, 19 27 Research Article doi: 10.1111/j.1467-9256.2012.01451.x National News Attention to the 106th Senate Brian J. Fogarty University of Missouri St Louis Scholars have established that the national news media focus their attention on important actors in Congress such as party leaders, committee chairs and influential senators. However, researchers have yet to consider whether the median voter and filibuster pivot salient actors in the legislative process receive differential coverage by the news media. Examining the 106th Senate, I demonstrate that having a positive probability of being the median voter in the chamber affects the attention garnered from the national press. However, there is no significant difference observed for filibuster pivots. Keywords: US Senate; US news media; median voter; filibuster Introduction and background The importance of understanding how the news media cover Congress and its members has been strongly noted and argued by scholars over the past decades (Kuklinski and Sigelman, 1992; Sellers and Schaffner, 2007; Squire, 1988). Foremost, understanding how the media cover members allows researchers to understand what information voters are exposed to, which leads to a better understanding of the nature of congressional elections and electoral representation (Hibbing and Theiss-Moore, 1998; Kahn and Kenney, 1999). The study of the media s involvement in congressional elections has been extensively examined in the past (Goldenberg and Traugott, 1987; Kahn, 1991; Kahn and Goldenberg, 1991; Kahn and Kenney, 1999). However, scholarly attention to how the national press covers senators within Congress has mostly stalled over the past decade. This is for several reasons, including an apparent stability in coverage patterns, the newsworthiness of Senate elections and a shift towards understanding news coverage of members of Congress on the local level. Along with providing strong support for the traditional explanations of press coverage of the Senate, I argue in this research note that scholars must re-investigate national news attention to the Senate by taking into account other, previously ignored, important actors in the legislative process. Providing a broader accounting of national media attention to the Senate is important for understanding how the institution and its members are presented to the public. Using ideological placements and location probabilities in the 106th Senate from Joshua Clinton, Simon Jackman and Douglas Rivers (2004), I demonstrate that
20 BRIAN J. FOGARTY senators with a positive probability of being the median voter in the chamber receive different news attention from other members in the New York Times and Washington Post. However, having a positive probability of being the filibuster pivot has no effect on coverage. 1 The direction of coverage for the median voter is conditioned on whether or not it is an election year. Median members received significantly more coverage during 1999 when legislative activity in the Senate was high; 127 public laws passed the Senate that year. During 2000, a congressional and presidential election year and a year with light legislative activity, only 48 public laws passed the Senate, and median members actually received significantly less coverage than other senators. Since Clinton, Jackman and Rivers (2004) only provide estimates for the median and filibuster pivots in the 106th Senate, the scope of this article s analysis is constrained. However, it is to be hoped that this article can serve as a stepping stone for a broader consideration of the median voter and filibuster pivot in the fabric of national congressional news coverage. The article proceeds with brief discussions on the median voter and filibuster pivot in political science research, scholarship on national news attention to the Senate and the journalistic incentives for covering certain senators. This is followed by a look at the data and the presentation and discussion of national news attention to the 106th Senate. Median voter and filibuster pivot overview Beginning with the pioneering work of Anthony Downs (1957) and Duncan Black (1958), the median voter theorem (MVT) has played an important role in theories of elections and political institutions. Considering a uni-dimensional rank-ordering of voters along political ideology, the MVT argues that parties and candidates will converge to the median voter in the electorate to win elections (Downs, 1957; Grofman, 2004). Likewise, in a majoritarian legislature, policy will move to the chamber s median voter s preference in order to secure passage of legislation (Gilligan and Krehbiel, 1987 and 1990; Krehbiel, 1998). Specifically, proponents of new legislation will push their bill s location closer to the median