Issue Advertising and Legislative Voting on the Affordable Care Act

• DOI: 10.1177/1065912917724007
• Date: 01 March 2018
• Published date: 01 March 2018
• Subject Matter: Articles

https://doi.org/10.1177/1065912917724007

Political Research Quarterly

2018, Vol. 71(1) 102 –114

© 2017 University of Utah

Reprints and permissions:

sagepub.com/journalsPermissions.nav

DOI: 10.1177/1065912917724007

journals.sagepub.com/home/prq

Article

In the early 1990s, the health insurance industry ran a

series of issue advertisements as part of an effort to defeat

President Clinton’s proposal to reform the US health care

system. The characters in the commercials, “Harry and

Louise,” became household names, and in the aftermath,

several analysts attributed the failure of health reform to

the effect the ads had on the American public (Goldsteen

et al. 2001; Johnson and Broder 1996; West, Heith, and

Goodwin 1996). That allegation may well be overstated

(Brodie 2001; Jacobs 2001; Jacobs and Shapiro 2000),

but the prominence of the advertising campaign refueled

a longstanding controversy about the ability of resource-

rich special interests to influence members of Congress

(Falk, Grizard, and McDonald 2006; Seabrook and

Overby 2009).1

Neither Harry and Louise nor the controversy over

high-priced advocacy died in 1994. Fifteen years later,

the same characters played by the same actors would

reappear in the fight over the Patient Protection and

Affordable Care Act (ACA). This time, however, Harry

and Louise were featured in an expensive issue advertis-

ing campaign by interest groups on behalf of health

reform, while interest groups opposing reform launched

an expensive campaign of their own.

Issue advertising in the ACA fight broke all records. An

estimated $91 million was spent on issue advertising dur-

ing the first one hundred days of the Obama presidency,

most of that on health reform. That sum exceeded the

amount spent in the first one hundred days of George W.

Bush’s presidency by a factor of ten (TNSMI/CMAG

2009). In calendar year 2009, interest groups spent over

$200 million on ACA-related television advertisements.

By comparison, spending by all political action commit-

tees and 527 groups in the health sector combined was less

than $50 million over the preceding two-year election

cycle.2 But issue advertising did not dissipate thereafter,

nor has it been confined to health legislation. Total spend-

ing on issue ads was forecast to hit $2 billion in 2013,

double what was spent only four years before (Stilson

2012).

Like other forms of outside lobbying (e.g., Bergan

2009; Goldstein 1999, 2001; Kollman 1998), the central

purpose of issue advertising is to influence legislators on

an issue about which the interest group cares. If the spend-

ing statistics are any indication, an increasing number of

groups believe that it affects what legislators do. Hall and

Reynolds (2012, 895) show that in a fight over Medicare

724007PRQXXX10.1177/1065912917724007Political Research QuarterlyReynolds and Hall

research-article2017

1The Brookings Institution, Washington, DC, USA

2University of Michigan, Ann Arbor, USA

Corresponding Author:

Molly E. Reynolds, The Brookings Institution, 1775 Massachusetts

Ave. NW, Washington, DC 20036, USA.

Email: mereynolds@brookings.edu

Issue Advertising and Legislative

Voting on the Affordable Care Act

Molly E. Reynolds1 and Richard L. Hall2

Abstract

During the congressional fight over the Affordable Care Act (ACA), interest groups spent record sums on television

issue advertising in targeted efforts to influence members of Congress, but did the money make any difference? We

use the literatures on outside lobbying and legislative behavior to develop two hypotheses about issue advertising’s

effects on members’ voting decisions. We test the hypotheses using population-weighted, station-level advertising

data mapped into congressional districts. We find negligible evidence that issue advertising had a causal effect on

either House committee or floor votes on the ACA, even applying forgiving statistical standards. Neither do we find

evidence when we ignore the endogeneity bias that should inflate advertising’s effects, employ alternative measures

and specifications, or limit the analysis to legislators for whom the probability of vote change was highest. The results

justify skepticism that the millions of advertising dollars spent on the ACA had a net effect on members’ voting

decisions. In conclusion, we consider several reasons why our hypotheses are not borne out and suggest several

avenues for future research.

Keywords

issue advertising, legislative politics, health policy