Amateurism in the Statehouse

Document Sample
Amateurism in the Statehouse Powered By Docstoc
					Amateurism in the Statehouse: Political Amateurs and Becoming Governor from 1978-2002

John A. Hamman Department of Political Science Southern Illinois University at Carbondale Carbondale, IL 62901-4501

This paper sets out to use David Canon’s theory of political amateurism in Congress to better understand how amateurs become governor and subsequently behave in office. After questioning the stability and impenetrability of the gubernatorial career opportunity structure, and reviewing the basic model put forth in the literature to explain gubernatorial elections, the paper runs a series of equations testing the impact amateurism and candidate quality have on the percent of votes received in primary and general elections. The results show that most amateurs placing their names on primary ballots (a third of the candidates are amateurs), simply are not competitive, have no real chance of winning. They appear to be potential candidates for Canon’s hopeless amateur category. Amateurs that win primary and general elections are then examined and found to be more consistent with Canon’s policy and ambitious types. These amateurs in many ways are similar to career governors with the exception of prior experience in elected office.

Prepared for presentation at the 2005 Annual Meeting of the Southern Political Science Association, New Orleans, Louisiana, January 6-8.


Amateurism in the Statehouse: Political Amateurs and Becoming Governor from 1978-2002

Introduction According to ambition theory, opportunity structure manifests in recognizable career paths. Progressively ambitious politicians respond to this opportunity structure. The opportunity structure determines the elected offices progressively ambitious politicians seek as well as the political appointments they accept. This is evident in how presidents typically serve first as governor or senator, in how governors serve first as state legislator, statewide elected official, or congressperson. While professional politicians may follow numerous paths to higher office, they follow a few paths more frequently than the others because the structure is relatively stable. Most high-level elected officials first serve predictable apprenticeships in lower office. But what about those who win seats in Congress or the statehouse without previously serving in an elected office? By definition they are political amatuers and they work outside of this strategic framework. Instead, they run for highly coveted elected positions without prior experience in elective office. They seek to enter the system laterally. The salient candidacy for Governor of California in the fall 2003 recall election of Arnold Schwarzenegger is a case in point. He utilized his fame, family political connections, and personal wealth to defeat a crowded field of lesser-known candidates to become governor. Over the years there have been numerous celebrity actors, athletes, and astronauts like former Minnesota Governor Jesse Ventura, former U.S. Senator Bill Bradley and former Senator John

3 Glenn winning congressional and gubernatorial office. By working outside of the career structure, amateurs also forgo whatever experience “ambitious” politicians gain by first serving in lower elected office (Canon 1990). Canon addresses this problem and broadens ambition theory and the notion of opportunity structure to accommodate amateurs in Congress. His revised theory applies to the broader population of potential candidates and includes amateurs who he argues face different risks and payoffs than professional politicians. The theory relaxes assumptions like Rohde’s (1979) which hold that all officeholders are progressively ambitious in that they will always accept higher office if it can be obtained without cost or risk. Instead, he creates a typology of amateurism that distinguishes between ambitious amateurs, policy motivated amateurs, and hopeless amateurs who stand little chance of winning but somehow are swept into office by unforeseen events. The ambitious type more closely resembles the conventional career oriented professional politician. The investigation of the amateur’s place in the career structure of U.S. governors proceeds as follows. First, the frequency with which amateurs challenge for gubernatorial office and their ability to compete against “higher quality” politicians in gubernatorial primary and general elections is examined. Multivariate models test the ability of candidate quality measures to predict the share of votes candidates receive in primary and general elections. Then, the analysis weighs the extent to which the professional backgrounds and ambition of amateur governors mirror career governors. The paper concludes by considering the applicability of Canon’s theory of amateurism for understanding the place of amateurs in the gubernatorial career structure. Fitting Amateurs into the Career Structure of U.S. Governors It is not unusual to portray career structure in the United States as hierarchical and stable with the least enviable offices constituting the base of the hierarchy (Schlesinger 1966). The sequence

4 of offices that lead to the most desired posts such as governor, senator, or president then might be thought of as career “paths” or “ladders.” From this perspective, political amateurs are misfits, an anomaly that does not fit the conventional wisdom of how or why people become governor or president. Yet some accounts of career structure allow for some amateurs to win higher office. This is because some argue that career structures are better thought of as loose hierarchies. For instance, years ago Lasswell (1930) comments “in American politics the escalator to the top is not a regimented, orderly lift, but a tangle of ladders, ropes, and runways that attract people from activities at various stages of the process, and lead others to a dead end or a drop.” Over time shifts in the paths leading to the governor’s office are evident as well as the variability in the paths themselves at any given point in time. While successful careers generally advance from the least desirable to the most desirable offices, the sequences are not neat and tidy as the hierarchy or ladder terms suggest. For instance there is variation in the penultimate office for governor over time. Beyle (2004) shows that after a fairly stable pattern of penultimate office for governor over much of the 20th century, a noticeable takes place after 1981. Prior to the 1980s, about three-fifths of governors served in statewide elected office, the state legislature, or law enforcement in about equal proportions. Since then fewer serve in law enforcement and more in either statewide elected office or Congress. Experience from serving a statewide elected office, or a congressional office with a large constituency may be more important today than it once was. From 1977 to 2004, 233 governors follow 44 distinct, unique career paths. Some careers progress from local elected office, to a state legislative post, to a statewide office, and then to governor. Others take different paths. Within this tangle of career paths, most high-level elected