s preference than the status quo or an alternative piece of legislation. Since the US Senate often operates as a supermajoritarian institution, due to the filibuster, some have argued that the filibuster pivot, the 60th senator, is more important than the median voter in the ideological composition of legislation (Groseclose and Snyder, 1996; Koger, 2010; Krehbiel, 1998). Similar to the MVT, scholars contend that legislation will move to the filibuster pivot s preference, thus assuring no filibuster and the passage of a bill. Political science research has long detailed the importance of the median voter in the legislative process both theoretically (e.g. Cox, 2000 and 2001; Gilligan and Krehbiel, 1987 and 1990; Krehbiel, 1998; Snyder and Ting, 2005) and empirically (e.g. Grofman et al., 2001; Krehbiel, 1996; Schickler, 2000; Schickler and Rich, 1997; Wawro and Schickler, 2004). For example, Keith Krehbiel (1996) finds that the median voter was critical in explaining Congress s 1987 smoking ban on domestic flights. Additionally, Eric Schickler (2000) shows that changes in the median voter s position in the House accounted for significant rule changes during the nineteenth and twentieth centuries. Scholars have also empirically demon-
NEWS ATTENTION TO THE 106TH SENATE 21 strated the importance of the filibuster and filibuster pivot member for policy- and decision-making in the Senate (Groseclose and Snyder, 1996; Koger, 2010; Krehbiel, 1998). For example, David Primo, Sarah Binder and Forrest Maltzman (2008) demonstrate that including the preferences of the filibuster pivot (and median voter) in the Senate significantly improves scholars understanding of confirmation patterns for appellate and trial court federal appointments. National press attention to the Senate Research beginning in the 1970s and 1980s established several norms about how the national media cover Congress. First, the Senate, with its statelier, powerful individuals, garners more attention than the House (Cook, 1986; Kuklinski and Sigelman, 1992; Robinson and Appel, 1979; Squire, 1988). The Senate is a more prestigious institution than the House and its unique rules, such as the filibuster, allow individual senators to shape policy substantially (Koger, 2010). Second, within Congress and in the Senate in particular, members with important institutional positions receive more attention than others (Squire, 1988). National journalists focus on party leaders, committee chairs and more senior senators (Cook, 1986; Hess, 1986; Kuklinski and Sigelman, 1992; Sellers and Schaffner, 2007; Squire, 1988). Not only do these senators have greater ability to reach the press, they are also important actors in the legislative process and policy outcomes (Cook, 1998). Senators are also able to garner media attention through supply-side activities such as issuing press releases, sponsoring bills and holding press conferences (Sellers and Schaffner, 2007). Finally, senators running for higher office, specifically for the presidency, receive considerably greater coverage from the national media (Kuklinski and Sigelman, 1992). This is primarily due to the horse-race nature of presidential campaign coverage (Patterson, 1980). The journalistic incentives for covering important institutional actors within the Senate are clear. Party leaders and committee chairs are key players in the legislative process and thus are inherently newsworthy for understanding policy outcomes. Through sourcing these Senate elites, journalists make their stories appear more legitimate and add heft to the discussion of likely outcomes of policy and political manoeuvring. By focusing on these members, the media are easily able to craft and develop continuing storylines through the life cycles of congresses and presidencies (Cook, 1998). Given the need for senators to communicate with the public and with other branches of the government, the costs for the media of attaining information from important Senate actors is minimal (Cook, 1998). Party leaders and prominent members routinely and consistently issue press releases, hold press conferences and make appearances on news shows to express their opinions and explanations for legislative outcomes. In considering heretofore unexamined salient actors in the legislative process of the Senate, the median voter and the filibuster pivot stand out. In spite of the median voter s and filibuster pivot s apparent importance in the Senate, political communication scholars have not examined whether and how these members are treated differently by the press. If these members play a critical role in the legislative process and policy outcomes, we should expect the national press to increase their attention to these senators compared to the average senator.
