9.10 Myths about the Proliferation of Candidates and a Breakdown of the Two-Party System

Tara Ross, a lobbyist against the National Popular Vote Compact who works closely with Save Our States, speculates that a national popular vote would lead to a proliferation of candidates and Presidents who are elected with a tiny percentage of the vote:

“[The National Popular Vote Compact] is not even looking for a minimum plurality. Thus, a candidate could win with only 15 percent of votes nationwide.”[301] [Emphasis added]

Ross has also stated:

“The most likely consequence of a change to a direct popular vote is the breakdown of the two-party system.”[302] [Emphasis added]

Hans von Spakovsky of the Heritage Foundation has written:

“NPV could destabilize America’s two-party system.”[303]

If an Electoral College type of arrangement were essential for avoiding these conjectured outcomes, we should see evidence of this outcome in elections that do not employ it.

Evidence from plurality-vote popular elections for state chief executive

A nationwide campaign for President would have the same political dynamics as existing campaigns for state chief executive. In both cases, every voter is equal, and the winner is the candidate receiving the most popular votes from the jurisdiction served by the office.

When state chief executives are elected in statewide plurality-vote popular elections, there is no evidence of a proliferation of candidates, candidates winning with 15% of the vote (or any similar small percentage), or a breakdown of the two-party system.

In the 1,027 general elections for Governor in the United States between 1946 and 2015:

• 88% of the winners received more than 50% of the vote (908 out of 1,027).
• 97% of the winners received more than 45% of the vote (1,001 out of 1,027).
• 99% of the winners received more than 40% of the vote (1,013 out of 1,027).
• 100% of the winners received more than 35% of the vote.[304]

Table 9.27 shows the 26 general elections (out of 1,027) for Governor between 1946 and 2014 in which the winner received less than 45% of the popular vote.

Table 9.27 The 26 general elections for Governor between 1946 and 2014 (out of 1,027) in which the winning candidate received less than 45% of the vote

Winning percentage	Winner	State	Year
35.4%	Angus King	Maine	1994
36.1%	Lincoln Chafee	Rhode Island	2010
36.2%	John G. Rowland	Connecticut	1994
36.6%	Benjamin J. Cayetano	Hawaii	1994
37.0%	Jesse Ventura	Minnesota	1998
38.1%	John Baldacci	Maine	2006
38.2%	Paul LePage	Maine	2010
38.2%	George D. Clyde	Utah	1956
38.9%	Walter J. Hickel	Alaska	1990
39.0%	Rick Perry	Texas	2006
39.1%	Jay S. Hammond	Alaska	1978
39.1%	James B. Longley	Maine	1974
39.7%	Evan Mecham	Arizona	1986
39.9%	John R. McKernan Jr.	Maine	1986
40.1%	Norman H. Bangerter	Utah	1988
40.4%	Lowell P. Weicker Jr.	Connecticut	1990
40.7%	Gina Raimondo	Rhode Island	2014
41.1%	Tony Knowles	Alaska	1994
41.4%	Meldrim Thomson Jr.	New Hampshire	1972
41.4%	Don Samuelson	Idaho	1966
42.2%	Michael O. Leavitt	Utah	1992
43.3%	Brad Henry	Oklahoma	2002
43.7%	Mark Dayton	Minnesota	2010
44.4%	Tim Pawlenty	Minnesota	2002
44.6%	Nelson A. Rockefeller	New York	1966
44.9%	Jim Douglas	Vermont	2002

As a practical matter, it is generally easier for a minor-party or independent candidate to launch a gubernatorial campaign in a smaller state than a larger state. Indeed, more than half (14 of 26) of the Governors who were elected with less than 45% of the vote in table 9.27 were in states with only three or four electoral votes (Alaska, Hawaii, Idaho, Maine, New Hampshire, Rhode Island, and Vermont). That fact suggests that there would be fewer such winning candidacies in a larger (that is, nationwide) election.

