California is the home of direct democracy, recalls, Mickey Mouse, and a devilishly ingenious stratagem to neuter the electoral college and elect presidents by nationwide “popular vote,” or more accurately, by plurality rule. The ingenious thing is how it bypasses the need for a Constitutional amendment. It does that by exploiting the language of Article II:
“Each state shall appoint, in such manner as the Legislature thereof may direct, a number of electors, equal to the whole number of Senators and Representatives to which the State may be entitled in the Congress.”
Like every other state but Maine and Nebraska, California currently appoints those electors by the winner-take-all rule: whoever wins the most popular votes statewide wins all 55 of California’s electoral votes. But by a law passed in 2011, California has created a compact (joined by nine other states and DC) to award all its electoral votes instead to the national “popular-vote winner.” The compact would kick in once it had attracted states commanding at least 270 electoral votes (165 so far), a majority of the total. Footnote 1
The result would be a procedural calamity in pursuit of quixotic goals. After describing the mischief that the compact invites, we examine constitutional challenges and the arguments of proponents, then show how the supposed drawbacks of the current system can be remedied more effectively and less dangerously by other means—if indeed they are drawbacks worth remedying. The underlying problem is that it is hard to achieve the effect of a Constitutional amendment without amending the Constitution.
MISCHIEF
This is of five kinds.
Legal Instability
Its champions brag that the compact requires no Constitutional amendment, but for that very reason it would give us a political system wanting in durability and predictability. If states can join the compact by legislative act, they can as easily withdraw. A few withdrawals, even a single one, could break the compact by reducing its electoral votes to fewer than 270. Withdrawal would most likely follow a state election that replaced Democrats with Republicans as dominant party. Proponents of the compact tout bi-partisan support, but the states that have ratified it are all bright blue, and as tables 1 and 2 show, it is rather opposition to the compact that is somewhat bipartisan: the legislators who voted for it are almost all Democrats. The compact has been vetoed, moreover, by four governors, all Republicans, and signed by nine, all Democrats. Footnote 2 And the 2012 national Republican Platform opposes any “scheme to abolish or distort the procedures of the Electoral College.”
Table 1 Final Passage Votes for the National Popular Vote Interstate Compact, Lower Chamber
* Seven Republicans and five Democrats abstained from voting.
Table 2 Final Passage Votes for the National Popular Vote Interstate Compact, Upper Chamber
From the point of view of electoral advantage, a state might as well withdraw (or not join to begin with) because the compact would make a difference only in a pivotal state forced to elect a president opposed by most of its voters. True, having joined the compact to encourage others to join, a state might be reluctant to withdraw without knowing in advance that it is both pivotal and on the wrong side of the national popular vote. But it can always withdraw after popular votes are counted but before the Electoral College meets.
The compact does ban withdrawals during the six months before a president’s inauguration, but it provides no penalty. Interstate compacts can be enforced by Federal courts, but enforceable ones require Congressional consent according to Article I, Section 10 of the Constitution. Not only is that not in the offing but the compact itself does not make it a condition of implementation. Maybe the legislators and governors who originally ratified the compact feel bound to enforce it. But in all likelihood they would be long gone when the issue of enforcement arose.
One might contend that withdrawals that terminated the compact by reducing its electoral votes to fewer than 270 would merely restore the old system at no cost. But the looming prospect of withdrawal would create uncertainty: voters, parties, and candidates would not know for sure which set of rules would govern their game. Imagine the effects if a pivotal state withdrew during a presidential election campaign.
Uncooperative Electors
Faithless electors, ones who do not vote for their own party’s nominee, have been so rare that they have never come close to making a difference. Since 1796, there have been just nine clear cases of infidelity. Footnote 3 But consider what can happen under the compact. Say a state is bright blue but the Republican candidate for president noses out the Democrat nation-wide. Then the state’s chief election officer would have to appoint the Republican slate of electors. There is a fair chance some of those electors would not go along with the compact. Yes, they would prefer the Republican candidate. But they might have a stronger, principled preference for discrediting the compact. Could legislatures force electors to vote for their party’s nominee? No, all they could do is condemn rogue behavior (Ray v. Blair, 343 U.S. 154 [1952]). If the states in the compact commanded little more than 270 electoral votes, a handful of electors opposed to the compact could block its effect.
That prospect is quite realistic because of how the compact separates the two parties. Democratic supporters wish to assure us that Republican electors would adhere to the compact, but they cannot. Under the present system, electors are pledged to vote for their parties’ nominees, but if the compact took effect, some Republican electors, many or few, might feel obliged instead to uphold their national anti-compact platform plank. Republican state conventions might even free their electors to do so, if only as a threat to force Democratic agreement to withdraw. At the very least, Republican electors across the country might coordinate their votes to protest the compact without electing the Democrat. Because opposition to the compact is somewhat bipartisan some Democratic electors might do likewise.
