Leveling the Playing Field:

1
Leveling the Playing Field A Field Experiment
Testing the Impact of Radio Advertisements on
Electoral Competition Costas Panagopoulos costas
_at_post.harvard.edu Donald P. Green donald.green_at_ya
le.edu 2006 APSA Annual Meeting Philadelphia,
PA
2

• Motivation for the Current Study
• Impasse in the campaign spending literature
  Model identification in the shadow of strategic
  allocation of resources
• Competing proposals of Jacobson (1978, 1990),
  Green Krasno (1990), Levitt (1994), Erikson
  Palfrey (2000)

3

• Breaking the Impasse Through Experimentation
• Gerber (2004) assess the cost-per-vote generated
  by randomized campaign inputs and compare the
  results to the non-experimental estimates
• Current project is based on the same
  methodological strategy but attempts to assess
  the effects of a common input a nonpartisan ad
  that mentions the candidates names and, where
  applicable, party affiliations

4

• Hypothesis
• Jacobson (1990335) Because voters are
  demonstrably reluctant to vote for candidates
  they know nothing about, challengers have a great
  deal to gain by making themselves betterknown to
  the electorate. Their level of campaign
  activitythus has a strong influence on how well
  they do at the polls.
• Prediction Interventions that mention both
  incumbents and challengers will, on average,
  benefit challengers, who typically are less
  familiar to the electorate

5

• Subjecting This Hypothesis to an Experimental
  Test
• Mayoral contests in November 2005, similar in
  competitiveness to US House elections
• Defining the experimental population of cities
• Random assignment within radio cost strata

6

• Radio Intervention
• SCRIPT (Nonpartisan 60-second Nov. 1-7, 2005)
• Many people don't realize how important local
  government is.  But think about it. Your local
  government is in charge of things that affect
  your life every day police protection,
  transportation, garbage collection, tax
  assessment. From fire departments to libraries to
  safe drinking water -- it's all part of local
  government.
• Here's where you come in. Voting. If you're a
  registered voter in SYRACUSE, you have an
  opportunity to shape the direction of your city
  by electing the mayor and other local officials.
  On Tuesday, November 8th residents of SYRACUSE
  will vote to decide whether to RE-elect
  Democratic MAYOR MATTHEW DRISCOLL or to support
  his opponent Republican JOANNIE MAHONEY.
• Take part in shaping your city's future. Be sure
  to vote on November 8th.
• Paid for by the Institution for Social and
  Policy Studies, a nonpartisan organization that
  encourages citizens to take an active role in
  their communities.

7

• Regression Models
• Model 1
• Model 2

8
(No Transcript)
9
(No Transcript)
10

• Placing These Experimental Results into the
  Context of the Campaign Finance Literature
• Bayesian exercise, recognizing that expert priors
  are divergent
• A closer look at the implications of the leading
  models, however, reveals that there is more
  consensus than one might think about the effects
  of grants-in-aid to both candidates

11
(No Transcript)
12
(No Transcript)
13

• Implications
• A Bayesian perspective suggests that the current
  experiment affirms and sharpens priors from the
  campaign spending literature
• Although more data are needed to reduce sampling
  variability, the current study suggests that
  campaigns matter greatly in the context of
  asymmetrical competition the norm in US
  elections
• Provides a rigorous new research paradigm for
  reinvigorating the literature on the effects of
  campaign spending

14
(No Transcript)