Evidence from states that made the transition from an electoral college to popular election for Governor

At the time when the U.S. Constitution came into effect in 1789, Governors were elected by popular vote in only five of the original 13 states (Connecticut, Massachusetts, New Hampshire, New York, and Rhode Island).[305]

Seven of the original states had an Electoral College type of arrangement for electing their chief executive—either by means of specially elected gubernatorial electors or an election in which state legislators acted as an electoral college.[306]

In six of these seven original states, there was a transition before the Civil War from an Electoral College type of arrangement for electing the state’s chief executive to statewide popular elections.[307]

In addition, Kentucky was admitted to the Union in 1792 and subsequently made the transition from an Electoral College type of arrangement for selecting its Governor to popular elections.

These transitions provide further evidence concerning the incorrectness of speculations about popular elections resulting in a proliferation of candidates, candidates being elected with 15% or other small percentages of the popular vote, and the breakdown of the two-party system.

Let’s examine what happened in these transitions from an Electoral College type of arrangement to popular elections.

Under the 1792 Kentucky Constitution, the Governor was elected by a state-level electoral college. Seats in the lower house of the legislature were apportioned among the counties on the basis of population.[308] Every four years, the voters of each county were entitled to vote for a number of gubernatorial electors equal to the county’s number of members of the legislature’s lower house. Each voter[309] was allowed to vote for all of his county’s gubernatorial electors—that is, the election of electors was conducted on a countywide winner-take-all basis.[310] The winning gubernatorial electors then met two weeks later to choose the Governor.[311] This state-level electoral college was used to elect the Governor in 1792 and 1796.

When Kentucky’s constitution was revised in 1799, the gubernatorial electoral college was abolished and replaced by a statewide popular election for Governor starting in 1800.[312]

Were there any 15% governors after Kentucky transitioned from a gubernatorial electoral college to a popular election for Governor?

In 81% of the subsequent pre-Civil-War gubernatorial elections in Kentucky (that is, 13 of 16 elections), the winning candidate for Governor received more than 50% of the statewide popular vote. The winners of the other three elections received 49%, 39%, and 33%.[313] That is, there were no 15% winners after the abolition of Kentucky’s electoral college for choosing the Governor. During this period, nine of the 16 elections were two-person races; three were three-person races; two were four-person races; and there was no competition at all in two races.

In Delaware, the transition occurred in 1792. After Delaware’s transition, 91% of the pre-Civil-War gubernatorial races (21 of 23) were two-person races, and the winning candidate received between 50.1% and 55.2% of the popular vote. In the two three-person races, the winners each received 48% of the popular vote.[314] That is, there was no proliferation of candidates, and there were no 15% winners.

After Georgia’s transition in 1825, 100% of the 18 winners in the pre-Civil-War gubernatorial races received more than 50% of the vote. There were 16 two-person races and two three-person races. Again, there was no proliferation of candidates, and there were no 15% winners.[315]

After North Carolina’s transition in 1836, 100% of the 13 winners of the pre-Civil-War gubernatorial races received more than 50% of the vote. All of these general-election races were two-person races (usually between a Democrat and a Whig).[316]

After Maryland’s transition in 1838, 100% of the six winners of the pre-Civil-War gubernatorial races received more than 50% of the vote.[317] These general-election races were all two-person races (all between Democrats and Whigs).[318]

After New Jersey’s transition in 1844, 100% of the seven winners of the pre-Civil-War gubernatorial races received more than 50% of the vote. These general-election races were all two-person races (all between a Democrat and a Whig).[319]

After Virginia’s transition in 1851, 100% of the three winners of the pre-Civil-War gubernatorial races received more than 50% of the vote.[320] These general-election races were all two-way races.[321]

Table 9.28 shows that 94% of the 86 gubernatorial winners received more than 50% of the vote in the seven states that transitioned from an Electoral College type of arrangement for electing the Governor to statewide popular elections before the Civil War.

