The notion that relative political power or the dignity of voters depends on whether or not the vote of an individual voter has the exact relation to the number of electors of all other voters shows a lack of serious thinking, or perhaps just a myopic vision of how a great republic might be structured to balance the interests of all its citizens and protect minority rights. Sadly, that is also true generally of the proposal now circulating to defenestrate the Electoral College, a proposal known as National Popular Vote (NPV).
NPV calls for states to join an interstate compact in which signatory states would agree to award their electoral votes not to the winner of the popular vote in their states, but to the winner of the aggregated popular vote totals from across the 50 states and the District of Columbia. The compact would become effective once states comprising an Electoral College majority—that is, 270 or more Electoral College votes—sign on, thus guaranteeing that the candidate with the most popular votes as aggregated from state vote totals would win the presidency. As of the close of 2019, 15 states and the District of Columbia, with 196 Electoral College votes, had joined the NPV compact.
NPV is the brainchild of Akhil and Vikram Amar, brothers and law professors who proposed such a compact after the 2000 election in which George W. Bush won the presidency despite receiving fewer popular votes than Al Gore. John Koza, a California-based computer expert, took up the proposal and in 2006 published the first edition of the NPV manifesto, Every Vote Equal, which contained 350 pages explaining how the plan would work (even providing a procedure to break an exact tie in the nationwide, aggregated popular vote totals) and over 200 pages of appendices but barely 10 pages explaining why we would actually want to adopt such a plan or how we would benefit from it. Each successive edition of Every Vote Equal has become longer—the most recent is 1,118 pages—but still with little explanation as to why NPV is a good idea.
Rationale vs. Reality
To the extent that Every Vote Equal attempts to argue for the merits of NPV, it is riddled with confusion. For example, in one of a series of short commentaries that introduce the volume, former Senator Birch Bayh of Indiana argues that “[t]he President should be chosen by a majority of our citizens.” But the NPV compact does not require a majority of the popular vote (let alone a majority of all “citizens”) to win: A plurality of any size and any baseline will do, so a candidate in a hypothetical four-way race might win with 30 percent of the vote or less. Thus, NPV begins by immediately violating what it purports to be the cardinal principle that trumps all others: that the President be elected by a popular majority. Another contributor makes the mistake of assuming that the number of voters is synonymous with population.
The authors of Every Vote Equal—and indeed most NPV enthusiasts—seem to believe that the case for a straightforward popular-vote election is self-evident. Erwin Chemerinsky, dean of the law school at the University of California Berkley, claims that the United States “is the only country in the world where the candidate who loses the popular vote can be chosen president.” But this is emphatically not true—at least not if we replace the title “president” with “head of government” so as to include chief executives who are titled “premier,” “prime minister,” or something else.
The truth is that in the vast majority of democracies, a candidate who loses the popular vote can be chosen chief executive. Sometimes, typically in multiparty states operating under the curse of proportional representation, that has happened as a result of coalitions formed after the election. When no candidate gains a majority, the leaders of the various political parties gather after the election and negotiate a coalition designating a head of government (typically called a prime minister), whose party may or may not have won the most votes. In recent years, this has been true in such democracies as Sweden, Norway, Germany, Israel, and Italy. In other cases, countries have elected chief executives outright even though those heads of government did not win the popular vote. Since World War II, such situations have occurred in Japan (2003); the United Kingdom (1951 and 1974); New Zealand (1978 and 1981); and elsewhere.
Most significantly, such examples have occurred in the most populous of all democracies—India—in 1991 and 1999 and in each of the three great democracies that geographically span a continent: Canada in 1979 and 2019; Australia in no fewer than five elections since World War II, most recently in 1998; and the United States in 2000 and 2016. Why do these four democracies, each of which spans a vast and diverse geographic area, have systems that occasionally allow for the second-place finisher to win the executive office?
In 48 presidential elections since people began recording nationwide popular vote totals, the U.S. has had four in which the Electoral College winner did not win the aggregated state-by-state popular vote (1876, 1888, 2000, and 2016). In 1876 and 1888, however, the Democratic Party in the southern states engaged in so much fraud and violent vote suppression that it is hard to believe than a more honest election would not have caused the result in the Electoral College to match the result of the aggregated popular vote.
Leaving that aside, is it acceptable that, on an average of one in every 12 elections, the second-place finisher in the popular vote has won the executive branch? Why have such a system?
