Oh dear! It seems the cycling boom in Australia is a myth …Â Check out these links and see what you think:
Meanwhile, over in New York City, they are going great guns with some very simple and obvious solutions.
Thoughts?
- Home
- The Knowledge
Thinking about it?
Bicycles
Where
Links
- Local
Local
This is where you will find out what's happening locally - what's on, who's doing it and how you can join in.What’s happening
Take action
- Global
- Cyclopolitanise
Cyclopolitanise
Cycling in Auckland is in favour of:this
and this
and of course this
- Business case
- Shop
Ah Ha – Sydney Uni – my old Alma Mater.
This is more like the reason
lack of investment in cycling infrastructure. Qualitative research consistently reports that people not confident riding on roads prefer bike paths separated from traffic.”
Just got back from lovely holidays in and out of Broome (WA). Heaps of commuter cyclists there. Flat, and lovely shared bike/pedestrian paths. And when you add the tourists (like us) who hire bikes to go for a spin, the place is positively bristling with bikes.
Lying with statistics! They are comparing a 1986 and and a 2011 survey? And then make great news out of the fact that cycling declined? WE KNEW THAT. Everyone knows that.
If the study was, say, comparing 2005 to 2011, then it would have some worth. As it is, it’s pretty useless, just codified common knowledge. But watch the “cycling boom is a myth” argument being trotted out at talkshows, “based on university research”…
Helmets are also mandatory in parts of Canada, in Vancouver for instance i.e. British Columbia.
same point as Max. This is a weird ass study. It’s like me saying “Public transport use per capita in Auckland dropped between 1950 and 2011″ (an unarguably true observation), “therefore it is impossible that public transport use per capita could have increased between 2002 and 2012.” But, in fact, these two statements are not mutually incompatible. If you look at this graph, you can see that just because something goes down overall, on average, over a given time period, doesn’t mean it’s impossible that it also goes up at certain times WITHIN that time period
http://transportblog.co.nz/wp-content/uploads/2011/07/pop-pt-usage1.jpg
Lucy, I looked but did not find anywhere in the study or article that implied it was impossible for there to have been a cycling boom. Besides the “obvious” finding that cycling per capita was much lower in 2011 than in 1986, the study only asserts there is no evidence of a net growth behind the boom story (which is usually supported by unadjusted year-on-year increases or absolute numbers). The law of parsimony would have us reject the notion of a cycling boom without actually seeing sufficient evidence for it — even if it is conceivable.
The study methodically establishes a baseline, so it is hardly weird or useless. Its main contribution is not the findings themselves, but its review of the poor data available, and the demonstration of a repeatable method of evaluating and comparing these data. That its first result is consistent with something that “everyone knows” is an encouraging sign of its robustness.
The essence of it is that not all observed changes in cycling participation are made equal. An increase in absolute terms is not sufficient to support claims of real growth or decline or stability. Those numbers must be adjusted for population changes in the same period to begin to be meaningful for that purpose.
The study leaves us with two challenges. Firstly, as Max also suggests: to analyse more historical statistics where possible within smaller windows of time to draw out localised correlations and trends, perhaps around key policy changes or likely boom/bust periods. Secondly: to repeat the survey and analysis annually into the future, gathering comparable statistics that allow us to make confident, meaningful claims about cycling participation rates and changes over time.
So why the hostile reaction?
“So why the hostile reaction?”
Because I do not actually see much purpose in the study, if it is so limited. Because it opens itself up to misuse, with the use of “shouty” headlines saying the cycling increases of recent years were a fake, a lie, and cycle advocates were fools when they claimed so.
It is well known that such headlines and uses create the impression among even readers who do more than just glance at the headline and then read something else.
If you do a headline in the Herald saying: “Politician accused of murder”, then what most people take away isn’t the fact that the article later said that the accusations are baseless. What the average person remembers is “Hey, wasn’t so-and-so the guy they think murdered someone?” Basic human psychology. Mud sticks. The louder the better.
Similarly, any such shoddy study which claims “the cycling boom is a myth” on the basis of such misleading evdidence is anything from shoddy reporting to direct sabotage of the cycling transport method for me.