Table 9.28 Winning percentages in gubernatorial races in the seven states that transitioned from an Electoral College type of arrangement to statewide popular election before the Civil War

State	Year	Number of races	Number of winners with less than 50% of popular vote	Number of winners with more than 50% of popular vote	Percentage of winners with more than 50% of popular vote
Delaware	1792	23	2	21	91%
Georgia	1825	18	0	18	100%
Kentucky	1800	16	3	13	81%
Maryland	1838	6	0	6	100%
New Jersey	1844	7	0	7	100%
North Carolina	1836	13	0	13	100%
Virginia	1851	3	0	3	100%
Total		86	5	81	94%

Note that this 94% percentage is even higher than the percentage of post-World-War-II gubernatorial races in which the winner won with more than 50% of the popular vote that we mentioned earlier in this section.

In the five races where no candidate received more than 50% of the popular vote, the winning candidates received 48% and 48% in the two Delaware races and 49%, 39%, and 33% in the three Kentucky races. There were no 15% winners.

Table 9.29 shows that 87% of the gubernatorial races were two-person races:

• 2% of the races were uncontested;
• 87% of the races had two candidates;
• 8% of the races had three candidates;
• 2% of the races had four candidates; and
• no races had more than four candidates.

Table 9.29 Number of candidates in gubernatorial races in the seven states that transitioned from an Electoral College type of arrangement to statewide popular election before the Civil War

State	Year	Number of races	Number of races with 1 candidate	Number of races with 2 candidates	Number of races with 3 candidates	Number of races with 4 candidates	Percent of races with 2 candidates
Delaware	1792	23		21	2		87%
Georgia	1825	18		16	2		89%
Kentucky	1800	16	2	9	3	2	56%
Maryland	1838	6		6			100%
New Jersey	1844	7		7			100%
North Carolina	1836	13		13			100%
Virginia	1851	3		3			100%
Total		86	2	75	7	2	87%

The above pattern also applies to the one state that transitioned from an Electoral College type of arrangement to statewide popular election after the Civil War—South Carolina. All of South Carolina’s nine gubernatorial elections between 1865 and 1882 were two-person races (so that the winning candidate received more than 50% of the popular vote).[322]

The two-party system does not owe its existence to the Electoral College or the winner-take-all method of awarding electoral votes.

The two-party system first emerged at the national level in 1796—32 years before a majority of the states adopted the winner-take-all method of awarding electoral votes.

There were no political parties at the national level in the nation’s first and second presidential elections in 1789 and 1792 when George Washington won 100% of the votes in the Electoral College. See sections 2.2 and 2.4.

Because of inevitable differences of opinion on various policy issues, this unanimity ended with the first election in which Washington was not a candidate.

In the 1796 election, the congressional caucus of the Federalist Party and the caucus of the Democratic-Republican Party nominated candidates for President and Vice President. Both national parties ran slates of presidential electors at the state level supporting their nominees (section 2.5).

Given that only three states had winner-take-all laws in the 1789, 1792, and 1796 elections, it can hardly be argued that the Electoral College—much less the winner-take-all method of awarding electoral votes—created the two-party system in the United States.

Duverger’s Law

After studying election systems around the world, the French sociologist Maurice Duverger observed and explained the tendency of plurality-vote elections to prevent a proliferation of candidates and to sustain a two-party system.[323]

Duverger observed that voters tend to shy away from parties or candidates who have no chance of winning. Indeed, the effect of voting for a splinter candidate who cannot win is usually to help a candidate whose views are diametrically opposite to the voter’s own views.

For example, 97,488 Floridians voted for Green Party presidential candidate Ralph Nader in 2000. George W. Bush carried Florida by a mere 537 popular votes. The votes cast for Nader enabled George W. Bush to win the electoral votes of Florida and thereby win the presidency.[324]

Similarly, in 2008, votes cast for Bob Barr (the Libertarian presidential candidate) enabled Barack Obama to win North Carolina’s electoral votes.[325] In 2008, votes cast for Ralph Nader enabled John McCain to win Missouri’s electoral votes.[326]

In 2020, Libertarian presidential candidate Jo Jorgensen received considerably more popular votes in Arizona, Georgia, and Wisconsin than Biden’s margin over Trump in these states, as shown in table 9.30.