5 officials serve apprenticeships in lower level office prior to serving as governors, senators or presidents. They bring experience obtained through the years from winning and serving in lower level elected offices into office with them. In Table 1, the percentage of governors with experience in different elected office is given for two time periods between 1977 and 2004. About one-half of the governors served in the state legislature over this period, the most common career experience. However, considerably more (18 percent) have experience serving in state legislatures in the latter half of this time period. The next most common form of experience is in statewide elected office where about a third of the governors have experience. Considerably fewer governors have experience in administration, local elective office, or in Congress. Insert Table 1 about here Most governors serve in more than one elected office although there is also considerable variation in the number of offices they hold before being elected governor. While only about 8 percent of the governors elected from 1977 to 1990 held four or more offices before being elected governor, more than 15 percent did from 1991 to 2004. Whether these experiences amount to an apprenticeship which prepares governors to manage effectively in office has been left largely unanswered. Interesting, while there is the presumption that apprenticeships are an important condition for effective governing, little research actually substantiates this. although the sentiment against amateur politicians is widely held among scholars of state politics and administration.1 Beyle credits the celebrity status and ability of many amateurs to take advantage of the “spend your way into office” (p. 196) style of campaigning as one key to success. The frequency

A recent exception is an analysis of two Washington Post/ABC polls which showed governors with experience serving in statewide elected office and other public sector executive positions maintain higher job performance ratings as governor. See Hamman (2004).

6 of amateur governors may reflect the reality of candidate centered, buy-yourself-into-politics style elections where candidates count on less support from national and state party organization support apparatus. The next section of the paper examines the impact factors such as candidate quality and campaign spending have on vote margins in primary and general gubernatorial elections. Amateurs and Becoming Governor Ambition theory assumes professional candidates choose strategically when and for what office to run. Celebrity candidates are advantaged, in part, by media fascination with them. The attraction of celebrity name recognition also appeals to the party themselves. This was recently evident in the Republican Party’s effort to recruit Chicago Bear player and coach, Mike Ditka, to step in the U.S. Senate race for Jack Ryan. Ryan had withdrawn his candidacy following a scandal over allegations made in his divorce papers by his former wife about lewd activities at a sex club in Paris. Ditka ultimately turned down the Republicans, but the fact that they had solicited him points to just how attractive a well-known celebrity such as a successful professional football player and sports commentator can be to a party trying to come up with a competitive candidate at short notice. However, the phenomenon of amateurism encompasses more than celebrity candidates and office holders. It is not just high profile, celebrity amateurs who effectively run for higher office, an important point that may be easy to overlook. The more visible administrations of celebrity candidates like JesseVentura and Arnold Schwarzenegger are not necessarily typical of amateur governors elected over the past 25 years. Canon (1990) argues that in addition to lack of name recognition, the electoral fortunes of political amateurs are further hindered by the absence of a track record to run on, and/or their

7 lack of campaign experience. Amateur governors come from a variety of occupational and professional backgrounds. Most, though, exhibit poor candidate quality characteristics and have little success becoming governor. So while media may give a celebrity candidate for governor a considerable amount of attention, the candidate sill has to cope with the lack of a record and campaign inexperience. The other amateur candidates who are not celebrities likely have to work even harder to gain name recognition as well. The lack of amateur qualifications in relation to career politicians is reflected to varying degrees in candidate quality measures. A simple dichotomous variable to show whether or not the candidate has previously served in elected office is the most straightforward candidate quality measure. Jacobson (1990, 1992) argues that measures of whether a candidate has served in elected office tap into the most important dimension of candidate quality. Others studies use multidimensional measures that attempt to more fully account for variance in candidate quality. Bond (1997) argues that the concept conveys more about the nature of a candidate. Candidate quality is an individual’s experience and personal characteristics that contribute to running an effective campaign. Bond, Covington, and Fleisher (1985) argue that candidate quality measures should allow for the possibility that all elective offices may not provide the same visibility or develop the same kinds of organizational or fund raising skills. Also, some valuable experience might be gained from non-elective offices such as congressional staff work or positions in political parties. For instance, celebrities may bring considerable name recognition and possibly better potential for fundraising. Measures of candidate quality measures have also been used to assess candidates who run for governor. Squire’s (1992) candidate profile scale accounts for the level of a candidate’s recent elected or manifest office. U.S. senators and former governors are given a value of 6, U.S.