Under the present system, electors are pledged to vote for their parties’ nominees, but if the compact took effect, some Republican electors, many or few, might feel obliged instead to uphold their national anti-compact platform plank.
Close Votes and Recounts
Popular votes with margins of less than one percent have occurred surprisingly often: in 1880, 1884, 1888, 1960, 1968, and 2000. The compact would increase the likelihood of close popular votes because it is popular votes that candidates would seek, competing to attract support and turn out supporters regardless of location. Currently, by contrast, electoral votes must be won in state-wide lumps, and those do not correlate closely with popular votes. The trouble is that close votes prompt challenges and the demand for recounts. But as we saw in the Tallahassee tally hassle of 2000, when the original count is close, the recount has to be polity-wide: a fair one cannot be confined to counties or precincts cherry picked by one or both sides. Under the compact, any recount would have to be nationwide. It would have to wend its way through fifty-one election codes and ballot forms and thousands of county bureaucracies, and it would have to do so within 35 days (the legal deadline for state certification of elections) or at most 41 days (when electors meet to vote). That is almost certainly infeasible, and it would remain so even if Congress extended the deadline to January 19. Footnote 4 Congress would have to declare the winner or, blocked by incomplete information or internal divisions, allow the dispute to fester while the House speaker acted as president.
Far from overstating this problem, we have given proponents the benefit of the doubt by assuming that, unless all states joined the compact, the joiners could somehow compel the other states to send them certified and timely vote tallies and to recount them, or at least begin the process, whenever nationwide totals were close. Obviously no state has the power to do that. Nor could courts compel recounts unless within-state results were close or otherwise contestable. The problem of compulsion would disappear and efficient recounts would be easier if the federal government ran presidential elections. But that would require a Constitutional amendment—the very thing proponents brag about avoiding.
Manipulation of Vote Counts
The compact would magnify the incentive politicians have to manipulate vote counts. They do that already, Footnote 5 but their incentive is limited by the fact that few states are swing states. By making every vote count, regardless of location, the compact would lift that limit and encourage officials everywhere to play partisan accordion with the recorded vote. Local election officers could make it easier or harder for voters to register, allow more or fewer forms of identification, and accept or reject ballots cast late or in the wrong precinct. And it is they who decide the location and advertisement of polling places. At the legislative level, Democratic majorities could boost their party’s share of the vote by allowing convicted felons, sixteen-year olds, and noncitizens to vote (yes, states can let noncitizens vote, and in the nineteenth century some did). Republican majorities could in turn toughen voter-identification requirements, ban provisional and foreign-language ballots, fund fewer polling places, and increase penalties for fraud. Footnote 6 If partisan majorities do not do more of that already it is because they are not strong enough in closely divided states, and in solidly partisan states they have no reason to manipulate statewide votes (for president, governor, senator, or anything else). District votes are another matter, but they are controlled well enough by creative cartography.
Narrowed Support
The current system rewards candidates for winning pluralities in states that are numerous and populous but not for winning large majorities in those states. By contrast, the compact would let candidates profit from ever more support in fewer states: the victor would have no incentive to spread his support across territorial or other divisions of the population. Footnote 7 A nationwide plurality could represent a fair diversity of interests, but there is no reason why it should. Assuming that numerous states are a rough proxy for numerous interests, the compact would encourage candidates to appeal to narrower, less diverse interests than they do now.
CONSTITUTIONAL CHALLENGES
Opponents of the compact would naturally wish to challenge it in court. Who would have standing to do so? Most obviously, candidates and parties who lost an election under the compact but would have won under the old system. First they must lose, however, and that would give courts scant time to hear their petitions.
Are there grounds for suit? The words of Article II are more intriguing than instructive. States decide their methods of choosing presidential electors. At first they were chosen mostly by state legislatures, later by voters, now mostly at large, but there are variations. Congress may dictate times and prescribe equal-rights protections. But one might contend that the states themselves must still choose their electors, albeit by methods of their design, and that, in consequence, they cannot devolve the choice on other states—contrary to the compact.