That wasn’t meant to be a hostile reaction. I have no time to respond to this in detail but I just find the conclusion that he has drawn re helmet use based on this data a very odd one.
actually having read the study, I still think it’s weird in many ways. But no time to expand!
Max, I am on tenterhooks now. You have made all sorts of criticisms, but without connecting them to reality. I am now left wondering “how” as well as “why”.
I am surprised you do not see much purpose to such a study. It aims to collect and compare cycling data in a repeatable and meaningful way. As a cycling advocate, are you not interested in statistical facts about cycling? Even if you disagree with the subsequent policy recommendations or speculation about causes, do you not see value in the per-capita method and evaluation of historical and current data sources? Are you unaware or uninterested in the poor state of measurements regarding cycling, and how this can be improved?
Just on principle, saying an academic study is shoddy because it “opens itself up to misuse” is quite bizarre. How can a scholar determine when and how the public could one day editorialise their work? Chilling scientific, technical or statistical research on the basis of what someone else might eventually say about it is a scary proposal in an open society like ours.
I am also confused as to whether you are criticising the study or the separate article for being shoddy and based on “misleading evidence”. The study makes no statement about a “myth”. For your benefit, here is the conclusion quoted in full:
“Cycling participation has not kept pace with population growth, representing a per capita decline in cycling over the past 25 years. Reasons for this are unclear, although a historical lack of investment in cycling infrastructure coupled with mandatory helmet legislation may have contributed. It will be important to repeat the 2011 cycling participation survey.”
The whole report is prosaically titled, “Australian per capita cycling participation in 1985/86 and 2011″, and also includes discussion of its own limitations and the reliability and comparative quality of various data sources. This is academically sensible.
However, in the separate article, the author does add some commentary regarding the myth of a cycling boom and makes recommendations about helmet laws. You may question the nature of headlines and human psychology, but the general thrust of it is sound: there is no reason to believe there has been a cycling boom, especially in the context of a long-term decline in cycling. If you believe this is shoddy, or it misrepresents the cited evidence, then I would love to hear why.
Lucy, I apologise: I was only referring to Max’s response as hostile. I am also interested to hear your critique of the study, which I expect will be level-headed as always.
I still don’t have time to get into a detailed critique of this study, although I would highlight that in the first paragraphs he specifically talks about reports of an increase in bike sales since 1999 and in particular the fact that more bikes were sold in the last year than ever before. However, he then goes on to compare mode share change from 1986 to 2008.
I do think that by juxtaposing those statements in the introductory paragraphs with the results he subsequently presents he is trying to give a certain impression to the reader, i.e., reports of a recent cycling boom have been exaggerated. That he produced that impression is shown by the headlines this article generated.
I also think he is somewhat selective in highlighting changes in helmet law as a possible determining factor of travel mode choice without talking in great detail about any of the other (major) changes which occurred in our society, built form, design of our cities, transport funding balance, ideas about acceptable levels of safety and personal risk, etc etc in the same time period that may have driven changes in peoples’ travel mode choices. He talks about helmets quite a bit but he devotes just two sentences to those other factors like infrastructural changes.
to give one example, I just did 5 minutes of googling and found this report which shows that walking as a proportion of total mode share also dropped in NZ from the late 1980s to 2008. The drop hasn’t been as dramatic as for cycling, but it has certainly happened.
http://www.transport.govt.nz/research/Documents/How%20New%20Zealanders%20travel%20web.pdf
So we can see that walking and cycling rates dropped for all NZers, but the most dramatic drop off seems to have been with primary school age children. Rates of cycling to primary school in New Zealand dropped dramatically from the 1980s to 2008s – they went from 12% to 4%. But rates of walking to school dropped dramatically as well, from 42% of all trips in 1989 to 25% in 2004-2008.
There was a similar dramatic drop in cycling rates for high school children, but rates of walking remained almost unchanged.
Among adults, we went from 4% of trips in the morning peak being by bike to 2%. Walking rates dropped from 6% to 4%.
