Printer Friendly

Does Money Always Matter? Campaign Spending in Michigan State Ballot Proposal Elections.


Between 1963 and 2015, a total of 112 statewide proposals have appeared on the Michigan ballot. Of those, 75 were constitutional amendments, with 43 of those proposed by the Michigan Legislature and 31 proposed through petition. Since 1963, 24 referendums have been placed on the ballot, with 14 of those proposed by the Legislature and 10 placed on the ballot through petition. During that same time period, 13 state statute initiatives were placed on the ballot by petition. An additional six initiatives were enacted by the Legislature within a 40-day period after the initiatives qualified for the ballot, so those proposals became law and therefore did not appear on the ballot (State of Michigan Bureau of Elections, Initiatives and Referendums 2015).

Ballot proposal elections are typically considered "low information" races. They appear farther down the ballot and receive less media and public attention than top of the ballot candidate elections. However, ballot proposal elections have had a significant impact on the state of Michigan as a number of important public policies have been enacted through the ballot proposal process. For example, in 1976, about two-thirds of Michigan voters approved Proposal A, also known as the "Bottle Bill." This proposal placed a 10-cent deposit on bottles and cans, and mandated that they be returned for refund or recycled. Michigan has the highest return rate of bottles and cans of any state in the U.S., including those with their own deposit and return programs (Klemanski and Dulio 2017).

In 1992, almost three-fifths of voters in the state approved executive and legislative term limits for Michigan's elected officials. Term limits have had a major impact on how the business of state government is conducted. (Sarbaugh-Thompson and Thompson 2017). In 2004, about three-fifths of Michigan voters approved Proposal 04-2, which made same-sex marriages unconstitutional in the state. This proposal was part of a nation-wide effort to ban same-sex marriages in U.S. states, even as some states were specifically approving same-sex marriages and civil unions. This political and policy debate was not over, however, until 10 years after the Michigan ballot proposal was approved. At that time, the U.S. Supreme Court overturned the law in Michigan (and similar laws in other states) with its Obergfell v. Hodges (576 U.S.__ 2015) decision.

In 2006, another nation-wide movement came to Michigan in the form of Proposal 06-2, the Michigan Civil Rights Initiative (MCRI). This initiative sought to end affirmative action hiring and admissions programs in the state's public universities, state and local government, and other publicly-funded state institutions such as school districts, county government, and community colleges. Even before reaching the ballot, petition signature gatherers for the proposal were accused of fraud (in selected areas, signature gatherers falsely claimed that the proposal supported affirmative action). A lawsuit was brought against the MCRI, and a federal judge in Detroit determined that some voter fraud had occurred, but allowed the initiative process to continue. In the end, 58% of Michigan voters supported the proposal. A series of legal challenges followed that vote. First, a federal district court initially upheld the constitutionality of the MCRI, but a federal court of appeals later overturned it. Michigan's Attorney General Bill Schuette appealed the decision to the U.S. Supreme Court. In Schuette v. Coalition to Defend Affirmative Action (572 U.S. __ 2014), the Court upheld the constitutionality of the Michigan Civil Rights Initiative.

The importance of state policy-making through ballot proposals, as the above examples illustrate, can be significant to the state and its residents. With an appreciation for this process, this article examines the role of campaign spending in Michigan statewide ballot proposal elections between 1978 and 2015. The time frame selected provides for a useful analysis because it covers 86 ballot proposals over a period of 37 years, and includes every type of ballot proposal question that the state of Michigan allows. Moreover, this analysis allows us to identify a number of possible factors that could present themselves under different electoral conditions (e.g., timing of elections; controversial vs. non-controversial ballot questions). We also identify a number of trends in Michigan's ballot proposals over time with respect to spending and ballots cast. We begin, however, by examining the legal environment governing ballot proposal elections.


The Progressive era of the late 19rh and early 20th century has left a longstanding impact on U.S. politics in several ways. For example, one component of Progressive-era reforms was the introduction of "direct democracy" provisions by a number of states. These included citizen-initiated statewide ballot proposals (or initiatives), the use of referendums, and recall elections. All of these direct democracy reforms were intended to allow citizens a more direct influence on politics and public policy-making. While not all states have provided for these reforms, 24 states have some form of a citizens' initiative process. Moreover, Michigan is one of nine states that allow for all three forms of direct democracy (National Conference for State Legislatures).

While ballot proposal elections are considered different than partisan, candidate-centered elections, the possible influence of money in elections has had a long history. In the Watergate era of the early 1970s, Congress enacted the Federal Election Campaign Act (FECA), with Amendments. This legislation set limits on direct contributions as well as expanded disclosure and reporting requirements. The FECA Amendments in 1974 (Pub. L. 93-443) also created the Federal Election Commission, which was established to oversee federal elections and campaign finance activities.

Michigan followed federal action by creating PA 388 of 1976, the Michigan Campaign Finance Act (MFCA), which provided for contribution limits to candidates and disclosure requirements for candidate and ballot question committees. As with candidates' campaign committees, ballot question committees must file a series of regular campaign statements during each election cycle. These statements must itemize the committee's contributions, expenditures, and debts. For all contributions, regardless of the amount, the committee must report the amount and date of the contribution, along with the name and address of the donor.

Ballot measures in Michigan have several different forms: citizen-initiated state statute proposals (initiatives); constitutional amendments (which can be proposed by the Legislature or placed on the ballot via citizen petition); referendums (which also can be placed on the ballot by the Legislature or through petition); and automatic ballot referrals that call for a new Constitutional Convention and are placed on the ballot every 16 years.

