(Bloomberg Opinion) -- When it came time in 1787 to set the rules for choosing a president of the U.S., three of the principal authors of the Constitution — James Madison, Gouverneur Morris and James Wilson — argued that the best approach, the one most likely to inspire public confidence and national feeling, would be a nationwide popular vote.
All three also understood the prospects of this happening were, as Wilson put it, “chimerical.” It was obvious the method would instead have to reflect the two great (or awful, if you prefer) compromises hammered out at the Constitutional Convention over political representation. To keep the slave-holding states on board, the delegates had apportioned seats in the House of Representatives on the basis of a population count that considered slaves to be three-fifths of a person. And to assuage the smaller states they had created a Senate with two members per state, regardless of population.
States were thus allotted presidential votes on the basis of how many Senators and House members they had. At first the plan was simply to have the members of Congress vote. But fears this would make the legislature too powerful led the delegates to create what later came to be known as the Electoral College, with the manner of its choosing left to the individual states. If no candidate got a majority of the electoral votes, then the decision went to the House of Representatives — albeit with each state delegation getting only one vote, in yet another concession to the small states.
It’s fair to say the men who designed this system did not entirely think it through. Complaints began springing up almost as soon as it went into effect and have continued ever since. In 2004, the Congressional Research Service found “more proposed constitutional amendments have been introduced in Congress regarding electoral college reform than on any other subject.”
Only one amendment directly addressing the method of electing the president has ever been ratified (in 1804), while another (ratified in 1933) had some impact on the contingent election in the House.(1) There have been two periods, the 1810s-1820s and 1960s-1970s, when broader reform seemed possible or even likely, as Harvard Kennedy School historian Alexander Keyssar describes in his excellent new book “Why Do We Still Have the Electoral College?”(2) But most of the time, political factions convinced that they benefit from the system as-is have stood in the way of change.
As I learned last year while writing about the near-miss 1960s-1970s effort to replace the Electoral College with a national popular vote, these political factions and perceived benefits have not remained constant. The U.S. presidential election system definitely favors particular brands of politics at particular times, as it did in 2016, when it gave Donald Trump the presidency even as Hillary Clinton received almost three million more votes. But it can perhaps best be thought of as a sort of random-number generator, with unpredictably shifting biases, that usually churns out the same result as the popular vote but occasionally does not.
With an election coming up in which the makeup of the Electoral College appears to favor Trump again but probably not by enough to get him reelected (when last I checked, FiveThirtyEight’s election model gave him a 3% chance of winning the popular vote and an 10% chance of winning the electoral vote), a brief(ish) review of its post-Constitutional-Convention history seems in order.
Mistakes were made
The electoral-system amendment ratified in 1804 was the 12th, which arranged that vice presidents and presidents be chosen separately. In the original setup the vice presidency went to the runner-up in the presidential race, and in 1800 Thomas Jefferson and his running mate Aaron Burr ended up with the same number of electoral votes. The subsequent tiebreaker in the House of Representatives proved quite problematic, putting the election in the hands of a Federalist majority that had just been voted out of office in a Democratic-Republican landslide. Two state delegations couldn’t agree on how to vote until, as you may have heard set to music, Alexander Hamilton persuaded just enough of his fellow Federalists to let Jefferson win on the 36th ballot.
By changing when members of Congress and the president take office after an election, 1933’s 20th Amendment at least ensured newly elected House members, rather than lame ducks, would do the voting. But the threat of deadlocked state delegations and a deadlocked House remains.
The aspect of the real-world Electoral College that seemed most to surprise its designers, though, was how quickly it evolved into a rubber stamp for the partisan leanings of state legislators and voters. Many delegates at the Constitutional Convention assumed that — after voting for George Washington in the first election, of course — electors would choose independently for a variety of well-qualified men, leaving most elections (19 out of 20, predicted George Mason) to be decided in the House.
Instead, state political leaders quickly began insisting electors pledge to vote for specific candidates and increasingly arranged that all electors in a state vote for the same candidate, to maximize the national political clout of that state’s majority party. Maryland and Pennsylvania were the trendsetters, awarding all their electoral votes in the 1788-1789 election to the winner of each state’s popular vote (Washington, of course). By 1832 every state but two had adopted this winner-take-all approach. That’s where things stand now, with only Maine and Nebraska choosing some of their electors by House district instead.
