Jump to ContentJump to Main Navigation
Show Summary Details
More options …

The Forum

A Journal of Applied Research in Contemporary Politics

Ed. by Disalvo, Daniel / Stonecash, Jeffrey


IMPACT FACTOR 2018: 0.500
5-year IMPACT FACTOR: 0.623

CiteScore 2018: 0.83

SCImago Journal Rank (SJR) 2018: 0.595
Source Normalized Impact per Paper (SNIP) 2018: 0.631

Online
ISSN
1540-8884
See all formats and pricing
More options …
Volume 15, Issue 2

Issues

Small Donor Empowerment Depends on the Details: Comparing Matching Fund Programs in New York and Los Angeles

Michael J. Malbin / Michael Parrott
Published Online: 2017-07-29 | DOI: https://doi.org/10.1515/for-2017-0015

Abstract

Political campaigns have long been financed by people with well above average incomes, but the balance has tilted dramatically since the Supreme Court’s 2010 decision in Citizens United v. FEC. A number of jurisdictions have been looking to rebalance the incentives through new (or updated) public financing programs. Much of the discussion about their potential effects, however, has been sweepingly generic. But we know that these programs do differ from each other and have good reason to expect that “success” or “failure” will depend both on their goals and the programs’ details. This article focuses on one type of program that has become a model in recent years. Until recently New York City was the only jurisdiction with a multiple matching system explicitly designed to increase the role of small donors.Previous studies noted apparent successes, but it has been difficult to feel comfortable with only one jurisdiction to test. After Los Angeles revised its system in 2013, serious comparisons became possible. This article finds that New York City’s campaign finance matching fund program increased the number, proportional role, and diversity of small donors in city council elections but that the Los Angeles program was substantially less effective. The findings were confirmed through a difference-in-differences procedure that tested each city council over time against state legislative districts representing the same geographical space. A series of explanations relating to the programs’ details were tested, leading us to conclude that the policy details were affecting the results. The results were also different in both cities for mayoral and city council candidates. This suggests alterations may be needed if one were to consider the model for offices with larger constituencies, such as Governor or the US Congress. Finally, the article concludes with a discussion of major arguments for and against increasing small donor participation as a goal for public policy.

Introduction

Political campaigns have long been financed by people with well above average incomes, but the balance has tilted dramatically since the Supreme Court’s 2010 decision in Citizens United v. Federal Election Commission. A number of jurisdictions and advocacy organizations have been looking to rebalance the incentives through new (or updated) public financing programs or tax incentives to enhance the role of small donors. While a remarkable variety of such programs have been introduced or adopted in recent years, 1 one frequent starting point has been New York City’s innovative system that provides $6 in public matching funds for each of the first $175 that a city resident donates to a participating candidate. The program has been found to increase both the number of small donors and the proportion of funds that city council candidates raise from small donors (Malbin et al. 2012). It has also been found to diversify the types of donors that give money to city council candidates in local elections (Genn et al. 2012). Recent experimental research (resulting so far in two articles by the same research team) has raised doubts about whether matching fund programs can increase small donor participation by affecting donors’ behavior directly through nonpartisan informational messages (Green et al. 2015; Schwam-Baird et al. 2016). But as discussed below, the city foresaw the program having more of an indirect effect, shifting the incentives for candidates who would then mobilize the new donors. Whatever the causal mechanism, there can be no doubt about the results.

One problem with holding New York City up as an example, however, has been that it has stood alone until recently as the country’s only jurisdiction with an ongoing program that introduced matching funds at higher than a two-for-one rate explicitly to increase the role of small donors. Without comparative reference points, it has not been possible to sort out what might be idiosyncratic from what might be more general. That has changed. In 2013 and 2015 the city of Los Angeles held its first elections under its own new multiple-matching fund system. With two cities having relatively similar systems to compare, it is now possible to consider how program details and local conditions might affect the results.

A number of practitioners have proposed implementing the New York City matching fund approach. Serious proposals have been developed for local, state, and federal elections. Yet policy makers should not simply assume that variations of the New York City approach will work the same way in every context. While we know that the New York City matching fund program has increased small donor participation and diversity for city council-candidates (Genn et al. 2012; Malbin et al. 2012). We do not know if similar policy incentives are equally effective for candidates for higher office who run to represent larger numbers of constituents, or even for city council candidates in other locales. To better understand when policy incentives will, or will not, be successful at enhancing the role of small donors, we take a closer look at program results in New York and Los Angeles.

Matching Fund Programs and Public Financing

Most of the political science literature on public campaign financing to date has been about the impact of public financing in general, often without making finely tuned differentiations among programs and goals. (See Mayer and Wood 1995; Werner and Mayer 2007; Mayer 2013) In contrast, this article zeroes in on what works and does not for one prominent type of public funding regime with respect to its most distinctive policy goals.

Having said this, it is useful to explain where the programs tested in this article fit within the larger public financing universe. Roughly three dozen states and localities, as well as the federal government, have enacted programs that offer public grants, matching funds, or tax benefits to candidates, political parties, or donors. Most of the programs enacted in the 1970s and 1980s followed one of the two approaches taken for presidential candidates in the Federal Election Campaign Act Amendments of 1974. These included a simple one-for-one matching fund formula for the presidential primaries and a full public financing flat grant for the general election. There were slight variations among the states during this period: some offered two-for-one matching rates (Florida, Rhode Island, New Jersey, and Michigan in primaries) while others gave partial instead of full grants (Hawaii, Minnesota, and Michigan in the general election). Only two of the early programs (in Hawaii and Minnesota) included legislative candidates.

A second set of programs (dating from the 1990s) offered full public financing grants to legislative as well as executive candidates. This approach was adopted by initiative in Maine (1996) and Arizona (1998), and by legislative enactment in Connecticut (2006). All three of the states also offered extra grant money to candidates who were faced either with independent spending or high spending, non-participating opponents. The Supreme Court held the provisions for extra grant money to be unconstitutional in 2011 while upholding the basic provisions for public financing ( Arizona Free Enterprise Club’s Freedom PAC v. Bennett).

