Campaign Fundraising and Gender in Alabama. Laura Merrifield Sojka. University of Alabama

Transcription

1 1 Campaign Fundraising and Gender in Alabama Laura Merrifield Sojka University of Alabama Presented at The Alabama Political Science Association s Annual Conference Auburn, Alabama March 31, 2012

2 2 Preface This paper is a component of a larger research project on which I am analyzing a comparison between Alabama and Colorado to analyze whether the gender composition of the state legislature relates to the ability of female candidates to fundraise. In this particular piece, I am focusing strictly on Alabama but my ultimate objective is to create a more comprehensive analysis. I. Introduction In spite of the themes of American democracy that proclaim equality, the political arena has a long-standing reputation as being discriminatory and biased against minorities not conforming to the status quo. Gender, ethnicity, socioeconomic status, religion, and sexual orientation are far from representative at most state and federal levels of government compared to the composition of the electorate. Nearly one hundred years following the passage of the 19 th Amendment, securing for women the right to vote, not one state has a proportionately balanced legislator on the basis of sex. Yet, when one considers the elected officials chosen to represent the masses, similar concerns for equity seem to not apply. Over two-hundred years ago when the institutions of government were first conceived, the politicians serving the voting public appeared identical to them: just as voters were confined to the white, upper-to-middle class, well-educated male demographic, so were the officials they elected. Now surging forward in the 21 st century, women, African-Americans, and 18-year-olds are no longer disenfranchised and comprise a substantial segment of the voting eligible population. The candidates running and the politicians repeatedly obtaining elected office, however, still resemble a collective group more reminiscent of the post-colonial era than that of the post-hegemonic superpower one. On the federal level, men occupy 76% of the seats in Congress, 66% of the Supreme Court positions, and have thus far maintained sole dominance over the presidency (Center for American Women and Politics, 2009). The state level seems slightly more promising, but also embodies more variety that inherently limits its consistency. Men can claim an average of 77% of current state legislative seats and 88% of present governorships. It is evident that in our modern democracy boasting a fairly inclusive electorate and purportedly representative of the people, those who maintain the government and those who are the governed are two very distinct groups. It is important to clarify that the lack of a truly proportionate representation does not innately mean that it should be render ineffective. The theory of trustee leadership entails that an individual can govern with regards to the best interests of his/her constituents, regardless of his/her own similarities to the mass. In some ways, men and women serving in various capacities reveal little distinction that can be specifically attributed to their gender: voting records, for instance, do not necessarily differentiate a female legislator from her male counterpart. Focusing strictly on sex to conform to the parameters of the ultimate argument, if men can articulate the needs of women as clearly as women themselves, then the demands for more women in higher positions of government are more superficial and aesthetic than they are essentially helpful. I argue that the need for a more gender-balanced government is critical to the function and continuation of democracy, as it involves greater participation among the entire breath of the citizenry, empowering individuals not just to submit a ballot every