8 representatives a value of 5, statewide elected officials a value of 4, state legislators a value of 3, local elected officials a value of 2, and other political positions a value of 1. Squire then weights the level of the office by the office’s corresponding state constituency size. So, a U.S. senator or former governor’s weighted score is 600, a statewide-elected official’s score is 500, and so on. The scale then ranges from 0 for candidates without any experience (amateurs) to 600. Leadership positions in the state legislature are arbitrarily scored 100. The expectation is that experience in higher office obtained from serving larger constituencies will better prepare a candidate to campaign and win the governorship. Those with experience in an office serving the entire state have the highest potential for winning the election. Squire finds that candidate quality correlates positively with election results. This supports the argument that greater visibility, experience campaigning, and serving a larger constituency are important prerequisites for becoming governor. Conversely, amateurs have not held an elected office, claim a smaller percentage of the votes cast, and seldom win. Clearly the overwhelming expectation of those studying the relationship between political experience and election success is that amateurs will not do well in elections. Chances improve with greater experience in higher-level offices serving larger constituencies. The following analysis will test these expectations with primary and general election data from 1978 to 2004.2 Data The analysis uses Beyle’s gubernatorial expenditure data set, which contains over 2900 records of candidates who were on the ballot of gubernatorial primary and general elections over this time period. The data set also includes variables for each candidate’s party, the amount they reported spending in the primary and general election campaign, and the percent of the primary


The data are compiled from a number of sources including CQ Weekly Reports are available to the public at .

9 and general election vote they received. Amateurs make up a sizeable proportion of the candidates running for governor from 1978 to 2002. Of those running for governor over this time, (over 2000 are listed in the Beyle data set) 45 percent are on primary ballots and about 30 percent of those running in general elections are from private or non-profit sector occupations, by definition amateurs. Of these, just 16 percent win the primary election and less than 7 percent win in the general election.3 From a different perspective, just 17 percent of the more than 200 governors elected over this same period of time are amateurs.4 Dependent Variable The dependent variable for the analysis is the percentage of the overall vote a candidate receives in a state primary or general election in a given year. The equations test how much impact amateur status has on the vote margin received by different types of candidates in primary and general elections. Two sets of equations are estimated, two for primary and two for general elections with and without incumbent candidates. Candidates that did not receive any votes are not included in the primary election equations. Candidates without funding are excluded from the general election equations. Independent Variables Candidate Quality. The impact of candidate quality is assessed by an index of candidate quality similar to Squire’s candidate profile. The candidate quality index differs from Squire’s only in that I distinguish between lower and upper house members in state legislatures in states besides Nebraska by adding an extra level. That changes the maximum score from 600 to 700.

The percentages are likely substantially higher than this as many of the names listed on the ballot and included in the dataset are of relatively obscure candidates for which no occupational description is available. The data are compiled from a number of sources including CQ Weekly Reports are available to the public at 4 Data compiled by the author from American Almanac of Politics, the Biographical Directory of the Governors of the United States, and the National Governors Association internet site,1169,C_SEARCH_GOV,00.html.

10 This accommodates the different constituency sizes of the two legislative houses in many states. Otherwise it is the same index. Amateurs are at bottom of the scale. Campaign Spending. The amount of money that is spent by each candidate is entered into the equation. Spending is expected to correlate positively with the percentage a candidate receives. The expectation is that the more money that a candidate spends, the greater their percentage of the vote will be. The literature also finds that partisanship, evaluations of economic conditions, and the president’s job performance influence gubernatorial elections outcomes. Party Strength in the Electorate. The partisan and ideological disposition of the electorate likely will be a factor in determining the percentage of the vote candidates can expect to receive. Since candidates from the same party typically compete in primary elections, the EriksonWright-McIver (1993) measure of partisan advantage is included only in the general election equations. The variable is recoded to reflect each candidate’s partisanship. For independent and third party candidates, the score is recoded to equal zero. Controlling for other factors, the proportion of votes received in the general election should reflect the candidate’s party strength in the state’s electorate.5 Their measure for ideological composition is also entered in the equations and is similarly corrected for the candidate’s partisanship with liberal pointing to Democrats and conservative to Republicans. Economic Performance. The literature finds that economic performance impacts gubernatorial popularity and election results although there is not a consensus over whether the public considers national economic conditions, state level economic performance, or some relationship between two. A number of studies find that state economic performance has only a minimal impact on gubernatorial election outcomes (Kenney 1983; Chubb 1988; Stein 1990). Instead, the performance of the national economy seems to matter most. Voters in gubernatorial

This is available from McIver’s web page:

11 elections hold incumbent governors accountable for past national economic performance. While states may be placing higher priorities on economic development policies (Beyle 1992; Herzik and Brown 1991), and state economies are becoming more independent of national economic forces (Brace 1993; Hendrick and Garand 1991), it is equally true that state economies remain integrated in the national economy and that states lack policy making tools to fully manage the full extent of economic conditions in their state. For instance, Cohen and King (2004) observe that states are particularly deficient in their ability to influence inflation or money supply. Instead, they argue that since governors are the most visible leaders in the state (Squire and Fastnow 1994), and focus the public’s attention on economic issues (Brace 1993; Grady 1991), some concept of relative economic performance is what will affect public assessments of the governor. It is not clear what impact national and state economic condition measures will have on the percentage of vote each candidate receives. Svoboda (1995) argues that it is incumbents who to varying extents benefit or suffer economic conditions as the public rewards or punishes the incumbent seeking reelection accordingly. Alternatively in elections where no incumbent is running the impact of economic conditions is less clear. There would not seem to be any clear reason for why a challenger should either gain or lose votes unless the electorate associates them by way of their party affiliation to either the previous governor or to the president. So, in open races economic conditions would not be expected to differentially impact on the candidates. No effect is expected. Presidential Popularity. There is considerably more consensus that presidential popularity has an impact on the electoral fortunes of governors of the president’s party with gubernatorial elections serving as a referendum on the incumbent president’s policies (Chubb 1988; Simon,