All of the state and local matching fund or grant programs to date have required candidates to adhere to a spending limit as a condition for accepting public funds. Mandatory spending limits were held unconstitutional by the Supreme Court in 1976 (Buckley v. Valeo), but the same case upheld voluntary limits as a condition for public financing. Some newer proposals (not yet enacted) have eliminated spending limits, asking instead that participating candidates abide by lower contribution limits than the ones for non-participants. (For a more complete description of state and local public financing programs, see Malbin 2015.)

Most of the programs adopted since the 1970s have purported to be serving a broad range of potential outcomes, such as reducing public corruption or the appearance of impropriety, reducing the undue influence of wealthy donors, increasing electoral competition, or increasing the diversity of candidates and elected officials. New York City’s original description of the goals it sought to achieve through disclosure, contribution limits, and partial public financing resembled these broad claims. (The city adopted its earliest version of public financing in 1988.) The goals articulated by the chair of the New York City Campaign Finance Board (NYCCFB) in its first post-election report were as follows:

Probably the most important of these [goals] was the reduction of the influence of wealthy contributors on electoral campaigns and the attempt to address the perception of the public at large that large contributions to candidates purchased special access to elected officials and special privilege in the conduct of government business…. To restrict the influence of money on electoral campaigns, the Campaign Finance Program sets limits on contributions and expenditures and also imposes strict requirements for disclosure of campaign finances. By providing public matching funds to candidates who agree to observe these limits and requirements and who reach certain threshold levels in fund raising, the Program also intends to “level the playing field” for all candidates, whether or not they have access to substantial wealth (NYCCFB 1990, p. ix).

The matching fund rate in effect for that first election provided $1 in public matching funds for each of the first $1000 that a donor contributed to a qualified candidate. Beginning with this 1990 report, however, the board began asking the city council to increase the matching rate (eventually from $1 to $2 and then $3) while lowering the amount to be matched (from $1000 to $500 and then $250). To justify the recommendation, the board introduced a new goal. The purpose of multiple matching, the board said in 1998, would be to further “the Program’s goal of ‘democratizing’ fundraising by providing financial incentives for candidates to collect smaller contributions from City residents” (NYCCFB 1998, p. 132). That is, the goal was to provide incentives to candidates to mobilize new donors – to democratize the system by enhancing participation and “leveling up”, not just by “leveling down”. The city council was not convinced in the early and mid-1990s, but later adopted an enhanced 4:1 match in pursuit of the same goal. Also important for our purposes was that the Los Angeles City Charter of 2011 included similar language when it adopted multiple matching funds, describing one of the goals as being “to increase the value to candidates of smaller contributions” (Los Angeles 2011). (Similar language has been used in connection with the Seattle and Montgomery County programs, but these are too new to be part of the current research.) Some might question whether small donor participation is a goal worth pursuing. We address this at the end of this article. For now, it is sufficient to say that these two jurisdictions have claimed this to be among the major goals they seek to accomplish through multiple matching funds. This article will test whether they have done so.

Preview of the Findings

To preview the findings: The Los Angeles program seems not to have stimulated small contributions as much as New York’s, and within New York the program has had less of an effect on mayoral and other citywide elections than on races for the city council. Given this variation, we wondered whether these results were due to economic differences between the two cities, political differences among the offices, or program design features. The programs differ in the following potentially important ways:

  • They offer different public matching rates.

  • Public funds make up a higher proportion of a candidate’s spending in New York.

  • The requirements to qualify for public funds are different. New York City Council candidates have to raise significant amounts from donors who live within their council districts. There is no similar requirement for mayoral candidates in either city. There also was no such requirement for city council in Los Angeles in 2013, but there was one for the Los Angeles elections of 2015 (Los Angeles 2015).

To test the two programs’ stated goal of democratizing fundraising by increasing the incentives for candidates to raise small donations, the article explores the relationship for city council and mayoral elections between the above program features and the following outcomes:

  • The number of small donors per candidate (holding constituency size constant);

  • The proportion of candidates’ money coming from small donors;

  • The number and proportional importance of small donors, holding the underlying economic (and other) differences between the cities constant;

  • The proportion of small donors residing within the districts that city council candidates seek to represent; and

  • The representational diversity of the donor pool in both cities for mayoral and city council candidates, as indicated by the racial and income characteristics of each donor’s census block group.

Through this analysis, we find evidence that that not all matching fund programs are created equal. Differences in program design and constituencies affect the extent to which each regime accomplished its stated goals. We begin with an explanation of the salient differences in these policy components.

Background: Matching Rates and Program Design

We have identified three prominent components of the matching fund regimes in New York City and Los Angeles that may affect the impact of the two programs on small donor participation: matching rates, the maximum amount of public funding, and residential requirements for at least a portion of the donors a candidate needs to qualify for matching funds.

Matching Rates

The defining policy component of small donor matching fund programs is the strength of the matching ratio. These have increased over time in both cities. From 1989 to 2013 New York City increased matching rates twice. In the first several elections, the city matched the first $1000 in aggregated contributions from a donor to a single candidate with a one-to-one (1:1) match in public funding. For the 2001 election, the city implemented its first multiple matching rate. Up to $250 dollars in aggregated contributions from a donor to a candidate was matched with public funds at a 4:1 rate. For the 2009 election the matching rate was increased again. In 2009 and 2013, the most recent elections studied here, the city gave candidates $6 in public matching funds for each of the first $175 dollars that a donor contributed. The first $175 dollars from a donor was therefore worth $1225 to the candidate. A contributor could give more, but the city would only match the first $175. (For brief histories of the New York City system, see Ryan 2003; Malbin and Brusoe 2012).