3 other year but enabling them to take a more active role serving their community as well. Rather than relegate women and other underrepresented groups to simply preserving the integrity of free and open elections by participating as voters, it remains imperative that these groups are represented within the high ranks of elected office. Male politicians may be able to govern in their perception of the best interests for women, but women alone have the political and social agency to access their unique experience as women. Richard J. Fox (2000) notes that though women have excellent opportunities to run successfully for Congress men and women often have different experiences highly instructive of the gendered political dynamics that are still present in American politics (229). In a point I wish to explicitly reiterate, this need is not relegated solely on the basis of sex, but is equally applicable to race, ethnicity, religion, and socio-economic status. For the focus of the thesis and structure of this argument, however, sex will be the primary delineating variable of interest. Many reasons for this gender disparity exist but one of the most debated and critical is the ability of a candidate to raise enough funds to surmount an effective campaign. Fundraising proves to be an invaluable component in the modern era and without enough money needed to propel an aspiring (or incumbent) politician, the campaign is unlikely to be successful. Research on discrepancies between the sexes in fundraising is somewhat divided as some scholars argue a gross inequity between the two and other scholars maintain no such difference exists. The critical need for fundraising in a campaign has been examined thoroughly by literature in the field, but most research has concentrated on Congress or focused on large scale studies that summarize national data for all 50 state legislatures. All states, however, are not equal, particularly with regards to inclusion of women in political office. Some states boast almost proportionate male-female representation in their state legislative assemblies, such as Colorado with over 46% of its legislature seats held by women. Other states dip as low as the single-digits in female representation at the state level, such as South Carolina with a mere 9% of its House and Senate held by women. A number of factors contribute to the extreme variety of women in state legislative office amongst the states, but perhaps the most influential (and unfortunately, for a quantitative study, the most immeasurable) is the state s culture. Because the unique history of each state shapes its economy, its politics (institutions as well as policies), and its people, evaluating the influence of gender and fundraising at the individual state level is valuable. This analysis can offer critical insight into the particular facets that play a role in women s road to state-level political office. Alabama ranks on the lower end of women s involvement in the state legislature at around 13% and has some distinguishing institutional features that have been noted by studies centering on other arenas to make minority inclusion in government more challenging. Term limits are often cited as a beneficial mechanism through which more individuals from underrepresented groups (i.e. women, African-Americans, etc.) can gain entry as open seats are more prevalent following the conclusion of the incumbent s term(s) in office. Neither the assembly in the Alabama Legislature has term limits, which decreases the frequency of open seats and, given the high incumbency advantage, makes female candidate s opportunities more difficult. Likewise, both the House and Senate serve four year terms, an unusual attribute for lower assemblies, which in most states and at the federal level have much shorter two year terms. The longer term, like the infinite 3

4 4 term limit, means incumbents stay in office for a longer period of time, have more time to accomplish political objectives (and cultivate allegiance from their constituency) and the opportunities for challengers are inversely shortened. Though historically Alabama has not demonstrated progressive attitudes towards the inclusion of women in public office, the interactions of this low margin with the obstacles of fundraising (paramount to political success) are critical in understanding the challenges female candidates in Alabama face. This paper seeks to assess the implications of gender and fundraising at the state legislature and consider what factors may influence women s ability to raise funds in route to obtaining a seat in the Alabama House or Senate. When the underlying principles of democracy emphasize representation and the modern political climate demands adequate money to win, analyzing the effects that gender has on fundraising proves to be crucial in developing a greater understanding of female representation in elected office. II. Theory Any potential candidate debating running for elected office encounters a number of valid considerations that will ultimately lead to his/her decision that he/she has a viable chance at winning given the resources (physical and other) available at his/her disposal. Each hopeful politician carefully evaluates the political/social climate of his/her constituency and its relation to the larger scope, the popularity of his/her party, the status of the current officeholder (maintaining incumbency or leaving office), and the financial, media, and mass support predicted. This common political practice establishes a helpful context through which we can consider the challenging issues facing women deciding to run in the state legislature. In spite of the formal legal equality women have been afforded for nearly an entire century now, descriptive representation still lags. Women comprised only 24.3% of legislators at the state level in 1999, a number quadrupled from a meager 4.5% in 1971, but still disparaging nonetheless (Center for American Women and Politics, 2009.) In some states, the state legislature is almost proportionately representative of the electorate it serves and represents with regards to sex with one state (Colorado) just shy of an even mark. Other states have as few as 12% of women in the state legislature, a surprisingly low number given the recent surge of women in office within the past 20 years (Center for American Women and Politics, 2009.) At the state level, the disparity among female representation within the legislature within the 50 states is remarkable to say the least. The apparent gap that exists throughout the various legislatures poses an array of interesting and germane concerns as to why it exists, what it means, and what implications it has on women interested in running campaigns for public office. Why are more women not involved actively involved in the forefront of politics, particularly at the state level? If legal barriers are not the obstacle, then another factor must be in play to create this lag in gender proportionate representation. Three fundamental components could be affecting this inequity. The first point of consideration is that women could be less likely to run in campaigns and therefore less likely to obtain office (assuming that their chances of winning any given election is at least equivalent with men). A second possibility is that women are not as well perceived on the whole as their male opponents and voters thus prefer to vote for male candidates over female ones

