaining human electors instead of going to a completely mechanical process. It is true, after all, that circumstances could arise in which we might appreciate the presence of human beings who could make important choices, including, for example, who should receive the votes in the case of the death or discrediting of the initial nominee. My principal problem with his argument is that most presidential electors are selected to reward loyal servants of the party. It is hard to believe that one would look to such individuals for the requisite sagacity and wisdom to make such choices. Indeed, I strongly suspect that such decisions would be made by the national committees of the respective parties, who would expect (and generally receive) the literally thoughtless loyalty of “their” electors. Imagine, though, that a small number of electors reject the choice and coordinate their votes for a third candidate in circumstances that would then throw the election into the House.

I want to conclude with some comments about the most important single election in our history, the election of 1860 that put Abraham Lincoln in the White House. Mark Graber has demonstrated, in his brilliant book Dred Scott and the Problem of Constitutional Evil, both that Lincoln's election was an artifact of the Electoral College—he received, after all, only 40% of the popular vote—and that it triggered secession and then war. Perhaps this is a merit of his election, since the slaughter of two percent of the population did lead to the abolition of chattel slavery. But, at the very least, we must recognize both that elections can have the most fundamental of consequences and that their outcome may be dictated in large measure by the structure that the United States, alone among the nations, has adopted. Shouldn't we at least consider the possibility that the rest of the world might actually have something to teach us about the meaning of democracy-and possibly even of how to achieve stability-in the 21^st century?

Professor Levinson's reply does not show why the Electoral College fails to achieve the rather modest goals that a presidential election system can fulfill, such as excluding candidates without the popular support to govern. Moreover, he still does not even try to demonstrate systematically that the typical popular election system would have fewer defects than the electoral college. It is only though such comparisons that one can plausibly make the case for junking a structure which has gained legitimacy by its longevity. It is striking that Professor Levinson does not provide an example of a popular presidential electoral system in any foreign jurisdiction to show how that system has contributed to that nation's welfare in some concrete way that has eluded our electoral system.

To recap: Professor Levinson's principal argument remains that the Electoral College system misfires and elects individuals who do not enjoy majority support. My principal response was that in a democracy with a vacillating and fickle electorate, no electoral system can assure that in close elections the winner has a stable majority of the electorate. Professor Levinson concedes this point generally but suggests that third parties may make a difference. But in none of the elections he mentions does he show that the presence of a third party prevented the election of a president who would have enjoyed stable majority support. The elections were close enough that neither of the main party candidates enjoyed that level of support.

The difficulty of getting a stable majority in an election with a third party is not an accident of history. When third parties are strong, a large body of voters refuse to attach themselves to either of the two major party candidates, because there is something palpably unattractive about both of them to an important part of the electorate. This free floating-bloc of voters tends to turn the race even more volatile, making the exact time of the election and other accidents of campaigning even more crucial than in a two-person election. Thus, there is even less reason to expect that any system can assure the victory of a candidate with a stable majority of popular support when third parties are running.

Professor Levinson also fails to compare the perceived defect in the electoral college with the alternatives. The distortion of third parties affects popular elections no less than the electoral college. Witness the last French presidential election in which the first round second-place finish of the protest candidate Jean LePen prevented the electorate from being able to choose between Jacques Chirac and a contender of the center left, although the center left had far more support than LePen and perhaps more than Chirac. Professor Levinson does offer an idea for avoiding that kind of problem: a system of single transferable vote. But since no major democracy has ever adopted such a presidential system, it would seem unreasonable to jettison the electoral college on the supposition that the United States would choose his particular scheme. And, as I discussed before, the single transferable vote system is likely to confuse a large number of voters whose intentions have been defeated by even slighter complications.