As scholars started to put in mathematical terms in the 1940s but politicians suspected almost from the start, winner-take-all negates most of the small-state advantage the delegates to the Constitutional Convention had attempted to insert into the system. This is because when individuals or blocs control differing numbers of votes — as with corporate shareholders, the Council of the European Union or the Electoral College — the likelihood that a big bloc will cast the deciding vote is usually higher than its share of the overall votes (the exception is when there are multiple big blocs with identical numbers of votes).
Sure enough, four of the first five presidents came from Virginia, the most populous state in the 1790 and 1800 censuses and No. 2 in 1810. Not until New Hampshire’s Franklin Pierce in 1852 was a president elected from a state in the bottom half of the population ranking, and that didn’t happen again until Dwight Eisenhower of Kansas was elected a century later. (Since then there’s just been Bill Clinton from Arkansas, although Delaware’s Joe Biden would be the fourth.) Thirty-seven of the 44 men who have served as president have come from a state in the top 25% of the population rankings, 26 from a state in the top 12%.
The first near-miss: district elections
More people live in large states than in small ones, so it should be no surprise their residents get to move into the White House more frequently. Still, small-state politicians perceived early on that winner-take-all was not working to their benefit, and they and other reform-minded sorts began pushing for a requirement that electors be chosen by district rather than statewide. Big-state politicians, understanding such a change would reduce their clout, demanded in exchange that the fallback election in the House be altered to reduce its small-state bias, for example by giving every member of both the House and Senate a vote.
Congress came closest to striking such a deal during the “Era of Good Feeling” after the end of the War of 1812 and the collapse of the Federalist Party. Partisan considerations mattered less then because one party, the Democratic-Republicans, controlled almost everything. The Senate mustered more-than-two-thirds majorities year after year for amendments to require district elections, only for them to come up short in the House, where big states were better represented. In 1821, the House did vote 92-54 in favor, but that was still six votes less than the required two-thirds.
Not long after that, the Democratic-Republican Party split up into Democrats and Whigs. The prospects for reform faded, even as the 1824 election brought the system under fire again, when the House awarded the presidency to the candidate (John Quincy Adams) who came in second in both popular and electoral votes. There were some perfunctory attempts at reform during the presidency of Andrew Jackson, who lost that disputed 1824 election and then won in 1828. But then Congress more or less let the matter drop until just after the Civil War.
Lawmakers proposed multiple reforms both before and after the disputed 1876 election, in which the loser of a not-all-that-close popular vote, Republican Rutherford B. Hayes, ended up in the White House thanks to a political deal even some in his party found suspect. Most of these proposals were again along district-election lines, but some called for a national vote. None really went anywhere, which continued to be the case until after World War II.
The barrier to district elections was that the dominant political parties in large states didn’t want to cede clout, while a national popular vote was opposed by the former slave states. The three-fifths compromise had been repealed in 1868 by the 14th Amendment, but after a few years in which freed slaves voted in the South, most states in the region reverted to what became known as the “five-fifths rule,” in which Black citizens counted fully in apportioning House seats and electoral votes but weren’t allowed to vote. A national popular vote would effectively punish the South for this massive voter suppression, so it was a non-starter.
By the late 1940s, conditions started becoming more conducive to change. Partisan allegiances had weakened, with liberal and conservative wings in both major parties. The odds Southern states would soon be forced to confer political rights on their Black citizens seemed to be rising.
The first big reform effort was led by the odd couple of liberal Massachusetts Republican Senator Henry Cabot Lodge Jr. and conservative Texas Democratic Representative Ed Lee Gossett. They proposed allocating each state’s electoral votes in proportion to its popular-vote tally. This was approved by more than two-thirds of the Senate in 1950, but foundered in the House.
Gossett had made it quite clear he hoped the plan would reduce the power of the biggest states and cities and the Blacks, Jews, Italians, Irish, Poles and Communists he said held political sway there. The Czech-born Jewish Chicagoan who ran the House Rules Committee, Adolph Sabath, didn’t let it reach the House floor.