5 when every other aspect aside from gender is identical between the two competitors. Lastly, it could be the case that women do run a substantial amount of campaigns in comparison to men and they are also fairly favored by voters, but they face greater obstacles in trying to fundraise for campaigns and are then less capable of competing in a level race for office. Though the number of female candidates has been climbing significantly within the past forty years, the number of women running for elected office, particularly at the state and federal level, is still lagging behind the number of men for the same positions. Lawless and Fox (2004) note that the distance between the men and women interested in running for office is considerable (with 59% of men and 43% women), while the percentages that ultimately choose to run are 12% men and 7% women. It is worthwhile to note that of these numbers for men, they were roughly evenly split between interests in local office at 37% and interest in state-level office 36%, while for women the interest was significantly slanted to local office at 41% over state-level office at 27%. Women who do run for office actually favor comparably to their male opponents (Burrell, 1990; Dolan, 2004; Fox, 2006; Hogan, 2001), when all other factors are equal, though partisanship complicates the issue slightly. Matland and King (2002) found that for Democratic candidates, gender was for the most part irrelevant. For Republican candidates, a woman would generally incur a significant difficulty in obtaining their party s endorsement, but once running on the Republican ticket, they would gain support from moderate Democrats and independents, while retaining much of their Republicanbase. Partisanship is related to the public perception of female candidates and the individual voter s opinion of female candidates, both which also play a significant role in election results (Dolan, 2008). Numerous studies have examined this issue and determined that a variety of factors, including those that are institutional, social, personal, and economic, remain even as subtle obstructions preventing women from initiating campaigns in the first place. Earlier studies focused on gender socialization, noting that politics was often considered a man s game (Lee, 1977; King, 1977; Stoper, 1977; Welch, 1978; Flammang, 1997.) These contended that women were neither seriously considered as valid candidates worth election nor should women view politics as an appropriate arena for a career for their sex. Contributing to this impermeable socialization was the cyclical lifespan of politicians that required participating in the party beginning at a young age and serving in powerful positions. Women were expected to get married and raise a family and, if a woman subsequently cultivated an interest in the political arena, she was easily overlooked because she lacked the breadth and depth of service that men were able to accumulate (having been indoctrinated early in life about the value of politics). While the old adage does not necessarily apply in the 21 st century, social factors are still cited as impediments to women s involvement, particularly family and support structures that complicate the decision to run for a woman in compassion to a man (Lawless, Fox, & Freely, 2001; Carroll & Sanbonmatsu, 2009.) Women and men initially exhibit similar levels of interests in running for office, but when women are ignored for candidacy in a particular position, they will pass without ample encouragement and support. The decision made by a woman to run is further challenging, as she must consider the level of support she can potentially garner in order 5

6 to have a viable candidacy, as well as how she will balance her traditional female roles with that of a politician. Other studies have noted institutional features as inherently biased against women s involvement that could play a critical role in explaining the low percentages of women in office (Carroll, 1994; Carroll & Jenkins, 2001; Darcy, Welch, & Clark, 1994; Rule, 1981). When men face less socialized and economic obstacles towards conducting a successful bid for a position, women are more likely to struggle in single-member districts competing against men, rather than in multimember districts with more than two candidates. Term limits typically yield positive results in allowing more women to compete for public office, though these do not exist in the Alabama legislature. Incumbency, a powerful variable in American politics on nearly any level, is even more pervasive, as men already obtain an overwhelming majority of the elected offices and the benefits of this minimize the chances of a female challenger winning in a race. Finally, another reason women may be less inclined to declare candidacy for public office is due to the processes of the election and positions themselves and how they require women to make challenging decisions less likely to be incurred by their male competitors (Rule, 1988; Lawless & Fox, 2004). Though women enjoy fundraising and personal campaigning working on self-promotion as much as their male counterparts, they also must do more of it in order to achieve the same level of recognition with the voters and the time necessary to establish this is time that is needed to cultivate other aspects of the campaign as well. An additional deterrent for women is their perception (and one potentially held by men also) that a woman must highly educated and on the higher end of socioeconomic status in order to run a viable campaign. Men perceive no such limitations on themselves and have been demonstrated to obtain political aspirations regardless of educational attainment or personal economic level. All these factors may play critical roles in the issue of gender underrepresentation, but the central question upon which we will now focus is how fundraising affects women s abilities to successfully run for office. Money is a necessary component to the modern candidacy and quickly and easily stratifies potential frontrunners from the rest of the pack. For women, who already face a myriad of other potential obstacles, fundraising is a central factor explicitly related to their political success. Every component that can potentially explain the underrepresentation of women is important, but the value of fundraising, crucial in campaigning, will remain the focus of this paper. For female candidates who are more likely to come into the race at a disadvantage (by having worked in politics less than their male counterparts and encountering more social prejudices against her bid), partisanship and fundraising are essential. Individual and PAC contributions can make a significant difference in the ultimate success of a campaign. Given the prevailing importance of PAC endorsement, female candidates are in a particularly vulnerable state, where the support of a PAC, especially early in the campaign, can really make a needed difference in the eventual outcome of the race (Connor Green, 2004). Partisanship plays a role in PAC support, as well, particularly given the saliency of certain gendered political issues (Day, Hadley, & Duffy Brown, 2001; Day & Hadley, 2002). Donors, both as individuals and as heads of PACs, want to support candidates they believe have solid potential at being able to win. Funds are needed, of course, as a part of doing this, however PACs wait to review candidates carefully and precisely 6

