Survey of San Diego Region Voters Regarding the Potential Extension of the TransNet Program

Transcription

1 Survey of San Diego Region Voters Regarding the Potential Extension of the TransNet Program Conducted for SANDAG May 2001

2 Table of Contents Table of Contents Introduction...1 Introduction to Study...3 Key Findings in Comparison... Conclusions and Recommendations Methodology... 1 Summary of Results Issues of Importance - Transportation Only Measure First Ballot Test - Transportation Only Measure Tax Durations - Transportation Only Measure Information about TransNet - Transportation Only Measure Spending Projects - Transportation Only Measure Second Ballot Test - Transportation Only Measure... 3 Arguments Supporting the Transportation Only Measure Arguments Opposing the Transportation Only Measure... 1 Third Ballot Test - Transportation Only Measure... 5 Additional Information About Transportation Only Measure... 9 Preferred TransNet Funding Distribution - Transportation Only Measure Expanding TransNet Program - Transportation Only Additional Demographic and Behavioral Measures - Transportation Only Measure Issues of Importance - Combined Measure... 0 First Ballot Test - Combined Measure... 2 Tax Durations - Combined Measure... 7 Information about TransNet - Combined Measure... 9 Spending Projects - Combined Measure Second Ballot Test - Combined Measure Arguments Supporting the Combined Measure Arguments Opposing the Combined Measure Third Ballot Test - Combined Measure... 8 Additional Information About Combined Measure Preferred TransNet Funding Distribution - Combined Measure Additional Demographic and Behavioral Measures - Combined Measure Appendix A...see Appendix A Tab Appendix B... see Appendix B Tab Appendix C... see Appendix C Tab Appendix D... see Appendix D Tab Page i

3 List of Tables List of Tables Table 1. (Q1) Issues of Importance - Comparison... Table 2. (Q2) First Ballot Test - Comparison...5 Table 3. (Q3) Emphasize Extension - Comparison...5 Table. (Q) Tax Duration - Comparison... Table 5. (Q7) Spending Projects - Comparison...7 Table. (Q8) Second Ballot Test - Comparison...7 Table 7. (Q9) Positive Arguments - Comparison...8 Table 8. (Q11) Third Ballot Test - Comparison...9 Table 9. Methodology... 1 Table 10. Number of Voters by Supervisorial District Table 11. Subgroup Labels... 1 Table 12. Guide to Statistical Significance Table 13. First Ballot Test Transportation Only Version by Supervisorial District and County Area Table 1. Means Questions and Corresponding Scales Table 15. Issues of Importance by Supervisorial District and County Area Table 1. First Ballot Test by Supervisorial District and County Area... 2 Table 17. First Ballot Test by Party... 2 Table 18. First Ballot Test by Household Party Type and Gender Table 19. First Ballot Test by Age Table 20. First Ballot Test by Ethnicity... 2 Table 21. Support for Measure as an Extension by Supervisorial District and County Area Table 22. TransNet Information by Supervisorial District and County Area Table 23. Proposed Spending Projects by Supervisorial District and County Area Table 2. Second Ballot Test by Supervisorial District and County Area Table 25. Second Ballot Test by Party Table 2. Second Ballot Test by Household Party Type and Gender... 3 Table 27. Second Ballot Test by Age... 3 Table 28. Second Ballot Test by Ethnicity Table 29. Arguments Supporting the Measure by Supervisorial District Table 30. Arguments Supporting the Measure by County Area... 0 Table 31. Arguments Opposing the Measure by Second Ballot Test... 2 Table 32. Arguments Opposing the Measure by Supervisorial District... 3 Table 33. Arguments Opposing the Measure by County Area... Table 3. Third Ballot Test by First Ballot Test... Table 35. Third Ballot Test by Supervisorial District and County Area... Table 3. Third Ballot Test by Party... 7 Page ii

4 List of Tables Table 37. Third Ballot Test by Household Party Type and Gender... 7 Table 38. Third Ballot Test by Age... 8 Table 39. Third Ballot Test by Ethnicity... 8 Table 0. Effect of Funding Additional Projects by Third Ballot Test Table 1. Issues of Importance by Supervisorial District and County Area... 1 Table 2. First Ballot by Supervisorial District and County Area... 3 Table 3. First Ballot Test by Party... 3 Table. First Ballot Test by Household Party Type and Gender... Table 5. First Ballot Test by Age... Table. First Ballot Test by Ethnicity... 5 Table 7. Support for Measure as an Extension by Supervisorial District and County Area... Table 8. TransNet Information by Supervisorial District and County Area Table 9. Proposed Spending Projects by Supervisorial District and County Area Table 50. Second Ballot Test by Supervisorial District and County Area... 7 Table 51. Second Ballot Test by Party... 7 Table 52. Second Ballot Test by Household Party Type and Gender Table 53. Second Ballot Test by Age Table 5. Second Ballot Test by Ethnicity... 7 Table 55. Arguments Supporting the Measure by Supervisorial District Table 5. Arguments Supporting the Measure by County Area Table 57. Arguments Opposing the Measure by Second Ballot Test Table 58. Arguments Opposing the Measure by Supervisorial District Table 59. Arguments Opposing the Measure by County Area Table 0. Third Ballot Test by First Ballot Test Table 1. Third Ballot Test by Supervisorial District and County Area Table 2. Third Ballot Test by Party... 8 Table 3. Third Ballot Test by Household Party Type and Gender... 8 Table. Third Ballot Test by Age Table 5. Third Ballot Test by Ethnicity Table. Question 13a by 13b Page iii