The second near-miss: national popular vote
Still, the appetite for reform remained, whetted by the third-party presidential candidacies of conservative Southerners Strom Thurmond in 1948 and George Wallace in 1968, which threatened to deprive the major parties of Electoral College majorities and throw the decision to the House. After a series of Supreme Court rulings in the early 1960s asserted the principle of equal representation (before then, states were allowed to have House and state legislative districts with widely varying populations), and the poll-tax-abolishing 24th Amendment (ratified in 1964) and Voting Rights Act of 1965 spelled an end to the five-fifths rule, it became clear a national popular vote was the natural replacement.
None of the original justifications for the Electoral College seemed to apply any more. The idea of electors as independent decision-makers had been abandoned by the early 1800s, although there were still occasional “faithless electors” who reneged on their voting pledges. The accommodation made to the slave states in apportioning electoral votes had finally ceased to matter. And the accommodation made to the small states — long viewed as nearly worthless as long as winner-take-all voting was the norm — was made to look even worse than worthless in some calculations published in 1968 by young lawyer and engineer John Banzhaf III.
Banzhaf, now a professor at the George Washington University Law School, reckoned based on the bloc-voting considerations described above that voters in California and New York possessed more than twice the per-person voting power as those in the smallest states. Other scholars have since argued Banzhaf’s formula incorrectly assumed victory margins would be tighter in big states than in small ones, and that the biggest voting-power advantages are enjoyed by voters in whatever the current battleground states happen to be. But at the time Banzhaf seemed to have persuaded a lot of small-state lawmakers they’d be better off with the national popular vote.
His work also eventually seems to have persuaded some big-state lawmakers they’d be better off without it. But in 1969 the House of Representatives approved a popular-vote plan (with a runoff if no candidate got more than 40% of the vote) by a whopping 338-70 margin. Then a filibuster led by conservative Southerners Thurmond, James Eastland and Sam Ervin halted its progress in the Senate.
Sponsor Birch Bayh of Indiana was able to muster 54-36 and 53-34 majorities for cloture in 1970, but both fell short of the required two-thirds vote. They also fell short of the two-thirds majority needed for a constitutional amendment, but both Bayh and his opponents suspected a vote on the actual amendment might have come out differently. National polls conducted in 1968 and 1970 found nearly 80% of Americans supported switching to a national popular vote, so openly opposing it was a risky stance.
The 1976 presidential election, in which Democrat Jimmy Carter got 1.7 million more votes but Republican Gerald Ford would have won the electoral vote if just 4,000 votes in Hawaii and 6,000 in Ohio had shifted his way, sparked a renewed effort by Bayh. By this time the South had elected a few Senators who weren’t opposed to a popular vote, but newly elected Republicans Orrin Hatch of Utah and Alan Simpson of Wyoming were opposed, seemingly just because that seemed like the conservative thing to be.
The leaders of the National Urban League and National Association for the Advancement of Colored People also voiced opposition, worrying that urban Black voters might find their clout reduced under the new setup. Civil rights leader and future Representative John Lewis begged to differ, presciently describing such arguments as succumbing to the “temptation of temporary advantage.” But the proposal again fell short in the Senate.
The limits of resilience and inventiveness
Throughout the debates of the late 1960s and 1970s, the most consistent argument against replacing the Electoral College was simply that it wasn’t broken. There hadn’t been an electoral vote-popular vote disagreement since 1888 or a vote thrown into the House since 1824. The Electoral College often turned narrow victories into more-decisive ones, thus conferring legitimacy on the winners, the reasoning went, and helped the states and political parties retain a role in choosing presidents they might otherwise lose.
“The Electoral College was unquestionably intended to serve ends we no longer care to serve, and which it no longer serves,” legal scholar Alexander Bickel wrote in 1971, but “we do well to remain attached to institutions that … challenge our resilience and inventiveness in bending old arrangements to present purposes with no outward change.”
I happen to find that argument weirdly compelling, or at least charming. But in retrospect it was also quite naive. Between 1892 and 1956 there wasn’t a single presidential election with a popular-vote margin of less than three percentage points. Since then there have been six. The two Bickel was familiar with (1960 and 1968), delivered Electoral College results that arguably added legitimacy to razor-thin popular vote victories. But there was no guarantee that was going to keep happening (in fact there are those who now argue electoral-vote winner John F. Kennedy actually lost the popular vote in 1960).