Matching rates have been substantially lower in Los Angeles (Los Angeles City Ethics Commission 2015). The 2013 election was the first in which Los Angeles implemented a multiple match. Before this the city program provided a 1:1 match on the first $250 from a donor to a candidate. For 2013 and 2015, Los Angeles (unlike New York) offered different matching ratios for the two stages of its nonpartisan election process. In the two-stage election process, candidates can win by receiving 50% of the vote or more during the first round of elections. If no candidate does so, the top two compete in a runoff election. In first round elections, candidates may receive $2 for each of the first $250 they receive from a given donor. In runoff elections the matching rate increases to 4:1. Thus, a candidate could receive $500 dollars in matching funds per $250 donor in first round elections and $1000 per $250 donor in runoff elections, but most candidates (including winners) run in only the first round. In other words, New York City’s participating candidates receive nearly twice as much in a primary from a $250 contribution ($250 plus 6 times $175 equals $1300) as do Los Angeles candidates (three times $250 equals $750). Los Angeles candidates who run in a second round receive roughly the same amount from a $250 contribution as New York candidates (five times $250 equals $1250). We expect, therefore, that the matching rates in the New York City, especially in first round elections or primaries, should offer stronger incentives than the matching rates in Los Angeles for candidates to seek contributions from small donors.

Maximum Public Funding as a Proportion of Total Spending

In addition to the variation in matching rates, New York City and Los Angeles differ in the maximum amount of public funding they provide to candidates. The fundraising deficit that candidates face over and above the maximum amount of public funding may also influence whether they pursue a fundraising strategy targeted to smaller donations. Candidates who think they still need to raise large sums of private money after the maximum in public funds may pursue fundraising strategies that fill their campaign coffers faster, assuming the private contribution limit is high enough to let them do so (Miller 2013).

In New York and Los Angeles, competitive candidates often spend very close to the voluntary spending limits. For city council elections, the 2013 spending limit in first round (primary) elections in New York City was $168,000. In Los Angeles it was substantially higher at $480,000. The maximum amount of public funding that each city gave to candidates was, on other hand, nearly identical. Los Angeles first round candidates were eligible for up to $100,000 in matching funds and New York City candidates could receive slightly less at $92,400.

After receiving the maximum amount of public funding, city council candidates in Los Angeles thus could collect up to $380,000 in private finds before reaching the spending ceiling. In contrast, New York City Council candidates who received the maximum in public funds could collect only an additional $75,600 in private money – about one fifth the amount of council candidates in L.A. Interestingly, the comparison between cities is less sharp for mayoral races. The fundraising deficits are much higher for city-wide offices in both cities, and they are not very different in New York than in Los Angeles. The sheer amount of private money a candidate raises should work either to reinforce or counter the previous incentive. Just as New York’s higher matching rates should increase the candidates’ willingness to rely on small donors, so too should the larger role of private funds in Los Angeles’ city council races give the candidates a reason to look for large donors.

Residency Requirements for Donors Who Give Qualifying Contributions

One additional regulation seems important. New York’s city council candidates have long had to raise 75 contributions from individuals within their districts to qualify for public matching funds. In Los Angeles from 1993 through 2011, contributions from all “persons” counted from wherever they might come, including those outside the city or from political committees rather than natural persons. Candidates did not have to raise a minimum number of contributions from within-district to qualify. For the election 2013, Los Angeles contributions were matched only if they came from individual natural persons, but still with no geographic constraints. Starting in 2015, matchable contributions in Los Angeles had to come from city residents, and city council candidates had to raise at least 200 donations of $5 or more from residents of the districts they were running to represent. 2 New York’s mayoral candidates must raise 1000 qualifying contributions from city residents, but with no requirement that the contributions come from a minimum number of city council districts. Los Angeles had no geographic requirements at all for its matching contributions to mayoral candidates in 2013, as was also true for city council candidates. In 2015, Los Angeles began requiring all qualifying contributions for mayoral candidates to come from city residents, but with no further distribution requirements.

In theory, a geographic requirement should do more than just change candidates’ fundraising tactics. It should also help to change how candidates spend their time, build their volunteer organizations, and campaign. We do not have direct evidence about campaign organizations and volunteering. However, we are able to ask whether geographic requirements are associated with a greater reliance overall on small donors, donors within the district, and diversity among donors’ neighborhoods. By considering the differences in results for city council and mayoral candidates in both cities, we are also led to speculate as to why the incentives work differently for different offices, and not just for the two cities. These speculations in turn may help inform whether, or the conditions under which, programs that produce results for one set of offices may require adjustments to scale up to more powerful offices with larger constituencies and greater visibility.

Data and Methodology

It is important to note that we are interested in the role of small donors, not small contributions. A person who gives hundreds of $5 contributions to the same candidate is not a small donor in our understanding. As a result, our unit of analysis is the donor-candidate dyad, in which a donor’s contributions to a candidate are analyzed in the aggregate (see Malbin et al. 2012). We considered small donors to be those who contributed a total of $250 or less to a candidate.

The raw campaign finance data for municipal elections in Los Angeles and New York City were supplied by the Los Angeles City Ethics Commission and New York City Campaign Finance Board. Data for California and New York State Assembly candidates (used later) were provided by the National Institute on Money in State Politics. To ensure that we compared candidates who were running in similar electoral contexts, we limited the analysis to candidates who were at least moderately competitive. We defined these as candidates who received at least half of the winner’s vote total in either a primary or general election (i.e. one-third of the vote in a two-candidate race; less in a multi-candidate field). 3

To calculate the number of small donors, we had to account for an important difference in the disclosure rules among the four jurisdictions. New York City provides detailed information for all contributions and donors. The City of Los Angeles, State of California, and State of New York provide similar information only for donors who give $100 or more. This required us to estimate the number of small donors. Because New York City discloses all contributions, we had an actual count for that city. The disclosure of all contributions also meant we could determine that donors who gave an aggregate of less than $100 to city council candidates in fact gave an average of $33. The other three jurisdictions provide information showing the total amount each candidate received in unitemized contributions. Since the unitemized contributions only includes donations from individuals who gave less than $100 in the aggregate, we assumed that each of the unitemized donors in these jurisdictions gave the same amount as their counterparts in New York City – an average of $33. To arrive at the estimated number of unitemized donors, we simply divided the unitemized dollar total by $33. This is not exact, but the estimate should be close and should not bias the results in favor of any jurisdiction. These estimates were added to the actual number of disclosed donors who gave $101–$250 to arrive at the total number of donors who gave $250 or less.

