Friday, March 17, 2006

Book Review: Richard Fenno's "Home Style"


Although many political scientists wear their methodological badge on their sleeve, others are less obvious and more eclectic in their approach. While the former might be said of scholars like Anthony Downs or Kenneth Shepsle, the latter could be said of Richard F. Fenno. Fenno’s, Home Style: House Members in Their Districts, combines elements of thick description, behavioralism and rational choice to present a compelling account of how elected congressmen perceive their constituencies. By practicing a method he calls “soaking and poking,” Fenno influenced methodologists of all persuasions, and the book became a classic in congressional studies.[1] John R. Hibbing remarked that while he, “would not want all social scientists to adopt Richard Fenno’s [non-scientific] techniques,” he also encouraged social scientists to occasionally “dip their toe” in the pool of qualitative observation (Hibbing as cited in Fenno xii). The purpose of this paper is to examine how Fenno’s study of Congress would have been different had he not “dip[ed] his toe” in his study of congressional perception. I suggest that while Fenno’s study is not without methodological concerns, Home Style is a phenomenal example of good political science, and the rich details garnered by the “soak and poke” method would be lost by taking a methodologically purist stance.

Fenno’s Technique: Thick Descriptive, Behavioralist, or Rational Choice?

Fenno’s goal in writing Home Style, was to determine how congressmen perceived the political nature of their constituencies (Fenno 8). To try and understand these perceptions, Fenno accompanied congressional members up for reelection in their home districts, not Washington, and observed how members cultivated what he dubbed as their “home style” (Fenno 31-33). Members proceed, so says Fenno, to allocate resources, present themselves and explain their actions in Washington, to constituents. Moreover, Fenno suggested that there were four different constituencies congressmen perceived and had to contend with: a) geographic, or the territorial boundaries of the constituency; b) reelection, or those constituents perceived to vote for the congressmen; c) primary, or the core or “nucleus” voting block; and d) personal, or those constituents personally close to the congressmen (Fenno 1-18). Perceiving the differing constituencies as such enabled members to act accordingly in order to gain their vote. For example, relations with a personal constituent may be entirely apolitical because the vote is guaranteed, whereas relations with a reelection constituent whose vote is possibly up in the air, take on a different tone.

On the surface, Fenno’s research seems nothing more than what political scientists call “thick description.” Often criticized for its lack of theoretical rigor and merely journalistic accounts, thick description posits that it is the “details” in politics that are most decisive. Certainly Fenno’s method of “participant observation,” is rich in detail. As he noted, political scientists cannot know all the facts by reading election statistics and they must exhaust other forms of analysis — namely qualitative, perceptual types — to really understand the nitty-gritty of politics (Fenno 18). But Fenno’s Home Style is not just a collection of facts and details about congressmen. Although it does provide a more qualitatively-savvy approach to studying Congress, it also seeks to find “perceptual and behavioral patterns,” about politicians in their districts (Fenno 282). This is certainly more than just a journalistic account of life on the campaign trail.

Fenno’s search for “behavioral patterns” is akin to much of the behavioral focus in political science. Seeking to find “the empirical aspects of political life,” in the methods of “modern empirical science,” behavioralists dominated the debate for much of the 1950s, 1960s and 1970s (Dahl 767). Furthermore, the behavioral mood shifted the unit of analysis off large institutions and onto the individual (Dahl 766). These discoveries alone weighed greatly on Fenno’s study. The insights Fenno provides regarding how individual congressmen viewed their districts were invaluable. He also used scientific techniques to prove or disprove common assumptions political scientists’ had about congressional elections. One such hunch, those members who were in office for longer periods of time visited less often, proved true. But as he quickly pointed out, the correlation was “not as strong or consistent” as once imagined (Fenno 36-37). Still yet, to argue that Fenno was a behavioralist would be great folly. Ever critical of research that preferred “analytical range,” to “analytical depth,” Fenno stood at odds with some behavioral tendencies. He also knew that Home Style bucked the behavioral bias towards research “…amenable to statistical analysis” (Fenno 255).

The logical conclusion most political scientists have conceded about Fenno’s Home Style, and perhaps Fenno himself, is that the book situates itself inside the rational choice debate. Although a Harvard graduate, Fenno’s prominence at Rochester, the birthplace of rational choice, has been second only to William Riker. In some ways linking Fenno to rational choice is easy. His most general claim — that politicians singular goal is reelection — is the basic rational choice assumption (See Downs, 1961). John Aldrich and Kenneth Shepsle have argued that while Fenno suggests politicians might have other objectives, like actually shaping good policy or seeking power on Captiol Hill.[2] To quote Aldrich and Shepsle: “Fenno [suggests] legislators…really care about policy issues quite apart from their value from reelection” (Aldrich and Shepsle 9-10). Neverthless, Aldrich and Shepsle have acknowledged that Fenno’s underlying assumptions about reelection and perception ring overwhelmingly rational choice. Some students of political science have had other problems pinning Fenno down as a rational choice, namely his minimal use of models. If rational choice is about building theoretical models per Downs and Riker, how does Fenno fit-in? This is an important claim, but rational choice scholar Morris Fiorina notes that not all rational choice is about building mathematical models. Rather, rational choice methodology constructs “self-consciously theoretical” research programs of all types (Fiorina 4).

Different Methods: How Fenno Would Have Changed

Discerning how Fenno fits into the methodological order is a question of certain disagreement. What cannot be argued, however, is the profound impact Fenno’s little book had on a generation of political scientists. How would the book have been different if multiple-methods of description, observation, individualism and rationality were not imported? Simply put: the book would not have been nearly as profound. What I find intellectually satisfying about the “soak and poke” method is that it exhausted and tested many suppositions and claims political scientists had — more specifically congressional scholars — that could not be expressed in opinion polls or roll-call votes. By the time Fenno wrote Home Style, hundreds of polls asked how constituents viewed their respective legislator, but no poll could truly assess how legislators perceived their constituents. Secondly, Downs’ assertion that members’ ultimate goal was reelection proved generally true. But the degree to which congressmen, even those with safe seats, frantically concerned themselves with this end was an important observation. This type of research could only be done by “peeking over the shoulder” of members in their districts.

But Fenno’s “soak and poke” had its drawbacks. Fenno’s own admission, that the work was costly and hard to research, not to mention tiring, does not bode well for the assistant professor trying to balance teaching and earn tenure (Fenno 255, 283). There were other pertinent methodological issues, though. First, participant observation is hard to replicate. Only Fenno knows what he saw and observed, and any attempt to arrive at an interpretation of events different from his own would be difficult. A more statistically driven research agenda would produce conclusions amenable to interpretation. Second, participant observation is helpful in studying Congress, but this technique would be limited, if not impossible, in studies of other branches or institutions. I am pretty certain it would be impossible to “soak and poke” around with Supreme Court justices or CIA operatives. For this, political scientists will have to utilize other deductive techniques to find answers. Third, and perhaps the most damning critique, it is not clear that participant observation helped Fenno answer his question about constituency perception any better than giving members surveys to fill out about how they view the electorate. Moreover, why does that question really matter? As some have pointed out, Fenno merely reinforces Downsian assumptions about reelection and adds nothing more.

[1] The Richard F. Fenno Prize is given annually to the best book written about Congress.

[2] More on this point is found in Fenno’s (1973) Congressmen in Committees.