Bicycling is healthy, now prove it
One constant in policymaking, from the Progressive Era through the New Deal up to today, is the need to support policy proposals with data. Policymakers are expected to quantify the extent of a problem and measure the impacts of policy interventions. Just look at the curriculum of a public policy graduate program today and you will see that it is heavily weighted toward quantitative analysis. While this approach increases our knowledge and our ability to make informed decisions and generally leads to greater government accountability, it does present a challenge for bicycling given the current dearth of data collected on bicycling. (To give just one example of the feeble state on data on bicycling, bikes appear just twice in this 385 page catalogue of “transportation energy data.” Despite their limitations, the Census’ American Community Survey and the National Household Travel Survey remain the best sources on bicycling levels.)
For many of us, the health benefits of bicycling may seem self-evident. After all, traveling by foot and by bicycle requires a physical exertion not needed to drive a car. However, for policy purposes these benefits need to quantified and demonstrated empirically. Researchers Pucher, Buehler, Bassett, and Dannenburg, recently released a study called “Walking and Cycling to Health: A comparative analysis of city, state, and international data,” which uses existing data on health and active transportation to show that higher rates of walking and biking are correlated with lower rates of self-reported obesity at the city, state, and country level. At the city level, active travel has a negative relationship with diabetes, meaning that cities with higher biking and walking rates have generally lower rates of diabetes.
For those interested in statistics, in both figures above, the solid line is the log regression line for the fifty states and the dashed line is the regression line for 47 of the 50 largest cities. The direction of line shows the inverse relationship between biking and walking (the x-axis in both figures) and obesity (the y-axis in the top figure) and diabetes rates (the y-axis in the bottom figure). The relationship is statistically significant at p<0.001 for cities and states for obesity and for states in diabetes. The relationship is significant at p<0.01 for cities for diabetes.
The results effectively show the direction of the relationship. Cities and states with higher levels of biking and walking have lower levels of obesity and diabetes. This may be enough to encourage communities to promote walking and biking. However, the data constraints limit the conclusions that can be drawn from this particular study. First, the data are cross-sectional, a snapshot of one moment, and therefore they cannot tell us anything about change over time. Secondly, the data are aggregated, which means that they speak to population-wide characteristics, not individual results. Third, data were not available to account for other factors that may have contributed to biking and walking levels and health status. For example, they did not take into account diet or family medical history. Finally, even if such data had been available, the relatively small sample sizes would have made it difficult to get statistically significant results if they had added control variables . These data limitations reinforce the need for better data collection.
In a related study, “Evaluating Public Transportation Health Benefits,” Todd Litman of the Victoria Transport Policy Institute reports on the health benefits of living in transit-oriented and mixed used neighborhoods. Litman finds that high quality public transportation and walkable and bikable mixed-use communities located around transit stations reduce traffic crashes and pollution emissions, increase physical fitness, improve mental health, improve basic access to medical care and healthy food and increase affordability.