Analysis of Bicycling Trends and Policies in Large North American Cities: Lesson for New York

HOME          SOCIAL

John Pucher, Ralph Buehler

March 2011

University Transportation Research Center 2

This is another survey of statistics so typical of Pucher. The statistics are justified, made to seem more persuasive, by being submitted to a variety of statistical tests. Several of these tests are unknown to me, but my ignorance of these does not change my conclusions about the study, which are based on the content of the study and not about the reliability of the statistics presented, and which agree with Pucherís own conclusions.

Pucher has become much more cautious in his claims. He now admits that he presents nothing more than correlations between urban characteristics and amount of cycling. All his study shows is that those cities in which there is much governmental activity about bicycle transportation have more bicycle commuting, and even Pucher repeatedly states that the causal relationship, if any, may be in either direction and there is no way to separate the effect of any one program.

And a bit more besides. Dense urban cores have higher bicycle mode share, as do areas with many students. In these two cases, causal relationships probably exist, but these are not characteristics that can be changed to increase the bicycle mode share.

 Despite Pucherís presentation of statistics with repeated tests of statistical reliability, he has made at least one glaring mathematical error. He has presented a graph of two ratios. Fatalities per cyclist is graphed against cyclists per worker. Any graph of this form is meaningless, because, whatever numbers may be used, even random, they produce a declining pseudo-hyperbolic curve that looks as though increases in bicycle mode share produce decreases in cyclist death rate. That is, the Jacobsen error.

I need to emphasize that Pucher makes no claim that bikeways reduce cyclist crashes or how such an effect might be achieved. As he stated in his meeting in San Diego, when asked about this, he replied that he paid no attention to engineering but just did (in this case, reported) what was popular.

Home                                           Social