7 7 choose who will receive not only their hearty endorsement, but additionally their financial weight. It is a problematic cycle in one sense: as the more women who win, the more PACs will choose to support them, the more PACs contribute to women s races, the more women will win. Because of the recurring nature of this relationship, it is difficult for women running for office to secure PAC support and it is risky for PACs to contribute until they are confident in the potential of the candidacy. It is critical to note PACs that specifically focus on bundling campaign contributions towards female candidates, such as EMILY s List (for Democrats) and WISH List (for Republicans) do not devote their efforts to female candidates in Alabama. This is likely to do with the perceived viability of the candidates and the fact that few women have established political careers, making the prospect of future expansion uncertain. Partisanship has routinely demonstrated to have substantial effects on candidates effectiveness to fundraise during a campaign; the magnitude of its influence on female candidates in particular is critical. Democratic women generally are more successful in raising money for their campaign, benefiting from the power of Emily s List but also more implicitly, the correlation between women s issues and the Democratic party. Republican women can be tremendously effective fundraisers, as well, but they traditionally incur more challenges, in spite of having the aid of Wish List (a slightly smaller and less powerful PAC). Studies have found that while women running for the Democratic party actually attained a significant amount more money compared to Democratic men, while both men and women running for the Republican party achieved essentially identical amounts (Fox 2000.) The party for which a candidate is a member offers a number of benefits or disadvantages, depending on the larger circumstances, for the individual s campaign. In terms of garnering votes, party identification is an easy indicator on which a lesssophisticated electorate can use to determine vote choice. For prospective donors, given the increasingly ideological and polarized positions of the parties, the R or D label provides a helpful reference for the candidate s policy positions and assists them by providing preliminary information that serves as a basis for determining endorsement. III. Data and Methodology In order to examine the relationship between gender and amount of money raised for state legislative campaigns, a review of the public records on how much each candidate used and total number of donations is necessary. Considering a number of elections years on which we have sufficient data, the 2006 year was selected for a number of reasons. First, it served as the off-year congressional elections between two relatively substantial presidential races. Though the political actors and the public in 2006 could not have predicted the magnitude of the 2008 presidential race, the political climate was already changing. The 2006 race was also interesting given the economic and national circumstances of the time. The economy was still two years short of collapse and the nation was already five years from the 9/11 terrorist attacks that brought a sense of patriotism and unity to the country and subsequently propelled the invasion and invasion in Iraq. These provide a nice time from which to sample because the time was slightly more gender neutral given the diminished significance of hard policy arenas (economy and war) on which women are typically deemed worse to handle. It is probable that