First, there are numerous practical benefits to the Electoral College. For example, it prevents a race to the bottom as states try to gin up turnout for their preferred candidates by, for example, lowering the voting age to 16 or even 14 or allowing even a convicted felon who is still incarcerated to vote (as Vermont and Maine have done). If some states want to do these things, they can, but their doing so does not affect other states. The Electoral College reduces incentives for fraud by making much electoral fraud meaningless in the presidential race. There is no value in encouraging fraud in states that a party is likely to win anyway. Under NPV, every fraudulent vote anywhere in the country will have value. The Electoral College spares us the chaotic possibility of nationwide recounts. Other legal and practical weaknesses of NPV can be found elsewhere.
The point to emphasize is that NPV—and proposals for an actual constitutional amendment—strike at the heart of our constitutional system of government. Over more than 230 years, the United States has become the most durable, prosperous, and free society in the world. The one time our political system truly failed, resulting in the Civil War, it was not the Electoral College that was at fault. The Electoral College yielded a clear winner in the election of 1860: Abraham Lincoln. And while Lincoln won with the lowest percentage of the aggregated popular vote in history—39.8 percent—it should be noted that he actually met the NPV criteria in that he won the most popular votes and that, although Lincoln did not campaign in and received virtually no votes in the South, he was the most national of the four major candidates, including Stephen Douglas, John Breckinridge, and John Bell.
Breckinridge and Bell were distinctly regional candidates. Bell’s appeal was limited to the “border” states, and he won only Virginia, Kentucky, and Tennessee. The second-place finisher in the Electoral College, Breckinridge, won just 18 percent of the total popular votes and no states north of the Mason–Dixon Line. Douglas, like Lincoln a northerner, at least made an effort to compete in the South, but in eight of the 11 states that joined the Confederacy, Douglas got less than 10 percent of the vote, and in none did he even approach 20 percent. He won just 29.5 percent nationwide and carried only Missouri and New Jersey.
In short, Douglas ran a national campaign only in the sense that he was not popular anywhere. Lincoln, on the other hand, won 17 of the 33 states, winning states in New England, the mid-Atlantic, the Midwest, the Prairie, and the West Coast. One would have to conclude that the Civil War came about in spite of the Electoral College, not because of it.
In the four elections in which the winner of the aggregate popular vote winner did not win the Electoral College, the successful candidate averaged 47.4 percent of the aggregated popular vote, with Donald Trump’s 45.9 percent being the lowest. Seven times, however, the winner of the popular vote has received less than 47.4 percent of the aggregate popular vote; five times, he has received less than Trump’s 45.9 percent.
In fact, in 17 presidential elections, the winner of the popular vote received less than 50 percent of the aggregated popular vote, so Trump’s popular support in the election was not abnormally low for an elected President, including those who won the most aggregated popular votes. In other words, on those rare occasions when the Electoral College has not coincided with the aggregated popular vote, the winner in the Electoral College has still received popular support roughly equivalent to or better than that won by more than a third of those who have been elected President while winning the aggregate popular vote.
Under a national popular vote, candidates would need neither a popular nor an Electoral College majority—just more votes than anyone else. Thus, a candidate could win with as little as 25 percent or 30 percent of the vote, and we should expect to see such elections.
At Odds with Constitutional Principles
NPV would throw out our successful system and substitute a radically different one with no serious regard for the consequences and based on the abstract and highly contested principle that only the aggregated popular vote totals can bestow legitimacy on the winner. This principle is at odds not only with the Electoral College, but also with a host of other constitutional principles.
The United States is founded on the idea that the people—not some king or the sitting government—are sovereign. The Declaration of Independence proclaimed, contrary to all established political systems of the time, that “Governments…deriv[e] their just powers from the consent of the governed.” The signers of the Declaration made clear that they signed only as agents “of the good People of these Colonies.”
In 1787, “in order to form a more perfect union,” we adopted a new Constitution. The new Constitution was ratified not by state legislatures, but by conventions whose delegates were elected specifically for that purpose by “we the People of the United States.” But while legitimacy rested on popular consent, crude majoritarianism was never the design or intent of the people.
Our Constitution is democratic, but with numerous checks and balances to assure enduring majorities, moderation, and respect for minority rights and desires. These include separation of powers into three supportive but independent branches, a national government of enumerated powers, a bicameral legislature, staggered elections, presidential veto power, and federalism. Even then, the people ratified the Constitution only after assurances that further checks on the majority would be added to it. These additional checks became the Bill of Rights.