One possible aspect of campaign spending in ballot proposals is that there could be differences among the types of proposals, if only because some ballot questions (e.g., citizen-initiated constitutional amendments) require more petition signatures in order for a question to appear on the ballot. Legislative constitutional amendments (Michigan Constitution, Article XII, Section 1) and referendums (Article IV, Section 34), are placed on the ballot by legislative action. Finally, every 16 years, a ballot question calling for a constitutional convention is automatically (Article XII, Section 3) placed on the ballot without needing petition signatures. There are likely more costs associated with citizen-initiated proposals, since fees are often paid to petition gatherers or petition-gathering firms are hired to collect signatures. Since some ballot proposals require more signatures, more time and resources must be expended in order to obtain the higher number of signatures in order to qualify for the ballot.

For all types of elections, the MCFA regulates campaign contributions and provides for disclosure requirements for state and local elections in Michigan. The law makes a clear distinction between candidate committees and ballot question committees. For example, the relatively strict regulations (i.e., on source and amounts) regarding direct contributions to candidate committees do not exist for ballot question committees. The more relaxed standards for ballot questions is consistent with the view that the potential corrupting influence of contributions to an individual candidate or elected official running for office is not present with a ballot question. Moreover, spending on issue advocacy in the eyes of both federal and state law has tended to have fewer restrictions than money spent on express advocacy (i.e., direct advocating to elect or defeat a candidate).

The MCFA defines a ballot question committee:
as one acting in support of, or in opposition to, the qualification,
passage, or defeat of a ballot question but that does not receive
contributions or make expenditures or contributions for the purpose of
influencing or attempting to influence the action of the voters for or
against the nomination or election of a candidate (MCFA, PA 388 of
1976, 169.202).

Virtually any (legal) source may make a direct contribution to a ballot question committee--in any amount. Allowable sources include corporations and labor unions, as well as individuals. This regulation is unlike those for contributions to candidate committees in Michigan (and federally), where corporations and labor unions are required to establish legally separate organizations (political action committees) for the purposes of raising funds and making direct contributions to candidates. According to Michigan law, "independent expenditures" are expenditures made to influence a campaign but are not made under the direction of, or coordination with, another person or committee; they are not contributions given directly to a candidate or committee. Because of this independent status, independent expenditures can be spent in any amount, although any expenditures over $100 in a calendar year must be disclosed and reported to the Michigan Secretary of State's office.

As such, ballot proposal elections are different from candidate elections in one major way--campaign spending. Without the same kind of campaign finance limits that are required for candidate campaigns, it is possible for statewide ballot proposal elections to see many millions of dollars spent during a campaign. At the same time, ballot proposals that are not controversial or cover mundane topics might see little or no spending--by one side or even both sides. This wide variation in spending prompts research questions regarding the types of ballot questions that might see more versus less spending, as well as those elections where there are substantial spending differences between those who favor a proposal and those who are opposed.


Over the past several decades, scholars have taken a keen interest in the role of ballot proposals in making state public policy, as well as the role of money in ballot proposal elections. Previous literature has noted the policy areas most commonly on state ballots, which include tax policy, government reform measures, social and moral questions, and environmental policies (Magleby 1984; Tolbert, et al. 2001; Bowler and Donovan 2018). Other research has examined how different states impose different rules about how proposals qualify for the ballot (McCuan et al. 1998; Banducci 1998).

Scholars also have sought to determine the influence of ballot proposals on civic engagement and voter turnout among voters (Tolbert and Bowen 2008). Some research has argued that use of direct democracy policy-making has allowed electoral majorities to undermine the interests and rights of racial and ethnic minorities (Hajnal, et al. 2002). In a related vein, Bowler and Donovan (2018) and Qvortrup (2001) have explored the role of the courts in invalidating a variety of voter-approved ballot proposals, such as the same-sex marriage bans in states that ultimately were ruled unconstitutional by the U.S. Supreme Court.

Regarding campaign finance and ballot proposals, much of the scholarly attention has focused on the experiences in California (see Magleby 1984; Owens and Wade 1986; Matusaka 1993; Stratmann 2005). This interest has emerged because California is among the most active states in using ballot proposal questions to make public policy. For example, California had the most initiatives of all states between 2011 and 2015 (Bowler and Donovan, 2018)

An early research concern about the role of money in ballot elections was whether wealthier interests had a built-in advantage in helping to influence or determine election outcomes. Scholars such as Garrett (1999), for example, examined how spending by wealthy interests funded certain proposal attempts, and thus allowed greater access to the ballot by overcoming the relatively high costs of obtaining enough valid petition signatures. Moreover, research by Magleby (1984) and Lowenstein (1982) found that corporate interests were able to successfully oppose citizen-initiated proposals by substantially outspending the citizen groups that were supporting those proposals. In most of these early studies, the findings suggested that spending against a proposal was more effective than spending in favor of a proposal. These results have led some scholars to speculate that there was a built-in advantage to a "no" vote on ballot proposals, in part because voters would vote for the status quo rather than for any change about which they were unclear or uncertain (Owens and Wade 1986; Stratmann 2006).

