Americans, Congress, and Democratic Responsiveness: Public Evaluations of Congress and Electoral Consequences

5. Voting the Bums Out

[ 97 CHAPTER 5 Voting the Bums Out One of the most famous puzzles in the study of Congress is that while a majority of Americans frequently disapprove of Congress as a whole, Americans by and large continue to vote to keep their individual member of Congress in office. According to the time series data we compiled in chapter 4, public approval of Congress’s job performance averaged only 37 percent from 1974 through 2006, but during the same period, an average of two-thirds of Americans cast ballots to reelect their own House member.1 This seeming disconnect between attitudes toward legislator and legislature is so well known that is has acquired its own label, Fenno’s Paradox—so named for the classic exposition of the trend by political scientist Richard Fenno in the 1970s (Fenno 1975, 1978). This phenomenon, along with political science’s conventional wisdom that Americans do not know or think much about Congress, has led scholars to assume that Americans do not consider the performance of the collective Congress when casting their votes in congressional elections (Fenno 1978; Mayhew 1974; Stokes and Miller 1962). As a result, scholarly explanations of how Americans cast their votes in congressional elections normally do not consider views of Congress as a relevant factor (e.g., Jacobson 2004). We nevertheless believe that there are sufficient grounds to revisit the standard academic view that members of Congress are electorally insulated from performance evaluations of the institution. First, Americans do have meaningful opinions about the collective performance of Congress, as we have demonstrated in previous chapters. We know these opinions are not “nonattitudes” (Converse 1964), because they are empirically correlated with logically relevant events, circumstances, and other opinions.2 Second, a relationship between approval of Congress and congressional elections is not ruled out merely because a majority of Americans usually 98 ] americans, congress, and democratic responsiveness disapprove of Congress while most incumbents are easily reelected. The true test is whether variation in these evaluations of job performance can explain variations in voting in congressional elections. Certainly, both congressional approval and congressional voting exhibit significant variation, despite the aggregate trends. We have seen that within the aggregate figure of congressional disapproval, precisely which Americans approve and which disapprove changes according to various circumstances, including the policy direction of the majority party. Even in the aggregate, there are some elections during which— contrary to the usual pattern—more Americans approve of Congress than disapprove, including the elections of 1998, 2000, and 2002. Additionally, there is considerable variation in voting in congressional elections. While two-thirds of voters cast ballots for their House incumbent in the seventeen elections from 1974 through 2006, this means that one-third of voters cast ballots against their incumbent and for the challenger . Aggregate vote returns for incumbents ranged between 35 percent in some districts to 100 percent in others. Furthermore, 11 percent of House races in this time span had no incumbent running. In these elections, 42 percent of voters cast ballots against the candidate from the departing incumbent’s party, and 58 percent cast ballots for the departing incumbent’s party. Aggregate vote returns for these seats ranged from 16 to 96 percent for the departing incumbent’s party.3 It seems possible that differences in evaluations of Congress could help explain these differences in voting. retrospective voting theory and congressional evaluations V. O. Key was among the first political scientists to articulate a theory of how Americans’ perceptions of past government performance can affect their vote choice on election day. Key writes, Governments must act or not act, and action or inaction may convert supporters into opponents or opponents into supporters. Events, over which government may, or more likely may not, have control, shape the attitudes of voters to the advantage or the disadvantage of the party in power. (1966, 9) In other words, voters observe government actions and events in the period preceding the election, use these observations to evaluate the per- Voting the Bums Out [ 99 formance of government, consider which party is in power, then use the evaluations as a guide in determining that party’s relative suitability for governing. If voters’ observations lead them to feel more positively than negatively about a party’s ability to govern, they will be more likely to vote for candidates from that party. If voters’ assessments are more negative , they will be less...