Now, I can think of about a billion explanations for these figures. One is that parents became reluctant to let primary age school children walk to school alone for fear of stranger danger, a concern that was amplified by mass media coverage of a few very high profile cases of child abduction. We know that parents did become more concerned about these issues and start progressively limiting their childrens’ movement throughout the 20th century – plenty of research avialable about that.
At the same time, perhaps parents also became fearful of letting their children of ALL ages cycle to school and of cycling themselves because of heavier traffic of on our roads, largely caused by planning and funding decisions which heavily promoted car use over all other modes.
They also began to walk less themselves because they were working longer hours and often, due to the low density sprawling devleopment in our cities at the time, living further from their workplace.
Obviously, a mad keen helmet advocate would explain this as “teenagers and adults are fashion conscious so they stopped cycling when had to wear a helmet (no explanation for why they also stopped walking) but that change didn’t make any difference to primary school kids. Instead their rates of cycling dropped, as their rates of walking did, primarily because of parents safety concerns.”
But none of these explanations are probably right. There are so many other factors in here. Changes in employment for parents, changes in our built form, changes in petrol prices, changes in our culture around what level of personal risk is acceptable. We just don’t have enough accurate evidence to show that it was helmet rates that caused these changes and, in the absence of that evidence, and in the light of the fact that we know there was also a dramatic change in the use of other active modes, it is surely rather selective to give this one factor so much emphasis.
Also, you said. “Just on principle, saying an academic study is shoddy because it “opens itself up to misuse†is quite bizarre. How can a scholar determine when and how the public could one day editorialise their work? Chilling scientific, technical or statistical research on the basis of what someone else might eventually say about it is a scary proposal in an open society like ours.”
It is true that researchers can never guess exactly how their work will be interpreted or used. But I actually think this is a really important ethical consideration for scholars that should always be in the fore front of their minds. Researchers have a moral imperative before they conduct or publish any study to consider what the effects might be in the real world. Did you know, for example, that in his later life Einstein became a massive pacifist and deeply regretted the work that had been done to create the atomic bomb?
This is an issue that is particularly evident in my field, which is cross-cultural research. For example, there has been a huge amount of research done which focuses on highlighting inequalities between ethnic groups. Often that research highlights how some ethnic peoples compare badly on a range of measures to Pakeha. Quite often researchers don’t expend any effort in explaining why they do so badly on these measures (often structural reasons like poverty and discrimination etc) – they just present the inequalities and let readers draw their own conclusions. Very often in my area, the measures and comparisons are inappropriate anyway because they were developed and tested primarily on Pakeha NZers. Anyway, I digress.
To give an example that has nothing to do with my own work. It’s questionable whether yet another study, for example, that highlights how much higher rates of obesity and diabetes are among Maori and Pasifika NZers does an awful lot to improve the wellbeing of those ethnic peoples. We know their rates of obesity are very high, the government sure as hell knows, and another study which reiterates that fact probably isn’t going to get more funding for that cause. Instead it’s just going to feed into some really negative stereotypes about those ethnic peoples.
Wouldn’t it be better to put that energy into doing some research into the structural causes of high obesity rates? Or even better, what can we actually do to reduce obesity rates among Maori and Pasifika NZers? Are there approaches out there that are working? Or do a political study of WHY the government won’t adequately invest in reducing this inequality in health outcomes, even though we have known about it for years?
I don’t necessarily think it was inevitable that this particular piece of research would lead to negative headlines. It could have led to headlines which said, “Cycle rates plummet: More investment needed.” But I think the overall point about the ethical imperative to consider the real world consequences of your research is critical.
Thanks for your post, Lucy. I understand there is more to your critique, but that already touches on many interesting points. I agree with some of them, and disagree with others.
I agree that the remark about bicycle sales representing a growth in cycling is clearly the motivation for publishing the study: that is, improper (absolute) numbers are being used in the wild to support claims about cycling one way or another, and it should be rectified in the way the author goes on to demonstrate. It just so happens that a per capita adjusted figure is not as rosy as a year-on-year unadjusted figure. A charitable reader would not impute malice (like “direct sabotage” of cycling advocacy) as the intent here. But if you are a priori opposed to those findings for some reason, I guess you might be tempted to think so.