In addition to our written Constitution, we have an unwritten constitution: norms by which we operate. They include such anti-majoritarian devices as approval of legislation by congressional committees, the Senate legislative filibuster, and judicial review. These devices, too, often protect minority rights against a triumphant majority.
Our national legislative design also rejects raw majoritarianism. Not only do states have equal representation in the Senate, but in the House of Representatives, each state, no matter how small, is guaranteed at least one congressional representative. It is possible for a party to gain a House majority without winning a plurality of the aggregated national vote for House candidates (as happened as recently as 2012) and thus to elect the Speaker, who is second in line to the presidency. If the Electoral College is illegitimate simply because on rare occasions it yields counter-majoritarian results, so are the Senate, the House, and even the Bill of Rights.
Although our constitutional system draws its legitimacy from the people—again, it was ratified not by state legislatures, but by votes of the people in the states—it did not create and was not intended to create a system of pure majoritarianism. Rather, it was aimed at promoting good government based on a bedrock of popular consent for the form of government chosen.
The Constitution set up a structure intended to promote what the late Michael Uhlmann called “reasonable” majoritarianism. It does not just emphasize the values of majority rule; it creates a structure intended to lead to the construction of moderate, stable, durable majorities composed of a broad cross-section representing the great diversity of this geographically expansive nation. It is no coincidence that the other great transcontinental democracies of Canada and Australia have also adopted systems that require a party seeking the national executive office to build a broad coalition that crosses the nation.
Much unites us as Americans: More unites us, or should unite us, than separates us. But you do not see oil wells operating in downtown Washington, D.C., or New York City or Cambridge, Massachusetts, or South Bend, Indiana. You do in Oklahoma City. The United States is a country of great diversity, and if we are to hold this great nation together, we must have a system that requires a President to campaign all across the country, to accommodate a wide variety of lifestyles and subcultures.
Our constitutional system aims not merely at majority rule, but at creating ruling majorities that are respectful of minority interests and values. The Electoral College contributes to this by forcing a President to build a campaign around winning individual states. In winning the Electoral College, a candidate will usually (though not always) win the aggregated popular vote, but he absolutely must win a high percentage of the aggregated popular vote. Only Lincoln, at 39.8 percent, fell shy of 40 percent. Equally important, the winning candidate must also win in many states. All four Electoral College winners who did not win the aggregated popular vote nonetheless won in a majority of states.
Only three Presidents have been elected without winning a majority of states. In 1880, James Garfield and Winfield Hancock each won 19 states, with Garfield winning the aggregated popular vote by a margin of 48.3 percent to 48.2 percent but scoring a substantial Electoral College victory. In 1960, Richard Nixon won 26 states to John Kennedy’s 24, but JFK won the aggregated popular vote by 0.1 percent and a solid Electoral College majority. And in 1976, Jimmy Carter won 50.1 percent of the aggregated popular vote while winning just 24 states and the District of Columbia to Jerry Ford’s 26 states. In short, every plausible route to an Electoral College majority requires competing successfully across a wide variety of diverse states.
People counter that candidates need to compete only in a few swing states. This ignores the fact that states move from one column to another, sometimes quite rapidly. For example, any presidential candidate running in 1988 would have looked at West Virginia as a solid “blue” state and Vermont as a hard “red” state, yet just 12 years later, the roles were reversed. Texas has voted Republican in 13 of 17 elections since 1948 (two of the exceptions came when favorite son Lyndon Johnson was on the national ticket, and a third came when his handpicked successor, Hubert Humphrey, was the nominee); but it had voted Democratic in 16 of the 17 elections up through 1948.
All that aside, one might say “so what?” Look at the 17 states that were won by less than 10 points in 2016, meaning that a shift of 5 percent of the voters would have swung them the other way. They included:
- New England states (Maine and New Hampshire);
- Mid-Atlantic states (Pennsylvania);
- Southern states (Florida, Virginia, and Georgia);
- States from the industrial Midwest (Ohio, Michigan, Wisconsin, and Minnesota);
- Farm states (Iowa);
- Western states (Arizona, Nevada, and Colorado);
- Big states (Florida and Texas) and small states (Maine and Nevada);
- Three of the eight states with the highest percentage of African-American populations (Georgia, North Carolina, and Virginia); and
- Four of the five states with the highest percentage of Hispanic populations (Texas, Florida, Arizona, and New Mexico).