22 BRIAN J. FOGARTY Data The analysis examines the amount of coverage of all the members in the 106th Senate (1999 2000) in the New York Times and the Washington Post. 2 These two national news outlets are generally considered the standard-bearers of political news in the US and often serve as agenda setters for much of the American news industry (Fogarty, 2005; Sabato, 1991; Wu et al., 2002). Using multiple versions of the senators names as the search term in LexisNexis, and purging duplicate articles, there were a total of 11,633 stories in the Times and 13,126 stories in the Post during the time span. However, the unit of analysis was the senator. The key explanatory variables in this article are the probability of being the median voter and the filibuster pivot in the 106th Senate. This information was attained from Clinton, Jackman and Rivers s (2004) article on a Bayesian method for determining ideological positions for members of Congress. Using non-unanimous roll-call votes in the 106th Senate, Clinton, Jackman and Rivers fit a unidimensional model for uncovering senators ideal points and the proportion of votes in which each senator occupied the median or filibuster pivot positions. The benefit of their Bayesian method is that it can assess probabilities of different political actors, such as senators, holding certain ideological positions, such as the median voter in the Senate. Other non-bayesian approaches to ideal-point estimation, for example, DW-Nominate, are not equivalently equipped to make probabilistic statements about where specific legislators may be located. Since a senator is unlikely to be the median or filibuster pivot all the time, utilising probabilities is more preferable than using static placements such as those provided by DW-Nominate. Relying on figure 2 of the 2004 article, two senators had a positive probability of being the median voter (Susan Collins (R-ME) and Olympia Snowe (R-ME)) and 13 senators had a positive probability of being the filibuster pivot (Ted Stevens (R-AK) had the highest probability). I also included explanatory variables that have been shown previously to predict increased national press attention to the Senate (Sellers and Schaffner, 2007). This includes dummy variables for whether a senator was a leader, committee chair, ranking member, running for higher office, 3 up for re-election or served a state in or adjacent to the Times or Post, where 1 = Yes for all variables. The latter variable is important as national papers like the Times must also serve the informational demands of their local audience, thus potentially increasing attention to senators representing their readership. A variable for a senator s term is also included since more senior members are likely skilled at dealing with the media and therefore may receive more press coverage (Arnold, 2004; Cook, 1989). Finally, the number of bills sponsored is included as a supply-side variable for senators attracting coverage. 4 The median number of bills sponsored in the 106th Senate was 32, with a low of 1 from Lincoln Chafee (R-RI) in 1999 and Zell Miller (D-GA) in 2000, who both served partial terms, and a high of 317 from Peter Fitzgerald (R-IL) in 1999. Results Since the dependent variables are story counts and there is over-dispersion, negative binomial regression is used to analyse the data (Long, 1997). Again, the unit of analysis is the senator.
NEWS ATTENTION TO THE 106TH SENATE 23 Table 1: Volume of national coverage in the 106th Senate Variables New York Times Washington Post 1999 2000 1999 2000 Median voter.881** -.250*.650** -.820** Probability (.108) (.150) (.101) (.125) Filibuster pivot -.540-1.65* -.531 -.788 Probability (.767) (.848) (.758) (1.10) Bills sponsored.005.015**.006**.016** (.003) (.005) (.003) (.004) Term.084*.096**.054.021 (.046) (.047) (.042) (.057) Leader 1.62**.994** 1.42**.793** (.470) (.318) (.413) (.280) Committee chair.833**.256.779**.402* (.209) (.227) (.175) (.229) Comm. ranking.186 -.106.414.450 Member (.215) (.229) (.226) (.348) Higher office 1.77** 2.78** 1.40** 2.06** (.294) (.710) (.229) (.647) Same/adjacent 1.37** 2.22** 1.29** 1.46** State (.302) (.339) (.156) (.241) Election.343**.419** (.166) (.172) Constant 2.85 2.53 3.24 2.92 (.176) (.249) (.174) (.226) Log-likelihood -462.49-456.63-489.53-477.02 Alpha.499.572.416.463 # of articles 5514 6119 6842 6284 N 101 101 101 101 Notes: Cells are negative binomial regression coefficients with Huber-White robust standard errors in parentheses. *P <.10; **P <.05. Table 1 has four columns of results split between the Times and Post and 1999 and 2000. I find that the traditionally important explanatory variables are statistically significant, including whether the senator is a leader, committee chair, running for higher office or up for re-election in 2000. However, being a ranking member had no effect on coverage. Serving in the same or an adjacent state to the Times and Post also significantly increases press attention. This finding speaks to the needs of even national papers to fulfil their local readers interests. For example, the Post covered many, if not all, of Barbara Mikulski s (D-MD) and Paul Sarbanes (D-MD) votes in the Senate, giving them roughly 10 times the coverage they received in the Times.