Instead of Article II, a disadvantaged party might cite Article I, Section 10, the requirement for Congressional consent to interstate compacts. Yes, neighboring states have to make a host of agreements too trivial to warrant Congressional attention. So courts recognize minor compacts that bypass Congress: as long as they do not affect federal power or other states, tacit consent is presumed (Virginia v. Tennessee 148 US 503 [1893]). But the sweeping effect of the compact and the improbability of Congressional consent put it under the Section 10 umbrella. Footnote 8 Defenders might argue that the compact is a mere statute, passed separately in a number of states, not a compact except in name. But plainly it is a compact, for it affects the whole nation and it is framed as an agreement that makes state actions interdependent. And why use such a legally loaded name if it is not meant? Besides making the compact vulnerable to withdrawals, the omission of Congressional consent makes it unconstitutional.
Whether in the judicial, the legislative, or the public forum, opponents of the compact might cite federalism as a fundamental Constitutional principle deliberately built into the electoral college but flouted by the compact. It is a principle that constrains pure democracy in many ways that we take for granted. States have equal representation in the Senate, regardless of population. They may draw congressional districts to favor one party, giving it a majority of seats when it has won only a minority of votes. Such exercises of state sovereignty can impose minority rule on the whole country: in 2012, for example, the Democrats won a majority of votes for the US House but Republicans won a majority of seats. Footnote 9 Together the states create a cacophony of methods of choosing and binding national-convention delegates and nominees for Congress. They also design and run their own congressional and presidential vote-count procedures, ballot forms, and listing criteria. And of course they enjoy a great deal of self-government under their own constitutions. The federal element in presidential elections is a short movement in a long symphony, but it is a movement, scored in the Constitution, that the compact would silence: as in Senate votes on legislation, small states are supposed to wield disproportionate weight in the election of the president.
The federal element in presidential elections is a short movement in a long symphony, but it is a movement, scored in the Constitution, that the compact would silence: as in Senate votes on legislation, small states are supposed to wield disproportionate weight in the election of the president.
More fundamental even than federalism is the principle of governance by a written constitution and the rule of law. When fully realized, that principle blocks any ambiguity over who is in charge of what, especially when government changes hands. Dynastic disputes over monarchic succession once drew blood. By making it harder to tell who has won the presidency, the compact would create a republican analogue of those disputes. If courts were troubled by this they might cite “unconstitutional vagueness” in violation of “due process.”
SUPPOSED DRAWBACKS
Proponents of the compact base their case on three familiar complaints about the current system.
It is Indirect: Voters Choose Electors who Choose the President
However, the compact too would preserve the formality of indirectness. It is only a formality, but that is likewise true of the current system, where presidential electors vote as pledged. Popular votes currently determine outcomes in a way that is complicated but still direct in its effect. Uncooperative electors are possible, but they are likewise possible and even more likely under the compact – more likely because the compact opens the door to defection based on principled opposition to the compact itself.
The Current System Can Reject Popular-Vote Winners
But what does that mean? There are countless formulas for translating popular votes into “winners.” Which is the right one? Proponents of the compact may think they are enforcing rule by majorities, but what they have actually proposed is rule by pluralities: whoever wins the most votes wins the election, even if those votes are not a majority. Imagine a three-way election among Libby, Maude, and Connie, with the voting population partitioned into three minorities: liberals, the largest of the three minorities, who prefer Libby to Maude to Connie, moderates, who prefer Maude above all, and conservatives, who prefer Connie to Maude to Libby. Under nationwide plurality rule, advocated by the compact’s champions, Libby would win. But a majority (moderates plus conservatives) prefer Maude to Libby—and for that matter a majority (moderates plus liberals) prefer Maude to Connie. It was the Marquis de Condorcet who discovered this problem near the end of the 18th Century, and nowadays any candidate who, like Maude, is majority-preferred to every other candidate is called the Condorcet winner. A strict majoritarian would favor Maude, although the compact calls for Libby’s victory. It is rather the electoral-college system that, by demanding a certain breadth of support, is somewhat more likely to promote centrist or compromise candidates like Maude, everyone’s favorite or second favorite in our toy example.
To be fair, the proponents of “direct popular election” assume that the typical election has only two candidates, making the plurality favorite the majority favorite too. But has a majority favorite ever lost the electoral college vote? Look at the cases often cited as electoral-college reversals of the “popular choice.” When would the compact have made a difference?
-
• Not in 1824. Jackson did receive the most popular votes, but Adams, chosen by the House for want of an electoral-college majority, was strongest in states that had no popular vote: legislatures chose their electors. The compact assumes a popular vote in every state.
-
• Not in 1876. Congress intervened to declare Hayes the winner in a race-based dispute over competing vote counts. The compact assumes uncontested state results and no racial disenfranchisement.
