Brief comments and excerpts from 2009 Pucher et al paper on effective pro-cycling measures

Ray Ryan, Chairperson, Skerries Cycling Initiative, 1st Oct 2010


John Pucher, Jennifer Dill, Susan HandyInfrastructure, programs, and policies to increase bicycling: An international reviewPreventive Medicine 50 (2010) S106–S125.


Pucher et al’s review addresses a tough question : what measures are effective in getting more people to cycle? Many studies they looked at were poorly constructed or provided inadequate measurement of cycling before and after the measure(s) was implemented.  There are other problems.  Some studies started with cyclists having say 25% of the modal share of commuting traffic while others started at 0.75% !  How do you compare pro-cycling interventions that raise the former modal share from 25% to 37% (a marvellous result but it can be described as a 48% improvement) to those who raise the latter modal share from 0.75% to 2% (cycling is still marginal after the interventions but the result can be described as a 167% improvement) ? How does one allow for the effect of a country’s existing cycling ‘culture’?  Cycling has been a big part of Holland and Denmark’s culture for most of the 20th century and into the 21st; they have numerous cycling-related features but it is not always clear which of these, if applied to a low-cycling culture country, would produce an increase in cycling. Saturation cycling may be evident in Groningen in Holland, where modal share has been stable at ~40% for many years, despite various measures being introduced.  The good news in this paper for me are the examples of increased cycling when multiple interventions by individual cities were made, including Dutch and Danish cities. The paper also mixes in a lot of data on cycling safety as well. Here are some specific comments on individual measures:

CYCLE LANES : studies show that the majority of both cyclists and drivers say that they prefer to have cycle lanes. However outcomes are less clear as some studies showed increased cycling when the lanes went in but some did not. (Many cyclists don’t know the facts about the safe width of a cycle lane.  Research suggests 1.8m as a minimum. Personally I would not agree to less than 1.5metres. RR)

CONTRA-FLOW CYCLE LANES : found to be popular and safe when properly constructed.

SHARED BUS/BIKE LANES : these were popular with cyclists.  (In Ireland however, we have problems with bus lanes that encourage dangerous overtaking of cyclists and’s current position is to favour wide lanes (4.5m+) which allow space for the cyclist to be overtaken, or else 3/3.1m lanes where the cyclist commands the lane and no overtaking is allowed-RR)

BIKE TRACKS (PATHS) :  some studies showed increased cycling after path provision but some did not. Women and less-confident cyclists expressed higher satisfaction with them. 40% of cyclists would take a longer off-road track rather than the shorter route on the road. However this means 60% would take the road and Pucher et al comment that more experienced cyclists often prefer to stay on the road. Two-way tracks are usually 3m wide.

ADVANCED STOP LINES : research did not show much effect either way on conflicts between bikes and other vehicles.  Many cyclists don’t understand ASLs.

BICYCLE PHASE TRAFFIC LIGHTS : (cyclists get green light before other vehicles) one study in Davis CA, USA, showed that in 35 months before installation of the pro-cyclist light phase there were 10 crashes and none in 35 months afterwards. The junction connected cycling traffic to an off-road cycling track. Is this not a very important result for us to note ?

TRAFFIC CALMING : the few studies on this topic show increases in cycling and cyclist safety. This supports the Hierarchy of Measures idea.

BIKE PARKING FACILITIES, BIKE LOCKERS, SECURITY, SHOWERS, INTEGRATION WITH PUBLIC TRANSPORT ETC : a small number of studies show what Pucher et al say one would expect, namely that all these measures produce low but measurable increases in cycling when provided.  Pucher et al say that causation also works the other way i.e. more cycling drives more parking etc provision. Cyclists love bike lockers (Iarnrod Eireann have provided them on the Navan rail line and say they will re-instate them in Connolly station-RR).

COMPULSORY HELMET LAWS :  here is a quote :“Mandatory helmet laws …. reduce bicycling. Studies in Australia in the 1990s found declines in bicycle counts one year after the implementation of a helmet law of 36% in Melbourne, 36% in New South Wales, and 20% in Perth.” 

REDUCED SPEED LIMITS : widespread speed limit controls in Hilden, Germany led to a ‘significant’ increase in cycling. UK studies produced no evidence of increased cycling in 30kph zones, despite an increase in declared willingness to cycle. A 30kph limit in Graz in Austria reduced cyclist accidents by 4% (that’s good but on the negative side, in Graz cyclists are forced up onto pavements where the cycle track is often just a painted line and where old people, children and tourists are frightened by the cyclists-RR). .

PROMOTIONAL CAMPAIGNS, EDUCATION ETC : general campaigns to reduce driving don’t seem to impact much on cycling but when cycling itself is targeted better results are obtained e.g. one month after a bike-to-work campaign in San Francisco, bicycle counts remained 25% higher than before the event.


These have been very successful e.g.

  • Berlin quadrupled cyclist number from 1975-2001 and raised the modal share for cycling from 5% in 1990 to 10% in 2007 with a 37% decline in serious injuries from 1992-2006.
  • Amsterdam started in 1970 with a terrific 25% modal share but raised it to a staggering 37% by 2005, accompanied by a 40% decline in serious injuries from 1985-2005.
  • Copenhagen : bicycle share increased from 25% of trips in 1998 to 38% in 2005 for 40+ age group; 70% increase in total bicycle trips 1970–2006 (36% of work trips in 2006); 60% decline in serious injuries 1995–2006. 



Two things struck me. First was the absence of any reference to our much prized and now government-adopted policy of following the Hierarchy of Measures . Secondly, Pucher never seems to reference the research summary of John Franklin on the safety of cycle tracks (paths).  This may be because the summary itself is not a peer-reviewed paper – but there are peer-reviewed papers in it. Pucher also restricted himself to post 1990 papers and safety was not the primary concern of this paper anyway.

Referring to the NCPF’s emphasis on the “Hierarchy of Solutions”, I asked why this hierarchy was not discussed in his paper. He said :

“We did not seek out to prioritize measures, since it would have required far more rigorous studies of all the various infrastructure, programs and policies than were available to survey.  Indeed, we were surprised at how few rigorous studies exist, and many of them had contradictory findings.  Giving the lack of solid evidence and the sometimes conflicting evidence, it would have been risky trying to establish a priority ranking based on the studies we examined.  Moreover, we would have had to prioritize within categories as well, and it would have been a mess, I think.  Thus, we organized by categories of measures but did not prioritize.  I am not really sure it is possible anyway, since different things work better in different places.”

While accepting that the thrust of his paper was on successful promotion measures, I asked him about the controversy about cycle track safety in urban areas and the research of John Franklin. He said with reference to his own paper:

“Many studies did indeed include the safety impact, but just reporting on the impacts on levels of cycling was more than enough for our review, as you can tell from its length and the huge tables we had to create to summarize the impacts.”

What to conclude from this ?  For myself, his paper and correspondence reinforces my view that it is unwise to be too prescriptive about what works to improve cycling rates.  I support the ‘Hierarchy’ but I would not raise it to the level of gospel truth.  As for cycle tracks, I look forward to a paper from Pucher which does focus on safety – at which time we can scrutinise it for reference to Franklin’s research.

Leave a Reply

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