24 BRIAN J. FOGARTY Finally, senators were able to increase press attention significantly through bill sponsorship. Most importantly, I find that having a positive probability of being the median voter had a statistically significant effect on the level of press attention from the Times and Post. However, this is not the case for senators with a positive probability of being the filibuster pivot, save for a significant effect at the.10 level in the Times in 2000. Whether being the median voter in the Senate increased or decreased national press attention depends on whether it was an election year. 5 In 1999, having a positive probability of being the median voter increased the level of attention from the Times and Post, but in 2000 it decreased the amount of attention. Two factors appear to be at play. First, 2000 was a congressional and presidential election year featuring a historic race between Al Gore and George W. Bush. It is typical that the national press focused on Senate and presidential elections during these years, but 2000 may have been even more unique (Kahn and Kenney, 1999). Second, the amount of legislative activity in the Senate was considerably less in 2000 than 1999. Five thousand one hundred and fifty-seven bills were sponsored and 127 public laws passed the Senate in 1999, while only 3,117 bills were sponsored and 48 public laws passed the Senate in 2000. Further, in 1999 there were 38 votes in the Senate that passed with a majority of four or less, while in 2000 there were only 31. If the median voter is an important actor in the legislative process, we should expect him or her to receive more attention when the level of activity in the Senate is high. However, it is interesting that the median voter received significantly less coverage in 2000 and not simply the same level of coverage as other senators. It appears that when legislative activity is low and congressional and presidential elections are present, the national press substantially shifts its attention away from median senators. In analyses of press attention to political institutions, individuals, events or issues, it is useful to get a specific sense of the expected changes in the volume of coverage. To this end, Table 2 includes the predicted average number of articles in the Table 2: Average predicted number of articles in the Washington Post in 1999 Variables Minimum value Maximum value Median voter probability 53.86 103.20 Bills sponsored 40.53 254.04 Leader 51.25 211.12 Committee chair 47.54 103.62 Higher office 53.46 215.72 Same/adjacent state 51.50 187.58 Notes: Cells are average predicted number of articles for the minimum and maximum values of the statistically significant independent variables from column 3 of Table 1. The values were calculated using SPOST9 (Long and Freese, 2006) and holding the other variables at their means.
NEWS ATTENTION TO THE 106TH SENATE 25 Washington Post in 1999 for the minimum and maximum values for the significant variables. 6 The predicted average number of articles for senators who are the median voter with a probability of 1 is 103. This value is nearly identical to the predicted average number of articles for a senator who is a committee chair. This is an important finding for understanding press coverage of the Senate. In the eyes of the national media, it appears that the median voter in the Senate is as newsworthy in the legislative process as a committee chair. It is worthwhile to note that leaders and senators running for the president are expected to receive more than double the coverage of the median voter or committee chair. Still, having a positive probability of being the median voter increases the importance of the senator in congressional coverage. Conclusion The analysis suggests that the national media treat senators with a positive probability of being the median voter differently from other senators. Specifically, median members receive significantly more attention in non-election years and significantly less attention in congressional and presidential years. This result does not supplant traditional explanations of press attention to the Senate as seen, senators who are party leaders, committee chairs and running for the presidency continue to dominate coverage. Instead, this finding offers a new institutional actor in the Senate for political communication scholars to consider when examining press attention to the Senate. Clearly, the scope of this research note is limited and more research is needed to establish the reliability of these results. The first step is to expand the analysis to more sessions of the Senate. This requires not simply more data collection, but new estimates of median voter and filibuster pivot probabilities using roll-call votes in the manner of Clinton, Jackman and Rivers (2004). This will help us answer important questions such as whether the newsworthiness of the median voter varies by majority party size in the Senate, the partisan composition of the House, and/or the presidency. Many congressional scholars consider the filibuster pivot position as more important to policymaking in the Senate than the median voter. Therefore, the finding here that the filibuster pivot does not receive more national news attention is surprising. It is possible that the 106th Senate was simply a unique case. Not only would considering more Senates help answer the importance of the filibuster pivot in the national news, but also we could assess whether the filibuster pivot has become more newsworthy as the number of filibusters has risen in recent years. Even though the national press traditionally ignores most House members, further work could determine whether being the median voter in the House attracts national attention on the level of a committee chair in the House. Or, does the median matter less in the House due to the strict partisan hierarchy? It would be fruitful also to investigate whether there are trickle-down effects to local coverage of senators and House members.