8 promptly following the invasion or the economic recession, voters would have employed the stereotypical gendered associations and biased this experience. Finally, these elections occurred relatively recently enough to be relevant to the current political mood while they are still distant enough to enable us to consider the larger implications of their results. After collecting all the data on the campaign contributions for all male and female candidates running for state congressional office, I categorized gender (employed here as the dependent variable) and race. The entire candidate pool for both House of Representatives and Senate seats in Alabama totaled 320; in Colorado, it was 197. Because this includes those candidates who did not raise any money, who died during the electoral campaign, who withdrew, or who, for whatever reason, conducted a small (if existent campaign, some of the subsequent variables included in this analysis were not available. In spite of laborious research and probing, some candidates did not have identifiable election histories (to determine where they were viable candidates), or even gender or race. This unfortunate omission of data, however disappointing it is in an effort to conduct a comprehensive analysis, should not invalidate the ultimate findings, however, as the importance of gender and fundraising can still be determined by those many candidates whose gender and fundraising totals were capable of being determined. The independent variables included in this analysis include partisanship, incumbency, open seat, challenger, leadership position (within the assembly), candidate quality, and district partisanship. Partisanship can play a very substantial role in women s involvement, as the earlier literature review suggests, and was coded 0 for Republican and 1 for Democrat. Candidates identifying as independent or not running under either the Democrat or Republican parties were coded 2, but it is important to note that non-party aligned candidates in this dataset were rare. In the Colorado data set, this eliminated roughly 5% of the candidates; in Alabama, the margin was less than 2%. I do not believe the inclusion of third-party candidates will have much influence on the findings because, aside from their small population, they consistently raised much less money, tended to lose by large margins in the general election, and were dominated by men. Traditional variables noted for their relationship to fundraising were also incorporated. Incumbency, Open seat, and Challenger status were used individually as dichotomous responses, as well as intra-assembly leadership. If the individual previously held a high position within that particular house, such as Speaker of the House, then that was included as a control as well. A candidate running for reelection who is the current Speaker might garner more campaign funds for his/her higher position of power, but it is possible that by attaining such status, that legislator has a long political legacy, which would diminish the need for excessive fundraising. Opposition in either the primary or general election was important to include in the data set. Whether an opponent existed (within one s own party in the primary competition or on the opposing party in the general competition) could impact the total amount of money raised. Though this analysis looks strictly at the total amount of funds raised (disregarding the points of time in the campaign in which they were secured), the influence competition would have on the overall total is worth consideration. Primary competition and General competition were parsed into two variables to account for races in which the type of competition varied (again, affecting fundraising). 8

9 9 Finally, candidate quality, district partisanship, and electoral result were added to the dataset. A quality candidate can exert influence over how many other candidates enter the race, the ability to raise funds, and can be particularly valuable when considering gender, where previous electoral successes can be tantamount in both competition and fundraising. For the purposes of this study, candidate quality was assessed as a dichotomous measure based on prior success in any other public election. Obviously, incumbents are quality candidates, but this evaluation includes former City Commissioners and School Board Members who do not have state-level legislative experience, but are familiar with the inside of the elections process and, presumably, are well-liked (enough). This binary measure does not account for variety within quality candidates but it provides a sufficient, if rudimentary, description of the viability of a candidate. The partisanship of the district, that is, the way in which a district tends to lean, was calculated by examining the previous presidential election s returns for that area. Using the previous presidential election prevents the endogenaity that would have accompanied a calculation based on the 2006 mid-term congressional returns but would still allow a basis understanding of whether the district recently tended to lean red or blue. As with the partisanship variable, 0 indicated a Republican-leaning district while 1 implied a Democratic-leaning district. Lastly, the candidate s resulting outcome in the election was also noted. Whether the candidate won the general, lost in the general, lost in the primary, or did not finish (withdrew, died, etc.) was important to include because of its implications with the ultimate objective of this study. As noted in the literature review, female candidates who are able to raise as much money as men are as likely as their male opponents to win an election but this finding is concerned with national data, which may vary from the unique social/political climates of Alabama and Colorado. To determine whether statistical significance existed between gender and any of the independent variables, I utilized the Ordinary Least Squares analytic model. For a variety of factors, the OLS is best suited the data being used and the question being examined. The linear regression model is preferable because none of the Gauss-Markov assumptions are violated by the data set. The errors are uncorrelated, have equal variance, and an expectation of zero. Likewise, the dependent variables are all relatively random, identically distributed, and independent of each other. Homoscedasticity and autocorrelation should not be an issue. Time is not used, so the variables are static as well. To confirm the suitability of the OLS model, I utilized a number of diagnostic tests including hat values, studentized residuals, dfbeta, and Cook s D. The actual numeric results of these analyses are available in the appendices, but in summary, they revealed no reason that the OLS model would be ineffective with this particular data set and research question. IV. Findings The analysis generated from the OLS model yielded some surprising results. The variables representing general competition, leadership, and assembly all revealed highly statistically significant findings (with pr(> t ) less than ) and positive coefficients. The influence of general competition in the election on the overall amount of funds raised by candidates follows a logical understanding of the electoral process. Candidates who

