Political Science 17.20
Introduction to American Politics
Professor Devin Caughey
MIT Department of Political Science
Campaigns and Elections
Lecture 16 (April 9, 2013)
1 / 13
An Anecdote Relevant to the Paper
From a friend in the Dept. of Health & Human Services:
Bureaucrats are most likely to “fill in the details” when
Congress doesn’t have the time/expertise to do so.
Section 4302 of the ACA requires HHS to collect
demographic data relevant to health disparities (“as
deemed appropriate by the Secretary”)
The Office of Minority Health within HHS interprets this as
requiring collection of data on LGBT population (not
mentioned in ACA) → eligible for extra funding as “at risk”
Controversial within HHS b/c other groups were ignored.
[Consistent with Obama’s desire for LGBT support.]
2 / 13
Outline
1
The Puzzle of Predictability
2
Campaign Effects
3 / 13
Outline
1
The Puzzle of Predictability
2
Campaign Effects
4 / 13
The Predictability of Elections
Elections are predictable on the basis of “fundamentals”:
Internal (e.g., party ID)
External (e.g., economy)
5 / 13
Incumbent share of two-party
vote (%)
Structural Forecasts: The “Bread and Peace” Model
65.0
1996
55.0
1960
50.0
45.0
1980
2008
1968 1976
1956
1984
1988
2004
2000
1992
1952
40.0
1964
1972
60.0
2012 prediction for Obama: 47.5%
-2 -1 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Real income growth and military fatalities combined
Image by MIT OpenCourseWare.
6 / 13
If Elections Are Predictable, Why Are Polls Unstable?
Why are polls so variable?
Do campaigns matter at all?
7 / 13
Outline
1
The Puzzle of Predictability
2
Campaign Effects
8 / 13
How Campaigns Still Could Matter
Even if elections were completely predictable, campaigns could
still matter if they:
Inform voters about the fundamentals (“enlightening”)
Have large but counterbalancing effects
Assumes optimal campaign and balanced resources
But elections are not entirely predictable. . .
9 / 13
Campaign Effects
Campaigns can affect outcomes in the following ways:
Reinforcement (bringing partisans back to the fold)
Persuasion (issues, attributes)
Priming (criteria of evaluation)
Mobilization (convince supporters to vote)
10 / 13
Presidential Campaign Strategies
Clarifying Campaign: Candidate advantaged by
fundamentals (prosperity, peace) emphasizes (primes)
those issues and clarifies his connection to them
Examples: Johnson in 1964, Reagan in 1984
Counterexample: Gore in 2000
Insurgent Campaign: Disadvantaged candidate
emphasizes issue on which their have an advantage and
which their opponent’s position is unpopular
Winners: Kennedy (1960), Carter (1976), Bush (2000)
Losers: Stevenson, Goldwater (1964), Dole (1996),. . .
11 / 13
Dynamics of Presidential Campaigns
Early polls not very accurate
Over time, polls become
less variable (fewer swing voters)
more even
more accurate and closer to forecast (informing)
Some campaign effects persist (“bump”) but most effects
dissipate quickly (“bounce”)
Short-term campaign effects can still matter if occur late
→ deluge of late ads
12 / 13
Do Campaigns Matter?
Affect outcomes, esp. if candidates not well known (e.g.,
primaries, open seats) or resources are unequal
We don’t observe “non-optimal” behavior very often
Campaigns affect candidates
→ Learn from voters, challengers
13 / 13
MIT OpenCourseWare
http://ocw.mit.edu
17.20 Introduction to American Politics
Spring 2013
For information about citing these materials or our Terms of Use, visit: http://ocw.mit.edu/terms .