26 BRIAN J. FOGARTY About the Author Brian J. Fogarty is an Associate Professor of Political Science at the University of Missouri St Louis. His interests include political communication, legislative politics and political methodology. Brian J. Fogarty, Department of Political Science, University of Missouri St Louis, St Louis MO 63121, USA. E-mail: fogartyb@umsl.edu Acknowledgement I thank David Kimball for comments and suggestions. Notes 1 Several requests to the authors for additional information on replicating their study for other Senates went unanswered. Hence, I am only left with information on the 106th Senate. 2 This includes senators who only served partial terms in the 106th Senate, including John Chafee (R-RI) who died in 1999 and who was replaced by his son Lincoln Chafee (R-RI), and Paul Coverdell (D-GA) who died in 2000 and was replaced by Zell Miller (D-GA). 3 Orrin Hatch (R-UT) and John McCain (R-AZ) ran in the 2000 Republican presidential primary. Robert Smith (R-NH) announced that he would seek the nomination, but withdrew several months later and before the primary elections. Joseph Lieberman (D-CT) was the Democratic vice-presidential nominee in the 2000 general election. 4 Bill sponsorship data were obtained from the Library of Congress s Thomas website. Press releases were excluded from the analysis because data could only be found on 71 senators. In a separate analysis on those 71 senators, press releases did not have a statistically significant effect on coverage. 5 It is important to note that Olympia Snowe (D-ME) was also up for re-election in 2000. 6 This same procedure can be done for the other three regressions. Based on the results in Table 1, it should be clear that one gets roughly the same predicted averages in the Times in 1999. For example, the predicted number of articles in the Times in 1999 for a senator who is the median voter with probability 1 is 94. References Arnold, R.D. (2004), Congress, the Press, and Political Accountability, Princeton, NJ: Princeton University Press. Black, D. (1958), The Theory of Committees and Elections, Cambridge: Cambridge University Press. Clinton, J., Jackman, S. and Rivers, D. (2004), The Statistical Analysis of Roll-Call Data, American Political Science Review 98(2), pp. 355 370. Cook, T.E. (1986), House Members as Newsmakers: The Effects of Televising Congress, Legislative Studies Quarterly 11(2), pp. 203 226. Cook, T.E. (1989), Making Laws and Making News: Media Strategies in the US House of Representatives, Washington, DC: Brookings Institution. Cook, T.E. (1998), Governing with the News: The News Media as a Political Institution, Chicago, IL: University of Chicago Press. Cox, G.W. (2000), On the Effects of Legislative Rules, Legislative Studies Quarterly 25(2), pp. 169 192. Cox, G.W. (2001), Agenda Setting in the US House: A Majority-Party Monopoly?, Legislative Studies Quarterly 26(2), pp. 185 210. Downs, A. (1957), An Economic Theory of Democracy, New York: Harper and Row. Fogarty, B.J. (2005), Determining Economic News Coverage, International Journal of Public Opinion Research 17(2), pp. 149 172. Gilligan, T.W. and Krehbiel, K. (1987), Collective Decisionmaking and Standing Committees: An Informational Rationale for Restrictive Amendment Procedures, Journal of Law, Economics, and Organization 3(2), pp. 287 335. Gilligan, T.W. and Krehbiel, K. (1990), Organization of Informative Committees by a Rational Legislature, American Journal of Political Science 34(2), pp. 531 564.