Levinson continues to complain that the electoral college reduces the influence of midsized states vis-à-vis smaller states. I conceded as much, but noted that it actually gave larger states disproportionate influence in the presidential election, and that additional influence was important in counterbalancing their gross underrepresentation in the Senate, thus making the electoral system fairer as a whole. But Professor Levinson fails to address this point at all. If one wants to look at the fairness of the electoral college, one needs to consider its effects on the entire political system and not focus on the fairness for one group of states without reference to the rest of our structure of political governance. Professor Levinson's approach is like that of a economist who compares the well-being of individuals in a society by looking at only certain sets of individuals and only one kind of income. Just as no scholar would do welfare economics so myopically, no one should evaluate electoral fairness by looking at only some of the states and part of the political process.

Professor Levinson also continues to worry about the electoral college's tendency to focus attention on swing states. Here he does at least compare the additional influence the electoral college gives swing states and the additional power a popular election system would give to sectors of society, such as cities, with low turnout costs. But he is plainly wrong that cities are more heterogeneous and representative of America than swing states. Cities are overwhelmingly Democratic and disproportionately minority. Moreover, they have a similar focus on urban issues and reflect, as they have since the rise of Babylon, more cosmopolitan values. (There is nothing wrong with these characteristics: I count myself a hyper-urbanite because of some of them.) In contrast, a swing state like Ohio has rural and urban voters, cosmopolites and yeomen. If some sectors are going to inevitably get more attention, it is best that they look more like America, to quote reference a recent president, because they are less likely to distort outcomes that the median voter would choose. It is absolutely true that the president will do particular favors for swing states under the electoral college system, but he would do similar favors for cities under a system of popular elections. (That is why some previous popular electoral systems were malapportioned to favor rural counties.)

Finally, Professor Levinson notes that polls suggest that a substantial portion of Americans did not think President George W. Bush won the election of 2000 fairly. But such polls do not suggest that people did not regard him as a legitimate president just as the doubts of many about aspects of the election of Rutherford Hayes or John Kennedy did not translate into general assaults on those presidents' legitimacy. In a prominent law review article Professor Levinson himself referred to George W. Bush as the “new occupant of the White House,” consciously withholding the title of President. His was indeed a denial of legitimacy, but I continue to believe that such extreme views were largely limited to the academy. And no democratic system can prevent unhappiness with the fairness of certain elections or even preclude small minorities from deriding the legitimacy of the winner.

Levinson's continuing discomfort with the results of the 2000 election is an example of one impetus for criticizing the electoral college. That kind of criticism dissipates as the memory of a particular election fades. The other source of criticism stems from social democratic enthusiasm for popular majoritarianism. But that kind of criticism may also dissipate as political scientists make clear the limits and tradeoffs of any electoral system. Thus, the electoral college will likely remain our system of presidential election for the foreseeable future and for good reason: it has served the nation well. It has resulted in the elections of presidents with enough popular support and legitimacy to govern. And its refusal to follow simple majoritarianism reminds us that popular consent is not the end of government, but simply a means to assure that government respects the liberties of life and property that make us a free and prosperous people.

Closing Statement by Daniel H. Lowenstein

In his opening statement, Professor Levinson based his argument against the Electoral College primarily on its deviation from pure majoritarianism. He pointed out that many American presidents have been elected with less than a majority of the popular vote and a few have been elected without a plurality. Professor Levinson was right to make this his main argument. I believe the deviation from pure majoritarianism is far and away the most important cause of opposition to the Electoral College and the argument has undoubted appeal to many Americans. Indeed, at first blush it probably has significant appeal to almost all Americans, who are brought up to believe the majority should rule.

In my opening statement I argued that the Electoral College yields results that are consistent with Madison's “republican principle,” in that it does not represent a major departure from majoritarianism, whatever the results may be in a particular close election. I rejected the idea that abstract, mathematical niceties should prevail over the practical workings of democratic government. Though Professor Levinson had denied that there were any arguments for the Electoral College other than fear of changing the status quo, I described five practical benefits, any one of which seemed to me to outweigh the purely abstract drawback that, when the election is close, on infrequent occasions the winner of the election may not be the winner of the popular vote.

In his response, Professor Levinson does not restate his abstract majoritarian argument against the College, presumably because he is confident—and perhaps rightly so—that more readers will share his mathematicism than my pragmatism. Be that as it may, I am content to join him in leaving that discussion where it stands.