To determine the proportions of candidates’ total funds for which small donors were responsible, we allocated public funds to the specific donors responsible for them. We did not have to estimate small donor money in this calculation because all of the unitemized money is eligible to be matched. However, because Los Angeles applied different matching rates for the primary and general election in 2013 and 2015, we separated the contributions for each candidate in that city into primary and general election receipts. Contributions to participating candidates were matched at a 2:1 rate in the primary and at a 4:1 rate in the general election.

Studying the demographic diversity of donors, as well as determining whether or not the donors were among a candidate’s prospective constituents, required the use of GIS. To match donors with demographic data, we used donor addresses to place them into the most granular geographic unit with available census data, the census block group. According to the US Census Bureau, block groups “are generally defined to contain between 600 and 3000 people” 4 This is much smaller than a zip code; this many people could live on one city block of densely populated apartment houses or a cluster of blocks with smaller dwelling units. It would be preferable to know the income and racial identity of individual donors, but this is not possible. Therefore, knowing something about the fine-grained neighborhoods in which donors reside will have to stand as a reasonable surrogate. The census block groups for small donors were then compared with citywide census data to examine the representativeness of the donor pool. All census block group demographic data were based on the most recent 5 year estimates from the 2007–2011 American Community Survey (US Census Bureau 2007–2011). To determine whether each contributor was an in-district or out-of-district donor, we utilized GIS to fit donors into New York and Los Angeles districts. Donors who lived in the districts of the candidates to which they contributed were categorized as constituents.

The Number and Importance of Small Donors to City Council Candidates

We use two different techniques to study the role of small donors. The first is a descriptive account of small donors to city council candidates in each of the cities before and after multiple-matching funds. This second is a test described below that compares city council to state assembly candidates from the same city over time.

Descriptive Account

Los Angeles

The analysis begins with a straightforward comparison of city council elections in the years immediately before and after multiple-matching funds. Because Los Angeles holds elections in only half of the city council districts at a time, we paired two sets for the pre- and post-reform comparisons (2009 combined with 2011 and 2013 with 2015). While the same district numbers were up for election in 2009 as in 2013 (and in 2011 as in 2015), redistricting came between 2011 and 2013, thus changing the shapes of each numbered district as well as the populations within them. The results therefore are subject to fewer distortions if we pool them over the 4 year cycles to let us include all of the city’s districts combined, pre- and post-reform. The results are summarized in Table 1. In that table, the first pair of columns presents the average number of small donors per candidate, shown as the number per 100,000 constituents to allow for later comparisons with mayoral candidates and with candidates from New York City. The second pair shows the percentage of total receipts from private sources (excluding self-financing) that came from small donors. The third pair shows the percentage of candidates’ money that results from small donors after the public matching funds generated by each donor has been allocated appropriately. The final columns show the number of participating and non-participating candidates. (This column was not used for New York City because most candidates have participated.) Candidates were considered participating if they said they would accept public funds, whether or not they actually received public money.

Table 1:

Los Angeles City Council Results for Incumbents, Non-Incumbents, and all Participating Candidates, Before and After Matching Fund Changes.

Beginning with the right hand side of the table (“Candidates”), we see an increase in the percentage of candidates participating in the system. This is a crucially important first step, because no voluntary system can accomplish anything unless candidates opt into it. Most viable non-incumbents participated in all years, but the percentage of incumbents went up from fewer than half of the total running (6 of 13) to two-thirds (4 of 6). It is not clear that the new matching fund rates were responsible but, if so, then to that extent the new rates could be called a success.

However, the other columns tell a different story. In the columns toward the left that present the average number of small donors per hundred thousand constituents, we can see that all participating candidates in Los Angeles under the old 1:1 system had an average of 174 small donors per 100,000 constituents. We noted that one reason offered in the Los Angeles City Charter for increasing the matching rate was to increase the importance of small donors. Surprisingly, the actual results were the opposite. All categories of candidates had fewer small donors after the reform than before. The percentage of private money coming from small donors also went down, while the percentage including public funds essentially stayed even. The percentage differences to all candidates and for non-incumbents were not statistically significant. The decline for incumbents was statistically significant, but the matching rate is not likely to have been responsible. (They should have been a force in the opposite direction.) We can only conclude from this descriptive account that Los Angeles’ modest, two-tiered increase in matching fund rates had little to no impact on the number and proportional role of small donors. This will be tested more rigorously below. It is worth noting, however, that the Los Angeles City Ethics Commission in 2015 recommended that the city adopt a 6:1 matching rate for both the primary and general election, which would make it the same as New York’s (Los Angeles City Ethics Commission 2015, p. 16). As of this writing, the Los Angeles City Council had not acted upon this recommendation.

New York City

These impressions are reinforced when we compare Los Angeles to New York City Council elections. Two considerations about New York City seem important for this comparison. First, New York increased the matching fund rates in two steps, starting with the same 1:1 rate as Los Angeles, then 4:1, and then to 6:1. Second, New York has long allowed the maximum amount of public funding to equal more than half of a candidate’s spending limit. While it is not possible from available evidence to sort out the effects of each change separately, the New York comparison suggests that they have a strong effect combined. The experimental studies mentioned earlier showed that matching funds do not stimulate small donors spontaneously (Green et al. 2015; Schwam-Baird et al. 2016). They seem more likely to work indirectly, by giving candidates a stronger incentive to seek out contributions from small donors. This was the causal path suggested by the NYCCFB’s explanatory language quoted above (see Rosenstone and Hansen 1993; Verba et al. 1995 on the importance of donors being asked.) With most candidates in Los Angeles receiving only a 2:1 match (the rate in effect for the primary) the previous tables showed that this was not enough to increase the incentives beyond what they were with a 1:1 match. In contrast, shifting from a 1:1 match to 4:1 in both primary and general elections (and then shifting again to 6:1) did result in a noticeable change for candidates to the New York City Council.