12 Ostrom, and Marra 1991). A variable measuring the average popularity of president’s in the month preceding the gubernatorial election is corrected to reflect the party of the candidates. Control Variables Other election specific factors also can matter. Squire (1992) hypothesized that other things being equal women candidates will fare less well in the election. There is also a variable for the number of candidates in the election. Particularly for primary elections, but also potentially a factor in general elections, the more candidates that enter an election, the lower a particular candidate’s share of the vote likely will be. So the basic model to be estimated is: Vote Sharei,t = ∃0 + ∃1 Candidate Qualityi,t + ∃2 Spendingi,t + ∃3 Party Strengthi,t + ∃4 Ideologyi,t + ∃6 Economic Performancei,t + ∃7 Presidential Popularityi,t + ∃8 Genderi,t + ∃9 Number of Candidatesi,t + ,i,t Results In Table 2a and Table 2b regression results from three different estimations, (OLS, Random Effects, Fixed (state election) Effects) are reported for primary elections with and without an incumbent candidate.6 The dependent variable is the percentage of votes received by each candidate receiving votes in primaries held between 1978 and 2002. The model performs particularly well for races with an incumbent running. The adjusted R2 of the equation estimated with OLS for candidates in primaries with an incumbent is .75. Considerably less variance is explained by the equation for open races. The candidate quality index is statistically significant and positive in both the equations showing that candidates having served in higher office serving larger constituencies receive a larger proportion of primary votes. In addition, the variable

The results are essentially the same for OLS, Random Effects, and Fixed Effects. The effects are fixed on states in the third estimation. The models are estimated with STATA xtreg procedure. OLS results are discussed unless otherwise noted.

13 indicating that a candidate is an amateur is statistically significant for the open elections equation (i.e., about 11 percent fewer votes) but significant at the .10 level in the incumbent election equation. However, amateur variables are statistically significant and have a similar impact when estimated as a cross sectional time series model with fixed effects. Candidate spending in the primary election is only statistically significant in the open elections equation. This is the case regardless of how the equation is estimated. Relative unemployment is not statistically significant either. Candidates in the president’s party tend to receive 3 percent fewer votes. Clearly, when controlling for other predictors of primary election voting, amateurs fair very poorly relative to those who held conventional positions within the career structure.7 Insert Tables 2a and 2b about here Tables 3a and 3b analyze how amateurs fare in general elections. The equations contain two more variables: the electorate’s partisanship, and ideology. The equation for elections with incumbents running perform well, but the adjusted R2 of the equation for open elections is just .15. So, the model does not work as well as it does for the primary elections. The amateur variable is not statistically significant in the open race equation but is in the incumbent race one. Amateurs receive about 6 percent fewer votes. Alternatively, the candidate quality index is only statistically significant in the equation for incumbent races. Spending has a positive impact in both of the equations. Party strength in the electorate is statistically significant with a large impact in the open races equation, 13 percent, but not a factor in incumbent races. Relative unemployment is not statistically significant in either equation. Presidential popularity is statistically significant in the incumbent equation only.


Both random and fixed effect models (fixed for state) were estimated. The results are essentially the same (not shown).

14 Insert Tables 3a and 3b about here Candidate quality, incumbency and amateurism are important predictors in primary elections as well as general elections in which incumbents run. Interestingly, they have far less explanatory in general elections without an incumbent candidate. Amateurs who survive to run in open general elections appear to do so on a more equitable basis. Indeed the analysis of the amateur governors winning primary elections (not shown) confirms that almost 75 percent win in open general elections. So, the general election has a somewhat different story to tell about amateur candidates who survive the primary contest to run in the general election. Particularly in open races, they appear to be at considerably less disadvantage than amateurs filtered out by the primary election process. Given what is important in campaigning and contesting elections, it is not surprising amateurs perform so poorly in primaries. Election success is explained, in part, by successful navigation of the gubernatorial career ladder with the reputation, name recognition, and the campaigning experience it brings with it. Still, some amateurs do go on to win the governorship. Addition factors obviously matter. In the next section of the paper I discuss the characteristics of successful amateurs, candidates without having held elected office or climbed the ambition ladder, to begin to better understand resources and skills they bring to the table that may help them successfully capture higher elected office. Characteristics of Amateurs Who Become Governor Canon (1990) observes that conventional wisdom places amateurism largely in the Sun Belt where there is less political tradition in the resurgent Republican Party. He quotes a Minnesota journalist to characterize the prevailing wisdom, “there’s little political tradition more ancient