Professor Levinson does reiterate another point, for the good reason that I did not address it in my opening statement. The Constitution provides that if no candidate receives a majority of the electoral votes, the president is selected by the House of Representatives, in which the delegation from each state is apportioned one vote. Never mind that the last time this happened was in 1824. We are taking the long view, and Professor Levinson is right to claim that it might happen again. It is a prospect he views with horror. I do not.

Referring back to his opening statement, we find that there are two reasons Professor Levinson fears selection of a president by the House. The first is that even if each House member's vote counted equally, the winner in the House might not be the winner in the popular vote. The state-by-state voting makes that possibility more likely to occur. But this is simply a reiteration of Professor Levinson's abstract mathematicism, which for reasons I have explained should be given little or no weight. The situation we are considering will arise rarely if ever and only when there has been no majority winner. Under those circumstances, there has to be some way to choose the president, and selection by the House seems as good as any other. Nor do I think it makes much difference which voting formula is used in the House.

Professor Levinson's second reason for fearing the House is that we cannot know how the House members would vote. In the foreseeable future, I think it is a pretty safe bet they would cast their votes on strict party lines. Professor Levinson thinks they might not and, of course, we cannot be certain he is wrong. In that case, the House would function more or less in the deliberative manner that the Framers expected the electors themselves to act. Call me complacent, as Professor Levinson has already done, but I simply do not react with fear and trembling to the (improbable) idea that House members will bolt party lines to vote for the person they or their constituents think would make the best president.

Professor Levinson also objects to some of my five reasons for supporting the Electoral College. Let us review them.

1. My first point was that the Electoral College turns popular vote plurality winners into majority winners and magnifies the margin of victory for winners who receive a slight majority of the popular vote. Professor Levinson concedes that in the past this effect enhanced presidents' ability to lead but regards it as a conjurer's trick that could work only when Americans did not understand how it was performed.

If the Electoral College were a conjurer's trick, it would be a very bad one, because it is quite obvious how it works. Indeed, I suppose it is still the case that it is taught in high school or earlier, and Professor Levinson does not explain what Americans understand about it now that they did not understand in the past. In fact, the College is not a conjurer's trick. Professor Levinson is fixated on the national popular vote—a mere statistical artifact, though of course an interesting one—that has no bearing on the election of a president. He therefore does not see that the Electoral College is the real way in which a president gets elected. The majority of electoral votes that the many plurality popular vote winners received were real majorities, not phony ones.

2. My second point was that the College causes presidential elections to be oriented around the states. Professor Levinson does not deny that this is the case. How could he? Rather, he denies that orienting presidential elections around the states has the effect of strengthening the states. His only support for this denial is the irrelevant claim that the institution of the Senate does not strengthen the states. It is therefore difficult to be sure what he has in mind, but I suspect that he interprets me as having argued that the Electoral College increases the power of the states in national politics relative to other interests. In that case, his denial is quite reasonable, but that is not what I claimed. I claimed, and continue to claim, that by orienting our most visible and consequential political contest around the states, the College modestly “maintain[s] the states as strong and vital elements of our system, both in practice and in public understanding.” This is not a claim about relative power. It is a claim about the actual and perceived importance of the states in our system of government.

3. My third point was that the Electoral College has produced good presidents. I named eleven presidents who probably could not be matched for distinction and achievements by a similar list of national leaders over the same time period in any other country on earth, with the possible exception of the United Kingdom, which also selects its leaders indirectly. Professor Levinson has not responded to this point.

4. My fourth point was that we ought to have learned from the 2000 post-election contest that the Electoral College confines any such controversy to one or, at most, a few states. A similar controversy in which every vote cast in the entire country could be questioned would have been many times more difficult. I conceded that an offsetting consideration is that a virtual tie in the national popular vote is statistically less likely than in the popular vote of any state, even acknowledging that the state has to be pivotal in the electoral college. However, the difference in absolute probabilities is very small (because the probability of either happening is very small). Therefore, that difference is strongly outweighed by the far more dangerous crisis that would be created if presidential elections were not organized by states.