In 2000 it finally did stop happening, with not only an electoral victory for the popular-vote loser but also a drawn-out battle over the crucial Florida result that was finally decided along partisan lines by the Supreme Court. Not a great way of conferring public confidence or legitimacy on the office of president! Still, President George W. Bush did at least strive to appeal to those who didn’t vote for him once in office, and won both popular and electoral votes in 2004.
President Trump lost the popular vote in 2016 by a much bigger margin than Bush in 2000. He subsequently showed little interest in winning over doubters and uniting the nation, instead claiming his popular-vote loss was the result of fraud and trying to politically punish states that didn’t vote for him.
Trump’s behavior has at times seemed almost a parody of past defenses of the Electoral College. Without the Electoral College, political scientist Judith Best argued in 1975, “demagogues, self-nominated, individualistic leaders of impermanent factions, charismatic leaders riding a single issue might replace the candidates presently recruited because of their moderation, experience, records of electoral success, and service to permanent party organizations.” Oh, really?
Still, the odds of doing away with the Electoral College anytime soon seem slim. Past close calls at reform happened in low-partisanship environments, which is not what we have at the moment. And Republican politicians and voters seem to have convinced themselves the current electoral system is permanently skewed in their favor.
I’m pretty sure they’re wrong about this: most indications are that in the 2004, 2008 and 2012 elections it was aligned to deliver the election to the Democratic loser of a close-enough popular-vote loss. But good luck changing enough minds about that to get a constitutional amendment ratified. While only 38% of U.S. adults in the latest Gallup poll on the topic want to keep the Electoral College, 77% of Republicans do.
Possible paths for change
Electoral College opponents have in recent years tried an end-run of sorts with the National Popular Vote Interstate Compact, in which states pledge to award all their electoral votes to the popular-vote winner as soon as states representing an Electoral College majority of 270 votes are on board. This does not seem at odds with the Constitution, which lets states select electors by any manner they choose. It’s not an approach the framers of the Constitution envisioned, but neither is the current setup.
The weakness of the NPVIC might be its fragility. States can back out at any time, and basing everything on the national popular vote when the counting of those votes and the determination of voter eligibility remains entirely in the hands of the states seems potentially fraught. Still, it’s a lot easier to achieve than a constitutional amendment, and is currently only 75 electoral votes short of going into effect.
Another reform that could shake up although not replace the Electoral College would be expanding the House of Representatives. Keeping its membership constant at 435 since 1912, even as the U.S. population has more than tripled, has been an affront not only to representative government but to George Washington, whose only speech at the Constitutional Convention was a plea for each House member to represent no more than 30,000 people.
At the current U.S. population of 330 million, this would mean a House membership of 11,000 and an Electoral College with 11,103 votes. A constitutional amendment that was thought to fall one state short of ratification in 1792 but maybe actually didn’t put the maximum at 50,000 constituents, which implies a House membership of 6,600. That’s a lot! But doubling or tripling the House membership seems like a reasonable reform that could be achieved without a constitutional amendment. Given that it would further diminish the voting power of smaller states in the Electoral College, it might force some movement toward a popular vote.
“The history recounted here has a Sisyphean air,” Keyssar concedes near the end of his book. But Sisyphus had to keep rolling his boulder up a hill for eternity. We’ve only been trying to fix the Electoral College for about 230 years!
(1) The 22nd Amendment, ratified in 1951, limited how many times a person could be elected president, and the23rd(1961) gave the District of Columbia electoral votes for the first time, but neither really changed the electoralsystem.
(2) It is the source of most of the historical information in this column, with Michael Klarman’s 2016 book “The Framers’ Coup” also providing a few details.
This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.
Justin Fox is a Bloomberg Opinion columnist covering business. He was the editorial director of Harvard Business Review and wrote for Time, Fortune and American Banker. He is the author of “The Myth of the Rational Market.”
For more articles like this, please visit us at bloomberg.com/opinion
Subscribe now to stay ahead with the most trusted business news source.
©2020 Bloomberg L.P.