Since public election financing was first implemented during the wave of post-Watergate reforms, the burning question has been, "does it work?" Eval uations of public financing have focused on its primary objectives, which are de signed to address familiar grievances: Elections are too expensive and not com petitive enough. Corporate PACs and other "special interests" contribute dis proportionately to incumbents because they are interested in purchasing influ ence. Candidates must devote so much time to fundraising that little is left for other campaign tasks. Lost in these con siderations, however, is the fact that mandated financial parity changes the strategic environment candidates function in, altering their decision making and potentially changing the nature of elec tions. As fully subsidized elections gain increasing ubiquity in the United States, reformers must decide whether this is a cost worth bearing.' Given its status as the most gener ously funded system of public election finance, Arizona is an ideal location to look to for unintended consequences. Some states, such as Hawaii and Wiscon sin, have long employed public financing schemes with subsidies that partially cover the cost of a campaign, but Arizona's Citizens' Clean Elections sys tem fully funds eligible candidates. Passed as a public ballot initiative in 1998, the Citizens' Clean Elections Act was implemented in time for the 2000 election. As it is in all subsidized pro grams in the post-Buckley regulatory environment, participation in Clean Elec tions is voluntary. However, participating candidates agree to abide by strict spend ing limits. Legislative candidates are provided with a subsidy equal to the spending limit, so long as they are able to demonstrate viability by successfully soliciting at least 210 contributions of exactly five dollars.2 In other words, by accepting public financing, candidates agree to forego any additional sources of finance and to spend only the sum of the grant. Subsidies covering the entire cost of a campaign are themselves unique, but the availability of matching funds marks the Arizona system as a paragon of generos ity. Matching funds allocations are grants given to participating candidates when they are outspent; the extra subsidies mandate financial parity even when not all actors participate in the program. In primary elections, matching funds are triggered when money is spent. In gen eral elections, participating candidates are matched when either their opponent raises money or an independent expendi ture is made.3 In either instance, expen ditures beyond the publicly financed candidate's subsidy amount are matched dollar for dollar to an aggregate limit of three times the original allocation. From the perspective of a non-participating candidate, once the original subsidy threshold is breached, a dollar spent is in effect a dollar contributed to the other side. The importance of matching funds is perhaps best understood by an exam ple of publicly funded candidates in par tially and fully funded states running against an opponent who raises and spends $75,000 (Table 1). In a system of partial funding, in which a candidate re ceives a subsidy equivalent to 45% of the spending limit (as is the case in Wis consin), a publicly funded candidate would still be facing a spending deficit of nearly $64,000.4 In Arizona's Clean Elections system, however, opponent's expenditures would trigger matching funds allocations, preserving funding equality. The elimination of the funding gap is the keystone of Clean Elections's prom ise. Arizona's program diverges from previous attempts at public financing, which have employed only partial subsi dies and have been far from overwhelm ingly successful. Partial subsidies have proven ineffective in slowing spending inflation in New York City municipal elections (Kraus 2006) and Minnesota state campaigns (Schultz 2002), but have shown some promise in Wisconsin (Mayer and Wood 1995). In terms of enhanced competition, 13 years of guber natorial elections from 1983 to 1996 yielded no measurable difference in com petition levels of publicly-financed elec tions compared to those funded by private sources (Malbin and Gais 1998, 136). Two studies have found little change in competitiveness for legislative candidates in Wisconsin (Mayer and Wood 1995) or Minnesota (Jones and Borris 1985). However, a subsequent examination of Minnesota races found significant relationships between public money receipts and challenger vote to tals, indicating a positive relationship between subsidies and competitiveness (Donnay and Ramsden 1995). As to the effectiveness of public money in dimin ishing the role of special interests, an early look at Minnesota found that public funds had helped individual interests (private contributors) to gain an aggre gate dollar advantage over special inter ests (Jones and Borris 1985). Seventeen years later, however, Schultz (2002) found that Minnesota's PACs simply channeled their money through soft money and lobbyists, resulting in little net difference. While it is unclear whether partial subsidies have had any real effect, it is worth noting that the basic candidate experience in those systems remains rela tively unchanged. Candidates in partial funding systems still must raise money from external sources, and given the well-documented funding gap between challengers and incumbents (e.g., Herrn son 2004, 160; Cassie and Breaux 1998), it should not be terribly surprising if the former still find themselves well behind on Election Day. Clean Elections, how ever, not only removes fundraising from the candidate experience with full subsi dies, but the matching funds provision also means that participating candidates can be outspent in only the most excep tional cases. A guarantee of adequate funding, combined with mandated finan cial parity between participating and nonparticipating candidates, makes Clean Elections an entirely different policy. Compared to evaluations of its predeces sors, studies of Clean Elections have been markedly positive, even if the program's youth has precluded broad study. A 2002 General Accounting Office study was cautiously optimistic as to the Michael Miller is a Ph.D. candidate in the department of government at Cornell University. His dissertation examines the manner in which public election financing alters campaign dynamics, citizen engage ment, and political participation. He can be reached at mgm44@cornell.edu. more...