Other than a brief reference to the 2000 controversy and its aftermath, which does not bear on the substance of my argument, Professor Levinson has not responded to my fourth point.

5. My fifth point was that in light of the New Jersey Senate controversy in 2002 and the inability of Texas and Florida Republicans to replace House candidates on the ballot who resigned in scandal in 2006, we should appreciate the Electoral College's ability to provide a political remedy if we should ever be confronted with the death, disability, or unsuitability of a major party presidential candidate before the election or of the winner after the election. My only proposed reformation of the Electoral College—which I believe could be effected by statute, though that is not clear—is that after the electors have met and cast their votes, signatures of a high percentage of the electors would result in a reconvening of the Electoral College at any time before the inauguration. I suggested 90 percent in my opening statement, but perhaps 75 or 80 percent would be more prudent. Whatever the required percentage, the procedure would extend the protection offered by the Electoral College through the entire period of vulnerability.

Professor Levinson is good enough to describe this as a “genuinely interesting” point. Since the problem it addresses—admittedly not especially probable, but by no means impossible—would unquestionably be serious if it arose, why is this not alone a sufficient reason for preserving the Electoral College? Professor Levinson's answer is that electors are “selected to reward loyal servants of the party.” He ought to have added that they are also selected because they are loyal supporters of the party's presidential candidate.

As I shall demonstrate shortly, these are reasons to favor the electors as the group to select a new nominee or president in the event of emergency, not to reject them. First, let me address Professor Levinson's dismissal of the electors as a group lacking in sagacity and wisdom. He provides no support for his attack on the electors. A list of 2004 electors can be found at http://en.wikipedia.org/wiki/List_of_United_States_presidential_electors,_2004. The great majority are identified by name only, but the information provided for some does not seem to match Professor Levinson's characterization. Rather, it suggests that many electors—easily enough to provide the leadership that would be necessary in an emergency—are experienced persons in responsible positions. The three electors from South Dakota were the governor, lieutenant-governor, and attorney general of the state. Identified electors from other states such as Alabama, New Jersey, New York, Pennsylvania, and Tennessee include state legislators and legislative leaders, mayors of major cities, district attorneys, party national committee members, and so on.

It is true, of course, that they are chosen for loyalty to party and to the presidential candidate. That is a good thing, not a reason for rejecting the Electoral College. If the time ever comes when overturning the result of a presidential election is to be seriously considered, it should not be done lightly. The fact that the electors are loyal to the party and to the candidate provides assurance that if they in fact do overturn the election it will be only because it is truly necessary. Professor Levinson says that in fact the decision would probably be made by the pertinent national committee. That might well be the case, and there would be nothing wrong if it were. Professor Levinson says the national committee would expect the electors' “literally thoughtless loyalty” (what would figuratively thoughtless loyalty be like?). The committee might “expect” such loyalty, but to get it I think they would have to act responsibly.

Finally, Professor Levinson says that a small number of electors might act independently, preventing any individual from getting a majority and throwing the election into the House, which, as we have seen, is for Professor Levinson a fate worse than torture by the Spanish Inquisition. For the rest of us, the House would simply be another locus for the political solution that would be sorely needed. The essential point is that if the country ever finds itself with a major party presidential nominee or, worse, a presidential winner who dies, is disabled, or otherwise becomes unsuitable, it will need a political means of obtaining a substitute. The Electoral College assures that such a political means will be available. If we repeal the Electoral College, we eliminate that solution and open ourselves to a potentially serious crisis.

In sum, Professor Levinson has not succeeded in undermining any of my five reasons for supporting the Electoral College. The issue comes down to how one thinks about political institutions. For those who think abstractly and mathematically, the Electoral College is offensive to democratic dogma. All the practical benefits in the world will fail to ameliorate that offensiveness. Those of us who see government as a practical enterprise will resist tearing down an institution that, however surprisingly, fits well into our system and fortifies it in numerous and diverse ways.