NEWS ATTENTION TO THE 106TH SENATE 27 Goldenberg, E.N. and Traugott, M.W. (1987), Mass Media in US Congressional Elections, Legislative Studies Quarterly 12(3), pp. 317 339. Grofman, B. (2004), Downs and Two-Party Convergence, Annual Review of Political Science 7, pp. 25 46. Grofman, B., Koetzle, W., Merrill, S. and Brunell, T. (2001), Changes in the Location of the Median Voter in the US House of Representatives, 1963 1996, Public Choice 106(3 4), pp. 221 232. Groseclose, T. and Snyder, J.M. Jr (1996), Buying Supermajorities, American Political Science Review 90(2), pp. 303 315. Hess, S. (1986), The Ultimate Insiders: US Senators in the National Media, Washington, DC: Brookings Institution Press. Hibbing, J.R. and Theiss-Moore, E. (1998), The Media s Role in Public Negativity toward Congress: Distinguishing Emotional Reactions and Cognitive Evaluations, American Journal of Political Science 42(2), pp. 475 498. Kahn, K.F. (1991), Senate Elections in the News: Examining Campaign Coverage, Legislative Studies Quarterly 16(3), pp. 349 374. Kahn, K.F. and Goldenberg, E.N. (1991), Women Candidates in the News: An Examination of Gender Differences in US Senate Campaign Coverage, Public Opinion Quarterly 55(2), pp. 180 199. Kahn, K.F. and Kenney, P.J. (1999), The Spectacle of US Senate Campaigns, Princeton, NJ: Princeton University Press. Koger, G. (2010), Filibustering: A Political History of Obstruction in the House and Senate, Chicago, IL: University of Chicago Press. Krehbiel, K. (1996), Committee Power, Leadership, and the Median Voter: Evidence from the Smoking Ban, Journal of Law, Economics, and Organization 11(1), pp. 237 259. Krehbiel, K. (1998), Pivotal Politics: A Theory of US Lawmaking, Chicago, IL: University of Chicago Press. Kuklinski, J.H. and Sigelman, L. (1992), When Objectivity is Not Objective: Network Television News Coverage of US Senators and the Paradox of Objectivity, Journal of Politics 54(3), pp. 810 833. Long, J.S. (1997), Regression Models for Categorical and Limited Dependent Variables, Thousand Oaks, CA: Sage. Long, J.S. and Freese, J. (2006), Regression Models for Categorical Dependent Variables in Stata (2nd edn), College Station, TX: Stata Press. Patterson, T.E. (1980), The Mass Media Election, New York: Praeger. Primo, D.M., Binder, S.A. and Maltzman, F. (2008), Who Consents? Competing Pivots in Federal Judicial Selection, American Journal of Political Science 52(3), pp. 471 489. Robinson, M.J. and Appel, K.R. (1979), Network News Coverage of Congress, Political Science Quarterly 94(3), pp. 407 418. Sabato, L. (1991), Feeding Frenzy: How Attack Journalism has Transformed American Politics, New York: Free Press. Schickler, E. (2000), Institutional Change in the House of Representatives, 1867 1998: A Test of Partisan and Ideological Power Balance Models, American Political Science Review 94(2), pp. 269 288. Schickler, E. and Rich, A. (1997), Controlling the Floor: Parties as Coalitions in the House, American Journal of Political Science 41(4), pp. 1340 1375. Sellers, P.J. and Schaffner, B.F. (2007), Winning Coverage in the US Senate, Political Communication 24(4), pp. 377 391. Snyder, J.M. Jr and Ting, M.M. (2005), Why Roll Calls? A Model of Position-Taking in Legislative Voting and Elections, Journal of Law, Economics, and Organization 21(1), pp. 153 178. Squire, P. (1988), Who Gets National News Coverage in the US Senate?, American Politics Quarterly 16(2), pp. 139 156. Wawro, G.J. and Schickler, E. (2004), Where s the Pivot? Obstruction and Lawmaking in the Pre-cloture Senate, American Journal of Political Science 48(4), pp. 758 774. Wu, H.D., Stevenson, R.L., Chen, H. and Guner, Z.N. (2002), The Conditioned Impact of Recession News: A Time-Series Analysis of Economic Communication in the United States, 1987 1996, International Journal of Public Opinion Research 14(1), pp. 19 36.