On possible explanations for the decline, the author actually gives equal time to investment and infrastructure as well as helmets. In fact, for every mention of helmets in the study and in the article, it is preceded by noting a lack of investment and infrastructure. The article has a subheading, “Three explanations” — and it is definitely not followed by “helmets, helmets, helmets” (which, incidentally, is how many times the word even appears in the study).
Admittedly, there are more external citations on the subject of helmets, but I suspect that is due to its controversial nature. It is prudent of the author to support those claims thoroughly. The investment and infrastructure issues are something that — to borrow a phrase — “everyone knows”, but helmets not so much.
The correlation between walking and cycling among school children is very interesting. I think you’re entirely correct in saying that broader investment and infrastructure issues could be a common cause for the shared decline. Obviously there are plenty of possible explanations, but I struggle to see why the author is at fault for merely exploring some of them. He uses appropriate qualifications like “may” and “might”, and offers cautious and well-demarcated hypotheses. Is that so bad?
In fact, given that there are so many uncertainties and quirks like the different rates of decline for walking and cycling, surely this is justifiably fertile ground for research — particularly with respect to measuring and analysing scant data from the past, and establishing a repeatable process for the future.
I also agree on the principle of being ethically conscious in research, and I think it is compatible with academic freedom.
However, on the specific point about data, I simply disagree that it is inherently less worthy to establish a better, more meaningful way to measure something (even if we already know the general state of affairs). Knowledge is crucial to identifying, testing, solving and then confirming or maintaining solutions to problems. Improving measurement and analysis in non-trivial ways is an indispensable part of it.
Even if you reject the novelty of this study, another under-appreciated virtue of rediscovering known results is in repetition or confirmation of past work — scientific and technical research depends on this as much as finding glamorous contrary results. This does not make the work “shoddy” or malicious.
I also do not see academic research of this kind as a zero-sum game. A scholar merely conducting a study, reporting their findings, and making recommendations does not preclude the government or others from taking useful action to improve matters. If anything, it could only contribute to taking a better-focused, evidence-based approach. Could the scholar have done something more useful with their time and energy? Maybe, or maybe not — that is such a nebulous thing to determine; but again, it does not detract from the worth of the work they did.
“Just on principle, saying an academic study is shoddy because it “opens itself up to misuse†is quite bizarre. ”
You made sme fair points – and I’ll acknowledge that it’s much more the “cycling boom is a myth” summarising of the study – picking sensationalist and even in my opinion outright wrong conclusions – that pisses me off. I admit that I obviously didn’t go back far enough to look at the study itself, and that may have well been much more rigorous, and less sensationalist in conclusions.
But the following statement comes straight from the linked article, which is on an .edu domain website – so is more that just some reporter misrepresenting the study. Obviously, even academia likes to make their statements a bit “edgier” to get attention:
“Australian cycling boom? Nope – it’s a myth”
I simply argue that that is a very inappropriate & much over-blown conclusion to draw from the data they have, and that still makes me angry.
“I simply argue that that is a very inappropriate & much over-blown conclusion to draw from the data they have, and that still makes me angry”.
Contrast that with the message ‘cycling is booming in Auckland’ that has been regularly touted…based on what evidence/data?
‘I saw more cyclists on the road this morning’ has been regularly used as a basis for claiming that cycling is booming in Auckland.
‘Having your cake and eating it too’, comes to mind.
Max, you are right that the article is substantially more editorialised than the study (and it is the same author). The .edu website is a kind of academia-plus-media web magazine project, not an ordinary journal, so you should probably expect a bit of that. Of course you are entitled to dislike its style for any reason. However, the substance still stands:
“I simply argue that that is a very inappropriate & much over-blown conclusion to draw from the data they have”
This is the one thing I have been asking you to explain since your first post. Can you say why it is inappropriate and over-blown?
My understanding is that there are two legs to the article’s argument rejecting the cycling boom:
1) As the study establishes with accepted data and the newly-applied method of per capita analysis, the big picture is a long-term decline in cycling on a per capita basis. Even if everyone knew that in a general sense, we now know it with the right kind of metric — by which we can sensibly measure and compare further changes in other appropriate intervals of time.