As Table 2 shows, city council candidates in New York in 1997, with a 1:1 matching rate, had an average of 176 small donors per 100,000 participants, almost exactly the same as the 174 for Los Angeles. But the number of small donors in New York increased with a 4:1 match (2001–2005) and again with a 6:1 match (2009–2013). The increases between 1997 and 2009–2013 (6:1 match) were statistically significant for incumbents and for all candidates. The candidates also raised a higher proportion of their money from small donors under both a 4:1 and a 6:1 match. The increase was statistically significant for non-incumbents and for all candidates, whether looking at private money alone or private and public money combined. It was statistically significant for incumbents at both 4:1 and 6:1 with public funds included. This latter point is important because it means that office holders (incumbents) depended on small donors for a majority of their funds (and were reciprocally less dependent on large donors). As was stated in the NYCCFB list of purposes quoted earlier, one goal of the campaign finance program was to alter the influence (or perceptions about the influence) of wealthy donors over government policy as well as in the electoral process.

Table 2:

New York City Council Before and After Matching Fund Changes.

Residential Requirements for Qualifying Contributions

We noted earlier that one additional regulation may be important for explaining the results noted so far. New York’s City Council candidates have long had to raise 75 contributions within their districts to qualify for matching funds. Los Angeles had no such rule until 2015, when the law began requiring at least 200 contributions from city council district residents to qualify. Table 3 shows the apparent effects of the requirement on small contributions within the district.

Table 3:

What Percentage of the Average City Council Candidate’s Itemized Small Donors are Constituents?

The table shows that the percentage of small donors coming from within a constituency in Los Angeles increased significantly between 2013 and 2015. The results are not as strong as New York’s, even when we limit the analysis to New York donors who gave $100 or more. However, New York’s programs had at least three different incentives working in the same direction: higher matching rate, higher maximum level of public money (as a percentage of total spending), and the residential requirement. That Los Angeles showed a significant increase at all between 2013 and 2015 with only one of these incentives suggests the potential importance of residential qualifying rules as an under-appreciated contributor to results. We return to the donor residency requirement in later sections of this article.

Difference in Differences Test

This article so far has compared the Los Angeles and New York City Councils to themselves over time, or to each other. However, there is a potential vulnerability with this analysis. What if the results attributed to multiple matching funds really stem from other underlying differences between the two cities that in turn affect contributions? For example, what if an in-migration of young professionals and the growing importance of the financial sector meant that more people had disposable income to contribute in one city than the other? Fortunately, there is a way to test for this possibility. Instead of directly comparing one city council to the other, we can compare each set of city council elections over time to elections for other offices that represent the same geographical space. This section compares city council elections to elections for districts in the State Assembly that are largely within city boundaries. Theoretically, the procedure we use would also let us compare the city councils to other offices, but we chose State Assemblies because of the similarities in their district sizes, campaign spending, and visibility of the two offices. If the claims we made earlier about the impact of matching funds on the role of small donors were brought about by something other than campaign finance policy (such as the economy or housing values), then we should see the same effects for state assembly candidates as we do for city council. But if the two offices show different results, then it makes more sense to look for an explanation that applies to one office but not the other.

The technique we use is called a difference in differences test. It is a bit like a quasi-experimental research design. (For other works using this approach, see Erikson and Minnite 2009; Hanmer 2009; Miller 2013, 2014; Miller, Tuma, and Woods 2015.) As used here, the procedure measures the difference between small donor participation in both city council and state assembly elections before and after a change in the matching fund rate was implemented for the city council. The policy change is the equivalent of a treatment, with participating city council candidates being the treatment group. Assembly candidates running to represent districts within the city are the control group. For this analysis, the absolute role of small donors is not important. What counts is the relative change across the two offices. If the role of small donors went up or down across the two offices at more or less the same rate, then the reform introduced for only one set of offices cannot be the explanation. We can only be confident that a reform may be the source if there is positive change in the treatment group relative to the control group.

The first results therefore compare changes in small donor fundraising for the Los Angeles City Council and California State Assembly over time. As we have done throughout this article for Los Angeles, adjacent city council elections were combined to cover the whole city. Assembly districts were chosen only if the bulk of the district fell within the city’s borders. We compared:

  • Los Angeles City Council candidates in 2009 and 2011 (1:1 matching) with the California State Assembly candidates of 2010 (no matching funds); and

  • Los Angeles City Council candidates of 2013 and 2015 (two-tiered matching) with the State Assembly candidates of 2014 (no matching).

Similarly, we compared:

  • New York City Council candidates of 1997 (1:1 matching) with New York State Assembly candidates in 2000 (the first year for which we could get electronic records);

  • City Council candidates of 2005 (4:1 matching) with State Assembly candidates of 2006; and

  • City Council candidates of 2013 with State Assembly candidates of 2014.

If higher matching fund rates had a positive impact, controlling for other factors, we should see the gap between the two offices change over time in a way that is significantly more positive for the city councils than for the state assemblies. The results portrayed in the following figures confirm the claim that the reforms were having a positive effect for New York City Council, but not for Los Angeles. As with previous analyses, we estimated the number of donors who gave unitemized contributions using the same approach we detailed above.

In Figures 13 below, the number of small donors declined at essentially the same rate over the years for Los Angeles City Council candidates as for State Assembly candidates who represented districts in the city of Los Angeles. The small donors’ proportion of private money declined less for the city council than state legislature. With matching funds included, the proportional importance of small donors did increase slightly in Los Angeles City Council elections. However, none of the three difference-in-differences tests reached the level of statistical significance.

Avg. Number of Small Donors per Candidate (per 100 K Constituents) in City Council and State Assembly Races over Time.
Figure 1:

Avg. Number of Small Donors per Candidate (per 100 K Constituents) in City Council and State Assembly Races over Time.

Avg. Percent of Total Private Fundraising from Small Donors per Candidate in City Council and State Assembly Races over Time.
Figure 2:

Avg. Percent of Total Private Fundraising from Small Donors per Candidate in City Council and State Assembly Races over Time.