However, Bowler and Donovan (1998) found that campaign expenditures had little effect on voter opinion regarding ballot measures. Gerber (1999) found little evidence that corporate or wealthy interest groups were able to buy policy outcomes by spending larger amounts of money. Broder (2000) countered that interest groups who spend significantly more will be more likely to win, regardless of whether that spending is in support or in opposition. In a later study of spending on ballot proposals, Stratmann (2005) made a strong case that both spending in favor of and in opposition to a ballot question has statistically significant effects on election outcomes. He examined county-level spending for and against statewide ballot questions, and found that increases in spending for either side can increase the total vote for that side. However, fairly large additional amounts of money must he spent in order to achieve relatively small vote increases of between about 0.5 percent and one percent. Given the lack of consensus on campaign spending effects in ballot proposals elections, we hope to add to the knowledge base of campaign spending effects by examining the state of Michigan's recent history of ballot proposals and spending patterns. We trace all ballot proposals from 1978 through 2015 (there were no proposals on the Michigan ballot in either 2016 or 2017).


Compared to California, Michigan is more modest in the number of ballot proposals voters see each election. At the same time, Michigan has had a steady stream of ballot proposals since 2000, including some that were high-impact and controversial. Moreover, campaign spending on state ballot proposals has increased over the past 20 years. This approach offers some interesting research questions as well. At first blush, ballot proposal campaigns appear to be different enough from candidate campaigns to warrant further exploration--after all, there is no incumbency advantage or name recognition of candidates, no party label, and little to no party mobilization. Money potentially has a meaningful impact in ballot proposal campaigns, especially as a means for increasing voter knowledge and interest in what are regarded generally as "low information" elections.

There are two main reasons we begin this study in 1978. First, the 1978 election was the first election after enactment of the Michigan Campaign Finance Act, which changed state law regarding campaign spending and disclosure. Second, the Michigan Secretary of State has compiled campaign spending amounts for all ballot proposals going back to 1978. We recognize that some changes have occurred in how campaign spending is disclosed and reported in recent years, but we believe this time period provides for the greatest consistency and reliability of reported campaign spending in Michigan on statewide ballot proposals.

While the bulk of our interest centers on the role of money in ballot elections, there are other questions related to spending that are worth investigating. For example, what are the vote totals for ballot proposals (as compared to other races) over time? What percentages of ballot roll-off (i.e., when voters do not indicate a choice in lower ballot elections) exist in these elections, and how do those compare with ballot roll-off in other statewide races? Are there any differences in roll-off across different types of elections?

In general, we consider what role money has played in ballot proposal elections when much of what occurs in candidate-centered elections does not exist (e.g., incumbency advantages in fund-raising and name recognition, the importance of spending for challengers, spending by parties for voter mobilization efforts). However, while factors such as name recognition and party label are not present, other factors--such as whether a proposal can be considered "controversial" or "non-controversial"--might have an effect on how much money is raised and spent on these campaigns.

Finally, we examine what the overall spending amounts have been since 1978, and whether there are any patterns of spending (e.g., toward opposing proposals or in support). Part of our interest in campaign spending also sought to identify the cost per vote figures for elections over time, and whether there were any identifiable or explanatory patterns in terms of cost per vote. To help answer these questions, we collected information and statistics from the sources summarized below.


We used information from the Michigan Secretary of State's website to determine the number of statewide ballot proposals in a given election and the amount spent on each ballot campaign. We also determined what type of ballot proposal (i.e., citizen-initiated state statute, legislative-referred or citizen-initiated constitutional amendment, veto referendum) each was considered. Campaign spending amounts were determined using the total expenditures reported by registered committees that supported or opposed each ballot proposal. All committees in support of a proposal were totaled ("Support Spending"), and opposition spending was totaled similarly ("Opposition Spending"). These figures were taken from "Expenditures by Ballot Question Committees" reports that were provided by the Secretary of State through its Campaign Finance Disclosure website. Ballot questions were categorized by type using the information provided by the Secretary of State.

The number of votes received for each proposal was taken from the relevant pages on the secretary of state's website ("Previous Election Information"); this figure was divided by population estimates taken from the United States Census reports to determine the percentage of the population that voted for each ballot proposal. Total votes for each proposal in each decade were averaged out to determine average votes in that decade except in the case of the 1970s where only one election (1978) was available for analysis (State of Michigan, Secretary of State, Report Description Expenditures, various years). Total votes were also used to determine ballot roll-off compared to presidential, gubernatorial and/or US Senate races that were at the top of the ballot on that given election. These figures, combined with expenditures by committee, were used to determine overall cost per voter. Cost per vote for each side (i.e., support or oppose) also was calculated.

The controversial status of each ballot was determined by the authors who independently rated the questions as controversial or non-controversial. In order to provide more rigor to our analysis, the authors rated each proposal before we investigated how close or competitive the final outcome was for each proposal. In this way, we were attempting to make predictions and test hypotheses rather than simply review each election outcome. While the authors independently defined "controversial" and "non-controversial," we were very consistent in our respective views on this question. For example, one decision rule was that all tax proposal questions were in the controversial category. In addition, issues that struck core "values" of voters were also categorized as controversial. Examples of core values-related proposals would include "morality" questions such as those involving same-sex marriage, public funding for abortions, and affirmative action. Non-controversial proposals included questions involving the general administration of government and protection of the state's natural resources. Ratings by each of the individual authors were compared and a majority decision reached if there was no consensus among the three raters. Unanimity among all three raters was achieved in a large majority of cases (80%).


During the time investigated, there were 86 proposals appear on the ballot for voters to consider. The vast majority of these (70.9%) have been questions related to the state constitution ("constitutional amendments") while 19.8% have been matters of statute ("statutory initiatives") and only 9.3% have been veto referendums. A recent example of a veto referendum is Proposal 1 from 2012. This proposal asked voters if Public Act 4 of 2011, the Emergency Manager statute, should be rejected (and which voters successfully vetoed).

Figure 1 shows how the proportions of these different types of propositions have changed over time. Constitutional amendments have always been the most prevalent type, but more interesting is that veto referendums have seen their share rise through the decades. Statutory initiatives peaked in the 1990s and have decreased since then. Those looking to change policy through ballot proposals may prefer constitutional amendments over initiatives because even though it is more difficult to secure more petition signatures, a constitutional amendment is much more difficult for future voters or legislatures to change.

In Michigan, initiatives and referendums can appear only on general election ballots--so for example, with no general election scheduled in 2015, only a legislatively-referred question such as the "roads tax" ballot proposal was possible. In this period of study, the November general election was by far the most frequent election for ballot propositions with nearly 92% of all proposals appearing then; the August election (traditionally a primary for congressional seats and other offices) only saw 3.5% of all proposals (see Table 1).

The timing of ballot proposals can involve strategic decision-making on the part of ballot proposal advocates. For example, a ballot proposal likely to favor more liberal interests would probably benefit from being on the ballot during a presidential election year (since these types of elections typically see a higher percentage of Democratic voters compared to other election dates). Ballot proposals occurring in March, May, or August are likely to have a very different profile of voters than one in a November presidential election given the far lower voter turnout during those times. It is worth noting also that ballot proposals can be used as a policy-making tactic when supporters do not believe they will be successful in obtaining favorable policy through the traditional policy-making process--through the legislature and governor. For example, most of the ballot proposals in Michigan since 2011 have been supported by interests closer to the Democratic Party because the state's government has been dominated by Republicans since that time.

Votes Cast and Roll-off for Ballot Proposals

Over time, Michigan has seen both more raw votes and an increasing percentage of its population cast ballots in proposal elections. Table 2 shows an average of nearly 800,000 more people voted on proposals appearing on the ballot between 2010 and 2015 compared to those on the ballot in 1978. Even as the state grew in population (save for a slight decrease between 2000 and 2010, according to the U.S. Census), a higher proportion of Michiganders cast ballots; on average, between 1978 and the early 2010s, Michigan saw over a 5.5% increase in ballots cast.

The increases since the 1970s in the average percent of the Michigan population voting on proposals is arguably tied to increased voter interest in ballot proposals, so we used another measure in an effort to confirm voter interest. One common assumption is that many voters only vote in "top of the ticket" races and do not continue completely down the ballot to where ballot proposals and other "down ballot" contests are located. This behavior is known as ballot roll-off--when voters vote for major offices at the top of a ballot, but then skip the lower-level races. Often this occurs because voters are not as informed about these lower-level races. Ballot roll-off can sometimes be sizeable--earlier research has estimated ballot proposal roll-off in other states at between 20% and 25% (Nichols 1998). In the case of Michigan's ballot proposals, the roll-off has been surprisingly low. Over the period of this study, the average roll-off in all ballot propositions was less than 6.5%. A couple of caveats apply here, however. In some instances (N=7), the ballot proposition was the highest office on the ballot, meaning that no roll-off statistic was possible to calculate. When these cases are removed, roll-off still rises to only 7.04%. This figure is much smaller than state Supreme Court roll-off, which has been estimated to be 25% - 30% in recent Michigan Supreme Court elections (Klemanski 2013).

As another point of comparison, we can examine roll-off from the highest office to a U.S. Senate race in Michigan. There were 56 instances in our dataset where there was a statewide ballot proposition, a U.S. Senate race and another race above those (i.e., governor or president). The average roll-off from the highest office to the U.S. Senate race was 1.65%. The average roll-off from the highest office to the ballot proposal was 6.12% in these instances.

Table 3 shows that since the start of the 1980s, there has been a steady drop in the percent of voters who do not mark their ballots on statewide proposals--falling from, on average, nearly 9.4% in the 1980s to about 3.5% in the 2010s. In other words, the general trend has been for more voters to make sure they vote on ballot proposals.

We also sought to determine whether there were distinct differences in average number of votes cast in support or in opposition to ballot proposal elections over time. There is little difference in the average number of votes cast in proposals that pass (3,247,713.43) and those that fail (3,300,557.57), although it is interesting that measures that fail see, on average, a larger number of total votes cast. There is, however, also a small difference in the roll-off--raw or percent--between those that pass and those that fail--but in the opposite direction. The average roll-off for passing proposals was 211,707.37 while the percent roll-off was 6.00%, compared to an average roll-off of 270,548.71 and 7.86% in proposals that failed. There may be something strategically compelling here for those who are involved in the effort to pass or defeat a ballot proposition. While proposals that fail typically have a few more voters making a choice in these down-ballot races, there is a slightly larger number of people who do not complete their ballot. In close contests, roll-off may be the difference between a measure passing and failing.

Ballot Proposal Spending Patterns

Ballot proposals have seen a great deal of change with respect to the resources that are spent to wage these campaigns. In terms of dollars spent in support of a proposal, dollars spent in opposition to a proposal, and total dollars spent during a ballot proposal campaign, each has seen a dramatic increase over time. Figure 2 shows the averages for each of these variables during the period of study. The average total spent in support of a ballot proposal has increased from just over $83,000 in 1978 to over $8.1 million in the 2010s. Interestingly, in the 1978 proposals over $100,000 was spent in opposition--more than was spent in support. This pattern also held through the 1980s and 1990s--on average, only about $267,000 was spent in support of ballot proposals but nearly $375,000 was spent in opposition during the 1980s while on average, $932,000 was spent in support compared to over $1.33 million in opposition during the 1990s. Some of these figures must be tempered by the fact that there are ballot questions from the 2000s and 2010s where little or no money is spent--just as in the 1980s and 1990s. Moreover, some of the average increase in recent elections is due to a few high-spending campaigns. This trend of increased average spending may likely continue as recent election cycles have seen more ballot proposal advocates seek a deep-pocketed benefactor to fund their efforts.

Regarding the balance or imbalance in spending for opposition to or support for an issue, Michigan has seen a shift in spending in support of a proposal compared to spending in opposition. In the 1980s and 1990s, on average, spending in opposition outpaces support. In the 1980s an average of $267,000 was spent in support of all proposals while $373,000 was spent against; in other words, 28% more money was spent in opposition compared to in support of proposals. This gap widened slightly to just over 30% in the 1990s with $1.33 million spent, on average, against while only about $932,000 was spent in support. In the 2000s, this pattern changes to support spending outpacing that in opposition to ballot proposals. The average spending in opposition to a ballot proposal was over $1.54 million while spending in support averaged over 1.4 times more at $3.75 million. This pattern continued in the 2010s, although with a smaller percentage gap between them, with over $5.7 million spent on average in opposition to proposals and over $8 million spent in support. Figure 2 illustrates the tremendous increase in total ballot proposal spending in the decade of the 2010s.

Over the time frame of this study, there were more dollars, on average, spent in support of proposals that were adopted ($2,691,367.96) compared to those that failed ($1,522,536.97). The same is true for the average dollars spent in opposition to those that win ($2,036,267.33) and those that lose ($881,256.97) as well as total spending ($4,727,635.29) for those that are adopted to $2,403,793.95 for those that are not.

These figures include the instances where $0 were spent either in support or opposition to a ballot proposal. There are more of these instances than one might think. Over the time period under study, 19 ballot proposals saw $0 spent in support while 39 had $0 spent in opposition. Considering proposals that were on the ballot before 2000, nearly 25% had $0 spent in support and nearly 51% had $0 spent in opposition. This trend changes dramatically after 2000 when only 17% had $0 spent in support while only 34-5% had $0 spent in opposition. In this regard, money has taken on a greater importance in ballot proposal campaigns simply because it has a presence in a greater share of these decisions in recent years compared to earlier ones.

On those proposals where there was $0 spent in support of the proposal (N=19), 4 managed to pass in spite of no reported spending, while 15 failed; 15 were constitutional amendments (78.9%), 1 was a statutory initiative (5.3%), and 3 were veto referendums (15.8%). Where there was $0 spent in opposition, 23 passed--as one might expect--but 16 others still failed. Of these, 29 (74.4%) were constitutional amendments, 9 were statutory initiatives (23.1%), and 1 was a veto referendum (2.6%). Table 5 shows the differences among the types of ballot proposals that saw $0 spending in support or opposition. A very high percentage--about three-quarters--of all constitutional amendment proposals had $0 spent in support or in opposition.

In a further effort to determine whether money matters in ballot proposal elections, we also determined how many elections there were when the higher spending side won or lost. In ballot proposal elections the "winning" side is not as clear as it is in a candidate election where the winning side is simply the candidate that receives the most votes. In ballot proposals, "winning" can be on either side. Indeed, the side that wants to see the proposal fail wins when the proposal is defeated, or, more clearly, the side that wants to see the proposal approved wins when it is adopted. In short, the "winning" side achieves its desired which could either be adoption or failure. As with several of the campaign spending measures we calculated, the results since 1978 were quite mixed. When looking at the results by decade, we found that in the 1978 election, and then again in the 1990s and 2010s, the winning side spent the most money in a majority of the elections. However, in the 1980s and in the decade of the 2000s, the losing side spent the most money in a majority of those elections. Table 6 provides a summary of these elections by decade.

As Table 6 illustrates, there is no clear pattern or suggestion that money is the deciding factor in determining a ballot proposal election outcome. In the decade of the 1980s, the number of elections in which the winning side spent the most money was about the same as when the side with the most money lost the election. The same was true for the decade of the 2000s. In the 1990s, the winning side did spend the most money in a majority of elections, but again, in only a small majority (58%). In the most recent decade, the effect of spending does seem to be clearer - in a solid two-thirds of those elections, the winning side spent the most money. Since we've found both large increases in spending and increased success by bigger spenders in the decade of the 2010s, these findings suggest that money is becoming more effective, perhaps with sufficiently large enough thresholds of spending.

There has been a dramatic increase in other measures of money's role in ballot proposal campaigns in Michigan. As seen in Figure 3, the average dollar amount spent to deliver a "yes" vote on a ballot proposal in Michigan has risen from literally pennies--7 cents in 1978--to nearly $8 per vote in the 2010s. This increase in cost per vote shown in Figure 3 parallels the increases in total spending noted in Figure 2. The average dollar amount spent to yield a successful "no" vote has also increased, but has gone up less dramatically--from 11 cents to nearly $3.20.

To further explore ballot proposal campaign spending, Table 7 presents the averages of the three spending variables (cost per vote, cost per no vote, and cost per yes per capita) by type of proposal. Not surprisingly, higher costs are associated with constitutional amendments, in part due to the higher thresholds for petition signatures required for these proposals to qualify for the ballot. It is also possible that constitutional amendment proposals are typically more controversial or important to voters. After all, they would change the Michigan Constitution and are much more difficult to change--compared to a state statute--once the amendment has been approved by voters. It is also possible that the higher costs per "no" vote (compared to a cost per vote per capita and a cost per "yes" vote) for veto referendums are due to the very nature of a veto referendum--such a ballot proposal is making every effort to promote a "no" vote on that ballot question.

Controversial Vs. Noncontroversial Ballot Proposals

As we noted in the section above devoted to data and variables, we examined whether a proposal was controversial or not. Our analysis of the 86 total proposals between 1978 and 2015 uncovered 40 controversial proposals and 46 noncontroversial proposals. Controversial ballot proposals were those that dealt with taxes, "moral" issues (e.g., same-sex marriage, public funding for abortions, and affirmative action), or similar topics that could be considered "hot-button" while noncontroversial proposals were those that dealt with government administration or state parks and recreation or natural resources.

Here we examine the differences between those proposals deemed controversial and those not in terms of whether a proposal wins or loses, the number of votes cast, and the amount of money spent. We expect controversial proposals to have less success, but more votes cast, and more money spent than noncontroversial proposals--all due to the increased competition surrounding controversial proposals. In terms of success at the ballot box--the major measure that matters in elections--noncontroversial proposals, on average, get more support from voters (53.33% 'yes' votes) than controversial ones (44.55% 'yes' votes); it follows that there is a difference between them on 'no' votes as well (controversial = 55.45% vs. noncontroversial = 46.67%). Whether a proposal is controversial or not may also impact how many votes are cast. For example, there have been, on average, more than half a million more votes cast in ballot proposal questions deemed controversial (3,622,800.65) than non-controversial ones (3,015,984.02). This difference is also expected, as it is likely controversial proposals will generate more interest, attention, voter mobilization, and spending on advertising than those considered non-controversial.

In terms of money spent, we also see a number of differences over time. As expected, on average, controversial proposals during our period of study saw more spending than noncontroversial proposals. The average total spent in support of controversial measures was much greater than measures that were deemed non-controversial--$4,035,036.50 to $557,404.50. These differences are to be expected since more work needs to be done by those looking to pass something controversial comparted to something that is not. Similarly, on average, there is more money spent in opposition to controversial proposals ($2,539,356.86) than noncontroversial ($644,659.17). The same pattern holds for average total spending (total spent in support plus total spent in opposition): $6,574,393.38 on controversial proposals to $1,202,063.67 for non-controversial proposals.

In the earlier section on spending, we noted that some proposals saw $0 spent on one side or the other. Of the 39 proposal questions where $0 were spent in opposition, 12 (30.8%) of those were controversial while 27 (69.2%) were noncontroversial. Again, this result should not be surprising, as there is less need to spend money against a policy question that is not controversial.

The controversial nature of some ballot proposals also has an impact on the dollars spent to secure a 'yes' vote, since it would increase the cost per vote. Large differences were also uncovered between the average cost per "yes" vote in controversial proposals ($3.2618) compared to those judged as non-controversial ($0.3769). As noted above, it takes more electioneering to convince voters to adopt something controversial than something that is not. Similarly, differences were found on cost per "no" vote on controversial ($1,474) vs. non-controversial ($0.4481) proposals.


We began this investigation seeking to determine the role that campaign spending may have played in ballot proposals elections over the past several decades. We found some interesting--and sometimes surprising--results based on our research of Michigan's ballot proposals. First, spending on ballot proposals has seen substantial increases in campaign spending over time, especially since 2000. While some of the average increases were fueled by a few notable elections, we also found that more spending does not guarantee victory. For example, over $33 million was spent by the losing side on the 2012 "The People Should Decide" ballot proposal, which would have required state and local voter approval prior to construction of a second bridge between Detroit, MI and Windsor, Ontario. Over $15 million was spent on the 2015 roads tax ballot proposal in Michigan, which also failed. By comparison, average spending on ballot proposals had not even reached $2 million in the 1970s and 1980s (barely breaking the $2 million mark in the 1990s). Average spending then jumped to over $5 million in the 2000s, and reached almost $14 million in the decade of the 2010s.

At the same time, a large percentage of ballot proposal sides (in support or in opposition) spend little or no money at all during an election. A high percentage--about 75% of both those in support and those opposed--reported spending $0 on all of the constitutional amendment proposals since 1978. It does appear that more money is spent on the other types of proposals --statutory initiatives and veto referendums. This finding should not be a surprise, since citizen-initiated statutory initiatives may need a fair amount of spending just to gather enough valid petition signatures.

More money is spent on controversial versus non-controversial ballot proposals. The controversial nature of a ballot proposal would suggest that there is more inherent voter interest, but also more interest on the part of individuals or groups who make contributions to these campaigns. It is also not surprising that such increased interest and attention results in more votes cast in elections with a controversial ballot proposal.

One shift in campaign spending patterns has been the trend toward greater spending in support of ballot proposals. During the 1980s and 1990s, more money on average was spent in opposition to ballot proposals--although the differences were not huge. Since 2000, campaign spending in support of proposals has outpaced spending by those in opposition. Overall, spending in support of proposals exceeded spending in opposition by about 1.5 times.

The relatively low ballot roll-off trend over time also was a bit of a surprise. Ballot proposal questions are commonly thought of as low information races, with less media attention and less voter interest. However, ballot roll-off for Michigan ballot proposals has consistently been under 10% for these elections since the 1970s, and has been under 5% since 2000. This figure puts ballot proposal roll-off much closer to roll-off for U.S. Senate races than for typical low information races, such as for state supreme court--which have closer to 25 % roll-off..

The average cost per vote has skyrocketed since 2000. The cost per vote, the cost pet a "no" vote, and the cost per vote per capita all generally increased in the past 16 years. This trend in increased spending is especially true of the cost per vote, which rose from $.07 in 1978 to almost $8.00 in the 2010s.

Despite the increases in campaign spending over the years, the role of money in affecting the outcome of elections is not consistent or obvious. We discovered that only a slight majority of elections since 1978 (55%) saw the winning side spend the most money. This finding suggests there are other factors that help determine ballot proposal election outcomes. We can speculate on some of these factors. For example, there are a number of ballot proposal topics covering what we have called "morality" issues earlier in this article. For issues such as same-sex marriage, gambling casinos, or affirmative action, many voters already have formed an opinion about the proposal. For other voters, their views on issues such as tax increases also will be relatively immune to advertising and persuasion--or what political scientists would call "campaign effects." Of course, there remain many ballot proposals where voters have not made up their minds and therefore can be persuaded by an effective campaign.

The higher level of campaign spending in support of ballot proposal elections since 2000 suggests that at least those directly involved in these campaigns believe that money matters--and is in keeping with Stratmann's (2005) findings about the effectiveness of increased spending by either side. Large increases in spending in the decade of the 2010s, along with increased success by the larger spenders, does suggest some campaign spending effects for ballot proposal campaigns. As such, it appears that money does matter in ballot proposal elections in Michigan, but it depends on the kind of proposal and the circumstances surrounding the election. In short, spending money is not a silver bullet that will guarantee one's preferred outcome once all the ballots are counted. There are notable ballot proposal elections where larger amounts of spending resulted in a loss for the big spender. As such, those who fund proposal questions in Michigan must make strategic decisions about how much--and under what circumstances--they will invest in a ballot proposal campaign.


BANDUCCI, SUSAN A. 1998. "Direct Legislation: When Is It Used and When Does It Pass?" in Citizens as Legislators: Direct Democracy in the United States, edited by Shaun Bowler, Todd Donovan, and Caroline Tolbert. Columbus, OH: Ohio State University Press.

BOWLER, SHAUN, AND TODD DONOVAN. 1998. Demanding Choices: Opinion and Voting in Direct Democracy. Ann Arbor, MI: University of Michigan Press.

BOWLER, SHAUN, AND TODD DONOVAN. 2018. "The Initiative Process," Chapter 6 in Politics in the American States, 11th edition, edited by Virginia Gray, Russell L. Hanson, and Thad Kousser. Thousand Oaks, CA: Sage/CQ Press.

BRODER, DAVID. 2000. Democracy Derailed: Initiative Campaigns and the Power of Money. New York: Harcourt.

GARRETT, ELIZABETH. June 1999. "Money, Agenda Setting, and Direct Democracy," Texas Law Review, 77(7): 1845-1890.

GERBER, ELISABETH R. 1999. The Populist Paradox: Interest Group Influence and the Promise of Direct Legislation. Princeton, NJ: Princeton University Press.

HAJNAL, ZOLTAN L., ELISABETH R. GERBER, AND HUGH LOUCH. 2002. "Minorities and Direct Legislation: Evidence from California Ballot Proposition Elections," Journal of Politics 64(1): 154-177.

KLEMANSKI, JOHN S. 2013. "Trends in Michigan Supreme Court Elections 2000-2012," Michigan Academician, 41: 289-309.

KLEMANSKI, JOHN S., AND DAVID A. DULIO. 2017. "Ballot Proposals in Michigan." In Michigan, Government, Politics, and Policy, edited by John S. Klemanski and David A. Dulio, chapter 11. Ann Arbor, Ml: University of Michigan Press.

LOWENSTEIN, DANIEL H. 1982. "Campaign Spending and Ballot Propositions: Recent Experience, Public Choice Theory and the First Amendment," UCLA Law Review 29: 505-641.

MAGLEBY, DAVID B. 1984. Direct Legislation: Voting on Ballot Propositions in the United States. Baltimore, MD: The Johns Hopkins University Press.

MATUSAKA, JOHN G. 1993. "Election closeness and voter turnout: Evidence from California ballot propositions," Public Choice 76(4): 313-334.

MCCUAN, DAVID, SHAUN BOWLER, TODD DONOVAN, AND KEN FERNANDEZ. 1998. "California's Political Warriors: Campaign Professionals and the Political Process," in Citizens as Legislators: Direct Democracy in the United States, edited by Shaun Bowler, Todd Donovan, and Caroline Tolbert. Columbus, OH: Ohio State University Press.

Michigan Campaign Finance Act, PA 388 of 1976. Michigan Secretary of State,; accessed September 22, 2016.

Michigan Constitution, Article II, Article XII;; accessed October 2, 2016.

National Conference for State Legislatures. September 20, 2012. "Initiative, Referendum, and Recall,"; accessed September 13, 2016.

NICHOLS, STEPHEN M. 1998. "State Referendum Voting, Ballot Roll-off, and the Effect of New Electoral Technology." State and Local Government Review 30: 2, 106-117;; accessed May 12, 2017.

OWENS, JOHN R., AND LARRY L. WADE. 1986. "Campaign Spending on California Ballot Propositions, 1924-1984: Trends and Voting Effect," Western Political Quarterly, 39 (4): 675-689.

QVORTRUP, MADS. 2001. "The Courts v. the People: An Essay on Judicial Review of Initiatives," in The Battle Over Citizen Law Making, edited by M. Dane Waters. Durham, NC: Carolina Academic Press.

SARBAUGH-THOMPSON, MARJORIE, AND LYKE THOMPSON. 2017. Implementing Term Limits: The Case of the Michigan Legislature. Ann Arbor: University of Michigan Press.

STATE OF MICHIGAN, BUREAU OF ELECTIONS. March 2015. "Initiatives and Referendums Under the Constitution of the State of Michigan 1963,"; accessed October 15, 2016.

State of Michigan, Secretary of State. "Report Description Expenditures by Ballot Question Committees on Statewide Ballot Issues, 1978-2004";; accessed October 14, 2016. For 2006 election:; for 2008 election:; for 2010 election:; for 2012 election:; for 2014 election:; for 2015 election:; all accessed October 14, 2016.

STRATMANN, THOMAS. 2005. "The Effectiveness of Money in Ballot Measure Campaigns," Southern California Law Review, 78 (4): 101-124.

STRATMANN, THOMAS. 2006. "Is Spending More Potent for or against a Proposition? Evidence From Ballot Measures," American Journal of Political Science 50(3): 788-801.

TOLBERT, CAROLINE, JOHN A. GRUMMEL, AND DANIEL A. SMITH. 2001. "The Effects of Ballot Initiatives on Voter Turnout in the American States," American Politics Research, 29(6): 625-648.

TOLBERT, CAROLINE, AND DANIEL C. BOWEN. 2008. "Direct Democracy, Engagement, and Voter Turnout," in Democracy in the States, edited by Bruce E. Cain, Todd Donovan, and Caroline Tolbert. Washington, DC: Brookings Institution Press.


Oakland University
TABLE 1. Share of Ballot Proposals Appearing on the Ballot, by Month of
Election, 1978-2015.

March   May    June   August   November

1.2%    2.3%   1.2%    3.5%      92%

Note: Percentages do not add to 100%, due to rounding

TABLE 2. Average Vote Totals and Percent Voting in Ballot Proposals by
Decade, 1978-2015.

Decade   Average percent of MI population voting   Average total votes
                in proposals                            in proposals

1970s            29.91                                2,757,024.55
1980s            31.74                                2,907,789.11
1990s            33.32                                3,226,291.74
2000s            35.06                                3,510,120.65
2010s            35.5                                 3,5045,96.08

TABLE 3. Average Roll-Off for U.S. Senate Races and Ballot Propositions
by Decade, 1978-2015.

Decade   Average roll-off to   Average roll-off to U.S. Senate contest
          ballot proposals

1970s        3.84%                        0.72%
1980s        9.38%                        2.59%
1990s        8.47%                        1.93%
2000s        4.31%                        1.31%
2010s        3.55%                        1.51%

TABLE 4. Average Dollars Spent For and Against Ballot Proposals, by
Outcome, 1978-2015.

                              Adopted         Failed

Total spent in support      $2,691,367.96   $1,522,536.97
Total spent in opposition   $2,036,267.33     $881,256.97
Total spent                 $4,727,635.29   $2,403,793.95

TABLE 5. Ballot Proposals with $0 Spent in Support or $0 Spent in
Opposition, by Type and Outcome, 1978-2015.

Type of proposal           $0 spent in support   $0 spent in opposition

Constitutional amendment      78.9% (N=15)            74.4% (N=29)
Statutory initiative           5.3% (N=l)             23.1% (N=9)
Veto referendum               15.8% (N=3)              2.6% (N=l)
Total                        100% (N=19)             100% (N=39)

TABLE 6. Percentage of Ballot Proposals Where the Side Spending More
Money Was Successful, 1978-2015.

Decade         Percentage of elections   Percentage of elections where
               where winning side        the losing side spent the most
               spent the most money                 money

1970s (N=11)          82%                       18%
1980s (N=27)          44%                       56%
1990s (N=19)          58%                       42%
2000s (N=17)          41%                       59%
2010s (N=12)          67%                       33%
Total (N=86)          54.5%                     45.5%

TABLE 7. Spending Differences by Type of Ballot Proposal.

                            Constitutional   Statutory       Veto
                            amendments       initiatives   referendums

Cost per vote                 $1.96            $1.33         $0.69
Cost per no                   $0.94            $0.81         $1.08
Cost per 'yes' per capita     $0.25            $0.18         $0.10
COPYRIGHT 2018 Michigan Academy of Science Arts & Letters
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2018 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Klemanski, John S.; Dulio, David A.; Cogo, Haris
Publication:Michigan Academician
Article Type:Report
Geographic Code:1U3MI
Date:Jan 1, 2018
Previous Article:Pais Intimo: Identity Formation of a Generation that Inherited the Trauma of the Spanish Civil War.
Next Article:Student Learners and Preferred Agents in Mental Health-Seeking: Assessing Higher Educational Services in a Pilot Study.

Terms of use | Privacy policy | Copyright © 2021 Farlex, Inc. | Feedback | For webmasters |