Local Media Coverage of Candidates for State Supreme Courts

Abstract

What explains media coverage of state supreme court candidates? While the scholarly literature thoroughly examines media coverage of judicial decision-making, little is known about how members of the press cover candidates for the bench. This is an important omission as the campaign trail is one of the few places where voters and candidates directly interact with one another. In this research, I present new data regarding local press coverage of candidates for state supreme courts between 2002 and 2014. I find that, similar to coverage of judicial opinions, members of the press are largely drawn to conflict and spectacle. Specifically, journalists are more likely to cover expensive, churlish, and competitive campaigns for the bench. Nevertheless, I also find that the predicted rate of candidate coverage has fallen dramatically between 2002 and 2014 (approximately 62 percent).

Keywords:

• State supreme courts
• judicial elections
• local media coverage

Notes

1 Unfortunately for Roy Moore, the room soon turned hostile, engaging in a biting series of questions about his philosophy regarding the separation of church and state (Reeves 2000).

2 Storyboards from this campaign are available from the following address (last accessed 16 December 2019). https://www.brennancenter.org/sites/default/files/analysis/Buying_Time/STSUPCT_TN_TENNESSEEFORUM_MOST_LIBERAL_15.pdf

3 This storyboard is available from the following address (last accessed 16 April 2021). https://www.brennancenter.org/sites/default/files/analysis/Buying_Time/STSUPCT_TN_WADE_TRUTH_TEST.pdf

4 Data are available from the following address (last accessed 16 April 2021). https://www.brennancenter.org/sites/default/files/analysis/Buying_Time/TV%20Coding_FINAL_web.xls

5 The storyboard for this advertisement is available from the following address (last accessed 16 April 2021). https://www.brennancenter.org/sites/default/files/analysis/Buying_Time/STSUPCT_GA_HUNSTEIN_PROFILE.pdf

6 This storyboard is available from the following address (last accessed 16 April 2021). https://www.brennancenter.org/sites/default/files/analysis/Buying_Time/KY%20Defeat%20storyboard.pdf

7 This storyboard is available from the following address (last accessed 16 April 2021). https://www.brennancenter.org/sites/default/files/legacy/Buying%20Time%202010/STSUPCT_AL_PARSONS_MEET_TOM.pdf

8 See, however, work from Vining and Wilhelm (2010), which finds no clear pattern of media reporting across heterogeneous institutional designs.

9 This is especially true among third-party candidates.

10 Below, I further differentiate some the different types of judicial elections that occur absent the party label (nonpartisan, hybrid, and retention elections).

11 A similar logic holds for elections to legislative positions such as the U.S. Senate. For example, Kahn (1991) finds that incumbents in uncompetitive races receive significantly greater newspaper coverage compared to challengers.

12 Pew data available from the following address (last accessed 16 April 2021). https://www.journalism.org/fact-sheet/newspapers/

13 Data are available at the following address (last accessed 16 April 2021). https://www.pewresearch.org/fact-tank/2019/07/09/u-s-newsroomemployment-has-dropped-by-a-quarter-since-2008/

14 I do not include in this analysis judges on either the Oklahoma or Texas criminal court of appeals. Nor do I include individuals who run uncontested for partisan, nonpartisan, or hybrid positions.

15 Data are from the Pew Research Center at the following address (last accessed 16 April 2021). https://www.pewresearch.org/fact-tank/2019/07/09/u-s-newsroomemployment-has-dropped-by-a-quarter-since-2008/

16 Consider, for example, the case of the Los Angeles Times (LAT). In 2000, the Chicago-based Tribune Company acquired the LAT, beginning a saga of unrest at the news organization. In 2007, the company sold itself to Sam Zell, who promptly saddled the organization with $13 billion of debt. Within a year, Tribune Company dismissed scores of reporters and filed for Chapter 11 bankruptcy as it sought to protect itself from the cratering American economy. After emerging from bankruptcy and consolidating its publishing business, Tribune rebranded itself Tronc in what would become a series of public relations disasters. In 2017, the LAT fired its senior editors. In 2018, one of its senior officials was accused of sexual harassment and anti-Semitism. That year, Tronc unsuccessfully opposed LAT employee efforts to unionize. It finally sold LAT to a private owner and changed the company name back to Tribune Publishing. The Chicago Tribune provides a detailed chronology of the unrest at Tribune Publishing (last accessed 16 April 2021). https://www.chicagotribune.com/business/ct-viz-chicago-tribunepublishing-ownership-saga-htmlstory.html

17 Hughes (2020) finds that news reporting in the 12 months leading up to an election is highly correlated with coverage in the three and six month periods before an election, but the 12 month period offers greater variation.