Table 9.30 Libertarian vote in Arizona, Georgia, and Wisconsin in 2020

State	Electoral Votes	Biden	Trump	Jorgensen	Biden margin over Trump
Arizona	11	1,672,143	1,661,686	51,465	10,457
Georgia	16	2,473,633	2,461,854	62,229	11,779
Wisconsin	10	1,630,866	1,610,184	38,491	20,682
Total	37	5,776,642	5,733,724	152,185	42,918

Specifically, Jorgensen received more than three times as many votes (152,185) as Biden’s combined margin over Trump in the three states (42,918). These three states together possessed 37 electoral votes. Without the 37 electoral votes from these three states, there would have been a 269–269 tie in the Electoral College. On January 6, 2021, the Republican Party had a majority of the House delegations and would have been in a position to choose Trump as President.[327]

Because of the severe penalty that plurality voting imposes on third-party and independent candidates, political groups with broadly similar platforms tend to coalesce behind one candidate in order to enable that candidate to win the most votes—and thereby get elected to office.

The result of Duverger’s worldwide study of voting systems is often called “Duverger’s Law.”

Footnotes

It has been claimed that:

“The current system quarantines … spoilers … within a small number of states.”[328]

In fact, far from quarantining spoiler candidates, the current state-by-state winner-take-all method of awarding electoral votes greatly increases the payoff to potential spoilers and, therefore, increases their incentive to launch such campaigns.

The current system offers a potential spoiler the alluring prospect of finding one or more states where the spoiler’s narrow appeal can flip all of a state’s electoral votes and thereby possibly flip the national outcome.

In 2000, for example, George W. Bush won a 537-vote plurality in Florida. Ralph Nader’s 97,488 popular votes in Florida in 2000 were more than sufficient to flip all of the state’s electoral votes to Bush and thereby decide the presidency in an election in which 105,396,627 votes were cast nationally.[329]

In 2020, Libertarian presidential candidate Jo Jorgensen received considerably more popular votes in Arizona, Georgia, and Wisconsin than Biden’s margin over Trump in these states, as shown in table 9.30. Without the 37 electoral votes from these three states, there would have been a 269–269 tie in the Electoral College. On January 6, 2021, the Republican Party had a majority of the House delegations and would have been in a position to choose Trump as President.[330]

Segregationist Strom Thurmond had a strong regional appeal and won 38 electoral votes in 1948 by carrying Alabama, Louisiana, Mississippi, and South Carolina.[331] Similarly, segregationist George Wallace won 46 electoral votes in 1968 by carrying Alabama, Arkansas, Georgia, Louisiana, and Mississippi.

In 2024, an NBC News story entitled “Operatives with GOP ties are helping Cornel West get on the ballot in a key state” said:

“Democrats fear West’s potential to siphon votes from President Joe Biden in places where he is on the ballot in a close election, and some Republicans are publicly discussing ways to boost West and other minor candidates like Robert F. Kennedy Jr. and the Green Party’s Jill Stein in the hopes of splitting the anti-Donald Trump coalition.”[332]

Under the current system, minor-party and independent candidates have significantly affected the outcome in 42% (eight out of 19) of the presidential elections between the end of World War II and 2020 by switching electoral votes from one major-party candidate to another, including:

• Henry Wallace in 1948
• Strom Thurmond in 1948
• George Wallace in 1968
• John Anderson in 1980
• Ross Perot in 1992 and 1996
• Ralph Nader in 2000
• Ralph Nader in 2008
• Bob Barr in 2008
• Gary Johnson in 2016
• Jill Stein in 2016
• Jo Jorgensen in 2020.

The current system also entices potential spoilers with another disproportionate payoff, namely denying an absolute majority of the electoral votes to any candidate. That result would either throw the choice of President into the U.S. House of Representatives or enable the spoiler to use his or her presidential electors to bargain with the major parties. This latter prospect has proven especially alluring to regional candidates, such as segregationist Strom Thurmond (who won 39 electoral votes in 1948 with only 2.4% of the national popular vote) and segregationist George Wallace (who won 46 electoral votes in 1968 with 13.5%).

The National Popular Vote Compact eliminates the possibility of a presidential election being thrown into Congress (section 1.6).

Footnotes