Can an Independent candidate win the
White House?

Voter/ConsunderResearch

JanvanLohuken

The Question

 Could an independent candidate actually win a race of the White House.
 As of today there are good odds that both the GOP and the Democratic
nominee will be extremely unpopular, that both will enter the general
election with higher negative impressions than positive impressions.
 If ever there were a circumstance under which an independent could win,
this is it.
 The question is, is it possible.

 Polling this far out cannot provide the answer
 This lays out some other ways to look at that question

2

 Polling in March out
does not provide
good forecasts for
November:
The further from the election
(horizontal) the less polling
results correlate with actual
outcomes (vertical)

3

 Vote
share

History:
3rd Party /
Independent
candidates who
topped 5%

EV

1912

Teddy Roosevelt

28%

88

1924

Robert LaFollette

17%

13

1948

Strom Thurmond

3%

39

1968

George Wallace

14%

46

1980

John Anderson

7%

0

1992

Ross Perot

19%

0

Since Perot
• Electorate has changed dramatically: sharp increase in
Independent registration
• Pervasive anger at all things Washington

4

 Gallup reports a sharp increase in Independent voters.
Voter registration statistics in states that register by party
show the same.

Sharp increase in
independents:
Gallup

5

 Approach #1: Back
of the envelope:
Electoral College
Votes in states that
have voted for an
independent

1. Look at statewide races that have elected Independent candidates in statewide
races

2. Take the most Republican state to do so (Alaska) and the most Democratic state
to do so (Rhode Island) and consider all states in between (on a Tilt R to Tilt D
ranking) in play.
3. This generates an estimated 421 electoral college votes
 Excludes 3 states more Democratic than Rhode Island
 Excludes 12 states more Republican than Alaska
 Note: Partisan ranking of states obtained from Cook Political Report

6

  States that have elected Independents for Governor (relatively recently)
 Ventura (MN)
 Walker (AK)
 Weicker (CT)
 King (ME)
 Chafee (RI)

Approach #1
Details

 States that have elected Independents for US Senate
 Sanders (VT)
 Lieberman (CT)
 King (ME)
 Note: most of were already quite well known before they decided to run as
Independents

7

  Elasticity is a measure reflecting how many swing voters a state has.
 You can have very partisan states that are very elastic:
 Rhode Island is both very Democratic and very elastic; even with a lot of
elasticity Republicans have a very tough time winning. But note that
Rhode Island has had both Republican and Independent Governors
recently.
 Make the assumption is that with more elasticity, and therefore fewer straight
ticket voters, ‘elastic’ states might be more likely to vote for an Independent.

Approach #2:
elasticity

 Scenario 1: If you pick the most elastic and center most states (see next page:
middle cell of the top row), the number of electoral college votes is 66
 Scenario 2: If you pick the high and medium elastic and center most states (middle
cell of top two rows), the number of electoral college votes adds to 169

 Scenario 3: all top swing states, plus medium swing of center states (entire top
row, plus middle cell of the 2nd row); this yields 205 electoral college votes.
 Note: any of these scenarios might be enough to prevent a Republican or a
Democrat from getting an absolute majority of the Electoral College votes,
throwing the election to the House of Representatives. See last page on that.
Note: Data is taken from a the 538 column a couple of years ago (NYT, May 21, 2012).

8

 Approach #2:
details

Partisanship D states
Elasticity

Center states

R states

High

HI, MA, RI, VT

ME, OR, WA,
AK, ND, SD, WV
CO, IA, NH, NM,
WI, AR, MT

Medium

CA, CT, IL

NJ, FL, MI, MN,
NV, OH, MO

AR, ID, KS, KY,
NE, TN, TX, UT,
WY

Low

DC, DE, MD, NY

NC, PA, VA, IN,
GA

AL, LA MS, OK,
SC

9

  Ross Perot got 13% of the Democrats, 17% of Republicans and 30% of Independents (exit
polling); he got fewer Independents than Clinton or Bush, so not a good model.

Approach #3:
Perot / Ventura /
King

Clinton

Bush

Perot

Democrats

77

10

13

Republicans

10

73

17

Independents

38

32

30

 Reflecting the increase in Independents, if you apply these numbers to Gallup’s estimates of
R’s, D’s and I’s, Perot would have gotten 21.2% of the vote, instead of 18.9% he did get.