-
• 1888 is better. Democrat Cleveland won a popular plurality but lost the electoral college to Republican Harrison. Footnote 10 Even so, the electoral college overcame massive electoral fraud based on race and party (and implemented by terrorism). In South Carolina, for example, Harrison reportedly won 20 percent of the vote although a majority of residents were black and, therefore, Republican. Would the compact have helped Cleveland? Only if election officials backed by legislators in Republican compact states (there would have to have been some) accepted vote tallies from South Carolina and other southern states that had blatantly disenfranchised most Republicans.
-
• 1960 is ambiguous. To give Kennedy his oft-reported 118,574 popular-vote victory over Nixon, you have to ignore vote fraud in Illinois and Texas and count all those who voted for the victorious Democratic electoral slate in Alabama as Kennedy voters, although that slate had only five Kennedy electors. Six more electors were pledged instead to Senator Harry Byrd of Virginia. To make things more complicated, the eleven electors ran separately, not as a slate: to vote for the whole slate, one had to cast eleven votes. If we give Kennedy five-elevenths of the Democratic vote, Nixon noses him out in the nationwide popular vote (Gaines Reference Gaines2001). Under the compact, the outcome would have depended on how state election officials interpreted the vote tally from Alabama, whose ballots did not even list Kennedy or Nixon, only the party nominees for elector. The compact assumes that every major candidate is on the ballot in every state.
-
• 2000 saw a near tie in both popular and electoral votes but also the clearest case of reversal. This time the compact would have made a difference: most likely the compact states would have given all their electoral votes to Vice President Gore, who, before a joint session of Congress, would have declared his own victory, or if not then acting President Hastert would have presided over an agonizing nation-wide recount, a temporary administration, a recession, and maybe 9/11.
In each of these cases, there were more than two candidates (including minor ones). In none did the “popular-vote winner” receive a majority of popular votes, only a plurality. And in the close-vote cases we cannot be sure who really won even a plurality: the influence of counting errors, ballot-marking errors, and uneven election administration dwarf any perceived margin of victory.
Owing to the Winner-Take-All Rule, Solidly Red or Blue States are not Worth the Investment of Campaign Resources to Either Party. Instead the Campaign is Conducted and the Outcome Decided in a Handful of Swing States, Such as Florida and Ohio. This has Disenfranchised all Voters in Monochrome States (Hertzberg Reference Hertzberg2010).
That charge is fatuous. True, in solidly red and blue states, no individual voter can make a difference in a presidential election. But no individual voter ever makes a difference anyway. One might as well say that all lopsided elections disenfranchise voters. That would cover a great many statewide elections (for president, governor, US Senate, and whatnot) and almost all Congressional and state-legislative elections. Are voters “disenfranchised” when they like an incumbent so much that no one runs against him, or when they vote almost unanimously for a ballot proposition to build new schools? True, less is spent on campaign advertising in solidly red and blue states. But voters probably like that. Besides, if the winner-take-all rule hurts the citizens of any state, why is it on the books? Far from hurting them, it has long been thought to benefit them by increasing their electoral influence.
Is there a better reform? A constitutional amendment to abolish the electoral college is unrealistic, and it is far from clear what its replacement should be.
CONCLUDING ADVICE
Is there a better reform? A constitutional amendment to abolish the electoral college is unrealistic, and it is far from clear what its replacement should be. There is no such rule as “direct popular vote,” unless that is a euphemism for nationwide plurality rule, but as we saw, plurality rule can work in an anti-majoritarian way. Runoffs can work better: in our liberal-moderate-conservative example, a runoff might choose Maude, the Condorcet winner. But maybe not: the runoff might pit Libby against Connie. If we had preferential ballots, ones which rank candidates in order of preference, we could spot and choose any Condorcet winner. But sometimes there is none to spot. Sometimes every candidate is opposed by a majority. Let the electorate be partitioned again into three minorities, but now let one of them prefer Xavier to Yolanda to Zeke, another Yolanda to Zeke to Xavier, and the third Zeke to Xavier to Yolanda. Then majorities prefer Xavier to Yolanda, Yolanda to Zeke, and Zeke to Xavier. This cyclic majority preference (or Paradox of Voting) is another, more famous discovery of Condorcet. Whoever is chosen, some majority prefer a different candidate. The paradox has generated a vast amount of thought for over 200 years, with no consensus on the best way to count votes. The immediate lesson is that it is not at all clear how to reckon the “popular will,” the “popular vote winner,” nor whether such a thing always exists. There is no warrant for the claim, made by advocates of the National Popular Vote, that popular plurality rule is “an elementary democratic principle” (Hertzberg Reference Hertzberg2007).