Avg. Percent of Total Private + Public Fundraising from Small Donors per Candidate City Council and State Assembly Races over Time.
Figure 3:

Avg. Percent of Total Private + Public Fundraising from Small Donors per Candidate City Council and State Assembly Races over Time.

In contrast, all of the three lines in Figures 46 – indicating the average number of small donors per candidate, the percent of private money from small donors per candidate, and the percent of private and allocated public money from small donors per candidate – moved in the expected upward direction in New York City Council races, while the lines for the State Assembly declined or stayed flat. Moreover, the difference-in-differences measures relating to the proportional role of small donors (Figures 5 and 6) were statistically significant. The upward shift for city council elections in New York City therefore was not because of conditions that also applied to state candidates from the city. The gap between the two offices grew significantly. We conclude, therefore that the descriptive results we presented earlier remain robust, after controlling for differences between the cities’ economies and between the two sets of offices (city councils and state legislatures).

Avg. Number of Small Donors per Candidate (per 100 K Constituents) in City Council and State Assembly Races over Time.
Figure 4:

Avg. Number of Small Donors per Candidate (per 100 K Constituents) in City Council and State Assembly Races over Time.

Avg. Percent of Total Private Fundraising from Small Donors per Candidate in City Council and State Assembly Races over Time.
Figure 5:

Avg. Percent of Total Private Fundraising from Small Donors per Candidate in City Council and State Assembly Races over Time.

Avg. Percent of Total Private + Public Fundraising from Small Donors per Candidate City Council and State Assembly Races over Time.
Figure 6:

Avg. Percent of Total Private + Public Fundraising from Small Donors per Candidate City Council and State Assembly Races over Time.

City Council and Mayoral Candidates Compared: Different Offices May Need Different Rules

Earlier we discussed the potential importance of residential requirements for qualifying contributions. This is supported when we compare mayoral to city council races in both cities. However, other important differences between the offices also need to be taken into account. Consider the following table, which shows some of the same information we have been seeing so far about small donor participation, but with mayoral elections for both cities included with those for city council. As with previous tables, the money from unitemized Los Angeles donors is included in the two columns at the right, with the number of unitemized Los Angeles donors estimated in the column toward the left, using procedures described earlier.

Table 4 shows that city council candidates in both cities have more small donors (per 100,000 constituents) than mayoral candidates and raise proportionally more of their money from small donors. With about only 10% of their private money coming from small donors, the mayoral candidates look more or less like the average state legislative candidate nationally without public financing. (Small donors accounted for 12% of candidates’ funds in the median state in 2014. See Campaign Finance Institute 2015.) Small donors become relatively more important with matching funds included, but they still represent a quarter or less of the mayoral candidates’ funds.

Table 4:

Los Angeles and New York: City Council and Mayor.

There are a number of plausible explanations, but it is not possible to weigh their relative importance with the information available. Since we expect that several factors contribute, we simply list three.

  • Mayors are more powerful and more visible than individual city council members. Their decisions are more consequential for potential large donors, who therefore feel more of a stake in the election results and are more willing to give.

  • Neither city requires mayoral candidates to raise qualifying contributions from geographically dispersed neighborhoods. This frees mayoral candidates to focus their fundraising on wealthy neighborhoods or downtown business sectors. The matching fund rate is not sufficient to counter this incentive.

  • While the maximum public funds for New York’s mayoral candidates is the same 55% of the spending limit as it is for New York’s city council candidates, the sheer amount of money that mayoral candidates have to raise is much greater. With the contribution limit for mayoral candidates in both cities set at roughly double the limit for city council candidates 5, it is easier for the mayoral candidates to raise what they need from large donors. This also helps shift the incentives toward larger donors in both cities.

There is one caveat to these conclusions about mayors and city councils. Preliminary reports, based on mapping done by the Center for Urban Research of the City University of New York for the New York City Campaign Finance Board, suggests that the incumbent Mayor Bill de Blasio may be raising more of his money in small contributions for the 2017 election than 2013, and it may be coming from a broader swath of neighborhoods (Neuman 2016; NYCCFB 2016). Even if this continues through the election, however, the fact that one mayoral candidate can use the matching fund system as do most city council candidates, does not tell us whether the incentives are strong enough to have a similar impact for mayoral candidates generally.

Donor Diversity in City Council, Mayoral and State Assembly Elections

The evidence so far tells us that the rules for matching funds can affect the number and proportional role of small donors. It does not tell us whether the donors are any different in kind. To get at that question, we geo-coded all disclosed donors’ addresses and placed them into their respective census block groups (CBGs), as described earlier. The results are shown in Table 5. Because Los Angeles and the two state assemblies only give address information for donors who give $100 or more, the information for New York City is presented in separate columns showing all donors who gave $250 or less, and also the results for donors who gave $100–$250. The bottom three rows show the median household incomes, poverty rates, and nonwhite population rates for small-donor CBGs as percentages of the citywide mean for each variable. We did not present information for donors who give more than $250, though our data indicate that they typically come from wealthier and disproportionately white neighborhoods.

Table 5:

Small Donor Diversity: Comparing LA and NYC Census Block Groups (CBGs) with Small Donors in City Council, Mayoral, and State Assembly Elections.

The first row in Table 5 shows that more than half of Los Angeles’ CBGs had $100–$250 city council campaign donors living in them in 2013–2015. The percentage of CBGs went up during the multiple matching fund years, but not to the 73% level for comparable donors in New York. Importantly, New York shows nearly blanket coverage when all donors are included. (Comparable numbers for all donors are not available for Los Angeles.) Coverage is not as broad for mayoral elections as it is for city council in either city, but both sets of offices in both cities show the candidates for city office raising their funds from a much broader set of CBGs than the candidates for either state assembly.

