A Closer Look at Success Rates for U.S. Transportation Tax Ballot Measures

Todd Stone, the Minister of Transportation, has sought to dispel pessimism about the outcome of a referendum on TransLink funding by pointing to success rates for other referenda. As reported in the Vancouver Sun, the Minister noted in an interview last November that “across North America since 2012, there have been (dozens) of initiatives, ballots, or referendums that were asking the voters to approve new spending measures to expand transit. Seventy-three per cent of the time, these referendums were successful.”

During the April 10 committee discussion of Bill 22—the legislation that establishes the framework for funding referenda—the Minister stated that in the last couple of years in the U.S., 79% of funding referenda were won.

But do these success rates really provide a basis for optimism?

In her M.A. thesis, Taxing for Transit: An Exploratory Analysis of Local Option Transportation Taxes, Lydia Rainville analyzed 274 U.S. transportation ballot measures with a transit component, voted on over the period from 2000 to 2011. These involved two types of taxes: property taxes (46%) and sales tax (54%). In aggregate, 70% of the ballot measures were successful.

Approximately half of the ballot measures were to extend an existing transportation tax, either at its current rate or at a higher rate . The other half sought to introduce new transportation taxes.

Ms. Rainville computed the success rates for various categories of the ballot measures. Of particular interest is the success rate for ballot measures seeking approval of a new tax. The rate was only 53%. This is in contrast to an 86% success rate for ballot measures to extend existing taxes.

In Ms. Rainville’s words:

The policy implication of this finding is that passing the first tax measure is the primary hurdle. Once a locality successfully passes one of these LOTT [local option transportation tax] ballot measures, it is more likely to remain in existence.

A caveat must be added. In California, a two-thirds majority is required to pass a transportation tax ballot measure. If a bare majority had been required, additional ballot measures would presumably have been successful. However, it is unlikely that this would have made much difference to the 53% success rate for new taxes. Only 44 of the ballot measures in the dataset were from California. Furthermore, according to Ms. Rainville, the success rate for these was relatively high.

The Minister has not provided any information about the referenda for which he has quoted the success rates. How many involved an extension of existing taxes and how many involved the introduction of new taxes? What was the success rate for the measures of the latter type? Were they essentially a coin toss, as Ms. Rainville found for the ballot measures she examined? Given the absence of specifics, no comfort can be drawn from the Minister’s numbers.

It is debatable whether a simple average number—the success rate of U.S. ballot measures, even when limited to measures that introduce new taxes—can be used to predict the probability that a referendum on new funding sources for TransLink will be successful. Many factors can affect whether a particular referendum is successful or not and can lead to a conclusion that the referendum is more likely to fail or to succeed than a “typical” referendum. To name a few that may be important for our referendum: the magnitude of the additional revenue (common sense says that the larger the amount of revenue for which approval is sought, the less likely that approval will be given); the type or types of new revenue sources; voters’ attitudes towards transit; the extent to which voters see this as a vote on TransLink itself; and the strengths of the campaigns for and against giving TransLink any new revenue sources.

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>