2) For anyone to claim there is currently a cycling boom, the burden is on them to give contrary evidence of a per capita rise in cycling, calculated in a similar way to what the study demonstrates but for a more recent time period. The author merely asserts that there is no such evidence being put forth by supporters of the view that there is a cycling boom, and cites an illustrative example of the wrong kind of metric being used (bicycle sales). It is hard to prove a negative like “there is universally no per capita measurement being used in public discourse to support the claim”, so we can either trust the author on this point or produce a counterexample.
So the study establishes a minimum standard for evidence to argue that cycling has grown, declined or remained stable. If we did not believe there was a cycling boom before anyone told us there was (and we should not, given the long-term decline that everyone knew), then we should still not believe it after hearing such a claim, unless they provided evidence of a sufficient standard (which they have not).
Yes, it is not the strongest possible argument against the proposition — in particular, it does not rule out the possibility that there could have been a cycling boom, and that evidence of it has yet to surface. Indeed, the study recommends that future work should seek out those trends. But surely it is a reasonable position to take that there is no cycling boom, until further information appears.
Reizar, Auckland numbers are actually counted systematically (yes, seeing more cyclists is the business end of all these studies). While they are not per capita adjusted in monthly or yearly reports, the usual windows of time are not normally wide enough for population changes to be a huge factor. It gets fuzzy as you approach intervals similar to the census period, or if there are rapid changes such as due to altering zone boundaries or some substantial migration (e.g. from Christchurch), but those factors will come out in the wash in a longer-term review. But are the short-term figures enough to claim there is a cycling boom? Probably not.
Tim – I disagree that it is useful for researchers just to measure the same thing again and again (although in this case, I agree that he wasn’t doing that, but in my example of Maori & Pasifika obesity rates, I think the argument is sound). I mean you always have to make a choice when you conduct a study as to a) what your main focus will be and b) whether you will actually do it all. Rsearchers should often think harder before doing studies about what the ethical consequences of their choices are.
I also don’t think it is wrong or incorrect for poeple to talk about a cycling increase or “boom”, if the per capita numbers do indicate that there has been a gentle increase in cycling rates since say 2006 in Australia. I mean sure, it’s a shorter time frame than what this researcher is focusing on but that doesn’t mean it isn’t true. I can’t be arsed looking up lots of Australian reports on cycling to see if there actually HAS been such an increase, however, because it’s dinner time and I’m hungry
I suppose my main objection to his giving equal space to helmets as compared to infrastructure etc is that I just don’t think there is evidence to support that helmets had as dramatic an effect on behaviour as those other factors. We know from studies of public transport, walking rates etc that they all changed during this same time period, and we know that a lot of those changes were due to infrastructure. So I think that it is selective of him to give so much attention to helmets. Personally, I would probably just have given it a paragraph or so, if I was writing this study, which I’m not.
It’s a bit unclear Reizar, what is happening here in Auckland, because we only started doing a monthly cycle count in late 2010. Those monthly counts suggest cycling numbers are increasing but they are selective because they are only placed in certain sites around the city. You can see the latest ones here: http://www.aucklandtransport.govt.nz/about-us/publications/Reports/Pages/Public-Transport-Patronage-Reports.aspx
We also have a manual annual count which shows there has been about a 30% increase in total cyclists counted between 2007 and 2011.
http://www.aucklandtransport.govt.nz/about-us/publications/PlansProgrammes/Pages/Regional-Cycle-Monitoring-Plans-2007-2010.aspx
This is a much faster rate of increase than the population grew during that period in Auckland (about 10%) and is also different from the vehicle count which on a lot of our big state highways has been dropping or plateaued in Auckland since about 2008.
http://transportblog.co.nz/2012/03/29/unprecedented-change-kiwis-driving-less-and-loving-it/
HOWEVER, the manual cycle count is a bit of a dodgy source to base your conclusions on because the data is very variable from year to year. For example, apparently the numbers in the manual cycle count were a little bit lower this year (although the full report hasn’t been published) which they attribute to the fact that the count was done during a very rainy two months in March.