10

 Ventura got elected with the following distribution of partisan support (exit
polling)

Approach #3:
Perot, Ventura,
King

Humphrey

Coleman

Ventura

Democrats

56

10

33

Republicans

5

66

29

Independents

17

28

52

Angus King has been elected as an independent 3 times; the first was his election
to the Governorship in 1994

Brennan
(D)

Collins (R)

Carter
(Green)

King (I)

Democrats

61

10

7

23

Republicans

13

41

2

45

Independents

26

23

11

40

Note I could not find exit polling the Linc Chafee race for Governor of Rhode Island.
11

  Apply the Ventura / King model with some changes. At this point there is no reason
to believe an Independent candidate would draw more Democratic than
Republican votes for instance. Also anti DC anger is higher than it was in ’98.
 The model: the Independent candidate gets 22.5% of GOP and Dem votes
and 55% of Independent vote

Approach #3:
Independent voters

% of all voters

% of group going to
Independent

D’s

30

.3

I’s

43

.55

R’s

26

.3
Total

Multiply
and add
9%
23.65%

7.8%
40.45%

 If these assumptions are correct, it is not hard to see an Independent get a larger
vote share than either the Republican or Democratic candidate
 Note: this only works IF Independents actually bother to show up: the 2012
national exit poll pegged the number of Independents on election day at 29%,
while Gallup reports that 40% considered themselves Independents in 2012.
 Below we use exit polling, reflecting counts of Independents who actually do
show up.

12

  It’s difficult to find a source of data measuring independent voters the same way in
every state

Approach #3:
Independent voters
by state, assign
Electoral College
votes

 Exit polling is one source: uses party ID
 One problem is that exit polling is not conducted in every state every two
years
 Data below goes as far back as 2008 exit polls to get statistics
 But: relatively few states, the bulk are ‘12 and ’14 data
 But: mixing Presidential and non Presidential elections

 Step 1: apply the Ventura / King model from the previous page and add models for
D and R candidates
 Dem model: 22.5% of I vote, 60% of D vote, 10% of R vote
 Rep model: 22.5% of I vote, 10% of D vote, 60% of R vote
 Step 3: calculate who will win each state and allocate EV’s.
 Result:
 Independent: 443 EV
 Democrat: 42 EV
 Republican: 30 EV

13

 There are several problems with this model

Approach 3#: grain
of salt

1.

Partisans have become more partisan in the last few years. Defection rates of 40%
among today’s R’s and D’s strike me as high.

2.

Under this scenario 23 states containing 302 EV’s would go to the Independent with a
margin smaller than 5%. That is a lot of close races that would have to fall in place.
 Only 17 states would go to the Independent by a margin of more than 5%,
amounting to 145 electoral college votes (AL, AR, AZ, FL, ME, MA, MO, MT, NV,
NH, OR, TN, VT, WA, WI)
 Ignoring for the moment that Maine could split its Electoral College vote

3.

The model does not recognize that some states have more ‘swing’ or ‘elasticity’ than
other states. The number of Independents does not reflect this entirely.
 In Kentucky, for instance, rural Democrats routinely vote for Republicans for
federal office. In California cross-over voting is a rare event.

4.

In a lot of the very close states either Democrats or Republicans or both have a solid
infrastructure in place to get their people out. In close races that matters.
 In many southern states for instance, an independent might split the White vote
while the African American vote would probably stay solidly Democratic,
potentially moving some Republican states into the Democratic column.

 I do think it is very possible that a strong independent challenge could prevent the major
party candidates from getting a majority of the Electoral College. There are enough
states on the list of 17 above that are ‘must get’ states for either party.

14

  The numbers suggest it is possible that an independent candidate could win
outright, but I think it is unlikely.
 A lot of closely divided states would have to fall in line

Bottom Line

 Independents would truly have to behave as Independents.
 In a number of states people register as Independent who are quite liberal
and vote Democratic nearly all the time (Massachusetts for instance).
 The model assumes equal defection rates among Republicans and Democrats.
This far out that is a reasonable assumption. In November things may work out
differently.
 Trump for instance, with many primary voters to his left, might suffer from a
higher defection rate than Clinton who is closer to the center.

15

  To be elected President requires an absolute majority of the Electoral College vote.
 The Twelfth Amendment provides for what happens if the Electoral College fails to
elect a president or vice president:

EV absolute
majority

… if no person have such majority, then from the persons having the
highest numbers not exceeding three on the list of those voted for as President, the
House of Representatives shall choose immediately, by ballot, the President. But in
choosing the President, the votes shall be taken by states, the representation from
each state having one vote. if no person have such majority, then from the persons
having the highest numbers not exceeding three on the list of those voted for as
President, the House of Representatives shall choose immediately, by ballot, the
President. But in choosing the President, the votes shall be taken by states, the
representation from each state having one vote
 Note: 33 state delegations are majority Republican, although that count could change in these
elections; but it won’t change by much.
… And if the House of Representatives shall not choose a President
whenever the right of choice shall devolve upon them, before the fourth day of
March next following, then the Vice-President shall act as President, as in the case
of the death or other constitutional disability of the President.

 So we’ll have to see who Trump picks as his VP nominee.

16