This report summarizes the results of our survey conducted over the past year as part of the Candidate Emergence Study. It has taken us somewhat longer to prepare the data for analysis than we anticipated, and we apologize for the delay in getting this report to you. In this report, we provide a brief summary of the design of the study, followed by a summary of some of our results thus far, and a discussion of some of the implications of our findings. We are posting this report on our web site ( Those who are interested can consult the web site for additional information about the study as it becomes available.

The survey to which you responded is part of a project that seeks to understand how individuals make decisions about running for public office. Our primary focus is on the factors that influence potential candidates in their decision making about whether to run for the U.S. House of Representatives.

Study Design

The study was carried out in a random sample of 200 congressional districts throughout the nation. Because the sample was randomly selected, the congressional districts were representative of the nation as a whole in their demographic makeup, the partisanship of the Member of Congress, and a host of other characteristics.

The study involved two surveys. In the summer of 1997, we began with the Informant Survey, which we sent to people in the 200 sample districts whom we felt had good information about the politics of their districts. These included delegates to the two parties’ national conventions, county chairs, and academics known as experts in American politics. In the first two categories of district informants, we attempted to survey an equal number of Democrats and Republicans in each district. We asked informant-respondents to provide information about the incumbent representative in their congressional district, and to identify potentially strong candidates for the U.S. House in their district.

The second survey, the Potential Candidate Survey, was sent to all potential candidates identified by informants in the first survey for whom we had valid addresses, and to all state representatives and state senators whose districts overlapped with our sample congressional districts, whether or not they had been named by the informants. We included state legislators in the Potential Candidate Survey even if they had not been mentioned as potential candidates because that office traditionally is one from which many U.S. House candidates emerge, though this pattern varies substantially from state to state. In the Potential Candidate Survey, we asked questions about the decisions individuals were making about whether or not to run for the U.S. House, either in the 1998 election or in the foreseeable future. The Potential Candidate Survey was sent out in three waves, more or less timed according to each state’s filing deadlines and legislative sessions. Our response rates exceeded 40% for the informant stage of the study, and were about 33% for the potential candidate surveys.

Results from the Informant Survey

Most of our informants were chosen because they were party activists, so we were not surprised to see they were strongly affiliated with their political party. About 54% identified themselves as Democrats, and 46% identified as Republicans; fully 87% said they "strongly" identified with the party of their choice. Fifty two percent of the informants who responded were affiliated with the same party as the incumbent Representative from their district, and 48% were from the opposite party. Over half said they had been extremely active in their party’s campaigns. Informants were also diverse in their ideological inclinations: About 38% identified themselves as liberals, 47% percent were conservatives, and 14% were moderate in their political philosophy.

In addition to asking informants to identify potential candidates for the U.S. House of Representatives, we posed a series of questions about the incumbent in the informant’s district. We asked informants to rate the potential candidates they named on similar or identical characteristics. The data in Table 1 are for informants who are of the same party as the incumbent or potential candidate in order to control for the effects of partisan bias in the ratings. Strategic resources are characteristics that relate directly to an individual’s chances of getting elected in the district. On these characteristics, informants uniformly saw potential candidates as disadvantaged relative to incumbents. Incumbents have a substantial head start in potential support from voters in their own party, visibility, and fund raising that makes them difficult to unseat.

Table 1
Comparisons of Incumbent and Potential Candidate Ratings on Selected Characteristics (percent giving "Extremely Strong" or "Strong" rating)
Incumbent Rating by Informant in Same Party Potential Candidate Rating by Informant in Same Party
Strategic Resources:
Potential support from voters in own party 86 64
Name recognition in district 90 47
Ability to raise money 84 53
Personal Qualifications
Personal Integrity 84 82
Ability to solve problems 71 71
Ability to work well withother leaders 86 71
Dedication to serving public 78 84
Grasp of the issues 83 72
Number of cases 748 1708

In contrast to the strategic ratings, informants saw incumbents and potential candidates as much closer to being equal in their qualifications for holding office. Informant ratings of potential candidates on integrity and ability to solve problems were virtually identical to the comparable ratings of incumbents. Incumbents had a slight edge in their ability to work with other leaders and in their grasp of the issues, but potential candidates were actually judged as more dedicated to serving the public. Certainly the advantage enjoyed by incumbents on strategic resources was nowhere near as great on the characteristics that describe an individual’s abilities as a legislator.