10 10 failed to win at the primary level never even competed outside of their party for the office and also had far shorter campaigns compared to those who gained the primary nomination and continued on to the November general election. The value of leadership reinforces prior understanding of campaign fundraising. Though this variable was narrowly construed for the purposes of this study (looking at individuals in the highest levels of leadership within the assembly, such as Speaker of the House), a candidate running for reelection who held a leadership role in the previous term likely has an easier time securing funds. These individuals are obviously also incumbents, so they benefit from being quality candidates and having viable political careers with proven track records. Furthermore, the PACs, interest groups, industries, and individuals who benefit either directly or indirectly from the leader s position have a stake in assuring that he or she retains their seat. The scope and nature of this particular study does not parse out the total dollar amount of donations between particular industries or PACs, but it is likely that those who felt compelled to give perceived high expected utility and benefits in the future. Table 1. OLS Model: Effects on Fundraising Residuals Minimum 1 st Quartile Median 3 rd Quartile Maximum Coefficients Estimate Std. Error T-Value Pr(> t ) (Intercept) Gender Party * Won Quality ** Primary Com ** General Com e-05 *** Leadership e-12 *** Incumbency Open Seat Challenger Assembly < 2e-16 *** District Party * Signif. codes 0 *** ** 0.01 * Residual standard error: 142,400 on 292 degrees of freedom (14 observations deleted due to missingness) Multiple R-squared: 0.51, Adjusted R-squared: F-statistic: on 12 and 292 DF, p-value: < 2.2e-16

11 The specific assembly for which the candidate was running achieved a highly statistical significant outcome and a positive correlation as well. The distribution of offices in the House is far more generous than the Senate, as is typical in bicameral legislatures that tend to have larger lower-houses and smaller upper-houses. For Alabama, the House of Representatives includes 105 members compared to the Senate s 35. Though fewer people accordingly ran for Senate seats, they yielded a higher peroffice spending rate in comparison to the House, a fact that can at least minimally be explained by the larger area of representation (and number of constituents) employed in Senate districting. Candidate quality and competition in the primary election both attained moderate levels of statistical significance with pr(> t ) less than 0.001) with positive coefficients. Having previous experience to demonstrate a viable candidate enabled the perceived quality candidates to garner more campaign contributions than their counterparts who lacked experience in politics. Primary competition, similar to the variable that captured general competition, demonstrated a positive relationship to fundraising which follows general precepts of campaigning. In districts where there was no primary competition for the seat (typical for incumbents of that party), the campaigning season began later and was restricted to the general election. Partisanship demonstrated a prominent role in the variables capturing partisan affiliation of the candidates as well as the partisan leanings of the candidate s district. Both generated a slight statistical significance at pr(> t ) less than 0.01) but what was most interesting was the negative relationship between the candidate and the overall partisan preference of the district. It is admittedly challenging to generalize party identification among an entire district comprised of thousands of votes but the negative relationship is nonetheless surprising. As I continue to develop on this research, I plan on restructuring the measure of the variable to analyze partisan preference on a scale based on the percentage yielded in the previous presidential election for the preferred candidate (i.e. if Bush received 57% of the votes, the variable would read 57 on the 100-point scale) to clarify the measure and make its implications more apparent. Having described the variables that obtained various levels of statistical significance, it was disappointing (albeit surprising) to note that gender did not achieve a statistically significant correlation with fundraising. The relationship between gender and the total amount of funds raised did yield a negative coefficient (-3,355), demonstrating that women on the whole raised less money than men in their campaigns. Yet it was far from statistically significant with a substantial pr(> t ) value of Though this result initially seems unfortunate with regards to the prescribed hypothesis governing this study, it actually opens the possibility of a variety of interesting behaviors (both in the actual world and the statistics constructed to resemble it). Many variables captured the key facets determined to generally be needed in order to raise money (and subsequently win the general election): candidate quality, competition at the primary and general levels, incumbency, and district partisanship. Yet I suspect that many of these variables also place women in a disadvantaged position given the political climate and history of the environment. Very few women are likely to be incumbents, as the present composition of the Assembly reveals only 13% of the seats are held by women. Incumbency itself cannot be ignored as a critical variable in not only winning re-elections, but garnering electoral and financial support as a convenient 11