18 More specifically, I entered into LexisNexis’ search field, “Candidate Name,” followed by the Boolean operator “AND,” followed by “State Name.”

19 Warren McGraw, a Democratic incumbent on the West Virginia Supreme Court, faced Republican Brent Benjamin in a highly contested, bitter campaign, which Benjamin ultimately won.

20 Candidates in partisan elections average 16.40 stories before an election. Candidates in nonpartisan elections average 17.56 stories. And candidates in hybrid elections average 17.05 stories before an election.

21 This difference of means is statistically significant ($t=5.57$).

22 Elections in odd-numbered years are grouped with the even-numbered year following them.

23 Baum and Klein (2007) provide an in-depth analysis of this race. https://www.followthemoney.org/

24 Campaign finance data are available from the National Institute on Money in State Politics at the following address (last accessed 16 April 2021). Television advertising data are available from the Brennan Center for Justice’s annual series on judicial elections, Buying Time, available at the following address (last accessed 16 April 2021). https://www.brennancenter.org/our-work/research-reports/judicial-electiontelevision-advertising-data-all-cycles

25 For a more detailed account of the Wisconsin Supreme Court’s scandals, see Caplan (2015).

26 For a lengthier examination of the politics of the Alabama Supreme Court, see Hughes (2018).

27 Bonneau (2017) argues that campaign fundraising and spending in state supreme court elections are nearly indistinguishable.

28 Advertising data and coding for their tone come from the Brennan Center for Justice’s series on state high courts, Buying Time. Data are available here (last accessed 16 April 2021). http://t.ly/edLnE

29 It is worth noting that the dependent variable records the total number of news stories covering state supreme court candidates in the 12 months leading up to their elections. But the campaign variables outlined in the section also occur during the year in which stories are recorded. Therefore, there may be an endogeneity concern to the extent that campaign factors like television advertisements could theoretically occur after a news story has already appeared. If the dependent variable precedes observations in the independent variables, then we cannot have a causal association. In the supplementary appendix, I provide robustness checks to the results presented in Table 2 and conclude that there is little concern for issues like reverse causality.

30 Partisan elections are like those for Congress or the presidency—candidates run under a political party’s banner and face one another in a competitive general election. Nonpartisan elections are similar insofar as they are competitive, but party labels are removed from the ballot. Hybrid elections blend features of partisan and nonpartisan elections. Political parties select nominees for the bench, but party labels are removed from the ballot. Retention elections are generally associated with states that use merit selection for their judiciaries. Under this scheme, elites appoint judges to the bench for a probationary term. After that term is up, incumbents face a retention election, allowing voters to decide to either return the judge to office or have elites appoint a replacement. Retention elections are nonpartisan and uncompetitive. Voters can merely vote to “retain” or “not retain.” They cannot vote for challengers or pick a judge’s replacement.

31 These states are Arkansas, Georgia, Idaho, Tennessee, and Wisconsin.

32 The data are available from the following internet address (last visited 16 April 2021). https://www.bls.gov/oes/tables.htm

33 I gather population figures from the Statistical Abstract of the United States for each year prior to 2010 (data available from the following internet address (https://www.census.gov/library/publications/time-series/statistical_abstracts.html), last accessed 16 April 2021) and from the U.S. Census Bureau for years 2010 to 2014 (data available form the following internet address (https://www.census.gov/data/tables/time-series/demo/popest/2010s-state-total.html), last accessed 16 April 2021).

34 Specifically, the Poisson regression model assumes a dispersion parameter, α, equal to zero. Testing, however, indicates a statistically significant dispersion parameter. Because we can reject the null hypothesis that $\alpha =0,$ a Poisson regression model is inappropriate.