Yet they also included:
- Three of the five states with the highest percentage of white voters (Iowa, New Hampshire, and Maine);
- Some of the most heavily unionized states (Michigan, Minnesota, and Iowa) and states with exceptionally low unionization rates (Texas, Georgia, Virginia, and North Carolina);
- Manufacturing states (Ohio and Michigan) and states with very little manufacturing (Nevada and New Mexico); and
- High-tax states (Minnesota and Iowa) and low-tax states (Florida and New Hampshire).
In other words, to win the Electoral College, a candidate must appeal to a wide array of geographic, economic, and social interests. It requires a coalition that is broad, moderate, and respectful of minorities. The emphasis on building such a broad coalition is a vital component of our national system of government—a system that has proven durable, has avoided wild swings in policy that create unrest, and fosters local solutions and respect for minorities.
Rejection of Popular Consent
Ultimately, NPV rejects the popular consent at the core of our Constitution, because it attempts to alter a key feature of that Constitution—the Electoral College—without going through the prescribed procedure for amending the Constitution. If asked, would the people go along with such an amendment? Almost certainly not—which is why NPV tries to end-run the amendment process.
And for what? Beyond the benefits it brings in protecting minorities and fostering moderation and consensus, the Electoral College has produced a string of executives that no other country can match, except possibly the United Kingdom, where it is also possible to win the executive while finishing second. Americans like to criticize our politicians, and that can be healthy. But leaving aside policy disagreements, what other country can produce a list of chief executives matching in ability the likes of Washington, Jefferson, Jackson, Lincoln, Cleveland, Coolidge, the Roosevelts, Truman, Eisenhower, and Reagan?
So often in debates such as this we get bound up in the partisan politics of the day and lose sight of the big picture. We disagree with policies, so we discount the abilities of Presidents associated with those policies. The purpose of our constitutional system is not to elect Republicans or Democrats: It is to elect good Presidents. Most of us will have reasons not to fancy some of these Presidents on policy or other grounds, but taken as a whole, they are a pretty distinguished group of chief executives, and they have represented well the people who elected them—through elections in their states and the selection of presidential electors to the Electoral College. As the great political scientist Walter Berns put it, “In all the years I have engaged on this issue, I have yet to encounter a critic of the Electoral College who argues that a president chosen directly by the people is likely to be a better president.”
Nevertheless, we sometimes hear an argument that goes something like this: Why don’t states elect their governors with an Electoral College? Would anyone advise a new democracy to adopt an Electoral College? If you could go back in time, would you adopt the Electoral College?
Why do our states not use an electoral college–like system? First, because the Supreme Court of the United States has held—some say wrongly—that doing so would violate the Fourteenth Amendment. Before that Supreme Court ruling, at least some states had indeed used mechanisms with some similarity to the Electoral College. More important, many of the reasons for adopting the Electoral College for a large national government simply do not apply to the states. States are, by definition, considerably more compact geographically; the diversity of interests in a particular state will usually not be as great as that for the national government; and citizens and officeholders are therefore more likely to have a natural familiarity with the entire state.
For similar reasons, it should be left to the new democracies to decide how best to structure a government for their unique circumstances. An emerging democracy considering its constitutional design should consider the benefits of both direct and indirect election that may be applicable to their situation, as well as the fact that most democracies have systems that account for values other than pure majoritarianism and thus allow, on relatively rare occasions, the second-place finisher to win the executive office.
Most of all, however, if one could be transported back in time to Independence Hall in September of 1787, it would probably be good sense not to lecture the likes of George Washington, Ben Franklin, James Wilson, Alexander Hamilton, and James Madison about how to structure a successful republic and how derelict they had been in not anointing direct election by a plurality of the electorate as the sole legitimizing principle of government. We should not fear change, but when something has a long, distinguished pedigree—when it has, in fact, worked—we might be a little more humble.
Bradley A. Smith is Josiah H. Blackmore II/Shirley M. Nault Professor of Law at Capital University School of Law and a former chairman of the Federal Election Commission. He is a recipient of the Bradley Prize and the coauthor of Voting Rights and Election Law, a leading casebook in the field of election and campaign finance law.