So let us start again. Justly or not, what most exercises enemies of the electoral college is the prospect of reversal, a discrepancy between the electoral-college winner and the “popular-vote winner,” or less misleadingly, the popular plurality winner. The only reasonably clear reversal was in 2000. But that and the near reversals have never reversed popular majorities, only pluralities, and they have occurred when the popular vote was so close we do not know for sure who the true popular favorite was. Also there already exists a built-in incentive against reversals: besides electoral-college majorities, presidents profit from popular-vote majorities because that lets them claim a stronger mandate for their policies.
Even so, we address the avoidance of reversals in our four-part concluding advice.
-
1. To diehard proponents of the compact: Make three changes. To discourage tiny pluralities, require that the winner receive at least 40% of the popular vote, else the old system would still apply. To avoid nationwide recounts, create an interstate canvassing board to examine election returns and decide the winner – or decide that the vote is too close to call and revert to the old system. To enhance stability, or discourage defections, raise the threshold for implementation of the compact from 270 to 325 electoral votes. That would make compact compliance a Nash equilibrium by ensuring that no state would ever be pivotal in the electoral college – though pairs of states could be and states could still defect at any time out of principled disapproval or the desire to enhance their electoral influence. Manipulation, uncooperative electors, and especially the need for Congressional consent would still be problems. Besides attracting a large enough coalition of states and avoiding a Supreme Court veto, success would depend on that canvassing board – on how it was appointed, its standards and procedures, and its apparent legitimacy.
-
2. To those who still insist that plurality rule is the essence of democracy: Drop the compact – and with it the need for Congressional consent. Instead advocate state laws mandating the appointment of electors pledged to the nationwide plurality favorite. If even one Bush state had had such a law in 2000, Gore would have won. Again add a 40 percent threshold and now separate, state boards to judge the reliability of vote tallies from other states and revert to the old system in case those tallies are not reliable enough. The problems of Congressional consent and recounts are gone, but those of instability, manipulation, uncooperative electors, and narrowed support remain.
-
3. To less extreme advocates of reform: Plump for the proportional choice of electors within each state. This too can be done one state at a time: no compact is required. Maine and Nebraska already do something similar: they choose two electors (corresponding to their two US senators) by winner-take-all and the rest by election from congressional districts. But that leaves open four possible sources of reversal: (1) the two “senatorial” electors regardless of state population, (2) their choice by winner-take-all, (3) the distortion of congressional district maps, and (4) the possibility of third-party spoilers denying anyone an electoral-college majority by winning a handful of districts in one or more states. There is no remedy for (1). But consider (3): it is quite possible, even easy in some states, for a party to win a majority of votes in each of a majority of districts but only a minority of votes statewide. The prevailing incentive of state legislatures to gerrymander district boundaries would be magnified. Instead states can adopt the simple, statewide choice of all presidential electors in proportion to popular votes. That would eliminate (2) as well as (3). As for (4), we can set a high threshold, say 15 percent of the vote, to win any electors. We see no grounds for court challenges. Manipulation would be less than under the compact because great efforts would typically achieve no more than one extra electoral vote. The incentive to defect would be much less too because no state would ever be obliged to cast all or even most of its electoral votes against the expressed wish of most of its voters. Each time a state adopted this procedure, it would be contested thereafter by both candidates. And close tallies would rarely prompt even within-state recounts because they would change but one electoral vote.
-
4. To everyone: Be skeptical of any reform. The greatest merit of the electoral-college system as it has come to be implemented is the cost of replacing it. For all its antique oddity, it is the foundation on which the rest of our practices and procedures, our politics and our party structure have arisen: its removal would topple them. But they do more than rest on the electoral college: they exploit it in beneficial ways and offset some of its potential drawbacks. By the 1830s the move toward popular election of pledged electors allowed democracy, direct election, and party competition to evolve, replacing the initially unavoidable elitism of the ancien collège. The winner-take-all rule has helped entrench a two-party system. That in turn has encouraged responsible government – not so firmly, perhaps, as in a perfect Westminster system of parliamentary democracy, but parliamentarism is not in the offing anyway, and Westminster Palace itself has long housed more than two parties, from time to time and again recently in the government. Combined with national parties, the need to win numerous and diverse state-wide pluralities has doubtless helped to cement the Federal Union. It has also attracted candidates with broad more than intense appeal. And it has quarantined several problems, such as manipulation and recounts, inside one or a few states. Thanks to the much criticized winner-take-all rule, the electoral college has turned close and otherwise contestable nationwide pluralities into clear winning majorities, blocking dangerous disputes over who has the right to rule: this feature especially would not be preserved so well by any contemplated reform.
