In 2016, Donald Trump was elected as president of the United States, despite losing the popular vote to Hillary Clinton. This result, as well as the 2000 presidential election outcome, has led to criticism on the functioning of the Electoral College. Marc Hooghe, Dieter Stiers and Michael S. Lewis-Beck analyze the last 37 presidential elections, and find that that the functioning of the Electoral College has changed little since 1876. However, as the margin between the two major candidates becomes narrower, the odds the candidate who did not win the popular vote will win the Electoral College will be higher.

While the functioning of the Electoral College has not changed since the 12^th Amendment that was passed in 1804, the system to select a US President – where states’ electors appoint the president and vice-president (usually) based on the candidate which won the plurality of their state’s votes – increasingly has received criticism. Both in 2000 as in 2016, the US presidential elections led to a mismatch between the winner of the country-wide popular vote, and the candidate that obtained a majority in the Electoral College. This has led to an ongoing debate about the role of the Electoral College, and one of the recurring points of criticism is that it gives undue weight to some of the smaller states, at the cost of the most populous. California has almost 40 million inhabitants and is characterized by a solid Democratic majority. In contrast, Wyoming has around 600,000 inhabitants and is equally characterized by a solid Republican majority. The fact that Wyoming is overrepresented compared to California, therefore could provide an undue advantage to one specific party. While this observation is, by itself, correct, several points should be made here.

First, having a federal system in the first place implies that the smaller states of the federation feel well represented. If the popular vote would be enough to appoint the president of the United States, it would be sufficient to gather the votes of the largest states (California, Texas, Florida, New York), and this would imply that the least populous states would not really have a say in the way the federation is being governed. So, having an Electoral College means one wants to give more influence to the less populated states, rather than just having a fully accurate reflection of their population weight. Second, the first requirement for any electoral system is that it would be neutral from a partisan point of view. Even if the goal is to provide an advantage to the least populous states, for the legitimacy of the electoral system it is important this option does not lead to an undue advantage for either the Republican or the Democratic Party. There’s the rub, in this case, as in 2000 as in 2016 the Republican candidate clearly benefited from the current system.

How disproportionate is the Electoral College?

But just how disproportional is the current system, and if it is, does this benefit a particular party? If that would be the case, this clearly erodes the long-term legitimacy of the Electoral College as a venerable institution in the United States. As it is important not to be myopic and include only some current events, we go back to the era of the Civil War. Obviously the 1872 elections do not help much for our purpose as candidate Horace Greely died on November 29, 1872, so his results in the popular vote no longer could influence the final outcome of the Electoral College, duly electing Ulysses S. Grant for a second term. For all 37 elections since 1876, however, we compared the popular vote with the outcome of the Electoral College (Figure 1).

Figure 1 – Proportion of two-party popular vote and votes in Electoral College since 1876

The dotted line in Figure 1 represents a perfectly proportional association between the popular vote and the results of the Electoral College. As can be noted, all observations are above the dotted line, indicating that the candidate receiving most of the popular vote, receives a further boost in the vote share within the Electoral College. This is a phenomenon that is typically found in any level of aggregate vote count. This basically implies that if a candidate receives a clear majority of the popular vote, s/he will obtain an ever-larger majority in the Electoral College. In Figure 1, we only calculated the vote share of the two major parties, but if we include third party candidates, the results stay the same.

However, Figure 1 also shows four outliers: in 1876, 1888, 2000 and 2016, a candidate that did not receive a majority of the popular vote, did receive a majority in the Electoral College. So, while 33 elections produced the “right winner”, in four elections this was not the case. As we can see, in all four cases the winning candidate only had a very small, and in some cases heavily contested majority. Is that the reason why the Electoral College sometimes selects the “wrong” winner?

When the Electoral College goes “wrong”

To answer this question, Figure 2 shows the difference between the popular vote and the vote share in the Electoral College for the two major parties by year, and the best-fitting quadratic curve. Apparently, the most disproportional elections were the 1988 elections, where Michael Dukakis obtained less than 46 per cent of the popular vote and lost the Electoral College by a landslide. In 1936, Franklin D. Roosevelt obtained more than 60 per cent of the popular vote and obtained almost all seats in the Electoral College.

Figure 2 – Absolute difference between vote share and share in Electoral College by party

So, throughout the 20^th century, we do not see any major problem in the functioning of the Electoral College: the candidate receiving a majority of the popular vote, obtains a considerable boost in the vote of the Electoral College. But in the final quarter of the 19^th century and at the start of the 21^st century, however, we observe that the differences in the popular vote have been very narrow. In those cases, we clearly enter a place where the system starts to break down. When the difference between both candidates is limited, apparently the system of the Electoral System is not sufficiently sensitive to be able to always select the “right” winner.

“Electoral-College_DEC2020_0696” (Public Domain) by wastatearchives

What is important in Figure 2, is that the fitted curves for Republicans and Democrats are almost the same. This means that for both major parties, the Electoral College functions in the same manner. If a candidate obtains a clear majority among the popular vote, the Electoral College will confirm this. When there is a very slim majority on the other hand, we enter the margin of error, and it becomes more likely that the “wrong winner” is selected. The system, however, works in the same way for both political parties. Note that we also conducted a more thorough multivariable analysis, controlling for the spread of the population between states and whether there was a successful third candidate, and include linear and quadratic and both additive and interaction effects. All results point to the same conclusion: the quadratic curve shown in Figure 2 fits the data better than a linear pattern, and there is no discernible difference in “electoral boost” in the College between the two parties.

Changing election results, not a changing Electoral College

Basically, therefore, the Electoral System did not become disproportional during the past century and a half. We also do not observe any bias toward one of the major parties – while the Republican Party tends to have a stronghold in some of the rural states, we must remember that the Democratic Party often scores well in some of the smaller states. So, the Electoral College did not change, but the election results did. Since 1996, all presidential elections have been heavily contested, with a narrow margin between the two major candidates. In those cases, it becomes more likely that the system will select the wrong winner (but again: without a bias toward a specific party).

So, what can be done? Our goal is not to say how things should be, but in the current heavily polarized environment, it become unlikely that a single candidate would obtain more than 60 per cent of the vote, as FDR did in 1936. Most likely, elections will remain highly contested, with professionally run campaigns, resulting in narrow margins, and a substantial chance of error. Earlier proposals to “solve” the problem of the wrong winner did not have much success, for the simple reason that the least populous states that stand to benefit from the current system, would have to approve such an amendment. This is rather unlikely. A more pragmatic solution would be to keep the system as it is, but to add a limited number of seats to the winner of the popular vote. Adding, e.g., 30 seats to the Electoral College, would move the regression line, so the margin of error that is so clearly visible in Figure 1, in fact would not exist anymore. We do not wish to speculate on how realistic such a scenario is, but we can observe that selecting the wrong winner in a repeated manner may erode the legitimacy of the Electoral College as an institution.

• This article is based on the paper, “Has the electoral college grown more disproportional? An analysis of election results, 1876–2020” in Politics and Policy.
• Please read our comments policy before commenting.
• Note: This article gives the views of the author, and not the position of USAPP – American Politics and Policy, nor the London School of Economics. 
• Shortened URL for this post: https://bit.ly/3JFv5pF