15 that television and shopping centers.”8 However, it that were ever the case, it is much more widespread geographically and an occurrence in both parties. Of the governors serving between in office between 1978 and 2002, 41 are elected without following conventional career paths, 41 governors were amateurs.9 Only two had ever served in an elected post of any kind. There is considerable variability in the backgrounds and occupations of these governors prior to winning their state’s highest executive post. The most common profile builds upon private sector management experience. Twenty-seven of the amateur governors work in either corporate management or managing their own successful business enterprises prior to becoming governor. Many of amateurs utilize their access to business resources and/or family wealth to bolster their campaign efforts. For instance, West Virginia Governor Gaston Caperton used his personal wealth and his experience as President of McDonough Caperton Insurance Group to win office in 1988. Caperton defeated a long-term incumbent, Arch Moore. Many criticized the millions of dollars spent by both candidates in the race. Similarly Kirk Fordice, an engineer and executive officer for Fordice Construction Company, defeated incumbent Mississippi Governor Ray Mabus after taking on a slate of tough opponents in the Republican primary election including the politically well-connected State Auditor Pete Johnson. Others fitting into this mold include Fife Symington of Arizona and Bob Kerry of Nebraska. Twenty percent of successful amateurs were celebrities and may have drawn upon their status as former professional athletes and television commentators in their campaign for governor. In additional to the more sensational examples such as Schwarzenegger, less notable celebrities would include Alabama Governor Fob James who gained some notoriety as an All

Kevin Phillips, “At Election Time, Celebrities Can Make Political Party’s Day,” Minneapolis Star and Tribune, 18 June 1985, 11A. 9 Nine were first elected between 1968 and 1976 subsequently serving additional terms.

16 American halfback and pro football player before he founded and managed a very successful athletic equipment manufacturing business. He switched to the Democratic Party in 1977 ultimately beating the statewide elective office holder, State Attorney General Bill Baxley, in a runoff primary. Lamar Alexander had never served in elected office prior to losing a bid for governor in 1974. However, he did have strong ties to Senator Howard Baker serving him as a campaign aid and legislative assistant. He also managed the successful gubernatorial campaign of Winfield Dunn in 1970. Following his failed attempt, Alexander gained recognition as a political commentator for a Nashville television station and walked over 1000 miles across Tennessee in his successful bid to become governor in 1978. Other successful amateurs, like Lamar Alexander, may have benefited from experience campaigning, party involvement, as well as family and business ties to political party notables. For instance, former Governor John Brown of Kentucky was the son of a well-known Kentucky political figure and was extremely successful in his fast food business ventures. Although his father was repeatedly defeated in bids for the United States Senate, Brown, no newcomer to political campaigns when he entered the race in 1979, beat former Governor Louis Nunn in the general election. Likewise, George Bush joined a lifetime of exposure to campaigning and public life with a successful business venture to beat a tough incumbent, Texas Governor Ann Richards. Finally, almost half served in some form of public sector non-elected office before becoming governor. Some amateurs, like former Alaska Governor William Sheffield, served in local appointive positions such as the Anchorage City Planning Commission or former Massachusetts Governor Edward King who held various positions with the Massachusetts Port Authority in Boston. King was also a former professional football player having played guard for the Buffalo

17 Bills. Table 4 lists the frequency with which amateur governors exhibit other factors which may enhance their prospects for election including previous campaign experience, party involvement, personal and family ties to politicians, and experience in non-elective office. Insert Table 4 about here The Amateur Typology and the Amateurs Who Become Governor Ideally, biographical data on all amateurs who run for governor from 1978 to 2004 could be scrutinized and classified in terms of an amateur classification scheme like Canon’s to test its ability to predict which candidates win gubernatorial election, how they perform in office after being elected, and whether they win subsequent elections. However, the fact that over 1000 amateurs run for office over this time makes such an analysis beyond the scope of this paper. The lack of such systematic information for the governors that lost in the primary and general elections makes the ensuing observations somewhat speculative for now. However, the analysis will attempt to “back out” insight about the plausibility of classification scheme and reflect on the amateurs that have won the governorship and how they compare with career governors. Canon’s (1990) typology of amateurs enables him to distinguish between more ambitious amateurs, those motivated by policy agendas, and hopeless candidates who stand little chance of winning but somehow are swept into office by unforeseen events. The picture that emerges in the previous section is one where exceptional individuals mobilize resources and competitively campaigned to win the state houses in their respective states. They typically are well financed, sometimes well known, often having demonstrated ambition and talent in their private sector business endeavors. They share many things in common with career politicians with the exception of previous elected office experience so they then seem likely to belong in Canon’s

18 ambitious and/or policy categories. The data allow comparisons on a number of background characteristics related to occupation, level of education, and party affiliation. Both amateurs and career governors are as likely to have experience in small business, although amateurs are more likely to have served in corporate management or to have built large financially successful businesses. There is also little different between them in terms of their previous involvement in interest groups, professional or charitable associations. Likewise, there is no difference in education level or their age when first elected. Amateurs and career governors on average receive similar vote margins in winning election and reelection and enjoy similar tenure in office when taking into account term limits. Political scientists question whether amateurs possess the experience necessary to govern effectively. By definition, amateurs do not serve an apprenticeship in elected office prior to becoming governor. Canon (1990) argues that the prospect for how amateurs perform once in office can be taken from at least three different perspectives. First, amateur governors will adapt to politics and the ways of compromise and cooperation to eventually perform at levels comparable to seasoned political executives. Second, amateurs will not adapt but perform effectively in their newfound political careers utilizing other skills and resources they bring with them into office. Third, amateur governors will not adapt, perform inadequately, and fail to continue in elected office. One political scientist elaborated on this third perspective by commenting that “(t)o suppose that one can run a complex system without first learning the trade is, as Plato pointed out, as silly as to suppose that one can be a doctor or a carpenter without prior training” (Dahl, 1957). There is not a systematic account of amateur governors failing in office, of amateurs maintaining abysmally low job performance ratings, or of amateur governors failing to be reelected or