The three remaining rows tell a more interesting story. These show household incomes, poverty rates, and nonwhite residents in the CBGs with small donors – all expressed as percentages of the citywide rate. New York City’s small donor CBGs almost perfectly reflects the city’s diversity as a whole when all donors are included. This strikes us as important, but we have no basis for comparison at this stage. When the analysis is limited to donors who gave $100–$250, the CBGs represented in both cities were less diverse, but the relative distances from the citywide means are interesting and suggestive. In Los Angeles, the CBGs for $100–$250 city council donors were substantially less diverse in the elections with 1:1 matching funds (2009–2011) than in the elections with higher matching rates (2013–2015). During the latter years, the diversity of the Los Angeles’ CBGs approached New York’s for donors who gave $100 or more. In all three rows, the percentages for mayoral donors’ CBGs diverged from the citywide mean more than did the percentages for city council donors. In State Assembly elections, the donors’ CBGs diverged even more from the citywide average for median income, but were roughly the same as mayoral CBGs with respect to poverty and race. All rows for mayoral and state assembly candidates were further away from the citywide means than the CBGs for city council donors. Therefore, if one is interested in greater donor diversity, one should be asking how to make the geography of giving in mayoral races (which do have matching funds), and in state assembly races (which do not), look more like the geographic diversity we see for city council elections with matching funds – especially New York’s.

Summary of Results and Implications

This article has furthered our knowledge about the ways different public matching fund programs work out in practice. Previous research had shown that New York City’s matching fund system increased the number and proportional importance of small donors to city council candidates, and that it furthered the reach of candidates into a broader and more diverse set of small donor CBGs (Genn et al. 2012; Malbin, Brusoe, and Glavin 2012). This article replicated the New York City Council results for an additional election cycle. More importantly, it showed that the Los Angeles program, with lower matching fund rates, did not have a similar impact. These results were confirmed through a difference-in-differences procedure that tested both city councils over time against state legislative districts representing the same geographic space. Additional results – comparing Los Angeles in 2013 with 2015 and then comparing each to New York – suggest that requiring candidates to raise a significant number of qualifying contributions from within the constituency has an independent effect on the level and diversity of small donor participation.

The results were quite different for mayoral candidates than for candidates running for city council. This finding is particularly important if one is considering using New York City as a model for offices with larger constituencies, such as the US House or Senate. Based on the necessarily partial evidence available, we suspect that the following three recommendations would bring the mayoral (and other large constituency) results more in line with those for the city council. The recommendations are based on a system that would use New York City’s six-for-one matching rate for the council (or legislature).

  • Introduce geographic requirements for qualifying contributions for all offices. The presidential public financing system requires qualifying candidates to raise money from 20 states (40% of all states). A proportionally equivalent requirement would be to require mayoral candidates in Los Angeles to raise a threshold number of qualifying contribution from six of the council’s 15 districts. In New York, they would have to qualify in 20 of the city’s 51 council districts.

  • Both cities set higher contribution limits for mayor than for city council. This increases the incentive for mayoral candidates to raise large contributions. Therefore, one could lower the contribution limit for mayor to a level nearer (or equal) that for council candidates. To make up for the candidates’ lost revenues, this recommendation could be coupled with the next.

  • Increase the matching fund rate and public fund maximum for mayoral and other citywide candidates. If the matching rate is to be 6:1 for city council candidates, one might consider 9:1 for mayoral candidates. There might be resistance to increased public spending in some quarters, but the total cost of the campaign finance program in each city was a minuscule portion of each city’s budget. 6

We also need to acknowledge that offices with the same names can be quite different in substance. The Los Angeles city council is smaller and collectively more powerful than New York’s. In addition, term limits apply to both state and city offices in California, while they apply only to city offices in New York. Between term limits and institutional power, the two city councils occupy different spots on political career ladders. About half of the Los Angeles city council has served in the state legislature; few city council members in New York have done the same. Different career paths mean the candidates have different fundraising capacities. Because of this, we would not expect that the Los Angeles results would look quite the same as New York’s even if it adopted the same law. Our difference-in-differences procedure controlled for these effects, but we need to be aware that having the same program in two places might well produce results that look different even if the direction of the causal relationships were broadly the same.

Why Care?

Beyond these matters, we would emphasize one major point. Too often scholars and practitioners ask whether a broad policy approach, such as public financing, “works”. Yet we know that programs may accomplish some of their original goals better than others. In this article we argue that even when one is clear about goals, it is crucial to remain sensitive to the programs’ particularities. The article offers the first systematic comparison of the impact of two matching fund programs on the specific goals that relate to small donor participation. We took this approach because we believe that many of the generalizations we hear in policy debates are, well, too general. To speak about public financing in the abstract, whether negatively or positively, might work for political campaigning but serious program design and analysis call for paying attention to details. And paying attention in turn calls for research. Neither theorizing nor slogans will do the job. We urge scholars to continue this effort in future years, when still newer forms of public matching funds and vouchers will be put through their first election cycles.

Finally, we end by discussing whether policy makers should care about small donor participation and, if so, whether the negative consequences of empowering small donors are likely to outweigh the potential benefits. To consider the potential negative consequences first: the major concern expressed to date has been that small donors are likely to be extremist or polarizing (Bonica 2011; Karpf 2013). However, there is no good evidence to support the claim that small donors are more polarizing than large donors (Malbin 2013). In fact, the one multi-state survey of donors to state elections found small donors to be less extreme than large donors on almost all issues. The major difference between small and large donors, this survey found, was that small donors were slightly less likely to be contributing for material motives and substantially less likely to lobby government officials after the election (Joe et al. 2008).

More recently, the concern about polarization has shifted from being about small donors to being about individual donors per se, who are said to be more extreme in their politics than organizational donors. Those who make this point sometimes will also argue that politics would be healthier if more contributions were filtered through political party organizations instead of going directly from individual donors to candidates (La Raja and Schaffner 2015). This raises a much larger set of questions about political parties – ones too large to address here. Whatever one may think about the parties does not settle whether public policy should seek to increase participation by small donors. After all, public financing or other policies could be used to support contributions from small donors to the parties as well as to candidates.