12 12 product deriving from the incumbent s unique ability to refer to his or her actions while in office (compared to the supposed actions that the opposition alleges they would do). Incumbency, though, did not achieve statistical significance. The phenomenon of few female state legislatures is not a unique trend related explicitly and solely to the assembly, and therefore quality would likely advantage men as well. The most pragmatic explanation relating to the construct of gender and fundraising lies in the candidate s partisan identification. In recent years. Alabama, like most of its southern neighbors, switched from its once famous and nearly exclusive Democratic party allegiance to affiliating more strongly with the Republican party. As noted earlier in the literature review, women in the Democratic party actually tend to receive slightly higher amounts of campaign contributions while their Republican counterparts experience the opposite effect. Many of the women who serve in the Alabama state legislature are Democrats, which might imply a greater ability to fundraise compared to men, though most of the members of the assemblies are Republican and male (which would suggest less). Examining this finding individually proves to be challenging, but it will contribute to the larger study at hand, of which this analysis comprises only a small part. Comparing Alabama as a state with low parity regarding gender representation to Colorado as a state with nearly equal proportions of gender representations will provide a greater comprehension of the variables that are truly influential and what the relationship between funding and gender in low- vs. high-equality states means. Utilizing an analytical comparison will enable me to evaluate whether my overall hypothesis (that states with more gender-equitable representation have less discrepancy between funds raised by women and men) has merit and validity. V. Conclusion The ultimate finding relevant to the original hypothesis of this paper revealed that gender was not statistically significantly related to fundraising in the 2006 Alabama state legislature elections. While this unanticipated result did not support the hypothesis, it does provoke some interesting points that I hope to address in future research. Methodologically, variables that capture a more accurate prediction of district partisan leanings (described in detail in the findings section) as well as variables that address opponent experience and partisan competiveness of the district. Additionally, I believe separating the data into two categories based on participants in primary elections and those in the general elections will enable a more accurate depiction of the implications these variables have on fundraising. Returning to the underlying normative theory guiding this research, political agendas can vary between male representatives and female representatives, depending on a number of other factors that come into play in the creation of policy. Studies have demonstrated that, following popular belief, female politicians generally do concentrate more on traditional women s issues compared to their male counterparts. Given the recent resurgence of women s health issues in the national spotlight, the inclusion of more women in the political conversation is imperative. The proportion of women in the state legislature can serve as a larger representation of the values and perceptions of women by that constituency, pending there is ample quantity and quality in the candidate pool.

13 13 The argument that states with less women in public office value women as leaders lies relies fatally on the assumption that there are equal opportunities and numbers of viable candidates of both sexes. Given the social and political barriers that have historically prevented or heavily discouraged female participation, I hesitate to assert that few women run at the state level for political office in Alabama because they are generally less valued. However, I want to explore the possibility that the obstacles hindering female participation are not dissolved but still very active. In my future analyses, I plan on incorporating ideas of social barriers (such as women in the workplace as a whole) and political barriers (such as women participating at local levels of office and women involved in local political office). Institutional barriers have conventionally been included in gender disparity explanations and though I regarded these indirectly (incumbency and term limits), additional inclusions may be forthcoming. VI. Appendix Plot 1. Cook s Distance

14 14 Plot 2. Hat Values Plot 3.

15 15 Plot 4. Plot 5.

16 16 WORKS CITED Burrell, B. (1990). The Presence of Women Candidates and the Role of Gender in Campaigns for the State Legislature in an Urban Setting: the Case of Massachusetts. Women in Politics, 10:3, Carroll, S. J. (1994). Women as Candidates in American Politics. Bloomington: Indiana University Press. Carroll, S. J. & Jenkins, K. (2001). Do Term Limits Help Women Get Elected? Social Science Quarterly, 82:1, Connor Green, J. (2004.) The Times Are They A-Changing? An Examination of the Impact of the Value of Campaign Resources for Women and Men Candidates for the US House of Representatives. Women & Politics, 25:4. Darcy, R.; Welch, S.; and Clark, J. (1994). Women, Elections, and Representation. Lincoln: University of Nebraska Press. Day, C. L. & Hadley, C. D. (2002). Who Contributes? Similarities and Differences Between Contributors to EMILY s List and WISH List. Women & Politics, 24:2. Day, C. L., Hadley, C. D., & Duffy Brown, M. (2002). Gender, Feminism, and Partisanship among Women s PAC Contributors. Social Science Quarterly, 82:4, Dolan, K. A. (2004). The Impact of Candidate Sex on Evaluations of Candidates for the U.S. House of Representatives. Social Science Quarterly, 85:1, Dolan, K. A. (2008). Women as Candidates in American Politics: the Continuing Impact of Sex and Gender. Eds. C. Wolbrecht, K. Beckwith, & L. Baldez. Political Women and American Democracy Flammang, J.A. (1997). Women s Political Voice: How Women are Transforming the Practice and Study of Politics. Philadelphia: Temple University Press. Fox, R. J. (2000.) Gender and Congressional Elections. Ed. S. Tolleson-Rinehart and J.J. Josephson. Gender and American Politics: Women, Men, and the Political Process. M.E.Sharpe. Fox, R.L. (2006). Congressional Elections: Where Are We on the Road to Gender Parity? Eds. S.J. Carroll & R.J. Fox. Gender and Elections: Shaping the Future of American Politics,

17 17 Francia, P. L. (2001). Early Fundraising by Nonincumbent Female Congressional Candidates: The Importance of Women s PACs. Ed. Karen O Connor. Women and Congress: Running, Winning, and Ruling. Haworth Press. Hogan, R. E. (2001). Campaign Spending by Men and Women Candidates for the State Legislature. Presented at the Annual Meeting of the American Political Science Association. Jenkins, S. (2007). A Woman s Work Is Never Done? Fundraising Perception and Effort Among Female State Legislative Candidates. Political Research Quarterly, 60:2, King, M. (1977). Cuba s Attack on Women s Second Shift, Latin American Perspectives, 3: Lawless, J. L.; Fox R. L.; & Freely, C. (2001). Gender and the Decision to Run for Office. Legislative Studies Quarterly, 26:3, Lawless, J. L. & Fox R. L. (2004). Why Don t Women Run for Office? Brown Policy Report. Lee, M. M. (1977). Toward Understanding Why Few Women Hold Public Office: Factors Affecting the Participation of Women in Local Politics. In Marianne Githens and Jewel L. Prestage, eds. A Portrait of Marginality, The Political Behavior of the American Woman. New York: David McKay. Matland, R. & King, D. (2002). Women as Candidates in Congressional Elections. Cindy Simon Rosenthal, ed. Women Transforming Congress. University of Oklahoma Press. Rule, W. (1981). Why Women Don t Run: The Critical Contextual Factors in Women s Legislative Recruitment. The Western Political Quarterly. 34:1, Sanbonmatsu, K. & Carroll, S.J. (2009). Gender and the Decision to Run for the State Legislature. Paper presented to the Midwestern Political Science Association Annual Conference. Stoper, E. S. (1977). Wife and Politician: Role Strain among Women in Public Office. In M. Githens and J. Prestage, eds., A Portrait of Marginality, New York: Longman, Welch, S. (1978). Recruitment of Women to Public Office: A Discriminate Analysis. Western Political Quarterly, 31: Women in State Legislatures: Fact Sheet. (2009). New Brunswick, NJ: Center for American Women and Politics

18 18