19 recalled from office on the basis of their performance in office. This does not pertain to amateurs who win office. Table 5 reports how amateur governors fare after their first term in office (five are still serving in their first term and three were legally prohibited from serving a second term). Just over half successfully win reelection. An additional one runs successfully for a different office. Four governors choose not to run. Six of the governors lose in their bids for reelection and one governor chooses not to run because of the poor prospect for reelection. Table 6 shows that these reelection rates are not statistically different from governors more generally although a larger percentage of the career governors serve additional terms of office. Twenty of the amateurs serve one term in office while 18 go on to serve two or more terms. Of the career governors, 74 serve a single term in office while 112 go on to serve two or more terms in office. So, Canon’s first two perspectives seem most appropriate. Insert Table 5 and Table 6 about here There are a few things that amateurs and the career governors differ on. The first is experience in previous office. As just pointed out, career candidates spend more time in other elected offices before being elected governor. The average tenure in public elected office for them is between 1978 and 2004 is about 15 years. Amateurs serve virtually no apprenticeship in elected office. Another difference regards political party. Amateurs more frequently are Republicans. While 45 percent of career governors are Republican, 57 percent of amateur governors are Republican. It is not entirely clear what this reflects although a few governors such as Alabama Governor Fob James and Arkansas Governor Frank White “switched” parties to run for governor in traditionally Democratic, southern states. Finally, amateur governors are more likely to have been employed in corporate management positions, large-scale development companies, and multi-state or national business. About a third of amateur governors have been

20 employed in this way compared to just about 6 percent of the career governors. Conclusion High profile amateurs running in high profile elections grab our attention and lead us to contemplate the impact of amateurs in American statehouses. The media has given considerable attention to high profile amateurs. But the analysis of amateurs has typically has been only a complement to analyses of political careers by political scientists. This paper has sought to address this by drawing upon Canon’s theoretical framework for understanding amateurism in Congress to better understand where political amateurs fit into gubernatorial politics. Beyle’s data show that many amateurs appear on the ballot for governor in primary elections, but very few move on to the general election ballot. Dahl (1957) may have characterized the plight of most of these candidates best when he wrote: “ . . . the intrepid amateur without experience in public life, whose eye is on the governor’s mansion and whose ears are carefully attuned to catch the first, faint cries of his friends and neighbors demanding the he be drafted into high office – this intrepid amateur, I say, will fall flat on his face, as he deserves . . . ” Faced with such prospects, an amateur running for governor is irrational and/or naïve. The hopeless amateur rarely survives the primary election. More importantly, while the analysis finds that the hopeless amateurs rarely if ever challenges and wins gubernatorial general elections, a minority of more ambitious and capable amateurs sometimes do win the primary election and go on to compete in open general elections. There are even a few cases where they have unseated incumbents. The concept underlying Canon’s typology of amateur congressional candidates, then, seems quite pertinent to understanding how amateurs become governor. With the exception of previous experience in elected office, these amateurs and career governors have similar career and personal background

21 characteristics. To varying extents they rival career candidates in name recognition, campaign financial resources, family political connections, and experience in campaigns. An interesting focus for future research may be to use the behavior of amateur governors to test Canon’s socialization and recruitment theory. An analysis could focus on the tenure (in years) and job performance ratings of governors serving over a similar time period as the current analysis. If recruitment theory is correct, the analysis should show significant differences in the job performance ratings and tenure of office of amateurs versus politically experienced governors reflecting their lack of apprenticeship in elective office. If socialization theory is correct, the converse will be true. Finally, it may be that the performance of amateurs varies. This would lend support to the idea that some amateurs govern effectively based on other skills. Corporate experience might be such a characteristic. It may turn out that the amateur capable of becoming governor is equally prepared or unprepared as the case may be to serve as governor.

22 References Beyle, Thad L. 1992. Governors and Hard Times Washington, DC: Congressional Quarterly Press. _______. 2004. “The Governors.” in Virginia Gray, Russell L. Hanson, and Herbert Jacob, eds., Politics in the American States. 8th ed. Washington D.C.: CQ Press. Beyle, Thad L. and Jennifer Jensen. Gubernatorial Campaign Expenditures Database. Bond, Jon R. 1997. “Partisan Differences in Candidate Quality In Open Seat House Races, 1976-1994.” Political Research Quarterly. 2(June): 281-299. Bond, Jon R., Cary Covington, and Richard Fleisher. 1985 “Explaining Challenger Quality in Congressional Elections.” Journal of Politics 47: 510-529. Brace, Paul. 1993. State Government and Economic Performance. Baltimore: Johns Hopkins Press. Canon, David T. 1990. Actors, Athletes, and Astronauts: Political Amateurs in the United States Congress. Chicago, IL: University of Chicago Press. Cohen, Jeffrey E. and James D. King. 2004. “Relative Unemployment and Gubernatorial Popularity.” Journal of Politics 66: 1267-1282. Chubb, John E. 1988. “Institutions, the Economy, and the Dynamics of State Elections.” American Political Science Review 82: 133-54. Erikson, Robert S., Gerald C. Wright, and John P. McIver. 1993. Statehouse Democracy: Public Opinion in the American States. New York: Cambridge University Press. Jacobson, Gary C. 1990. The Electoral Origins of Divided Government. Boulder, CO:Westview.