5 List of Figures List of Figures Figure 1. Issues of Importance Figure 2. First Ballot Test Figure 3. Support for Measure as an Extension... 2 Figure. Tax Durations Figure 5. Support for Permanent Measure* Figure. TransNet Information Figure 7. Proposed Spending Projects Figure 8. Second Ballot Test... 3 Figure 9. Arguments Supporting the Measure Figure 10. Arguments Opposing the Measure... 1 Figure 11. Third Ballot Test... 5 Figure 12. Additional Information About Measure... 9 Figure 13. Support for 3/ Cent Sales Tax for Transportation Projects Figure 1. Preferred TransNet Funding Distribution Figure 15. Effect of Funding Additional Projects on Likelihood of Support Figure 1. Support for 3/ Cent Sales Tax for Additional Projects... 5 Figure 17. Household Children Under Figure 18. Ethnicity Figure 19. Commute to Work or School... 5 Figure 20. Length of Commute... 5 Figure 21. Mode of Commute... 5 Figure 22. Household Income Figure 23. Gender Figure 2. Age Figure 25. Party Figure 2. Household Party Type Figure 27. Supervisorial District Figure 28. County Area Figure 29. Issues of Importance... 0 Figure 30. First Ballot Test... 2 Figure 31. Support for Measure as an Extension... 5 Figure 32. Tax Durations... 7 Figure 33. Support for Permanent Measure*... 8 Figure 3. TransNet Information... 9 Figure 35. Proposed Spending Projects Figure 3. Second Ballot Test Page iv

6 List of Figures Figure 37. Arguments Supporting the Measure Figure 38. Arguments Opposing the Measure Figure 39. Third Ballot Test... 8 Figure 0. Additional Information About Measure Figure 1. Agreement With Statements About the Measure Figure 2. Support for 3/ Cent Sales Tax for Additional Projects Figure 3. Preferred TransNet Funding Distribution Figure. Household Children Under Figure 5. Ethnicity Figure. Commute to Work or School Figure 7. Length of Commute Figure 8. Mode of Commute Figure 9. Household Income... 9 Figure 50. Gender... 9 Figure 51. Age Figure 52. Party Figure 53. Household Party Type Figure 5. Supervisorial District... 9 Figure 55. County Area... 9 Page v

7 Introduction Introduction (GRA) is pleased to present the results of a voter opinion research project conducted for the San Diego Association of Governments (SANDAG). This report is organized into the following sections: Introduction to Study This section briefly describes the background and motivation for the study. Key Findings in Comparison Two questionnaires were used in this study - one that presented respondents with a measure that will fund transportation projects only, and a second that presented respondents with a measure that would fund SMART growth, open space and stormwater projects in addition to transportation projects. This section presents key findings and a discussion of the results in a comparative format. Conclusions and Recommendations The Conclusions and Recommendations section details our recommended course of action based on the survey results. Methodology The Methodology section explains the methodology used to conduct this type of survey research. This section also explains how to use the detailed crosstabulation tables in Appendix C and Appendix D. Summary of Results In the body of the report, we present a question-by-question analysis of the two surveys. The discussion is organized into the following sections: Issues of Importance - Transportation Only Measure First Ballot Test - Transportation Only Measure Tax Durations - Transportation Only Measure Information about TransNet - Transportation Only Measure Spending Projects - Transportation Only Measure Second Ballot Test - Transportation Only Measure Arguments Supporting the Measure - Transportation Only Measure Arguments Opposing the Measure - Transportation Only Measure Third Ballot Test - Transportation Only Measure Additional Information about Transportation Only Measure Preferred TransNet Funding Distribution - Transportation Only Measure Expanding TransNet Program - Transportation Only Measure Additional Demographic and Behavioral Measures - Transportation Only Measure Issues of Importance - Combined Measure Page 1

8 Introduction First Ballot Test - Combined Measure Tax Durations - Combined Measure Information about TransNet - Combined Measure Spending Projects - Combined Measure Second Ballot Test - Combined Measure Arguments Supporting the Measure - Combined Measure Arguments Opposing the Measure - Combined Measure Third Ballot Test - Combined Measure Additional Information about Combined Measure Preferred TransNet Funding Distribution - Combined Measure Additional Demographic and Behavioral Measures - Combined Measure For the readers convenience, the reports were crafted so that the reader can turn directly to a section and understand the results without having to read prior sections. This results in considerable repetition of language throughout the report. Appendices We have included the following four appendices: Appendix A, which presents the Transportation Only version of the questionnaire with topline results. Appendix B, which presents the Combined version of the questionnaire with topline results. Appendix C, which presents the complete crosstabulations for the Transportation Only version of the questionnaire. Appendix D, which presents the complete crosstabulations for the Combined version of the questionnaire. Page 2

9 Introduction to Study Introduction to Study Prior to 1987, the San Diego region was largely dependent on state and federal funding to implement transportation improvements in the region, and these funding sources were not keeping pace with the increased demand and inflationary trends in construction, operation and maintenance costs. In 1987, voters approved Proposition A, which provided a 1/2 percent sales tax increase dedicated to transportation projects for a period of 20 years. The TransNet program is expected to raise $3.2 billion dollars over the 20-year period, with $1,000,000 per year devoted to bicycle facilities and the remaining budget split evenly between major highway projects, rail and transit, and local street and road projects. Assuming that the TransNet sales tax will expire in 2008 and that other sources, such as the gas tax, remain stable, SANDAG recently estimated that there will be a $12 billion deficit in funding for transportation improvements, operations and maintenance through 2020 (2020 Regional Transportation Plan, p.315). Accordingly, SANDAG has expressed a strong interest in extending the TransNet sales tax both for the funds that it will generate locally and the ability to leverage additional state, federal, local and private sector funding. Along with the need for additional transportation funding, SANDAG has identified funding needs for other regional infrastructure projects including funding to implement regional habitat plans and to improve water quality related to polluted runoff. The region has been engaged in habitat conservation planning for the past ten years. An unfunded need of over $1.25 billion has been identified to purchase habitat property that can not be protected in any other way and to monitor and manage the habitat land once acquired. Significant funding will also be needed to implement local jurisdiction water quality plans to reduce stormwater runoff in compliance with the Clean Water Act and Regional Water Quality Control Board municipal stormwater permits. The survey research was also intended to determine voters willingness to support an expansion of the TransNet program to include funding for these purposes in addition to transportation. To study the feasibility of extending the TransNet sales tax, SANDAG hired Godbe Research & Analysis to conduct focus group and survey research. The findings from the focus group research were presented in April, 2001 to SANDAG in the TransNet Focus Groups report. This report presents the findings from the survey research components of the study. Page 3

10 Key Findings in Comparison Key Findings in Comparison In addition to assessing the feasibility of extending the TransNet sales tax, one of the key objectives of this research was to determine whether the scope of projects that may be funded by the sales tax should be expanded. Currently, the TransNet program is devoted to transportation projects, including related environmental mitigation. Some have proposed that the scope of projects that may be funded by the sales tax be expanded to include the acquisition and preservation of open space and natural habitats, the reduction of pollution caused by stormwater runoff, and SMART growth. i In order to test whether voters prefer to extend the TransNet sales tax in its current form, or a version that would fund a wider scope of projects, GRA fielded two questionnaires. The first questionnaire (Transportation Only version) presented respondents with a measure that would fund transportation projects only. Alternatively, the second questionnaire (Combined version) presented respondents with a measure that would fund the acquisition and preservation of open space and natural habitats, stormwater improvements, and SMART growth in addition to transportation projects. The surveys were designed so that direct comparisons could be made between the two different measures on key questions. Accordingly, this section presents the key findings of the surveys on these central questions in a comparative format. For a quick overview of all of the survey findings, the reader should turn to the topline reports in appendices A and B. For a detailed question-by-question analysis of the results, the reader should consult the Summary of Results. Most Important Issues The first substantive question of the survey presented respondents with 13 issues facing residents of the San Diego region and asked them to rate the importance of each issue. This set of questions not only provides insight into how important an issue is on a scale of importance, it also provides a relative ranking among the issues. Table1 shows that the top four issues for each study are the same, although their relative ranking is somewhat different. Table 1. (Q1) Issues of Importance - Comparison Transportation Only 1. Improving the quality of public education 2. Reducing crime 3. Reducing traffic congestion. Maintaining local streets and roads Combined 1. Reducing crime 2. Reducing traffic congestion 3. Improving the quality of public education. Maintaining local streets and roads i It is worth noting that special enabling legislation is required before SANDAG could expand the scope of projects beyond transportation related projects. Page

11 Key Findings in Comparison First Ballot Test Early in the survey, respondents were presented with the ballot language for one of the measures and asked to indicate whether they would support or oppose the measure if the election were held today. Known as the first ballot test, this question assesses support for the measure without first priming the respondent with information beyond what is presented in the ballot language itself. As such, it represents a reliable measure of support for the proposed sales tax extension among an uninformed electorate. Table2 summarizes the results for the two different measures for the first ballot test. Overall, the measures received similar support levels from voters. The Transportation Only measure received 1 percent support, whereas the Combined measure received support from 58 percent of respondents. Given the statistical margin of error associated with the surveys, the results are a statistical tie. Table 2. (Q2) First Ballot Test - Comparison Transportation Only 33.2% definitely yes 27.7% probably yes 9.5% probably no 2.3% definitely no 5.3% DK/NA Combined 32.1% definitely yes 25.9% probably yes 12.7% probably no 23.7% definitely no 5.% DK/NA Emphasize Extension After reading the ballot language to each respondent in Question 2, the next question assessed the impact of emphasizing that the measure would simply extend the existing TransNet sales tax that was approved by voters in 1987 and that it would not raise the sales tax in the region. For respondents who received the Combined version of the survey, the question also noted that the measure would extend the scope of projects that may be funded by the TransNet sales tax to include open space, stormwater improvements and growth management. Overall support for extending the TransNet sales tax increased significantly for both versions of the measure. As shown in Table3, 70 percent of respondents stated that they would support the Transportation Only version of the measure, whereas 71 percent of respondents who received the Combined measure indicated that they would support the measure. Once again, the support levels for the respective versions of the measure at this point in the survey are, statistically speaking, the same. Table 3. (Q3) Emphasize Extension - Comparison Transportation Only 1.1% definitely yes 29.1% probably yes.9% probably no 18.0% definitely no.9% DK/NA Combined 38.8% definitely yes 32.% probably yes 8.7% probably no 15.1% definitely no.8% DK/NA Page 5

12 Key Findings in Comparison Tax Durations The ballot language tested in Question 2 indicated that the sales tax extension would not exceed 30 years. The purpose of Question was to estimate how support for the sales tax extension varies depending on the actual length of the tax extension. Question presented respondents with various tax durations in a Dutch Auction format, which involves reading the longest duration first (in this case, 30 years) and gauging willingness to support the measure, then the next longest tax duration, and so on. Naturally, the longer the sales tax extension, the less willing some voters are to support extending the tax. Table shows the percentage of respondents who indicated that they would definitely or probably vote yes on the measure if it were to last the indicated number of years. Comparatively speaking, the Combined measure received slightly higher support at the longer tax durations, whereas the Transportation Only measure received greater support at the shorter tax durations among respondents. Once again, however, the results between the two measures are, statistically speaking, not significantly different. The interested reader should refer to the Conclusions and Recommendations section for a more detailed discussion of issues surrounding the duration of the proposed tax extension. Table. (Q) Tax Duration - Comparison Transportation Only 30 years: 53% 25 years: 55% 20 years: 57% 15 years: 2% 10 years: 9% Combined 30 years: 9% 25 years: 8% 20 years: 51% 15 years: 57% 10 years: % Spending Projects As noted above, the central difference between the two questionnaires was the scope of the projects that may be funded by the measure. One measure would continue to fund transportation projects only, whereas the other measure would fund the acquisition and protection of open space and natural habitats, stormwater improvements, and SMART growth in addition to transportation projects. It was of particular interest, therefore, to discover which types of projects were most popular with voters. The Transportation Only version of the questionnaire presented respondents with nine specific transportation projects and, for each project, asked respondents to indicate whether they were more or less likely to support the measure upon hearing that some of the money from the TransNet sales tax could be used to fund the project. The Combined version of the survey presented respondents with most of the transportation projects tested in the Transportation Only version, as well as five projects that did not relate to transportation. Page

13 Key Findings in Comparison Table5 presents the overall ranking of the top six spending projects for both studies. It is worth noting that in the Combined version of the survey, all but one of the eight most popular projects are transportation projects. Table 5. (Q7) Spending Projects - Comparison Transportation Only 1.Improvements to region s highway system 2. Reduced fares for the elderly, students and disabled 3. Adding lanes to state highways. Adding carpool lanes and transit service on state highways 5. Transit operations to keep fares down for all passengers. Maintenance, improvement, and expansion of local streets and roads Combined 1. Improvements to the region s highway system 2. Reduced fares for the elderly, students and disabled 3. Controlling runoff of pollution into local water sources. Expansion of public transit system 5. Adding carpool lanes and transit service on state highways. Adding lanes to state highways Second Ballot Test After providing respondents with the ballot language, the possible durations of the tax extension, background information about the existing TransNet program, and the possible projects that may be funded by extending the TransNet sales tax, the survey once again presented respondents with the ballot language and asked if they would support the measure to extend the TransNet sales tax. The second ballot test is useful in assessing the way in which support for the sales tax extension may change once voters are presented with basic information about the existing TransNet program and what it may fund if extended. Table presents the results of the second ballot test for both measures. Once again, the results are essentially the same for both measures. Approximately 71 percent of respondents would vote yes on the Transportation Only measure at this stage in the survey. The corresponding figure for the Combined measure is percent. Table. (Q8) Second Ballot Test - Comparison Transportation Only 1.8% definitely yes 29.2% probably yes 5.7% probably no 17.8% definitely no 5.5% DK/NA Combined 3.% definitely yes 29.8% probably yes 8.3% probably no 17.3% definitely no 8.2% DK/NA Arguments about the Measure Ballot measures do not succeed or fail in a political vacuum. Proponents of a measure will present arguments to try to persuade voters to support a measure, just as opponents will present arguments to achieve the opposite effect. To simulate a campaign environment and to gauge the persuasive power of various arguments, the survey presented respondents with Page 7

14 Key Findings in Comparison strong arguments both in favor of, and in opposition to, the measure. Although some arguments were included in both versions of the survey, many of the arguments tested were specific to a single survey, which prevents an apples to apples comparison of the relative ranking of arguments. The arguments that generated the most support for the measure among respondents, as a whole, are shown in Table7 for each survey. One of the most important findings from both surveys is that despite presenting respondents with strong, negative arguments, most of the arguments designed to reduce support actually increased support for the measure overall. An explanation for this pattern can be found in the appropriate sections of the report, whereas the implications of this finding are discussed in the Conclusions and Recommendations section of this report. Only one of the arguments tested actually resulted in a net decrease in support for the measure: This measure will encourage growth and eventually make San Diego look like Los Angeles. It is worth noting, however, that the overall effect is quite small. Table 7. (Q9) Positive Arguments - Comparison Transportation Only 1. If the measure is extended, the funds will be focused on relieving traffic congestion throughout the County. 2. If the measure is not extended, traffic congestion in the County will become much worse. 3. All of the money raised in the County will be spent only in San Diego County.. Extending the tax will produce immediate benefits, and it will allow the County to fix problems before they become bigger and more expensive to fix. 5. San Diego s transportation system is a major factor in maintaining the County s economic health, and this measure will keep it from deteriorating. Combined 1.This measure is needed to clean up and protect San Diego s lagoons, creeks, bays and beaches. 2. If the measure is not extended, traffic congestion in the County will become much worse. 3. Traffic congestion would be a lot worse today without the improvements that were funded by the current TransNet sales tax.. All of the money raised in the County will be spent only in San Diego County. 5. There will be a clear system of accountability, including independent audits to ensure that the money is spent appropriately. Third Ballot Test After providing respondents with the wording of the proposed measure, the possible durations of the tax, background information about the existing TransNet program, a list of projects that may be funded by the measure, as well as arguments both in favor and against the measure, respondents were once again asked whether they would vote yes or no on the measure that would extend the TransNet sales tax. Because respondents learn a great deal more about the proposed measure during the survey process, the survey simulates a campaign environment. Therefore, the third ballot test is useful in assessing the way in which support for the proposed measure may change once voters are presented with basic information and arguments concerning the measure. Page 8

15 Key Findings in Comparison As with the previous ballot tests that were conducted at earlier points in the survey, the results of the third ballot tests indicate that respondents continued to be equally supportive of the respective measures. Table8 shows that support for the Transportation Only and Combined measures at this point in the survey can be found among 70 and 7 percent of respondents, respectively. Once again, given the statical margin of error associated with the two surveys, these results are a statistical tie. Table 8. (Q11) Third Ballot Test - Comparison Transportation Only 1.% definitely yes 28.% probably yes 5.7% probably no 18.1% definitely no.3% DK/NA Combined 35.7% definitely yes 31.1% probably yes 5.8% probably no 19.7% definitely no 7.8% DK/NA Page 9

16 Conclusions and Recommendations Conclusions and Recommendations Based on an analysis of the survey data, GRA offers the following conclusions and recommendations to SANDAG. Should SANDAG move forward with plans to place a measure on the ballot to extend the TransNet sales tax? The favorability of voters opinions regarding the TransNet sales tax represents just one of many factors that SANDAG will undoubtedly consider in deciding whether, and when, to place a measure on the ballot to extend the TransNet sales tax. However, to the extent that voters opinions weigh into SANDAG s decision calculus on these matters, the results of the survey suggest that SANDAG should move forward with plans to place a measure on the ballot to extend the TransNet sales tax. With nothing more than the information they had prior to the survey and the ballot language tested, a majority of respondents in both surveys were willing to support extending the sales tax. Once informed that the measure would simply extend the existing TransNet sales tax beyond 2008 and that it would not raise the tax rate in the County, support for the measure climbed to more than two-thirds for both versions of the measure. Support for extending the measure remained at or above the two-thirds threshold at each of the subsequent ballot tests in the survey, even after presenting respondents with strong arguments designed to persuade them not to support the measure. Barring changes in the economic and political arenas that lead to a much different election environment than the present, these findings indicate that the TransNet sales tax extension measure has a very good chance of passing in November 2002 if packaged correctly and coupled with an effective public information campaign. Should the TransNet program continue to focus on just transportation projects, or should the scope of projects be expanded to include projects like open space, stormwater improvements, and SMART growth? This is a difficult question to answer relying solely on the results of this study. On the one hand, this study makes it clear that it is not necessary to include issues like open space, stormwater improvements, or SMART growth in the TransNet extension measure in order to garner two-thirds support for the measure. Transportation issues such as reducing traffic congestion and maintaining streets and roads are among the most important issues to San Diego voters, on par with issues such as improving public education and reducing crime. Transportation projects, such as improving and expanding the region s highway system, expanding the public transit system, and maintaining local streets and roads, also represented seven of the top eight projects that voters were most interested in funding with Trans- Net revenues. Moreover, the Transportation Only version of the measure appeared to have a more even appeal to voters across partisan subgroups, with Republican respondents somewhat more likely to support this version of the measure when compared to the Combined version. On the other hand, voters -- as a group -- were equally supportive of the measure that would extend the TransNet sales tax and broaden the scope of projects that may be funded by TransNet to include open space, stormwater improvements and SMART growth. Of all of the Page 10

17 Conclusions and Recommendations arguments tested in the surveys, the most persuasive argument in favor of the measure was related to stormwater runoff improvements. Furthermore, the vast majority of respondents who received the Combined version of the measure agreed that transportation improvements should be balanced with the County s need to protect open space, improve water quality and manage growth. Thus, although it is clear that both types of measures can win if packaged correctly and combined with an effective public information campaign, the findings of the study do not provide a compelling case that one version of the measure is more likely to be successful than the other, all other things being equal. Of course, there are other factors that SANDAG will undoubtedly consider when deciding which approach is the best, including the need for enabling legislation before a Combined measure can be placed on the ballot, whether dividing the revenue across multiple project areas will allow sufficient funds to meet the transportation needs as well as the needs in these other areas, political considerations associated with building a coalition, and the availability of alternative sources of funding for the scope of projects being considered. Should SANDAG increase the tax rate to 3/ percent? The survey results indicate that raising the tax rate to 3/ percent will result in a substantial reduction in support for extending the TransNet tax, and is therefore not advised. Support for a 3/ percent sales tax, regardless of the scope of projects that may be funded by the measure, was far less than the needed two-thirds level. For how long should the tax be extended beyond 2008? One of the more interesting findings from this research focused on the discrepancy between voters willingness to support the TransNet extension to begin in the year 2008 and not to exceed 30 years and their willingness to support the tax extension when the actual length of the tax extension was specified. Whereas more than two-thirds of respondents were willing to support either measure when provided ballot language which stated that the extension would not exceed 30 years, ii many of these same respondents were unwilling to support the measure when the length of the measure was set at 30 years. In fact, less than 0 percent of respondents were willing to support the tax extension beyond 2008 if they knew that it would last for 20 years. There are two possible explanations for this pattern. One the one hand, ballot language always plays an important role in shaping voters willingness to support a measure. The ballot language that was used in this survey was very similar to the language of Proposition A, which established the TransNet sales tax in 1987, in that it also included the not to exceed language. If the language is indeed responsible for the discrepancies noted above, the appropriate duration for the tax extension depends critically on the wording of the ballot language ii These levels of support were found at the second and third ballot tests. Page 11

18 Conclusions and Recommendations selected for the TransNet extension measure, as well as the information that will be presented in the ballot arguments. On the other hand, the discrepancies may be due to artificially focusing respondent s attention on the length of the tax extension in Question when, in the case of a sales tax, voters are normally focused on the tax rate and the scope of projects that may be funded. Because the appropriate tax duration depends critically on which of these explanations is valid, further research on this matter is highly recommended. GRA will be happy to discuss these matters further with SANDAG and determine the best course of action. What project areas should be considered priorities to be funded by the TransNet sales tax? One of the principal tasks of the survey was to identify the projects that voters indicate are priorities for funding if the TransNet measure is extended. Once again, voter opinion is just one of the many considerations that SANDAG will likely consider when deciding the funding priorities in the region. The time and resources necessary to complete a given project, for example, must be balanced with voters desire for a project. Nevertheless, the results of the surveys certainly indicate which projects and project areas are popular with voters and generate support for the proposed sales tax extension. From an electoral perspective, the goal is to craft a plan that will receive the greatest level of support from voters overall. Generally speaking, this involves selecting projects that have widespread appeal across geographic areas as well as other subgroups of voters. Considering the results from the Combined measure as it presented respondents with transportation projects as well as open space, stormwater and SMART growth projects, it is clear that transportation projects are the highest priority for voters. Using the results from this survey, and assuming a Combined measure, the top priorities for TransNet funding should be: 1. Continued improvements to the region s highway system to help reduce traffic congestion. 2. Reduced transit fares for the elderly, students and the disabled. 3. Controlling runoff of pollution into streams, lagoons, bays and the ocean.. A significant expansion of the region s public transit system to help reduce traffic congestion. 5. Adding carpool lanes and expanding transit services on state highways such as Interstates 5, 15 and Adding lanes to state highways, such as Interstates 5, 15 and The maintenance, improvement and expansion of the region s local street and road network. 8. Transit operations to keep fares down for all passengers. 9. The preservation of wildlife areas to protect endangered species. Page 12

19 Conclusions and Recommendations 10. The acquisition and protection of lands in watershed areas. How might information affect support for the proposed measure? Voters opinions about ballot measures are often not well informed, especially when the amount of information presented to the public on a measure is limited. It was clear from the focus group research that awareness of the existing TransNet program and associated sales tax is very low. As a consequence, one would expect that many voters opinions about the TransNet sales tax measure are susceptible to change in the presence of additional information. One of the tasks of this survey was to identify how voters support for the proposed sales tax extension was influenced by a variety of information items pertaining to the measure, including strong arguments both in favor and against the measure. The most influential pieces of information about the measure are A) that the County already has a one-half percent sales tax for transportation projects called TransNet that was approved by voters in 1987, B) that the proposed measure would simply extend the existing TransNet sales tax, and C) that the proposed measure would not raise the sales tax in the County. Collectively, these information items, in combination with the ballot language tested in the survey, resulted in greater than two-thirds support for the measure without any additional information. Although subsequent information items did have an impact on voters opinions about the measure, for all but a small percentage of respondents the additional information simply reinforced their earlier voting intentions. As noted in the Summary of Findings, respondents appeared to form their opinions about the measure after being provided these initial items of information and were simply not receptive to subsequent information items that countered their established opinions. In fact, respondents opinions about the measure strongly colored the way they viewed the information presented. This was most clearly seen with respect to the arguments that were designed to reduce support for the measure. Whereas voters who were inclined to oppose the measure found these arguments as compelling reasons to oppose the measure, respondents who favored the measure at this point in the survey found the same information as compelling reasons to support the measure. This finding has important implications for those who intend to design a public education effort to educate, inform and/or shape voters opinions with respect to extending the Trans- Net sales tax.if voters opinions about the measure are resilient once provided with basic information about the measure, and their opinions about the measure strongly color how they perceive additional information about the measure, then subsequent efforts to persuade voters will meet with only marginal success. For those who support extending the sales tax, this is good news. More than two-thirds of voters were in support of the measure once provided with basic information, and the survey suggests that their support is not vulnerable to arguments against the measure. Page 13

20 Methodology Methodology Research Objectives At the outset of this project, SANDAG and GRA identified several research objectives for this study. Broadly defined, SANDAG was interested in using survey research to: Assess the feasibility of extending the TransNet sales tax Determine the optimal scope of projects to be funded by the measure Determine the specific projects that are most popular with voters Analyze how attitudes and opinions about the TransNet sales tax may change as a result of being educated on certain issues Identify the information items that increase and decrease support for the measure, and Estimate the optimal tax duration and tax rate for the measure Methodology Table9 briefly outlines the methodology employed in this project. Two versions of the survey were used -- one which presented a measure that would fund transportation projects only, and another which presented a measure that would fund open space, habitat preservation, SMART growth and stormwater improvements in addition to transportation projects. A total of 1,009 likely November 2002 voters in the San Diego region completed an interview in English or Spanish, representing a total universe of approximately 897,337 likely November 2002 voters in the region. iii Interviews were conducted on May through May 19, 2001, and each interview typically lasted 15 minutes. Table 9. Methodology Technique Telephone interviewing in English and Spanish Interview Length 15 minutes Universe Likely November 2002 voters in the San Diego Region Field Dates May through May 19, 2001 Sample Size 50 Transportation Only Version; 502 Combined Version Questionnaire Design After reviewing the findings of the focus group research and considering the various issues involved, particularly the need to identify whether voters prefer to dedicate the revenue to transportation projects only or expand the scope of projects that may be funded by the tax, GRA recommended conducting two surveys. The first survey presented respondents with a measure that will fund transportation projects only (hereafter referred to as Transportation Only Version or Transportation Only Measure). The second survey presented respondents with a measure that would fund SMART growth, open space and stormwater projects in addiiii San Diego region includes all incorporated and unincorporated areas within San Diego County. Page 1

21 Methodology tion to transportation projects (hereafter referred to as Combined Version or Combined Measure). Sample Choosing the appropriate sampling design for a study is a careful process that involves detailed consideration of the research objectives. In the present study, the main goal was to determine the feasibility of extending the existing TransNet sale tax beyond the year Because SANDAG was interested in the November 2002 election as the earliest possible date to place the measure on the ballot, it was important to assess the level of support for the measure among voters who are likely to participate in the November 2002 election. Based on historical data and the voting history of each registered voter in the region, GRA identified a total of 897,337 San Diego County voters as being likely to participate in the November 2002 election. This group of voters formed the sampling universe for the study. The universe of 897,337 likely November 2002 voters was stratified on the basis of Supervisorial District, gender, age, and party affiliation. Two unique and mutually exclusive samples were created, one for each version of the study. For each sample, five hundred clusters were formed and voters were randomly selected into a cluster based on their district-gender-ageparty profile. During data collection, individuals were sampled randomly from each cluster, and one interview was completed per cluster. A total of nine additional interviews were completed due to overlapping interviewing within clusters on the last night of interviewing. Table 10. Number of Voters by Supervisorial District Supervisorial District Total Registered Voters Likely November 2002 Voters One 20,05 123,753 Two 288, ,71 Three 39,70 2,935 Four 22,80 12,017 Five 25, ,11 TOTAL 1,335,392 a 897,337 b a. This total reflects the most recent (February 2001) data from the San Diego County Registrar. This purged file was not available in a usable format at the time of constructing the sample, so the universe was drawn from the December 2000 file which contained 1,33,27 registered voters. b. Based on the December 2000 file which contained 1,33,27 voters in San Diego County, the number of likely November 2002 voters represented 2.% of registered voters. Because of the research objectives motivating this study and SANDAG s interest in being able to make reliable estimates of opinions not only county-wide, but also within and between Supervisorial Districts, the sampling frame employed a strategic oversample of voters by Supervisorial District. Specifically, for each survey 100 interviews were completed in each of the five Supervisorial Districts. However, because voting registration and voting propensity is not evenly distributed across the Supervisorial Districts (see Table10), the final data were weighted to adjust for the strategic oversampling. The strategic oversample allowed for more reliable comparisons between Supervisorial Districts, whereas weighting the data ensures Page 15

22 Methodology that the final results are representative of the likely November 2002 voting universe for the region as a whole as well as within Supervisorial Districts. Subgroup Labels The following subgroup labels are used in the report and crosstabulation tables: Table 11. Subgroup Labels Age Respondents were grouped in the following age categories: 18-29, 30-39, 0-9, 50- and 5+ (5 and older). Children under 18 Interview Language Ethnicity Gender Household Income Household Party Type Party Commute Commute Length County Area Supervisorial District Voters were grouped based on whether or not they have any children under the age of 18 living in their household. Voters were classified as having taken the survey in English or Spanish. Voters were classified by their ethnic backgrounds: Latino/Hispanic, Asian, African American, Caucasian, and Other. Male and Female respondents were identified with separate labels. Voters were classified into the following household income categories: Under $25,000, $25,000-$0,000, $0,000-$50,000, $50,000- $0,000, $0,000-$70,000, $70,000-$80,000, $80,000-$90,000, $90,000-$100,000, Over $100,000. Individuals were grouped into the following household party types: Democrat 1 --one Democrat in the household, Democrat 2+ --two or more Democrats, Republican 1 --one Republican, Republican 2+ --two or more Republicans, Other --Other partisans, Mixed -- household members who differ in their partisanship. Individuals were grouped according to their political party affiliation: Democrat, Republican, Other or Decline to State ( DTS ). Respondents were classified according to whether or not they commute to work or school: Yes - No. Respondents were classified according to the length of their commute to work or school: less than 10 minutes, minutes, minutes, 31-5 minutes, 5-0 minutes, more than 0 minutes. Respondents were grouped according to the area of the County in which they live: North, Central, or East/South. Respondents were grouped according to the Supervisorial District in which they live: One, Two, Three, Four or Five. Randomization of Questions To avoid the problem of systematic position bias -- where the order in which a series of questions is asked systematically influences the answers to some of the questions -- several of the questions in this survey were randomized such that respondents were not consistently asked the questions in the same order. The series of items in Questions 1,, 7, 9, 10, and 12, as well as question sets 9 and 10, were randomized in both versions of the questionnaire. For the Transportation Only version, the items in Question 1 were also randomized. For the Combined version, the items in Question 13 and 15 were randomized Page 1

23 Methodology Understanding the Margin of Error Because a survey typically interviews a limited number of people who are part of a larger population group, by mere chance alone there will almost always be some difference between a sample and the population from which it was drawn. For example, researchers might collect information from 00 adults in a town of 15,000 people. Because not all people in the population were surveyed, there are bound to be differences between the results obtained from interviewing the sample respondents and the results that would be obtained if all people in the population were interviewed. These differences are known as sampling error, and they are expected to occur regardless of how scientifically the sample has been selected. The advantage of using a scientifically drawn probability sample, however, is that the maximum amount of sampling error can be estimated with a specified degree of confidence. Sampling error is determined by four factors: the size of the population, the chosen sample size, a confidence level and the dispersion of responses to a survey. Of the four factors, sample size is the most influential variable. Table 12. Guide to Statistical Significance Distribution of Responses n 90% / 10% 80% / 20% 70% / 30% 0% / 0% 50% / 50% % 2.8% 2.8% 3.03% 3.10% % 2.1% 2.99% 3.20% 3.27% % 2.77% 3.17% 3.39% 3.% % 2.9% 3.39% 3.3% 3.70% % 3.20% 3.7% 3.92%.00% % 3.51%.02%.29%.38% % 3.92%.9%.80%.90% %.53% 5.18% 5.5% 5.% 200.1% 5.5%.35%.79%.93% % 7.8% 8.98% 9.0% 9.80% % 11.09% 12.70% 13.58% 13.8% Table12 shows the possible sampling variation that applies to a percentage result reported from a probability type sample (clustering considered). If a sample of 500 voters is drawn from the universe of voters in the San Diego region who are likely to participate in the November 2002 election, one can be 95 percent confident that the margin of error due to sampling will not vary, plus or minus, by more than the indicated number of percentage points from the result that would have been obtained if the interviews had been conducted with all persons in the universe represented in the sample. As the table indicates, the maximum margin of error for all topline responses is between 2.3 and.38 percent. This means that for a given question with dichotomous response options Page 17