The questions are: is it legitimate for policy makers to be concerned about a political finance system dominated by wealthy donors (whether those donors give directly to candidates or through the parties) and, if so, is it appropriate and useful for the policy makers to address those concerns by increasing small donor participation? In the past, campaign finance reformers have tried to reduce undue influence by restraining how the wealthy spend their money on politics (“levelling down”). Their efforts have come under criticism for accomplishing less than they promised (Issacharoff and Karlan 1999). Whether one fully agrees with the criticism as originally made, it clearly applies more strongly after Citizen United. One response, already mentioned, is to accept the mix of participants as they currently exist, but filter their influence through the intermediation of political parties. There is a lot one can say in favor of enhancing the role of parties (Vandewalker and Weiner 2015), but this would not get at the central concerns behind small donor participation (Malbin 2017).

The core concerns stem from the belief that simply promoting intermediation does not get at one of the key pieces missing in Madisonian pluralism. Pluralism has many virtues, but it is essentially a system for promoting deliberation and compromise among those who already have resources to bring to the political bargaining table. Those who do not have the money or time, and who do not belong to pre-existing organizations with resources, are less able to participate and less likely to have their needs taken into account when policies are decided. The result is a system nominally based on equal rights but whose natural structure, without effort, will produce unequal participation and influence (Schattschneider 1960; Schlozman et al. 2012). Tools designed to bring more small donors into the system are meant to enlarge the table – to help give more people, and different kinds of people, a meaningful voice. They work by giving those who do have the resources to mobilize – candidates, parties and other donor mobilizers – an incentive to pay attention to those who do not. This concern goes to the heart of successful democratic representation. It should not be dismissed lightly. We owe it to those who try to address the concern to see whether and how their efforts bear fruit.

Acknowledgements

The authors would like to thank the Los Angeles City Ethics Commission, New York City Campaign Finance Board, and National Institute on Money in State Politics for the campaign finance data used in this article. Thanks also go to Brendan Glavin of the Campaign Finance Institute for his help in working with the data. Financial support has been provided by The Democracy Fund, William and Flora Hewlett Foundation, John D. and Catherine T. MacArthur Foundation, Mertz Gilmore Foundation, Rockefeller Brothers Fund, and Smith Richardson Foundation. Finally, we thank the anonymous reviewers for their good suggestions, as well as Michael G. Miller for his comments as a panel discussant that led to one of the main statistical approaches used in the article.

References

Footnotes

  • 1

    The following is a partial list: Los Angeles’ revised matching fund program (analyzed in this article) was put into effect in 2013. The Los Angeles Ethics Commission recommended in 2015 that the city council adopt a 6:1 system more like New York City’s (Los Angeles City Ethics Commission 2015, p. 16). Montgomery County (Maryland) adopted a new system in 2014 that includes as much as a nine-for-one match for candidates who accept a sharply lower contribution limit. Seattle’s voters in 2015 approved a ballot initiative to enact the country’s first voucher program, to be used for partial public financing for local office. South Dakota’s voters approved a broad Anti-Corruption measure in November 2016 that included vouchers, but the legislature overturned the initiative in 2017. (For a list of state and local public financing programs, see Malbin 2015.) 

  • 2

    Council districts in Los Angeles had populations of about 255,000 residents in 2013. New York’s city council districts averaged about 157,000. 

  • 3

    We chose to study candidates with one-third of the winner’s vote total to maintain consistency with previous work on small donor matching fund programs (Malbin and Brusoe 2012). Our results do not differ substantively if we relax the definition to include candidates with one-fifth of the winner’s vote total. Limiting the study to competitive candidates does include most of the money spent in the system. There were 190 candidates for the New York City Council who participated in the public matching fund system in 2017. Half (95) were competitive under our definition, but they accounted for 76% of the private money raised by program participants during the election cycle. With public money included, the 95 competitive candidates accounted for 84% of the money raised by all participating candidates combined. Similar trends occurred in each of the years studied in both Los Angeles and New York, with competitive candidates accounting for most of the money. 

  • 4

    U.S. Census Bureau, “Geographic Terms and Concepts – Block Groups.” Accessed July 27, 2016. Available at http://www.census.gov/geo/reference/gtc/gtc_bg.html. 

  • 5

    In 2013, the limits on contributions from individuals to candidates were as follows:In New York – $2,750 per election cycle (primary and general election combined) for city council and $4,950 for Mayor.In Los Angeles – For City Council: $700 per election (primary or general election; $1,400 combined for both elections); For Mayor: $1,300 per election or $2,600 per election cycle. 

  • 6

    New York City spent $38.2 million for matching funds in 2013. $14 million went to mayoral candidates (NYCCFB 2014, p. 1). Adding 50% to the mayoral match would increase the matching funds by $7 million, bringing the total to approximately $45 million. This is roughly one-half of one-thousandth of New York City’s $81.7 billion budget for fiscal 2016 (Council of the City of New York 2016, p. 1). Similarly, the Los Angeles City Ethics Commission estimates that $4.7 million in matching funds will be disbursed in 2017. It also estimated that the disbursement would have been $7.3 million with a six to one matching rate in both the primary and general election (Los Angeles City Ethics Commission 2015, p. 21). $7.3 million is less than a thousandth of that city’s $8.78 billion budget for fiscal year 2017 (Los Angeles City Administrative Officer 2016, p. 2). 

About the article

Michael J. Malbin

Michael J. Malbin is co-founder and Executive Director of The Campaign Finance Institute, a nonpartisan research institute based in Washington DC. He is also Professor of Political Science at The University at Albany, SUNY.

Michael Parrott

Michael Parrott is Lecturer-in-Discipline within the Quantitative Methods in the Social Sciences program and the department of Political Science at Columbia University. He was an American Political Science Association Congressional Fellow for 2016–2017.


Published Online: 2017-07-29

Published in Print: 2017-07-26


Citation Information: The Forum, Volume 15, Issue 2, Pages 219–250, ISSN (Online) 1540-8884, DOI: https://doi.org/10.1515/for-2017-0015.

Export Citation

©2017 Walter de Gruyter GmbH, Berlin/Boston.Get Permission

Citing Articles

Here you can find all Crossref-listed publications in which this article is cited. If you would like to receive automatic email messages as soon as this article is cited in other publications, simply activate the “Citation Alert” on the top of this page.

[1]

Comments (0)

Please log in or register to comment.
Log in