23 _______. 1992. The Politics of Congressional Elections. 3rd edition. New York: HarperCollings. Grady, Dennis. 1991. “Managing the State Economy: The Governor’s Role in Policymaking.” In, Gubernatorial Leadership and State Policy, eds. Eric B. Herzik and Brent W. Brown. New York: Greenwood Press, pp. 105-120. Hamman, John A. 2004. “Career Experience and Performing Effectively as Governor” American Review of Public Administration 34:151-163.

Hendrick, Rebecca M., and James C. Garand. 1991. “Variation in State Economic Growth: Decomposing State, Regional, and National Effects.” Journal of Politics 53 (4): 10931110. Herzik, Eric, and Brent Grown, eds. 1991. Gubernatorial Leadership and State Policy. New York: Greenwood. Kenney, Patrick 1983. “The Effect of State Economic Conditions on the Vote for Governor.” Social Science Quarterly 64: 154-62. Lasswell, Harold D. 1930. Psychopathology and Politics Chicago: University of Chicago Press. Mullaney, Marie M. 1989. Biographical Directory of the Governors of the United States 19831988. Westport, CT: Greenwood Press. _________. 1994. Biographical Directory of the Governors of the United States 1988-1994. Westport, CT: Greenwood Press. Raimo, John W. 1985. Biographical Directory of the Governors of the United States 1978-1983. Westport, CT: Greenwood Press. Simon, Dennis M., Charles W. Ostrom, Jr., and Robin F. Marra. 1991. “The President, Referendum Voting, and Subnational Elections, in the United States.” American Political

24 Science Review 85: 1178-1192. Squire, Peverill 1992. “Challenger Profile and Gubernatorial Elections.” The Western Political Quarterly. 45(1): 125-142. Squire, Peverill, and Christina Fastnow. 1994. “Comparing Gubernatorial and Senatorial Elections.” Political Research Quarterly 47: 705-720. Schlesinger, Joseph A. 1966. Political Ambition and Politics: Political Careers in the United States. Chicago: Rand McNally. Stein, Robert M. 1990. “Economic Voting for Governor and U.S. Senator: The Electoral Consequences of Federalism.” Journal of Politics 52: 29-53. Svoboda, C. J. 1995. “Retrospective Voting in Gubernatorial Elections: 1982-1986.” Political Research Quarterly 48 (1): 135-150.

25 Table 1 Percentage of Governors with Experience in Different Elective and Administrative Office1 Congress Statewide Legislature Law Local Administration Enforcement Elective Years N 19771990 17 35 42 17 13 22 104 19912004 14 33 60 26 15 25 129 19772004 16 33 52 22 15 24 233
Note, row percentages do not add up to 100 percent because many governors hold more than one office prior to becoming governor. The background data are based on each governor’s account in the Almanac of American Politics and the Biographical Directory of the Governors of the United States.

26 Table 2a Amateur Candidates and the Percentage of Primary Election Vote Open Races From 1978 to 2002 1 Variables b Amateur -11.20 Candidate .01 Quality Spending .00 Relative 1.08 Unem. Pres. Pop. .01 Number -3.60 Candidates Gender 3.05 Constant 46.10 R2 / Adj R2 - Within - Between - Overall N (obs) N (groups) Sigma_u Sigma_e rho

OLS se 1.36 .00 .00 .37 .01 .23 2.00 1.58 .31 p .000 .002 .000 .004 .292 .000 .128 .000 .30

Random Effects B se -11.37 1.36 .01 .00 .00 1.05 .013 -3.64 3.35 46.45 .00 .43 .011 .26 1.96 1.79

p .000 .001 .000 .015 .241 .000 .088 .000

Fixed Effects b Se -11.49 1.39 .01 .00 .00 1.01 .01 -3.59 3.72 45.93 .00 .50 .01 .28 1.97 1.79

p .000 .001 .000 .043 .194 .000 .059 .000

.28 .44 .31 1063 1063 50 5.10 18.6 0 .07

.28 .44 .31 1063 50 7.50 18.6 0 .14

Data are from each governor’s account in the Almanac of American Politics and the Biographical Directory of the Governors of the United States, Beyle and Jensen’s Gubernatorial Expenditure data set at , and U.S. Officials’ Job Approval Ratings

(JARs) website, .


