Section 508 Navigation
National Trails Training partnership Skip Navigation
HomeAbout usTrailsWhat's hotCalendarTrainingResources & libraryPartnersJoin usStore

Transportation and urban trails

Hosted by AmericanTrails.org

Tools for Predicting Usage and Benefits of Urban Bicycle Network Improvements

How do you evaluate which bike facilities would provide the greatest benefit, or would provide a specific desired result?

Authors: Gary Barnes, Kevin Krizeke

Map of Minnesota

INTRODUCTION

Bicycling is a very popular recreational activity, and is an important commuting mode in some congested, limited-parking areas, and for those with limited incomes. Making cycling safer and facilitating additional cycling has the potential to provide substantial public benefits. However, funds for facilities are limited and historically there has been no systematic, quantitative way to evaluate which facilities would provide the greatest benefit, or would provide a specific desired result.

The amount of cycling, and the size and type of benefits gained from it, should depend on a number of factors including the quality of the cycling environment in an area. The amount of cycling will depend (among other things) on demographics, the presence of significant cycling destinations, and on facilities. Benefits will depend on factors such as the purpose and location of the trips, their number and duration, and who is making them.

The critical question for planning purposes is how the cycling environment, as opposed to uncontrollable factors such as demographics, influences demand and the resulting benefits.

This project was part of a larger body of bicycling-related research that is ongoing at the Humphrey Institute at the University of Minnesota. This research has been funded by the Minnesota Department of Transportation (Mn/DOT), the National Cooperative Highway Research Program (NCHRP), and a variety of other sources. Because of the very broad nature of this research program, there was no single unifying question guiding the research that is reported here. Rather, the project essentially consisted of four essentially independent research questions, under a broad unifying theme. This theme was bicycling preferences and behavior with regard to bicycling facilities. The studies were also connected by the fact that they were all based on information from the Twin Cities of Minneapolis and St. Paul, Minnesota.

The theme of bicycling preferences and behavior with regard to bicycling facilities could be understood as an attempt to develop answers to the following three questions:

4) In what way and to what extent does cycling demand depend on the environment?

5) How does the environment, through its influence on demand, impact the size and types of benefits gained from cycling?

6) How would specific changes to the environment be expected to change the amount of cycling and the benefits gained?

In addition to working toward finding answers to these questions, another major focus of this project was developing research methodologies by which the questions could be addressed rigorously. We hope that this will inspire other researchers to use these or similar methodologies to study other places; a robust understanding of how facilities affect bicycling behavior must rest on evidence from more than one location.

Because the four parts of this project were basically independent, and because the resulting reports are in some cases very technical, this report is organized in a somewhat unconventional way. That is, the main body of the report is relatively short and provides 2 in essence a long executive summary of each of the four reports, as well as a general overview of the project and its findings. The reports themselves are then included as appendices for those who wish to view the detail.

The four reports in brief are:

1. Effect of Trails on Cycling. This is based on the 2000 Travel Behavior Inventory (TBI) and analyzes reported cycling behavior based on the distance of a person's home from the nearest cycling facility.

2. Value of Bicycle Facilities to Commuters. This is based on an original data collection; a computer survey asking people to choose between commutes of varying durations on bicycle facilities with different characteristics. The choices make it possible to deduce the value placed on various factors.

3. Effect of Facilities on Commute Mode Share. This study compared census bicycle commute-to-work mode shares in 1990 and 2000, and related changes to where new commuter-oriented bicycling facilities were constructed.

4. Cycling Behavior Near Three Minneapolis-Area Facilities. This is based on an original data collection; a mail survey to residents of areas near the Midtown Greenway, Cedar Lake Trail, and Luce Line Trail. The objective was to better understand the relationship between cycling behaviors, trail access, and various demographic and lifestyle factors.

CONCLUSIONS

While the results are not entirely unambiguous, the preponderance of evidence seems to support the hypothesis that the major bicycle facilities constructed in the Twin Cities during the 1990s did in fact significantly impact the level of bicycle commuting. The suburban parts of the region showed a decline in bicycle commuting, contrasted with a sharp increase in both central cities. Within the central cities, areas near bicycle facilities tended to show more of an increase in bicycle mode share than areas farther away, although this trend is less sharply defined. Trips that crossed the Mississippi River showed a much larger increase than trips that did not, seemingly demonstrating the impact of several major bridge improvements. Finally, trips into downtown Minneapolis and the University of Minnesota, where improvements were concentrated, showed substantial increases, while trips into downtown St. Paul, where few improvements were made, showed a slight decline.

The results also provide considerable support for the alternative hypothesis that facilities are the effect, rather than the cause, of high bicycle use. In the Twin Cities, the areas where major facilities were built already had bicycle mode shares that ranged from twice the regional average up to nearly 15 times the regional average. While the facilities did increase the bicycle mode share in their buffers by about 17.5% overall (from 1.7% to 2.0%), this is far from the factor of ten difference that is observed between the facility and non-facility areas when considering the year 2000 in isolation (2.0% compared to 0.2%). This highlights the risks inherent in trying to deduce the impact of facilities by trying to compare two different places.

There are a number of further lines of work that could add more insight to this analysis. One would be experimenting with different buffering methods. We defined our buffers somewhat arbitrarily in order to simplify the analysis. But in some cases TAZs that fall into the buffer for a facility would not necessarily be expected to use it much, because there are physical barriers to access or because there is a more direct route to the most likely destinations. We believe that this may be what is happening with some of the buffers that showed no increase in bicycle mode share. Conversely, there may be TAZs that are outside our buffer but that fall within the zone of influence, because the facility falls on the route to a major destination or because it can be easily accessed using existing facilities. For example, both the West River Parkway in downtown Minneapolis and the Kenilworth Trail seem likely to derive much of their value by providing needed links or extensions to already existing facilities.

Another improvement would be a more careful reckoning of new facilities in the area. Our accounting of new facilities in Minneapolis was perhaps more thorough than those in St. Paul due to the sources we were able to access. The large increase in bicycle mode share outside of the facility buffers in St. Paul leads us to wonder if there are important facilities that we failed to include in our analysis. A related improvement would be to extend the analysis some distance into the suburbs, again being careful to identify major possible bicycle commuting facilities.

There was one of our facilities that extended into the suburbs, and understanding the impact on bicycle commuting in this area compared with similar inner suburban areas without facilities would be interesting.

In this paper we did not try to control for demographic variables. The areas that we are studying seemed sufficiently large that major demographic shifts would be unlikely in such a short time, although they certainly could have had an impact on specific locations. Generally variables such as age and income are not as important as they are often believed to be (9); the differences across ages and incomes are only a small fraction as large as the differences across geographic locations. However, there would be value in confirming this point within the specific context of this analysis.

While there are many possible improvements to be made, the fact that this simple analysis seems to show a clear impact of bicycle facilities on the level of bicycle commuting is of considerable interest. Reliance on comparison of bicycling levels in different places is inherently subject to the criticism that no causality is implied by any observed relationship; facilities might have been built because many people already rode bikes, rather than the facilities causing the high levels of riding. This approach provides a method for demonstrating the effect that facilities have on the level of bicycling in an area in a much less ambiguous way.

Published by:

Minnesota Department of Transportation

Research Services Section

395 John Ireland Boulevard, MS 330

St. Paul, Minnesota 55155-1899

Related topics:

More resources:


page footer

Contact us | Mission statement | Board of directors | Member organizations | Site map | Copyright | NRT | NTTP