OLR Research Report

August 11, 2005




By: Kristin Sullivan, Research Analyst

In May 2003, the General Accountability Office (GAO) released a report on the 2000 and 2002 elections in Maine and Arizona—the first two states to offer full state funding for gubernatorial, statewide office, and legislative candidates. To draw conclusions about the impact of public financing, GAO identified five goals of those programs. Generally, the programs were intended to (1) increase voter choice, (2) increase electoral competition, (3) reduce the influence of special interest groups, (4) curb increases in the cost of campaigns, and (5) increase voter participation. GAO compared the 2000 and 2002 elections in Arizona and Maine to the 1996 and 1998 cycles in those states, focusing on the five goals.

In OLR Report 2005-R-0191 we summarized GAO's results, which generally indicated that it was too soon to determine the extent to which the programs met the five goals or to draw causal links to changes that resulted from public financing. This report surveys some of the limited scholarly research that is available on public financing's impact. But in some ways our results support GAO's finding. The articles, while drawing some conclusions, also indicate that not enough time has passed since Arizona and Maine implemented their programs to make concrete assertions about the impact of full public financing programs (clean elections). And they all recognize that variables outside public financing, such as term limits, could have influenced their results.



In their April 2004 paper, “Do Public Funding Programs Enhance Electoral Competition?” University of Wisconsin, Madison, political scientists Kenneth Mayer and Timothy Werner assess the impact of public funding on state legislative elections using traditional social science inference techniques. They emphasize that until recently, public funding analyses were confined to a comparison of Wisconsin and Minnesota. According to Mayer and Werner:

“As recently as 1998, Malcom E. Jewell and William E. Cassie concluded that since Wisconsin and Minnesota were the only two states that provide for public financing of legislative elections… it [is] difficult to assess the impact of public financing as it could apply to other states. In one of the few studies to even attempt to measure the impact, across states, of public funding on competitiveness, Malbin and Gais concluded that there is no evidence to support the claim that programs combining public funding with spending limits have leveled the playing field, countered the effects of incumbency, and made elections more competitive. However, this study, of necessity, was again limited to a focus on Wisconsin and Minnesota, and stands in contrast to research which concluded that public funding can make a difference.”

The authors argue that although there have been some attempts to apply social science methodology to public financing analyses, those evaluations are incomplete. They write:

“Some reports, especially those produced by advocacy groups that strongly support public funding, overstate the effect of the new law and ignore other factors (such as term limits or redistricting) that have without question shaped outcomes. Others, including the General Accountability Office's evaluation of the Maine and Arizona programs, understate the reforms' impact, in part by making some unusual methodological choices and jettisoning valuable data.”

To measure public financings' affect on electoral competition, Mayer and Werner calculated the percentage of incumbents who (1) faced a major party opponent (contestedness); (2) were in a competitive race,

defined as one in which the winner received less than 60% of the two-party vote (competitiveness); and (3) ran and were reelected to office (reelection rate). They used data from 1990 to 2002 for House elections in Arizona, Hawaii, Maine, Minnesota, and Wisconsin.


Generally, Mayer and Werner make the following conclusions:

“1. There is no question that public funding programs have increased the pool of candidates willing and able to run for state legislative office. This effect is most pronounced for challengers, who were far more likely than incumbents to accept public funding.

2. Public funding appears to have increased the likelihood that an incumbent will have a competitive race.

3. The reelection rate for incumbents in Maine and Arizona dropped significantly in 2002, even after controlling for the pairing effects of redistricting. While some of this can be attributed to public funding, in many cases other factors—scandal, anomalous earlier results, and the destabilizing effects of redistricting—offer plausible explanations for the results. In any event, the GAO's calculation of these reelection rates are simply wrong, since it did not count any incumbents who lost in primary contests.

4. Inadequately funded public funding programs are ineffective. If the public grants are not large enough to permit candidates (especially challengers) a reasonable chance to run a competitive campaign, the programs will have no appreciable effect on election outcomes, competitiveness, or candidate behavior.”

With regard to contestedness, the percentage of contested races in Arizona increased from about 40% in 2000 to nearly 60% in 2002. The authors indicate that they cannot attribute this jump entirely to public financing, which was also in place in 2000, but report that it played a key role. The impact on contestedness in Maine and Hawaii was less clear. The percentage of contested incumbents increased in both states, but the 2002 figures are lower than the highest pre-public financing ones. Minnesota and Wisconsin demonstrated a continuation of patterns that existed throughout the 1990s. (In Minnesota, uncontested House elections are rare and contested rates are almost always higher than 90%. Uncontested races are more common in Wisconsin where just over half of incumbents commonly face a major party opponent.)

The second measure, competitiveness, went up in Arizona, Hawaii, Maine, and Minnesota between 1998 and 2002. In Arizona, the percentage increase went from 22% in 1998 to 47.5% in 2002 and was the most substantial. Of all the states, Minnesota had the highest level of competitiveness with over half of incumbents facing viable major party opponents.

Finally, with the exception of Wisconsin, incumbent reelection rates appeared to decrease in all of the states. (In Arizona, in particular, the rate dropped from 98% in 1998 to 75% in 2002.) But the authors indicate that the changes appear to be within normal limits and are not radically different from the rates that existed in those states in the 1990s.



In their April 2005 paper, “The Impact of Public Election Funding on Women Candidates: Comparative Evidence From State Elections,” political scientists Kenneth Mayer and Timothy Werner examine whether public election funding increases women's ability to run for political office. They assess five states, Arizona, Hawaii, Maine, Minnesota, and Wisconsin, with public financing programs for state legislative races using regression model analysis.


Generally, the authors' indicate that female candidates running for office in the House are substantially more likely than their male counterparts to accept public financing. They found this holds true after controlling for a variety of partisan and candidate variables (e.g., experience). They therefore conclude that gender does have an independent effect on the probability that a candidate will participate in a public financing program. And they suggest that this finding supports those who claim that public funding serves to broaden the pool of candidates willing to run for office.

Based on their research, Mayer and Werner also make the following claims:

1. Democrats are more likely than Republicans to take public funding,

2. candidate experience plays a major role in the decision—both incumbents and Senate candidates who previously served in the House are less likely to take public funding, and

3. candidates who run in states with more generous or full public financing programs are much more likely to participate than those running in states with relatively trivial funding programs.



In his April 2004 paper “Clean Elections: An Evaluation of Public Funding for Maine Legislative Contests,” University of Massachusetts, Amherst, political scientist Raymond La Raja examines the first two clean elections cycles in Maine, 2000 and 2002, to determine the degree to which public financing advances certain policy goals or creates harmful consequences. To do this, he looks at (1) voter choice, (2) contested races, (3) financial parity, (4) interest groups, (5) independent expenditures, and (6) incumbent reelection rates, among others.


Voter Choice. According to La Raja, the total number of candidates increased steadily since Maine passed its Clean Elections Act in 1998, but it is “difficult to determine the degree to which the increase is attributable to the Act, term limits, or some other factor.” In 2000 there were 424 legislative candidates, up 29 from 395 in 1998. In 2002, the number declined by 11 to 413. But none of these figures surpasses the pre-public financing count of 459 candidates in 1996—a year when term limits forced several legislators to leave office. La Raja states that since 1998, the growth in the number of candidates showed a promising trend but he concludes that without additional longitudinal data, he cannot link this pattern directly to public financing.

Open seat candidates and challengers appeared to take advantage of the Clean Elections Act more than incumbents. This finding, according to La Raja, supports one goal of reformers: to help “outsiders” raise money for elections against incumbents. Among the 230 general election candidates who accepted public funds in 2002, 91 were open seat

candidates, 76 were challengers, and 63 were incumbents. Further, participation among all types of candidates increased. Between 2000 and 2002, the proportion of incumbents and challengers taking public funds within these respective categories almost doubled.

Contested Races. La Raja contends that the effect of clean elections on contested legislative races appears to be modest or negligible. But he attributes this to the fact that Maine had a high percentage of contested elections in the ten years before public financing. Even before the implementation of term limits and the Clean Election Act, Maine ranked 11th among 46 states in the percentage of contested legislative races. Between 1984 and 1994, the percentage of two-party contests for the Maine House of Representatives was 77%, compared with the 46-state average of 62%.

Financial Parity: Challengers vs. Incumbents. The average 2002 contribution total for House challengers slightly exceeded that for incumbents. House challengers collected on average $5,041, while incumbents collected $4,782. Senate incumbents collected on average $18,125, while challengers collected $17,589. But this gap diminished after 1996 when the incumbent average was $22,040 and the challenger average was $11,815.

La Raja makes the point that parity in fundraising between incumbents and challengers is highly unusual in American politics. But in Maine, where elections are relatively inexpensive, the financial gap between candidates has never been large. Nevertheless, the reforms appear to have helped reverse earlier financial distributions that gave an advantage to incumbents. Some of these changes may also reflect the effects of term limits, which may reduce incumbent influence in soliciting private contributions.

Interest Groups. La Raja also looked at the influence of special interest groups in elections and whether public funding reduces reliance on private funds. Overall, he found that the amount of private contributions diminished significantly, particularly from business firms. Total candidate receipts from private sources decreased from $3 million in 1998 to $1.6 million in 2000. In 2002, that amount decreased by about 50% to $834,000. The number of candidates who received public funds doubled to 230 between 2000 and 2002. And the Commission on Governmental Ethics and Election Practices distributed $2.1 million in public funds in 2002, which was more than four times as much as in 2000.

Under the Clean Elections Act, contributions to legislative candidates from business, labor, and political parties decreased. Contributions from business firms and PACs declined from $785,000 in 1998, to $418,000 in 2000, and again to $374,000 in 2002. The share from business PACs, in particular, went from 26% in 1998 to 13% in 2002.

Labor union contributions declined slightly, but not as much as business contributions. According to La Raja, this is because they started from a low pre-public financing threshold. In the two election cycles before the Act, labor unions provided 3% of legislative receipts. In 2000 they provided 2%, and in 2002, 1%.

In 2002, parties gave $25,000 to candidates, which represented a significant decline from the pre-public financing level of $362,000.

Independent Expenditures. La Raja found an increase in independent expenditures by non-candidate organizations in races where at least one candidate received public funds. According to the Maine Commission on Ethics and Election Practices, there was minimal independent spending prior to the 2000 elections, but news coverage of campaigns since that time suggests that independent expenditures and issue ads have become more pervasive. In 2002, there was a combined $196,200 in independent expenditures in 66 of 151 House contests (44%) and 21 of 35 Senate contests (60%).

The majority of these independent expenditures were concentrated in competitive electoral districts where the outcome was less than a 20-point margin. In competitive districts, independent spending occurred in 72% of the contests, compared with 18% in noncompetitive contests. Independent spending occurred most frequently in races involving one or more publicly funded candidates, even though it also occurred when only private funding candidates contested an election. It appeared in 77% of contests when both major candidates received public funds, 69% of contests when only one major candidate received public funds, and 58% of races when no major candidate received public funds. According to La Raja, these patterns indicate that the widespread appearance of independent spending cannot be attributed solely to the Clean Elections Act but the Act, which compels candidates to limit spending, appears to spur additional outside spending.

Incumbent Reelection Rates. Finally, incumbent reelection rates since the implementation of the Maine Clean Election Act reflect a mixed picture, going in opposite directions in each chamber. In Senate races, the rates initially declined from 100% in 1998 to 91% in 2000. But in House races they increased slightly from 88% to 93%. But in 2002, they declined from 93% to 85% in the House and increased from 91% to 93% in the Senate. La Raja therefore concludes that the Act did not appear to reduce incumbent reelection rates, but he also indicated that public funds might help improve this measure in the future.