Table 2b Amateur Candidates and the Percentage of Primary Election Vote Incumbent Races From 1978 to 20021 Variable b -6.61 .06 OLS se p 4.01 .10 .01 .000 .00 .074 .68 .858 .02 .517 .713 .000 Random Effects b se -7.65 4.04 .06 .01 .00 .02 -.12 -6.04 .00 .71 .02 .73 4.58 4.53 Fixed Effects p b se .059 -13.84 4.65 .000 .05 .01 .070 .977 .581 .000 .389 .000 .00 -1.27 .01 -4.59 -5.04 45.09 .00 1.14 .03 .96 4.99 5.32

Amateur Candidate Quality Spending .00 Relative .12 Unem. Pres. Pop. -.01 Number -6.11 Candidates Gender -3.75 Constant 45.09 R2 / Adj R2 - Within - Between - Overall N (obs) N (groups) Sigma_u Sigma_e rho

p .003 .000 .084 .270 .754 .000 .314 .000

4.58 .414 -3.94 4.50 .000 45.63 .76 .75

.69 .88 .76 207 207 45 2.84 16.37 .03

.70 .84 .73 207 45 12.95 16.37 .38

Data are from each governor’s account in the Almanac of American Politics and the Biographical Directory of the Governors of the United States, Beyle and Jensen’s Gubernatorial Expenditure data set at , and U.S. Officials’ Job Approval Ratings

(JARs) website, .


Table 3a Amateur Candidates and the Percentage of General Election Vote Open Races From 1978 to 20021 Variable b -2.46 .01 OLS se p 2.05 .233 .00 .045 .00 5.39 5.10 .503 Random Effects B se -3.50 2.02 .00 .00 .00 5.01 4.87 .56 .02 .22 2.70 2.72 Fixed Effects b se -3.29 2.14 .00 .00 .00 5.07 4.90 .69 .02 .29 2.80 3.27

Amateur Candidate Quality Spending .00 Party Strength 13.24 Ideol. Strength 6.09 Relative .467 Unem. Pres. Pop. -.016 Number 1.37 Candidates Gender -5.39 Constant 37.95 R2 / Adj R2 - Within - Between - Overall N (obs) N (groups)

p .083 .195

p .125 .193 .000 .000 .314 .857 .215 .96 .064 .000

.017 .00 .015 19.16 .233 5.11 .355 .09 -.02 -.33

.000 .00 .000 17.96 .293 4.95 .87 .13 .246 .141 -.02 -.01

.017 .363 .36 .000

2.76 .051 -4.25 2.04 .000 45.58 .16 .14

.12 -5.23 .000 41.70

.15 .06 .13 320 320 50

.16 .02 .11 320 50

Sigma_u Sigma_e rho

5.35 13.64 .13

8.24 13.64 .27

Data are from each governor’s account in the Almanac of American Politics and the Biographical Directory of the Governors of the United States, Beyle and Jensen’s Gubernatorial Expenditure data set at , and U.S. Officials’ Job Approval Ratings

(JARs) website, .

29 Table 3b Amateur Candidates and the Percentage of General Election Vote Incumbent Races From 1978 to 20021 Variable b Amateur -6.08 Candidate .03 Quality Spending .00 Party Strength 5.22 Ideol. Strength 5.79 Relative .72 Unem. Pres. Pop. -.04 Number 1.00 Candidates Gender 2.99 Constant 33.00 R2 / Adj R2 -Within -Between -Overall N (obs) N (groups) Sigma_u Sigma_e rho

OLS se p 2.28 .008 .00 .000 .00 5.38 4.62 .47 .041 .333 .210 .131

Random Effects B se -6.76 2.28 .02 .00 .00 7.39 5.90 .54 -.04 -.28 .00 5.33 4.65 .48 .02 .26 3.18 2.46

p .003 .000 .019 .166 .205 .26 .005 .279

Fixed Effects b se -5.67 2.47 .02 .00 .00 7.70 4.05 1.17 -.04 -.01 .00 5.30 4.67 .75 .01 .41 3.42 3.46

p .023 .000 .000 .148 .386 .125 .014 .983 .681 .000

.02 .007 .45 .027

3.16 .345 3.22 2.19 .000 38.05 .45 .43

.311 1.40 .000 34.23

.46 .39 .44 299 299 48 0 12.82 0

.48 .19 .41 299 48 6.99 12.83 .23

Data are from each governor’s account in the Almanac of American Politics and the Biographical

Directory of the Governors of the United States, Beyle and Jensen’s Gubernatorial Expenditure data set at , and the U.S. Officials’ Job Approval Ratings

(JARs) website, .


Table 4 Amateur Governors and Election Relevant Experience1

10 Campaign Experience Party Involvement Personal/Family Ties to Partisans Celebrity Status Worked in non-elected office N


11 18 10 18

25.6 41.9 23.3 41.9 43

The background data are based on each governor’s account in the Almanac of American Politics and the Biographical Directory of the Governors of the United States.


Table 5 Reelection Success of Amateur Governors1 Frequency Ran for governor and lost Ran for reelection and won Ran for different office and won Did not run adversity Did not run no adversity Could not serve second term Still in first term Total 6 21 1 1 4 3 5 41 Percent 15 51 2 2 10 7 12 99 %


The background data are based on each governor’s account in the Almanac of American Politics and the Biographical Directory of the Governors of the United States.

32 Table 6 The Number of Terms Served by Career and Amateur Governors1
Lateral entry to governor 0 Number of terms in office One elected term in office Count % within Lateral 39.8% entry to governor Two or more elected terms in office Chi-square = 2.14 sig. = .14 Count % within Lateral 60.2% entry to governor Count % within Lateral 100.0% entry to governor

1 74 20

Total 94













The election and background data are based on each governor’s account in the Almanac of American Politics and the Biographical Directory of the Governors of the United States.

Shared By: