a publication of the
Mineta Transportation Institute
College of Business
San Jose State University
San Jose, CA 95192-0219

Created by Congress in 1991

list of tables

Table 1-1. Within Community Differences in Voting Results 9

Table 3-1. Differences Between the 1999 and 1997 Elections
(Denver) 49

Table 5-1. Differences Between the 2000 and 1997 Ballot Measures
(Phoenix) 87

Table 9-1. Differences Between 1998 and 2000 Transportation Packages
(Alameda) 145

Table 10-1. Status of Each Case on Selected Critical Influencing Factors 172

Executive SUMMARY

The authors of this study participated in a prior project (Haas, Massey, Valenty, and Werbel, 2000) involving case studies of four local ballot tax measures for transportation packages with a substantial passenger rail component. In-depth interviews with knowledgeable individuals were used to obtain information about the process used to determine the transportation package, the nature of the package, and the communications campaigns used by both proponents and opponents. Although the study revealed quite a bit about how these factors influence the results of this type of ballot measure, the conclusions drawn were tentative at best, due in part to the small number of cases studied.

The current study uses the same case study methodology as the prior study to obtain data on recent ballot measures in eleven communities from the past four years. As in the prior study, each of the ballot measures had a substantial rail transit component. Four of the communities had two similar ballot measures within a three-year interval, with substantially different results between the two measures. The existence of two recent measures in these communities allowed systematic comparisons, facilitating an understanding of how the transportation package and the process used to determine it, as well as how the communications campaigns used by proponents and opponents may influence the outcomes of these measures.

The small number of cases included the qualitative nature of most of the potential explanatory variables, and the interactions and correlations among these variables prevent the use of rigorous quantitative analytical techniques that isolate the quantitative impact of explanatory variables on voting behavior. A simple quantitative analysis was conducted that consisted of identifying instances in which the result of the ballot measure was the opposite of what would have been predicted based upon the value of each of a series of explanatory variables. As an example, extensive stakeholder involvement in the development of the transportation package was predicted to be a component of a successful ballot measure. If a community was found to clearly exhibit stakeholder involvement, and yet the measure in that community failed, an "anomaly" was identified. Similarly, an anomaly was also indicated if a community did not clearly exhibit stakeholder involvement, and still managed to pass a measure. The number of anomalies was tabulated separately with each explanatory variable and the number was used as a general indicator of the extent to which each variable was a good predictor of success or failure. Most of the explanatory variables had only a small number of anomalies, suggesting that these variables may indeed have some explanatory and even predictive value. Some of the important conclusions are identified as follows:

The combination of energetic and credible opposition and a questionable reputation of the transit agency and/or the transit system make it extremely difficult for a ballot measure to be successful.

When a community has no prior rail transit system, a comprehensive rail-only package is extremely unlikely to be successful.

Without approximately $1 million or more to spend on a combination of direct mail and television advertising it is difficult for proponents to be successful. (The amount needed may be less in smaller communities with lower costs of local television advertising.)

It is extremely difficult to be successful in fundraising without having the business community and key elected officials enthusiastically support the ballot measure.

The degree of enthusiasm that the business community and key elected officials hold is apparently influenced by the degree to which they are involved in developing the transportation package. They are more likely to be enthusiastic supporters if they are involved in the development of the package.

Involving the business community, key elected officials, and possibly representatives from environmental organizations can make it difficult to develop a consensus about a specific transportation package that can be enthusiastically supported by all groups and key individuals. Each group often does not have a common vision on the desirability of different modes of transportation, let alone a common vision of the specific aspects of a package.

Developing a consensus transportation package depends on the specific details of the package and it is very difficult to generalize about the needed details. They will vary from community to community. As a general example, the inclusion of a highway component was important with some successful measures but this same component might have hindered success in other communities that were successful when a highway component was absent. An understanding of the importance of the specific dimensions of a transportation package can be gained only from reading each specific case analysis.

Under certain circumstances, voters do not appear to place significant importance on the existence or length of the expiration date of the tax used to fund the transportation package. Based upon the results of prior studies, it was predicted that an expiration date of 10 years or less typically was needed for success. This prediction was not valid with any of the successful cases.

Many aspects of the communications campaign do not lend themselves to even the simple analysis method used. As with the nature of the transportation package, the specific dimensions of the communications campaigns are extremely important and yet to generalize about the appropriate specific dimensions is very difficult. The information contained in each case analysis may provide ideas about the successful (and less successful) aspects of communications campaigns used by proponents and opponents. Proponents who conduct opposition research prior to the commencement of a campaign to try to predict both the nature of the opposition and the arguments likely to be used are more successful. This research is important largely because the opposition typically relies on communicating its message through the news media and short response times exist in dealing with these media. In addition, because news coverage of these ballot measures tends to be at least moderately intensive even during a general election, some of the more successful proponent campaigns took a proactive and energetic approach with the news media. Conversely, none of the unsuccessful cases took such an approach.

introduction
and description of methodology

brief description of Report topic

This research project seeks to identify factors contributing to the success, as well as those contributing to the failure, of local transportation ballot measures with a substantial passenger rail component. More specifically, the focus is on actionable success and failure factors related to the transportation package and the process used to develop it, and the communication campaigns employed by both proponents and opponents of each measure.

Recent ballot measures in eleven different communities have been studied, with data collected through interviews and written documentation. The communities are listed below, with both the result and year of the election in parentheses.

Alameda County, California (approved, 2000).

Austin, Texas and other smaller communities in the Capital Metro Transit District (rejected, 2000).

Charlotte, North Carolina and other smaller communities in Mecklenburg County (approved, 1998).

Columbus, Ohio (rejected, 1999).

Denver, Colorado and other smaller communities in its Regional Transportation District (approved, 1999).

Phoenix, Arizona (approved, 2000).

St. Louis County, Missouri (rejected, 1997).

Salt Lake, Davis, and Weber Counties, Utah (approved, 2000).

San Antonio, Texas and other smaller communities in its metropolitan transit district (rejected, 2000).

Santa Clara County, California (approved, 2000).

Sonoma County, California (rejected, 2000).1

Four of these eleven communities presented a prior measure to voters within three years of the most recent measure. Adding these four ballot measures results in a total of 15 ballot measures studied, with seven of the measures receiving support from over 50% of voters and eight receiving support from less than 50% of voters.

This study is a follow-up to a prior study conducted by Haas, Massey, Valenty, and Werbel (2000). Part of this prior study, which is briefly summarized later in this chapter in the "literature review" section, had essentially the same objectives as this current study. However, because the prior study included only four ballot measures, with two of them being in the greater San Francisco Bay Area, the current study includes a larger sample of ballot measures, with the hopes of identifying more nationally relevant findings.

importance of the research topic

The study of factors influencing voting patterns on local ballot measures involving funding for transportation improvements with a substantial rail component is important for two reasons. First, based upon prior voting results, it is difficult to obtain voter approval of these ballot measures, particularly on the first attempt, with the success rate being less than 50% on first-attempt measures. Those considering such measures need to have a good understanding of things to both do and avoid doing when only a few mistakes possibly can cause a measure to fail. Second, for reasons discussed below, it is likely that a large number of communities will attempt to obtain voter approval of ballot measures to provide funding for transportation packages with a substantial rail component in the next ten years or so.

A number of factors, including typical linking of federal funding to local matching of funding, have contributed to the need for a local funding source for transit improvements. Although a number of communities were able to obtain a dedicated source of transit funding a number of years ago, most of these communities did not have sufficient funding to construct, let alone operate, a substantial rail transit component prior to 1990.

Rapid population growth, particularly in many Sun Belt communities during the past ten or twenty years, has resulted in rail transit becoming a feasible option in a larger number of communities. This increased feasibility of rail transit in a larger number of communities, combined with the need to provide additional local funding for the construction and operation of rail transit systems, has resulted in a relatively large number of recent ballot measures. For example, based upon the above list, at least nine communities had a ballot measure to provide funding for a transportation package with a substantial rail transit component in 1999 or 2000 alone. Continued population growth in other communities, combined with successful ballot measures in midsize but high growth communities, such as Salt Lake County and Phoenix, probably will result in additional communities attempting to obtain voter approval of funding for rail transit systems. In addition, communities, such as Austin, San Antonio, Columbus, and Sonoma County, which have been unsuccessful with recent ballot measures, probably will attempt to obtain voter approval in the future. Even communities, such as Denver, which have been successful in the past, probably will try to obtain future funding to expand their current rail transit system.

literature Review

A literature search identified only three prior studies that have studied factors that influence voting behavior in local ballot measures providing funding for transportation. Each of these three prior studies is discussed in the remainder of this section.

The Prior Study (Haas, Massey, et al., 2000)

The prior study, upon which this project is based, consisted of both a case analysis approach, primarily using qualitative depth interviews with key participants, and regression analysis, using secondary data, to study factors influencing voting results with local transportation ballot measures.

The case analysis portion of the study included ballot measures in four communities. Each of the four communities had ballot measures with a substantial rail transit component, with the proposed funding source being a sales tax increase. Variables pertaining to a) the transportation package, b) the process used to determine the transportation package, with a focus on both the roles of research and coalition building, and c) communications campaign approaches used by proponents and by opponents. The four communities are listed below, with the year of the ballot measure and the result included in parentheses.

Denver, Colorado Regional Transportation District (1997, rejected).

Santa Clara County, California (1996, approved).

Seattle, Washington greater metropolitan area (1996, approved).

Sonoma County, California (1998, rejected).

Two different datasets were used in the regression analysis. One was a national dataset consisting of 57 local transportation ballot measures between 1990 and 1998. The second was a California only dataset consisting of 63 local ballot measures between 1980 and 1998. Only a limited number of the ballot measures in each dataset had a substantial rail component and some did not have a transit component. A wide variety of funding sources were used, and significant variance existed in the costs of the transportation packages. Of the nine independent variables included in the national dataset, only one involved a characteristic of the transportation package (the number of modes included in the transportation package). The remaining variables measured a) demographic characteristics, such as the percentage of people over sixty-five years of age in the community, b) transportation characteristics, such as travel time, and c) size of the community, population growth, and population density. The eight independent variables used in the California dataset were similar to those used in the national dataset with the biggest difference being the omission of the one variable in the national dataset that measured a characteristic of the transportation package.

Of the two approaches, the case analysis approach appears to generate more significant results because it focuses on variables that appear to have a greater impact on voting behavior than the variables studied with the regression approach. More specifically, voting behavior appears to be more sensitive to differences in both the transportation packages and the communication campaigns used by proponents studied with the case approach than to differences in community characteristics studied with the regression approach. The Seattle case study was particularly useful in demonstrating that characteristics of the transportation package, the process used to determine it, and the communications campaign used have a significant impact on results. This case allows a within community comparison, since a measure somewhat similar to the 1996 measure had been defeated in 1995. The 1995 measure received support from 47% of voters compared to a supporting percentage of 56.5% with the 1996 measure. Since the community characteristics in Seattle were virtually identical in 1996 as in 1995, this moderately large difference in results could not have been caused by community characteristics. Rather, the difference in results appear to have been caused by a combination of some differences between the 1995 and 1996 transportation packages and differences between the 1995 and 1996 communication campaigns used by proponents. Four communities included in this current project also allow a within community comparison because each one had two ballot measures within a three year time frame. As identified in Table 1-1 below, the within community difference between two ballot measures in the percentage of voting support is larger in each of the four relevant communities in the current project than the difference in Seattle. Thus, the results in Table 1-1 demonstrate that voting behavior is quite sensitive to differences in both the transportation packages and the communication campaigns used by proponents.

Table 1-1. Within Community Differences in Voting Results
(In communities with two ballot measures within a three year period)

	Supporting Percentage in First Ballot Measure	Supporting Percentage in Second Ballot Measure	Difference in Supporting Percentage Between Two Measures
Alameda County	58% (1998)	81.5% (2000)	23.5 percentage points
Denver	42.2% (1997)	66% (1999)	23.8 percentage points
Phoenix	49.9% (1997)	65% (2000)	24.9 percentage points
Sonoma County	47.6% (1998)	60.2% (2000)	12.6 percentage points
Seattle	47% (1995)	56.5% (1996)	9.5 percentage points

Although the differences in Table 1-1 above cannot be explained by community characteristics, none of the cases in the table were structured in a manner to measure the impact of community characteristics on voting results. An across community study would be needed to accurately access the degree of impact of community characteristics on voting results. The regression analyses used in the prior study did allow such an across community analysis, but the only variable that appeared to be correlated with voting behavior with both datasets was the percentage of the population in each community consisting of people at least sixty-five years of age. Moreover, the direction of the correlation with the California dataset was opposite the direction found with the national dataset.

In sum, the campaign and proposal characteristics appear to have more impact on voting behavior than do community characteristics. In addition, because the transportation package, the process used to formulate it, and the communications campaign used all are largely actionable, focusing on the impact of these characteristics on voting behavior will provide more useful information to those communities developing transportation ballot measure and attempting to obtain their passage than focusing on the impact of uncontrollable community characteristics on voting behavior. Accordingly, it appears to be more important to study the impact of transportation package and communications campaign characteristics on voting behavior than to study the impact of community characteristics on voting behavior.

Beale, Bishop, and Marley (1996)

These researchers also studied the impact of the transportation package and the process used to formulate it, along with the impact of the communications campaigns used, on voting results with local transportation ballot measures, using onsite visits to 22 sites. In addition to these 22 cases, the authors listed 82 additional cases that were included in the study. The authors did not describe the nature or amount of the information collected with these 82 additional cases.

The conclusions of the Beale, Bishop, and Marley study helped identify relevant variables for questioning in both our prior (Haas, Massey, et al., 2000) and current studies. Beale, Bishop, and Marley presented only two conclusions related to the nature of communications campaigns used. One of these conclusions discussed the conditions in which a special election is more likely to succeed vs. the conditions in which a general election is more likely to succeed. The second conclusion focused on ballot language recommendations. Our prior study, as well as the current one, have been designed to break more new ground concerning communications campaigns than Beale, Bishop, and Marley apparently learned about the nature of the impact of these campaigns on voting behavior.

Since Beale, Bishop, and Marley did not provide a categorization of the nature of the transportation options involved, the number of cases they studied that did not involve a substantial rail component cannot be identified. However their discussion suggests that many of their cases did not have a substantial rail component, or even a transit component. Some transportation package and funding issues probably are more difficult to resolve with transportation packages involving a substantial rail component. For example, since many communities considering a rail system have no current rail system in place and since rail systems are expensive to construct, it is difficult to propose a rail system that provides benefits to a majority of voters while not significantly increasing taxes for a lengthy time period. In contrast, many of the cases studied by Beale, Bishop, and Marley apparently involved funding only for road maintenance and improvements. It is relatively easy to provide these benefits at a low cost while, at the same time, making improvements throughout the community. In addition, the communication campaigns used by opponents, as well as the nature of the opposition itself, tend to be different when rail is a substantial component of the transportation package than when it is not a component. Since some of the issues involving ballot measures with a substantial rail component are different than the issues involving measures without this component, the Beale, Bishop, and Marley study has limited relevance to the focus of our research project.

Middleton (1998)

Unlike Beale, Bishop, and Marley's study, Middleton focused exclusively on ballot measures with a substantial rail component. Middleton's approach involved asking approximately six transit officials to identify the factors influencing voting behavior and to make recommendations likely to increase the chances of voter approval. The 1995 and 1996 ballot measures in the Seattle metropolitan area are the only ones described in any detail. The primary problem with Middleton's approach is that the recommendations are not systematically tested against actual ballot measures. In addition, the recommendations do not discuss a number of critical issues such as the existence and length of expiration dates, magnitude of tax increase, whether and how to respond to opposition, what combination of media and messages to use for advertising, and when and how to combine rail and highway components.

Summary of Literature Review

The literature review has demonstrated the need to generate a larger number of actual ballot measures with a substantial rail component than has existed in prior studies. In addition, as demonstrated previously by the large difference in the results of two ballot measures within a short time period in the same community (see Table 1-1), characteristics of the transportation package, the process used to formulate the package, and the communications campaign used by proponents can have a significant impact on voting results.

description of methodology

The two primary methodological decisions with this study were a) the number and identity of the ballot measures selected, and b) the nature of the information collected and the means of collecting it. Each of these two decision areas is briefly described below.

The Identity and Number of the Ballot Measures Selected

Each ballot measure selected had to meet three criteria. The first two of these three criteria were based upon the focus of this project, which is on ballot measures that involve funding for transportation packages with a substantial rail component. Thus, any ballot measure selected had to a) have a substantial rail component, and b) have a funding component. The third criterion, which involved eliminating any ballot measure prior to 1997, was needed because much of the information has been collected through qualitative depth interviews. Since a failure of interviewees to accurately recall all relevant information would compromise the integrity of the information, it was necessary to eliminate ballot measures prior to 1997. A systematic search of ballot measures identified 11 communities with ballot measures meeting these three criteria.

As a group, the ballot measures and communities included in this project have a number of desirable characteristics that make them appropriate for in-depth study.

As discussed previously, four of the eleven communities have had two ballot measures since 1997. Since this characteristic allows community characteristics to be held constant, it facilitates an understanding of the extent that changes in the nature of the transportation package, and/or changes in the communications campaign used by proponents can influence voting behavior.

A wide range of voting results exists. Supporting percentages range from a low of 30% in San Antonio to a high of 81% in Alameda County. Since the percentage of voters supporting a measure essentially is the dependent variable in this study, variability with the dependent variable is needed to derive an understanding of the variables that influence voting behavior.

More than token opposition existed to the ballot measure in eight of the eleven communities. Since opposition to passenger rail often exists, much can be learned from the strategies and tactics employed by opposition. In addition, much can be learned about the effectiveness of various means that proponents employ in attempting to counter opposition.

Variance exists in both the nature of the transportation packages and the process used to determine them. Some of the measures consisted only of passenger rail transit, other measures had both a passenger rail and a bus component, and still other measures had a passenger rail and/or bus component combined with a highway component, with the highway component either being in the same measure or in a companion measure. In terms of the process used to formulate the package, significant differences in the degree of both public and political involvement exist across measures. This variance is required to understand how the nature of the transportation package and the process used to formulate it influence voting behavior.

Variance exists in the communications campaigns employed by proponents across measures, particularly in the approaches employed to deal with opposition. In some instances the opposition was essentially ignored, and in other instances the opposition was attacked with differences in both the nature and intensity of these attacks.

Means of Collecting Information

The information to be collected included a) the process used to formulate the transportation package along with the characteristics of the package itself, b) the communications campaign used by proponents, c) the communications campaign used by opponents, d) relevant background information on the transportation system, the political and legal environment, and characteristics of the population, e) the nature and extent of coverage by the mass media, particularly local newspapers and radio stations, and f) research results identifying voter attitudes and voting motivations.

Much of the information could be collected only through qualitative in-depth interviews with different people within each community. Although the nature of the people interviewed varied somewhat from community to community, at least one person typically was interviewed from:

The transit agency;

; The transit Board;

An ad hoc committee or task force involved in formulating the transportation package if such an organization existed;

The business community, typically a Chamber of Commerce;

An environmental organization;

The group responsible for developing the proponents' communications campaign,;

The leaders of the opposition;

The local newspaper, and

The political structure, typically a local elected official.

Unless a person interviewed was a prominent elected official, the name of the person typically has been omitted from the case discussions in the remaining chapters.

In addition to the interviews with people from most or all of the above categories, documentary information was collected. Much of the documentary information was collected prior to conducting the interviews, since this information often was helpful in identifying specific interview topics and issues. The different types of documentary information collected were:

Newspaper articles and editorials;

Copies of ads used by proponents and opponents;

Long range transportation plans;

Major investment/corridor studies and other similar studies conducted. (For example, opponents often prepare reports or studies that discuss costs and ridership of light rail systems.);

The actual ballot measures;

Ballot arguments in voter information pamphlets;

Population, demographic, and transportation statistics; and

Survey and/or focus group results of citizen attitudes toward the transportation system, and the ballot measure.

The last item in the above list is one of the most useful types of information. In a study, such as this one, which attempts to understand voting behavior, information obtained directly from the voters, such as exit polling, is extremely useful. Results from surveys and tracking studies conducted prior to the election often were conducted and were provided to the researchers involved in this project. Unfortunately, exit polling and other types of post election voter research were not conducted with the vast majority of the ballot measures studied.

To insure that the description of each case is accurate and complete, at least two people interviewed were sent drafts of each onsite case for review and feedback and at least one person interviewed was sent a draft of each telephone case.

organization of remainder of this report

Chapters two through eight each focus on a single case with which onsite interviewing was used. Chapter nine includes four cases with which telephone interviewing was used. The results of the cases are integrated and synthesized, and conclusions are stated in chapter ten. Since insights can be drawn through contrast, the chapters are sequenced by whether or not they were successful, alternating between an unsuccessful one and a successful one.

To facilitate reading, the organization of each chapter is the same. After a short introduction, relevant background information is described about the transportation system, past ballot measures, population and demographic statistics, geography, and the political/legal environment. Following this background information, the process used to develop the transportation package along with a description of the package and its funding is discussed. The communications campaigns and strategies, along with the coverage of the measure in the media are then described. Each chapter concludes with an evaluative discussion of important and unusual issues involved with the case and a concluding discussion of factors contributing to the success or failure of the measure.

Since this is a lengthy report, readers will be tempted to read only the final chapter. This approach is not recommended as many specific ideas and insights will be lost with this reading approach. At a minimum, readers are encouraged to read the evaluation and conclusion sections of each chapter.

st. louis: A Setback for Measure M
and MetroLink (1997)

In November of 1997, voters in St. Louis County rejected, by a 58% to 42% margin, a 1/4-cent sales tax increase that would have primarily funded additional lines for an existing light rail system known as MetroLink. Voters in St. Louis County voted "no" on Proposition M, 58% to 42%. Although voters in St. Louis barely passed the measure, 50.4% to 49.6%, both jurisdictions needed to pass it for the tax increase to occur. Supporters of the measure estimated that the tax hike would have brought in about $43 million a year. They claimed that two thirds of the money would have gone to MetroLink expansion; the remainder would have helped fund existing bus and light rail service. Adding to the failure was the defeat of a similar measure in neighboring Madison (Illinois) County that would have funded expansion eastward. (See Table 10-1 in the last chapter).

For supporters of the measure, the defeat was ironic in that MetroLink is considered a popular and successful light rail system. Tracking polls conducted prior to the vote suggested that public support for the measure was substantial. How and why was the measure rejected by a substantial majority of the voters? This case represents an example of how fragile support for public transportation projects can be, and how proposals to improve transit can become mired in petty political squabbles.

background: Transportation and traffic

Located in both Illinois and Missouri, the St. Louis metropolitan region has a population of nearly 2.5 million people, including 1,324,442 registered voters. Eight major jurisdictions comprise the region, which covers 4,487.3 square miles. Within the region there are 904,743 households whose average size is 2.6 persons. 8.1% of the region's families have incomes below poverty level and 10.8% of the total population has incomes below the poverty level. There are 17,662 miles of roads in the region including 490 miles of interstate highways. There are 9,850 miles of urban roadways and 7,812 miles of rural roadways. Most people in the region drive to work and 79.9% of them drive to work alone. About 3% of the workforce uses public transportation to get to work. The region is home to 21 Fortune 1000 companies and ranks sixth in the United States as a headquarters location for Fortune 500 companies (East-West Gateway Coordinating Council, 2000a).

The City of St. Louis includes a population of 341,869. The City is poorer than the surrounding County, with a median household income of just $19,458 annually, versus $31,837 annually in the County. 20.6 percent of City families and 24.6% of the total population have incomes below poverty level. 66.5% of workers drive to work alone in the City and about 12.2% take public transportation. The City of St. Louis voted to separate itself from St. Louis County in 1876 and is the only city in Missouri not within a county. (East-West Gateway Coordinating Council, 2000b).

Other important components of the metropolitan area include St. Charles County in Missouri and Madison and St. Clair Counties, both located in Illinois across the Mississippi River. Thousands of workers commute into and around the metropolitan area from these jurisdictions and beyond. As the population and economy of the region continue to expand, traffic and congestion have worsened. Among the factors contributing to the region's traffic woes are the following:

St. Louis residents make 6 million vehicle trips each day. Of these 6 million trips, 5 million are single occupancy vehicle trips.

80% of the St. Louis workforce drives alone to work. In the City of St. Louis, 66.5% of all workers drive alone to work. In St. Louis County, 84.3% of all workers commute alone to work.

Only 3% of the work-related trips are made on local transit.

On average, people make 3.9 vehicle trips per day. (Citizens for Modern Transit, 2000)

However, St. Louis has been proactive in attempting to ameliorate its traffic problems. In 1994, MetroLink, the light rail system, was opened. Voters at that time approved a 1/4-cent sales tax increase, which helped to fund the system. Additional funding came from the Federal Government. MetroLink presently consists of an 18-mile line that runs from East St. Louis (across the Mississippi from St. Louis) to Lambert Field, the St. Louis airport. An expansion into St. Clair County opened in May of 2001; in 1993 St. Clair County voters approved a 1/2-cent sales tax to help pay for the new line, which is also receiving federal funding. In 1995, voters in St. Charles County (west of St. Louis) twice voted down proposals to fund a light rail extension into that part of the area.

In many ways, the MetroLink and the St. Louis area Bi-State transit system exemplify success in American mass transit. In 1995, MetroLink became the most popular light rail line in the United States; 54.2 million St. Louis residents rode public transportation in 1997-1998. MetroLink ridership continues to grow, and now averages 44,500 riders per day. On a single day (July 4, 1999) 160,883 rode MetroLink. The American Public Transit Association selected MetroLink as the first recipient of its Outstanding Achievement Award for Light Rail. The number of riders has exceeded projected ridership from the start, increasing every year.

Thus, when community figures began plans to ask voters for an additional 1/4-cent sales tax increase that would have helped expand the MetroLink system, optimism about the proposal's chances for success was justifiably widespread.

background: Political history

Transit in the St. Louis area is the responsibility of the Bi-State Development Agency, which operates the MetroLink system, as well as bus and paratransit van transportation. "Bi-State" refers to the States of Missouri and Illinois. In St. Louis (both the City and the County), Bi-State services are funded via the sales tax; in Illinois, individual counties contract with Bi-State for the provision of transit services.

A Mayor and a 28-member Board of Aldermen govern the city. A seven-member County Council governs surrounding St. Louis County. Local elections are partisan, meaning that partisanship can and does enter into public debate about local transportation issues. Additionally, the East-West Gateway Coordinating Council is the Metropolitan Planning Organization (MPO) for the area. Its six-member Board of Directors represents eight area counties, as well as the City of St. Louis and the State of Missouri. Additionally, Citizens for Modern Transit is a nonprofit organization that provides campaign support for transit-related issues. On the whole, according to interviewees, elected officials - as opposed to transit agency staff - have tended to be the instigating forces behind the push for expanded rail transit. Particularly, the East-West Gateway Coordinating Council has played a leadership role in developing plans for the transit system, although as a federally funded agency, its role in ballot campaigns is circumscribed.

According to interviewees, the original 1/4-cent sales tax for the MetroLink system (passed in 1994) was inadequate for the planned expansion of the system. The increases in the proceeds from the tax that had been projected were not materializing. In fact, the system was generating a shortfall in operating expenses without expansion. Additionally, federal funding for MetroLink operations had been slated to be drastically curtailed.

Discussions about the need for an additional source of funding for MetroLink expansion began in earnest early in 1997. Impetus for the tax increase proposal came from a variety of civic leaders, spearheaded by the St. Louis business community. Civic Progress, a group of the region's largest firms, bankrolled much of the pro-proposition campaign.

Focus groups and public opinion surveys were commissioned, both initially revealing reasons for optimism about the prospects for such a proposal. Although the proposal had the nominal support of the Mayor, he did not wholly embrace or actively campaign for it. The nonelected County Executive also supported the measure.

Public transit, and specifically MetroLink, is very popular in St. Louis. The existing MetroLink line is extremely accessible and useful to many areas of the city - it provides a direct link between the airport and downtown St. Louis, including the Kiel Arena and Busch Stadium, both professional athletics venues. The MetroLink is, in fact, a point of civic pride for the region, according to several interviewees. For example, the polling consultant used by the measure's supporters found that survey respondents tended to exhibit greater support for any tax measure that mentioned the MetroLink system by name.

Moreover, a 1994 survey found extremely broad support for not only MetroLink, but also the Bi-State agency and the City. Predictably, perhaps, the survey also found that those who rode the transit system were more likely to hold it in favorable regard. The poll further found that 45% of City respondents and 40% of those in the County had made at least one trip on MetroLink - numbers that were likely higher at the time of the 1997 vote. Support expressed for bus services was somewhat less, although still formidable with a majority finding bus services "excellent" or "good."

For these reasons, and perhaps because of the significant federal funds that had enabled creation of the system, the 1994 1/4-cent sales tax increase measure passed comfortably, with 66% of St. Louis, and 60% of St. Louis County residents voting in favor. In some respects, proponents of the 1997 measure had little reason to believe that passing an additional tax would prove problematic.

creating a transportation initiative

The ballot initiative that was advanced by the coalition of transit supporters was notable for its lack of specificity. It simply asked whether voters favored a 1/4-cent increase in the sales tax "for transportation purposes." In the early 1990s, the East-West Gateway Board of Directors agreed on a general plan that outlined which areas would get extensions first, although this plan was hardly set in stone.

By their own admission, supporters of the tax initiative did not seek to use a open, broad based approach to crafting a ballot proposal or an accompanying transportation plan; Proposition M was planned more or less behind closed doors. Given the looming threat of an operating funds shortfall, the emphasis was on ensuring that MetroLink was able to garner the additional funds (i.e., 1/4-cent sales tax) enabled by the state earlier in the decade. Although a general plan for the location of additional lines had been circulated for years, relatively little effort was paid to delineating the specifics of how the increased revenues would be spent. This lapse would prove to be potentially a costly one to the supporters' efforts to pass the tax measure.

The failure to open up the planning process was perhaps a result of the perceived contentiousness among factions in St. Louis. "We have one of the best, new light rail systems in the country. But now we have so many factions, we can't agree on how to expand it - so it may not expand," said Les Sterman, executive director of the East-West Gateway Coordinating Council.

Had the measure passed, the tax would have begun in January 1999 and was projected to generate $34 million per year, primarily to be put toward a MetroLink extension each year. The existing 1/4-cent tax generates about $33.5 million in St. Louis County and more than $8 million in St. Louis. Two-thirds of the money was planned to go to MetroLink expansion and one-third to help the Bi-State agency operate the bus and light rail service. Although Bi-State was not legally committed to spending the proceeds from the tax in a specific way, East-West Gateway Coordinating Council officials said the money would have financed MetroLink extensions in south St. Louis County and to suburban Florissant by 2010. The money might also have paid for one to three possible routes in south St. Louis and St. Louis County, north St. Louis and St. Louis County, or west St. Louis County by 2015.

use of Research Techniques

Supporters of the tax measure used market research extensively, primarily in the form of public opinion polling. Indeed, according to interviewees, favorable polling numbers in the months prior to November election were an important factor in convincing supporters to place the proposal on the 1997 ballot. Ironically, some observers believe that errors made interpreting the market research data may have led to strategic errors on the part of the transit tax measure's supporters.

Terry Jones, a political science professor at University of Missouri - St. Louis, conducted tracking polls. In April of 1997 (approximately six months before the election), a survey was conducted that tested support for increasing the public transit sales tax from 1/4- to 1/2-cent, preferences about how to spend the additional funds, and reactions to arguments for and against increasing the tax. The survey asked, "what if there was an election this year to raise the public transit sales tax in order to speed up the development and construction of MetroLink lines. Would you vote for or against raising the public transit sales tax from 1/4- to 1/2-cent to speed up the development and construction of MetroLink?"

Responses to this key question were overwhelmingly encouraging to supporters of a ballot measure: in both the City of St. Louis (70% yes, 26% no) and the County (68% yes, 28% no), imposing majorities favored the idea of a tax increase proposal. However, among those saying they were "very certain" about their vote, the margins were closer to 50-50. Interestingly, support for a tax increase expressed in this poll exceeded that measured in 1994 (just prior to passage of the initial Proposition M) by about ten percentage points. In its three years of existence, MetroLink has actually helped to increase support for public transit in the St. Louis area.

Analysis of the April 1997 tracking poll results provides insights into the nature of support for transit tax measures. The pollster (Terry Jones) highlighted the following patterns of support (Jones, 1997):

The best predictor of support for a transit tax increase was experience with the MetroLink system: the more a respondent reported riding it, the more likely he or she was to indicate support for the tax. For example, a 39-point difference existed between frequent and nonriders of the MetroLink system! Education was the second most important predictor, with differences as large as 22 points between the most and least educated voters in the sample.

Whereas patterns of support were found (such as those above), the most striking pattern was "the lack of any demographically or politically based source of opposition." Republicans were found to be about as supportive as Democrats were, for example.

Given these encouraging results, the decision to go ahead with the sales tax increase seems to have been made on a solid basis. Indeed, a series of tracking polls conducted during the campaign for Proposition M seemed to provide more ambiguous evidence that passing the measure would be a relatively easy task. The polls, conducted several times after the initial April survey described above, indicated relatively stable and strong support for the measure, ranging from 53% to 66% in favor in the City and from 50% to 67% in the County.

However, these results may have provided supporters of the measure with a false sense of security. This is because, whereas the survey questioned respondents about their willingness to support a tax increase to pay for expansion and development of the MetroLink system, the actual ballot proposition made no explicit mention of MetroLink. Thus, supporters may have been overconfident and perhaps too complacent about prospects for the measure at the voting booth. (In fact, opponents seized upon the wording of the ballot - and the omission of explicit reference to MetroLink - as a means of fomenting distrust about the measure among voters.)

devising and IMPLEMENTING A Communications Campaign Strategy

Supporters

According to interviewees, the basic strategy behind the supporters' campaign was to capitalize upon the existing positive affect for MetroLink identified in the public opinion polls, while avoiding the potentially conflictive mention of specific routes. This sort of strategy has been referred to as a "stealth campaign," in that it generally seeks to avoid specific issues that might generate controversy. The idea is to motivate supporters to come and vote for the proposition, while leaving potential opponents little cause to get interested in the campaign. Light turnout was forecast for the election, because the measure was the sole item on the special election ballot.

However, the stealth approach seems to have backfired against the measure's supporters. As questions about the routing of the already-funded Cross-County expansion line began to create controversy, the campaign opted to avoid the criticism. In the meanwhile, media attention to a small, poorly funded group of outspoken opponents escalated. The lack of competing campaign issues and candidates created a vacuum in which the media - in attempting to be fair to both sides - provided opponents with a bonanza of free coverage. In the view of measure supporters, opponents of the measure seemed to be receiving as much coverage for free as the supporters were able to garner with their relatively immense budget. Although the region's major newspaper, the St. Louis Post-Dispatch, endorsed the measure, Proposition M supporters were ultimately unable to "control the message" about the tax measure.

As suggested by the following figures, campaign resources were divided fairly equally between mass media and direct mail expenditures.

The supporters' committee - known as "Connecting St. Louis" - reported raising $922,095, with $500,000 of that total coming in the form of loans. Some of the major expenses were:

$229,739.36 for television and radio advertising production and television air time.

$199,738.20 for polling and a telephone bank.

$207,869.02 for direct mail expense.

Approximately $80,000 for campaign management expenses.

Approximately $25,000 for campaign contributions to local Democratic campaign committees.

The campaign had originally planned to spend $500,000 and limit its broadcast advertising to radio. When the opposition appeared stronger than had been anticipated, the campaign decided to raise additional funds and turn to television advertising (Post-Dispatch, Oct. 29, 1997). The ads began just days before the election, consisting of reactions to opposition from some elected officials and neighborhood organizations in various parts of the area. The campaign treasurer stated, "We had to increase the budget because of misinformation."

Opponents

By contrast, the opponents' committee, "Citizens Against Proposition M," reported raising a total of $13,900 and spending $14,123.02. More than half of the total raised ($7,700) was in the form of in-kind services from an advertising company owned by an opponent. The opponents' largest monetary donation was $100.

Opponents seized upon the vagueness of the ballot language - which did not specify that funds would necessarily go to MetroLink expansion - to tap into the larger public distrust of government. Additionally, the opponents were able to make the argument that the transit agency (Bi-State Development) had not delivered upon its promises made in the 1994 campaign. (Supporters concede that the 1994 campaign did overpromise on what could be delivered with a 1/4-cent tax level.) Opponents were thus successful in framing the issue as "trust in the transit agency" or "trust in government," rather than "support for MetroLink."

One opponent felt that the supporters failed to provide answers to basic questions. "People wanted exact answers to questions like where is MetroLink going, what route will be next, how much will expansion cost and when will it arrive. I don't think people were expecting it to be at their door tomorrow, but they wanted to know when they would see it, whether it was now or 20 years from now." Some perceived the vote as a demand for accountability, rather than as a vote against light rail. One County Council member said she commissioned a private poll among county voters before the election and found that 69% of those surveyed felt their input wasn't sought before the proposition was pushed. "They didn't feel involved...Slick advertising was not enough."

An editorial written by two opponents of the measure sums up the appeal opponents successfully made to voters:

Advertisements for Proposition M claim two-thirds of the revenue would go for MetroLink expansion. But that is a meaningless campaign promise. It has no legal basis. The legislation for Proposition M, which was passed by the St. Louis Board of Aldermen and the St. Louis County Council, does not require any of the funds to be spent on MetroLink.

Nothing is definite about how these funds would be used. There are no specific plans for MetroLink expansion and no certainty on what the expansions will cost to construct and operate (Post-Dispatch, Nov. 2, 1997).

Opponents also used the insufficiency of funds for all proposed routes in the 1994 measure to feed the cynicism of voters. One opponent stated that the tax would give the transit agency only enough to build one of the possible expansion routes: "Everyone needs to look closely at that map and see what they're not going to get...two groups are going to be suckers" (Post-Dispatch, Nov. 3, 1997).

evaluation

Ultimately, voters in St. Louis County disapproved of Proposition M, 58% to 42%. Voters in the City of St. Louis barely passed the measure, 50.4% to 49.6%. However, these percentages mask the extent of the defeat; the measure lost by nearly 22,000 votes in the County, and passed by only 300 in the City. Both jurisdictions needed to pass for the tax increase to take effect.

In hindsight, the defeat illustrates how precarious public support for transit taxes can be. Despite encouraging polling results that suggested an easy victory for the measure, a number of factors inveighed against a positive outcome for Proposition M. The following section provides a discussion of possible reasons for the failure of the measure, in no particular order.

Placement of the Proposal on a Special Election Ballot

Special elections are normally associated with a more conservative, anti-tax electorate, but supporters of Proposition M were confident that they could marshal a sufficient number of pro-transit voters. However, they did not anticipate the extent to which opponents would receive free publicity and, in effect, equal standing in the court of public opinion in the absence of other races and issues. Turnout for the election was just 26%, and supporters failed to attract adequate numbers of pro-transit voters.

Failure of Supporters to Identify a Specific Use of the Funds, Including a Transportation Plan

Some observers blamed the defeat on the failure of supporters to clearly define a transportation plan to provide voters with specifics about how the revenues generated by the proposed tax would be spent. Pollster Terry Jones noted that "many voters want a detailed plan about how money will be spent on which lines in what order before they will commit to a yes vote. Talk about a planning process is not sufficient." In Jones' view, opponents were able to use the lack of specificity in the proposal to tap into voters' more general distrust of government. The campaign had suggested general corridors for the planned expansion, but given the rapid timeline for the election, supporters did not have time to conduct a full-scale planning process with public participation. Again, the supporters were counting on the general public goodwill toward Metrolink to carry the day. As it happened, specifics about the use of the tax funds did appear to matter.

The lack of a well-publicized plan dovetailed with the ballot proposition language, which was also vague. Pollster Jones concluded that, "If the issue is whether the voters are willing to pay higher taxes for more Metrolink lines, over 60% are favorable. If the issue is whether the voters trust government institutions to spend the money on MetroLink expansion and operations, less than 50% are favorable." The failure to deliver on the promises made in the 1994 campaign may have contributed to the lack of voter trust in the claims made by proponents in the 1997 campaign. To some extent, the lack of specificity reflected a strategy that sought to mask disagreements among transit supporters about funding priorities for new MetroLink routes. As discussed below, additional routing disagreements developed nevertheless.

Controversy About Specific Routes for Extensions That Were already Funded

Approximately one month before the vote, controversy erupted with respect to the routing of an additional MetroLink line. Although this line had already received sufficient funds from other sources to begin construction and was therefore essentially irrelevant to the proposed tax increase, it was nevertheless a source of great controversy. Briefly, the conflict arose over the decision to take a route north of Forest Park from St. Louis to Clayton. Some community leaders had wanted this route to take a more southern approach, which would have involved building a rather expensive tunnel. The City's new Mayor, Clarence Harmon, was a vocal opponent of the proposed route for the new extension although he supported the tax measure. This placed him at odds with the (elected) County executive, Buzz Westfall. Thus political leaders were squabbling over the system at a time when a uniform stand might have benefited the cause of MetroLink expansion.

The debate over the route decision became extremely contentious, and also contained some seeds of partisanship because of the demographics of the neighborhoods involved. Unfortunately, the conflict deepened and lasted until the election was about to occur. Much of the dissatisfaction over the new route was centered in western St. Louis County, a wealthier area whose approval was viewed as critical by supporters of the measure.

Some supporters of the tax measure believe that this imbroglio poisoned the well for the fortunes of the proposal. Voters may have begun to associate the disagreement about the path of the extension with the funds to be generated by the new tax. In any event, the event comprised extremely negative publicity for MetroLink at a time when it needed to maintain its traditionally very positive image. This controversy might not have sunk the measure, had the vote occurred in a more complicated general election. However, in the absence of other issues, the routing conflict easily spilled over into the Proposition M campaign.

Partisan Opposition to the Plan

Although some of the polling data obtained by the measure's supporters suggested that partisanship did not affect support for Proposition M, the campaign did ultimately invoke a degree of partisanship. Among the more vocal opponents of the plan were several Republican members of the St. Louis County Council, whose opposition may have served to mobilize "no" voters from the western suburbs. On the Democratic side, the Mayor was at least a nominal supporter of the measure, but his involvement in the routing controversy (described above) tended to effectively nullify that support.

In any event, after the election officials from both parties accused one another of allowing partisanship to affect the outcome. Democrats contended that Republicans staged an orchestrated effort to defeat the proposal, and pointed to its especially weak showing in Republican strongholds in St. Louis County. Meanwhile, Republican opponents denied that and accused Democrats of abusing their contributions from Republican-leaning companies because some campaign spending went to Democratic operatives and party ward leaders (Post-Dispatch Nov. 6, 1997).

Dissatisfaction with Bus Service and the Transit Agency

Because Proposition M did not specify that the tax increase would be dedicated to the expansion of MetroLink, opponents of the plan were able to appeal to the relative unpopularity of the bus system. Responses to focus groups conducted after the election confirms that the positive feelings for MetroLink are matched by equally negative feelings about the bus system. Generally speaking, voters like, and are willing to fund, rail transit much more than bus transit. The transit agency has battled annual operating losses and suffers from a poor public image (Post-Dispatch, Nov. 6, 1997).

The Other Proposition M Loses

Across the Mississippi in Madison County, Illinois, another Proposition M was also defeated. The Madison tax measure would have raised the sales tax by 1/4-cent, raising an estimated $276 million by the year 2014. According to supporters, the tax increase would be used to fund MetroLink route extensions into Madison County, although no specific plan was offered to voters. In fact, officials said that receipts from the tax, had it passed, would have been held in escrow until planning began on the route.

Madison County's Proposition M suffered a fate similar to the more publicized St. Louis effort. Voters rejected the measure by a 62% - 38% margin. As in St. Louis, supporters of the measure were encouraged by pre-election polling that suggested an easy victory.

According to supporters, the Madison "M" campaign was only lightly funded (approximately $70,000 was spent on radio ads and a direct mailing). The local Democratic Party organization did not actively support the effort in the belief that it would pass easily. At the same time, the local Republican leadership vocally opposed the measure. The supporters' strategy was to tie in the vote to the St. Louis election, which backfired when the St. Louis measure lost.

conclusions

Because no exit polling data are available, the cause for the defeat of Proposition M may never be definitively known. However, the sum of the potential causes here seems quite plausible as an explanation. As the St. Louis Post-Dispatch (November 6, 1997) observed after the election:

Many [voters] wanted more specific routes, costs and construction timetables. Many wanted a better public process. One group wanted more money to go to bus service. Others wanted all funds to go to build MetroLink. Some thought the wording on the ballot - "for transportation purposes" - was too vague. Some wanted assurance that MetroLink would go underground in their neighborhoods. Others wanted light rail built on railroad rights of way above ground so more money would be left for extensions in their neighborhoods. Others were upset because nothing had been built since the first sales tax for public transit was passed in 1994. Some simply may think the sales tax rates are too high.

The decisive defeat of Proposition M provides a crystalline example of the fragility of public support for transportation related tax increases. That this defeat occurred in the context of widespread public support for the MetroLink system and apparent early support for the tax increase (as suggested by public opinion polls) serves to accentuate this fact. Indeed, supporters of Proposition M were perhaps blinded with overconfidence to the extend that they failed to fully recognize the difficult task it would be to succeed, given the unfortunate circumstances in which the election occurred.

Whereas it is difficult to assign accurate weight to each of the factors and events that appeared to impact negatively on the outcome of the ballot measure, a degree of consensus about the importance of several emerges from interviewees and other resources. No one could have predicted the negative atmosphere created by the strong and public disagreement over the route of the already funded MetroLink expansion, which every interviewee cited as a critical event. However, several lessons appear to emerge from this defeat.

First, although the controversy over the MetroLink expansion route was possibly unavoidable, it is possible that it would not have been quite so influential in the context of a general (and preferably presidential) election. Similarly, the arguments put forth by opponents, both organized and otherwise, might not have been nearly so salient were other issues and elections at stake. The entire strategy of waging a so-called "stealth" campaign during a special election is questionable.

Second, it appears likely that the lack of specifics provided by the tax measure's supporters harmed its credibility and popularity among some voters. Given that the decision to vie for the tax increase came only months before the election, this may have been more of a factor than the result of a hurried decision making process. In hindsight, the measure's supporters would have done well to take additional time to involve and educate the public about how the tax revenues would be spent. They clearly overestimated the value of the general popularity of MetroLink vis-à-vis deep-seated public skepticism about management by government and the transit agency.

Thus, two conclusions emerge from this defeat: (1) supporters of transit tax measures ought to weigh carefully the dangers of placing them on isolated special election ballots, particularly if they plan to run a low-key "stealth" campaign, and (2) although there may be a potential pitfall over specifying the details of a transportation improvement plan, avoiding them altogether may leave a campaign vulnerable to opponents' likely counterstrategy. Hindsight, of course, makes a complicated and difficult process seem much clearer and simple.

approval of a Light Rail Line in the
Denver Regional Transportation district

introduction and overview

In November of 1999, voters in the Denver Regional Transportation District (RTD) voted to allow the RTD to borrow $457 million with a repayment cost of $779 million to fund a 19.7-mile light rail line along the Southeast I-25 corridor. (Hereafter, this will be termed the "SE Corridor.")

Voters were not authorizing any change in the tax rate in this election. For quite some time, the RTD had been receiving revenues from a .006 sales tax (the equivalent of a 3/5-cent sales tax) within the District for the transit system. Due to both population growth and increases in spending resulting from a robust economy, the RTD estimated that the current tax rate would generate sufficient revenue to fund the SE Corridor light rail line in conjunction with $500 million of federal funding. Nevertheless, voter approval was required under the statewide Taxpayers Bill of Rights (TABOR) through a TABOR override election. TABOR override elections are needed in Colorado to determine if voters give permission to spend, rather than being refunded, tax revenue already collected in excess of TABOR limits.

The success of the 1999 measure, with 66% of voters supporting it, came on the heels of a defeat of a passenger rail measure in 1997, which received support from only 42.2% of voters. The primary difference between the two measures was the magnitude of the light rail system. The 1999 measure included only one light rail line along a single corridor, while the 1997 measure included a comprehensive passenger rail system along at least four corridors. Another interesting feature of the 1999 ballot measure is that it was accompanied by a separate statewide highway measure. Both the decision to include only a single light rail line, and the decision to pair the measure with a separate highway measure, were made to minimize effective opposition from Bill Owens, the conservative Republican governor and other prominent elected officials, one of the two leading newspapers, and others who opposed the 1997 measure. Since these decisions were extremely effective in minimizing opposition, this case demonstrates how a transportation package can be developed in a way that minimizes opposition in an environment in which it is extremely difficult to minimize opposition.

Although this chapter focuses on the 1999 light rail ballot measure, some relevant factors from the 1997 measure will be incorporated into this discussion. Readers interested in a detailed description of the 1997 measure should refer to Haas, Massey, et al. (2000).

background: The RTD and its Board

The RTD includes all or parts of six counties and covers a large geographical area. For example, Boulder, which is approximately 25 miles from Denver, is in the northwest part of the district, and a portion of Douglas County, including parts approximately 30 miles southwest of Denver, is also included.

The RTD Board consists of 15 elected members and is responsible only for transit. Each member represents a designated district and each district is constructed to have approximately the same population. As a result of an election in 1998, the composition of the Board was significantly different in 1999 than it was in 1997. In 1997, less than a majority of the 15 members were strong supporters of the GTR measure and one of the Board members led the opposition to the measure. After the defeat of the 1997 transit measure, an organization entitled Metro Transit, affiliated with the Chamber of Commerce, was formed to recruit and elect pro-transit people in the 1998 election. These results were successful as five of the seven people who were either opposed to GTR or were less than enthusiastic supporters of it, including the lead opponent, were replaced by pro-transit representatives in this election.

background: THe political Environment

The Denver RTD is quite diverse politically. Both Denver and Boulder tend to elect representatives from the Democratic Party while other portions of the district tend to elect representatives from the Republican Party. This political diversity also is represented at the statewide level, particularly in recent gubernatorial elections. The governor during the 1997 ballot measure was a Democrat. In 1998, Bill Owens, a conservative Republican, was elected as governor in a very close election. Transportation was an important issue in this gubernatorial election. Bill Owens' position on transportation focused on highway improvements and, along with other conservative Republicans, he publicly opposed the 1997 RTD ballot measure.

Largely because of the political diversity within the RTD, transportation tends to be a partisan political issue within the RTD, with many Republicans opposing rail transit, particularly when it involves a tax increase. In this environment in which transit is a partisan issue, and also one in which Republicans are influential in many portions of the RTD, proponents wanted to develop a package that would, at a minimum, prevent opposition from Governor Owens and other prominent elected officials from the Republican Party within the RTD. Of course, proponents hoped that they could obtain the endorsement of Governor Owens for the 1999 measure and this hope was fulfilled. For reasons discussed in a later background section, Governor Owens did endorse the 1999 light rail ballot measure.

background: Competing Newspapers

Two newspapers, the Denver Post and the Rocky Mountain News, are strong competitors within the RTD. As might be expected, particularly in a diverse political environment, each newspaper attempts to differentiate itself from the other based in part on its editorial positions. The Rocky Mountain News tends to take a more conservative editorial position than the Denver Post. The two newspapers took different editorial positions on the 1997 measure, with the Post supporting the measure and the News opposing it.

The objective of having the News take either a neutral position or a supportive position on the 1999 measure was not as important as accomplishing this objective with Governor Owens. However, it was perceived that accomplishing this objective with Governor Owens would, at the same time, accomplish it with the News. This perception proved to be accurate, as the News did endorse the 1999 ballot measure.

background: The Transportation System and Voter Attitudes

Prior to the 1999 election, the transit system consisted of buses, including some express buses using expressways and freeways, along with a 5.5-mile light rail system in Denver that had been operating for a few years. In addition, construction of a light rail line along the southwest corridor in Jefferson County, connecting to the southern end of the prior line, was close to completion. No new taxes were needed for the construction of this line.

The metropolitan area had a designated nonattainment status until the mid- 1990s, which precluded significant highway improvements. The only highway changes involved the addition of high-occupancy vehicle (HOV) lanes on a section of I-25 running north from Denver and on C-470, which runs in an east-west direction south of Denver.

According to all people interviewed, the public viewed traffic congestion as a high priority problem that was getting worse. In addition, the SE Corridor, which was the only district corridor identified for improvement in the 1999 transit and highway measures, was viewed as the most congested corridor in the district.

Voter research conducted after the defeat of the 1997 transit measure did not signify a general rejection of light rail. More specifically, the five most common reasons stated by those voting against the 1997 measure are listed below, along with the percentage of voters who stated each reason. Only the fourth of these five reasons represents a general rejection of light rail.

Plan too vague/don't know what it costs/costs keep changing (33%).

Too expensive/too big a project/taxes too high (31%).

Don't trust board to manage it/too much Board conflict/divisions (15%).

Won't be used/won't reduce congestion/won't reduce pollution (11%).

Our area doesn't get benefits/doesn't help suburbs/our area doesn't get light rail (9%).

background: The Southeast Business Partnership (SEBP)

A number of industrial parks are located along the SE Corridor and the number of employees who work on this corridor is approximately the same as the number who work in downtown Denver. One interviewee estimated that approximately 122,000 people work along the SE Corridor, compared with an estimate of 115,000 who work in downtown Denver (Neukirch, March 30, 2000). Many of the employees working along the SE Corridor are concentrated in 15 office parks within a 4.5-mile area.

With the large number of employers in the SE Corridor, it is not surprising that a business association was formed. Different specific associations and partnerships have existed in the recent past, not all of which have consisted strictly of employers. Only the two most recent associations are discussed here.

Shortly after the defeat of the 1997 transit measure, an association entitled the Southeast Corridor Mobility Coalition was formed. This coalition consisted of governmental representatives, as well as representatives from area businesses. It advocated a multimodal solution to the transportation problems on the SE Corridor and lobbied the federal government for funding of a multimodal solution.

The SEBP was formed in January of 1998. Although this organization focused on other issues, such as housing, it continued to advocate a multimodal solution to traffic congestion on the corridor. The SEBP played a major role in fundraising for both the highway and the transit ballot measures. According to a representative of this organization, it also had an important role in encouraging Governor Owens to enthusiastically support the transit measure. This representative argued to Governor Owens that a multimodal solution was needed to obtain federal support. The fundraising ability of this organization may have given it some leverage with the Governor.

background: Governor Owens and the Highway ballot measure

Governor Owens is not a strong supporter of light rail and he also is publicly opposed to tax increases. While he was State Treasurer, he publicly opposed the 1997 RTD transit ballot measure. During the hotly contested 1998 gubernatorial campaign, Bill Owens pledged to oppose tax increases and also pledged to widen I-25 along the SE Corridor, along with additional highway improvements in the state through highway bonding. He did not recommend a multimodal solution involving both highway widening and light rail construction during the campaign. In addition, Governor Owens announced in his inaugural address that he would withdraw $90 million of state funding for light rail promised by the prior governor.

Once elected, Governor Owens tried to move quickly to obtain authority for highway bonding. He thought that working through the legislature would be the fastest route and he was able to obtain legislative approval for his highway bonding proposal. The authority of the legislature to authorize this bonding was successfully challenged in the courts. The court decision was announced around May of 1999. This response from the judicial branch of government forced Governor Owens to try to obtain highway bonding approval through a statewide ballot measure and he moved quickly to get this on the ballot for November of 1999. A significant portion of the $1.7 billion in Governor Owens' statewide measure was proposed for highway widening along the same SE Corridor where the light rail line would be constructed. The highway measure also was successful, with a supporting statewide percentage of 62%.

Two reasons were identified in interviews for the Governor's decision to support the 1999 RTD transit measure even though he opposed the 1997 measure. First, he knew that an environmental impact study of the highway widening on the SE Corridor would need to be redone if the light rail project was not approved. This environmental study would have significantly delayed the highway widening and it also would have given environmentalists another chance to block the highway construction through the environmental impact process. The governor's awareness of the importance of the light rail project to the success of the highway project is demonstrated by his application of pressure on the RTD to act quickly on the SE Corridor light rail project. He exerted this pressure by having the state legislature pass a bill enabling him to control federal transit formula funds if the RTD did not act on the SE Corridor light rail project in a specified time period. Second, he expected the RTD to use their connections with environmental organizations to encourage environmentalists to remain neutral on the highway measure.

background: Voting Procedures

Voters in Arapahoe, Adams, Douglas, and Jefferson Counties were required to vote by mail. Only voters in Denver and Boulder Counties were allowed to vote at actual polling places on Election Day. Mail ballots were mailed to registered voters during the second week of October. Apparently, about 30% of those voting by mail return their ballots within a few days of having received them and about the same percentage tend to return them within a few days of the deadline. The remaining mail ballots are spaced roughly equally over the remaining three weeks or so.

the Transit package and the process used to determine it

The transit package in the 1999 ballot measure was a subset of the one in the 1997 measure. The package in 1999 involved a total of 19.7 miles of light rail. The proposed line started where the current line ends, which is about five miles south of downtown, and ran 15.2 miles in a southeast direction along I-25 through Arapahoe County and part of Douglas County. A 4.5-mile spur along 1-225, running mostly east from the main line also was included in the package.

Reasons for Proposing an Incremental Approach in the 1999 RTD Measure

Although the light rail line included in the 1999 RTD ballot measure was consistent with the recommendation in the SE Corridor MIS, this package was significantly different than the package in the 1997 RTD ballot measure. In addition to the light rail line on the SE Corridor, the 1997 transit package included rail lines on both a west and east corridor, along with unspecified transit improvements along two northern corridors, one through Boulder and the other through Adams County. The estimate was about $8 billion for the 1997 measure compared with an estimate of less than $1 billion with the 1999 measure. The general manager of the RTD still plans to develop a system similar to that in the 1997 measure. However, a decision was made to take an incremental approach rather than a comprehensive approach, with the first new increment being the SE Corridor. Three reasons existed for this decision and each reason is described in the remainder of this subsection.

First, in a post election survey cited previously, the comprehensive nature of the 1997 transit package was a significant concern to many respondents who did not support the 1997 measure. In this study, 31% of those surveyed who voted against the 1997 measure cited the high costs as a primary reason for their opposition. Based on this result, the Board realized that being able to present a transportation package with no tax rate increase would eliminate a primary reason for voter rejection. A tax increase could be avoided only with a single-corridor package.

Second, an incremental single-corridor package was required to obtain support from the Governor and other conservative Republicans who normally would not support an expensive light rail package. Since the Governor had publicly pledged in the 1998 gubernatorial election to avoid any new taxes, he was willing to support the Board's measure only if it did not involve new taxes. Having a transit package that did not require new taxes obviously was possible only with an incremental package. Having the support of Governor Owens was valuable for a few reasons. First, his support, which was mentioned in advertising, was said to be important in a pre-election survey by many voters who opposed the 1997 measure. Second, multiple interviewees indicated that the Governor was helpful in persuading conservative Republicans, who were opposed to the 1997 measure, to support the 1999 measure. Given that two of the main claims advanced by proponents, namely less highway congestion and less air pollution, were not convincingly supported by the MIS, minimizing opposition was important.

Although it probably was a secondary factor considered by the RTD in their decision to have a single-corridor package, this approach was needed to obtain the editorial support of the Rocky Mountain News. The Denver Post had supported the 1997 comprehensive measure in editorials. However, the Rocky Mountain News opposed the 1997 measure in their editorials, due largely to the high costs which required a significant tax increase for a lengthy period of time. Proponents essentially knew that this newspaper would support a transit package that focused only on the congested SE Corridor and did not involve a tax increase. Proponents used this support in one of their two television ads, as well as in their mailers and literature handouts. A lead political consultant indicated that they would not have identified the support of the Denver Post in these ads if the Rocky Mountain News had opposed the 1999 measure.

the communications Campaign used by Proponents

The Communications Budget

The total budget proposed by the political consultants in 1999, consisting of slightly less than $1 million, was very similar to the one proposed in 1997. This goal was met in 1999 while only about 50% of the goal was met in 1997. Somewhat surprisingly, the fundraising efforts were more successful in 1999 even though the single-corridor package seemingly reduced the pool of potential contributors. In addition, the 1999 fundraising efforts commenced in August, which allowed only about three months for these efforts. The fundraising efforts were organized and lead primarily by business organizations, namely the SEBP, the SE Metro Chamber of Commerce, the Denver Metro Chamber of Commerce, and the Downtown Denver Partnership. The fundraising efforts were more successful in 1999 than in 1997 for three reasons.

People who were less enthusiastic about transit improvements than highway improvements contributed to the transit campaign, as well as to the highway campaign, realizing that an additional environmental impact study of the highway widening on the SE Corridor could be avoided only if voters approved the transit measure. The fact that both campaigns shared their list of potential contributors indicates that fundraising efforts were perceived to be synergistic rather than competitive.

Lists of contributors from 1997 were used by fundraisers in 1999. This made fundraising efforts more efficient and focused. Efficiency was needed given the limited amount of time involved for fundraising.

Opponents were less successful in discouraging financial support for the 1999 ballot measure than for the 1997 measure. In 1997, the lead opponent, who also was a member of the RTD Board, was successful in his efforts to have the Board pass a resolution that prevented the district from entering into any financial relationship with any organization that contributed more than $100 to the 1997 measure. Even though this resolution was declared unconstitutional shortly before the 1997 election, it had a chilling effort on fundraising. This chilling effect was compounded by significant newspaper coverage of this issue, including publication of names of some contributors with potential vested interests.

Advertising

Television was the primary medium used for advertising. Radio, direct mail, and hangers also were used. Each of these four types is discussed below.

Approximately $540,000 was spent on television advertising. The consultants thought it would be most effective to use an emotional appeal in their advertising and concluded that television would be more appropriate in conveying emotions than other media would be.

Copies of each of the two television ads used were provided by the political consultants. The two ads had some common elements along with some differences.

Both stressed that no new taxes would be needed.

Both mentioned and showed highway congestion problems and linked these problems to a poor quality of life. For example, one ad said, "Congestion takes time away from our businesses and our families."

Both ads used positive visual symbols to implicitly link the transit measure with quality of life. For example, one ad showed a boy riding a bike at the beginning of the ad and said, "People live here for a better life." The other ad showed beautiful flowers in a mountain valley and said, "This is Denver, but what makes it great is what we do to keep it great."

Neither ad stated that the proposed light rail line would significantly lessen congestion. However, both implied this. For example one ad first presented the above quote about congestion and then immediately stated, "R

Factors Influencing Voting Results of Local Transportation Funding Initiatives with a Substantial Transit Component: Case Studies of Ballot Measures in Eleven Communities

FHWA/CA/OR-2001-26

table of contents

list of tables