Informant perceptions of incumbents’ chances of winning the House seat in their district were much higher than potential candidates’ chances of winning. On average, informants saw the chances of an incumbent winning as better than three times greater than the chances a potential candidate would win, assuming both decided to run. In other words, incumbents’ advantages in such strategic characteristics as name recognition and the ability to raise money translate readily into election victory in our informants’ minds. Characteristics such as integrity and ability to solve problems, on which potential candidates match incumbents in their qualifications for holding office, do not make up for the strategic advantages incumbents have over potential candidates.

The Potential Candidate Surveys

Table 2 presents data on the political and demographic backgrounds of our two samples of potential candidates: those identified as strong potential candidates by district informants, and state legislators whose constituencies overlap with the U.S. House districts in our sample. About 56% of the potential candidates named by informants hold an elective office, and 22% are state legislators. Of course, this means that there is some overlap between the two samples. Some of the potential candidates identified by informants were state legislators, and some of the state legislators surveyed because their districts fall into the sample U.S. House districts were named by informants as strong potential candidates.

Table 2
Selected Characteristics of the Named and State Legislator Potential Candidate Samples (in percent)
Potential Candidate Named by an Informant Potential Candidate Serving as a State Legislator
Political background
Identify with Democratic Party 59% 54%
Identify with the Republican Party 41 46
Identify as Liberal 30 27
Identify as Moderate 19 18
Identify as Conservative 51 55
Active in Presidential, US House, and Senate campaigns 54 48
Demographic Background: 29.8
Lived in Community More Than 20 Years 66 77
Male 77 78
White 92 93
Graduate or Professional Degree 63 47
50 Years or Older 47 59
Family Income $90,000 or Higher 54 38
Number of Cases 403 804

The data in Table 2 indicate that both samples of potential candidates were politically diverse, and quite similar to one another in their political and demographic backgrounds. The state legislator sample was slightly less Democratic in its party identification, somewhat more conservative in ideology, and moderately less involved in U.S. House, Senate, and presidential campaigns than the named potential candidate respondents. State legislators were also more likely to have been long-time residents in their communities, they were older on average, and somewhat less likely to have pursued a graduate or professional degree. Their family incomes were also slightly lower than named potential candidates. The profile presented in both samples suggests that potential U.S. House candidates are active in public affairs, relatively well educated and well off financially, and rooted in their communities. Men outnumber women three to one, but the percentage of female potential candidates still substantially exceeds the proportion of women who have seats in the U.S. House of Representatives.

We asked potential candidate respondents to give their best estimate of the chances that they would run for the U.S. House in the 1998 elections and sometime in the foreseeable future. Table 3 shows that the average chance of eventually running for the House among named potential candidates was almost fifty-fifty, which indicates a very high level of interest in seeking a congressional seat. Not surprisingly, potential candidates who were named by informants as strong possible contenders were more likely to consider running than state legislators, whose average chances of running sometime in the future were about one in three. The chances potential candidates gave of running in 1998 were much lower than their chances of running in the future. The lower chances of running in 1998 reflected the fact that, for most potential candidates in any given year, personal or political conditions are not conducive to running, even among those who are contemplating running for the House at some point in their political career.

Table 3
Potential Candidate's Chances of Running for the US House of Representatives
Potential Candidate Named by Informant Potential Candidate Serving as State Legislator
Chance of Running in the Forseeable Future .49 .32
Chance of Running in the Forseeable Future in 1998 .12 .07
Number of Cases 379 788

Our analysis indicates that potential candidates are much more likely to consider running when they think their chances of winning their party’s nomination and the general election are high. If they see their chances of winning as low, they tend to wait for another opportunity. One factor that bears directly on potential candidates’ chances is what the incumbent is likely to do. We find, for example, that potential candidates in seats that are expected to be open have a much higher tendency to run than if the incumbent is likely to run for reelection. In all districts, potential candidates in the party opposite the incumbent’s are more likely to run, in part because they believe their chances of winning their party’s nomination are relatively good, compared with potential candidates in the same party as the incumbent.

We believe there is much more to explaining why potential candidates consider running for a House seat than their chances of winning, as important as such strategic considerations are in their thinking. We are particularly interested in what makes the House attractive or unattractive to potential candidates as they think about their career options in the years ahead. For example, potential candidates may anticipate that running will disrupt their families and careers. Table 4 presents potential candidates’ reactions to a variety of possible deterrents to running for the House in the years ahead. Because responses do not differ significantly between named potential candidates and state legislators, we present only the named potential candidates.

Table 4
Deterrents to Running for the US House among Names Potential Candidates
Percent Strongly Discouraged
Loss of personal and family privacy 15%
Negative impact on political career if lose 2
Loss of leisure time 5
Having to endurie negative advertising in the campaign 12
Separation from family and friends 19
Having to raise large sums of money to fund campaign 34
Loss of personal/family income 9
Lack of support from political party 17
The chance that I might serve in the minority party 1
Having to give up my political career 8
Number of Cases 397

By far, the largest negative factor in potential candidates’ thinking about running for the House is having to raise money to fund their campaign. About a third of potential candidates indicate they are strongly discouraged by that prospect. This is also the only factor where there is a significant difference between the parties, with Democrats more discouraged at the prospect of having to raise large amounts of money than Republicans. Other factors that appear to deter potential candidates from thinking about running include loss of privacy, the expectation that they will have to endure negative advertising from their opponent, separation from family and friends, and the lack of support from their political party. Some of these deterrents to running for the House are intrinsic to running for an office that requires individuals to leave their home town and live in Washington DC for most of the year. But other problems, such as the need to raise large amounts of money and the prospect of facing negative advertising, might be addressed by the political parties or through public policy. To the extent that our research finds that significant numbers of well-qualified potential candidates are deterred by factors such as these, the current political climate may need to be changed to create a less formidable set of barriers to qualified potential candidates.

One of our interests is in comparing how attractive the U.S. House is compared with other institutions for pursuing one’s political career. Table 5 shows that the House was the most attractive institution to named potential candidates, which is not surprising since they were specifically identified by knowledgeable informants as strong potential House candidates. State legislators in both political parties were also quite strongly attracted to a career in the House of Representatives, although they were apparently quite satisfied with their current position. It is clear from these results that there is no shortage of interest among either named potential candidates or state legislators in a congressional career, despite the fact that prospective candidates appreciate the difficulties associated with running.

Table 5
Percent Highly Attracted to Holding Office in Selected Institutions, by Political Party
Democratic Named Potential Candidates Republican Named Potential Candidates Democratic State Legislators Republican State Legislators
US House of Representatives 61 56 59 67
US Senate 4.68 4.85 4.62
Governor 43 41 35 32
State Legislature 49 52 88 83
State Judicial System 15 10 13 7
Local office 37 30 20 15
Number of Cases 218 160 424 372


Our analysis in this report barely scratches the surface of questions that we will be investigating in this project. Already, however, the results point to several conclusions of interest. Informants and potential candidates themselves appreciate the strategic advantages that incumbents enjoy in House elections. These advantages deterred some potential candidates from running in 1998. The 1998 elections saw the largest number of uncontested seats for the U.S. House of Representatives since World War II. This lack of competition for seats is cause for concern because representative democracy depends fundamentally on a competitive electoral process.

As much as incumbents’ advantages limit the options of potential candidates in the short run, they are less important in potential candidates’ thinking about running in the longer term. Our survey uncovered a large number of highly qualified potential candidates whose interest in pursuing a seat in Congress is genuine. Moreover, the evidence provided by our informants indicates that most of these potential candidates in both political parties are extraordinarily well qualified to serve in the national legislature.

While potential candidates are attracted to a career in Congress, and are willing to wait until the time is right to run, they do realize that there are significant costs associated with mounting a House campaign. They are especially likely to pause because of the financial costs and the need to raise large amounts of money to run a successful campaign. Potential candidates are also influenced against running by the negative consequences of being a candidate on their privacy, their families, and their personal lives. Some of these problems are inherent to running for a highly visible office in our society, but some are also unfortunate aspects of our contemporary political culture.

We are extremely grateful to all of the participants in our study. We believe the Candidate Emergence Study can contribute significantly to our academic understanding of factors that bear directly on the quality of American democracy. This understanding would not have been possible without the time taken by our respondents to respond to the survey. As we continue our analysis of the data, additional reports will be available on our web site, and in various publications that we will produce from the project. We hope that one result of our research will be improved understanding of this under-studied aspect of American representative democracy.