But who knows? It could be because cyclists who were cycling in 2011 have thrown their bikes down in despair at having to wear helmets in early 2012
Or it could be because we have really shitty cycling infrastructure, or it could be a random blip because one of the counters was sleepy.
FINALLY, we have the household travel survey which gives us an idea of what mode share cycling has. This suggests that cycling mode share in Auckland has been increasing since about 2007 from about 0.25% to 0.85% of total travel but it is obviously still very low (i.e., under 1% of all trips).
Lucy, I am not sure who “Tim” is, but I will assume you were responding to me.
Regarding the shorter time frame, I was only referring to Auckland’s regular annual and monthly statistics, to which you linked, whose frequencies are too short to characterise a boom or bust within. Meanwhile, a comparison of something like 2006 to 2011 in Australia could be legitimate; the issue here is that is just not what the Australian-cycling boom proponents have been presenting, and so we do not know what those numbers show.
Naturally, no one is obligated to dig up this evidence (if you’d rather have dinner, say) but then as critically-thinking cycling advocates, we are also not obligated to believe poorly supported claims of a boom.
When the per capita rate of cycling has not kept pace with population growth, but the absolute numbers have grown, that is not a “gentle increase” but a “gentle decline” in cycling rates. If both values have shrunk, it is perhaps a dramatic decline. If the per capita rate has only kept pace, it is stable. Or if it has out-paced population growth, only then could there possibly be a “boom”. (All this is assuming that the population is generally growing.)
On helmets, you say you would only dedicate a paragraph or so towards it. As it happens, the author only dedicated one paragraph towards it in the article, while giving prior attention to motor vehicle prioritization, infrastructure and investment. The secondary picture caption also mentions it, which is more editorial poetry than anything (“Alone in the world”). In the study, he does not even give it a paragraph by itself! The only mentions of it are in the abstract, in one passage covering several possible causes, and in the summarised conclusion. In all these places, helmets are given the last spot, while infrastructure and investment are named first.
Beyond this study, another unspoken argument about cycling modeshare trends is that with such a small base, an unadjusted year-on-year growth of 30%, 50% or even 150% cannot be taken seriously — it is simply too close to the margin of error, too sensitive to random fluctuation. For example, in a different study about Australian cycling modeshare, optimistic researchers found that observed numbers of adult cyclists in Melbourne miraculously recovered in just a year after the helmet law was introduced (1992). The catch was that a large factor in the proportional increase was due to a “a bicycle rally passing through one of the [observation] sites” (p. 44), and everyone knows what happened next. Some boom.
Did someone mention my name? Aw Lucy, I had no idea you missed me so much.
While I have posted here on the premature, and misleading rah-rah-ing spun by Auckland Transport on cycle growth within a sub 1% modeshare composed largely of The Brave And Fearless, here I can only defer to the clearly perspicacious Non-Motorist in time honored fashion: yeah, what (s)he said.
I’m not going to argue that the numbers in New Zealand (or in Australia) aren’t particularly solid. But I’d love to see you present evidence that there has been a per capita decline in cycling rates in either New Zealand or Australia since 2006-7
Which is the period I was referring to for a gentle increase because it is about the time that we start getting at least some annual data from the Household Travel Survey and the manual counts in Auckland.
Just another comment – the same author also talked a lot about helmets in the associated press release that came out with this article. He said “”Well over half a million more Australians could be riding bicycles if we didn’t have mandatory helmet laws, according to research I conducted last year which showed one in five adults surveyed in Sydney said they would ride a bicycle more if they did not have to wear a helmet,” Professor Rissel said”
I still haven’t seen any conclusive evidence that helmet laws lead to a dramatic decline in cycling rates or, indeed, that removing them would lead to a dramatic increase in the absence of any improvements to cycling infrastructure. Find me some and I’ll be convinced
“Beyond this study, another unspoken argument about cycling modeshare trends is that with such a small base, an unadjusted year-on-year growth of 30%, 50% or even 150% cannot be taken seriously — it is simply too close to the margin of error, too sensitive to random fluctuation. ”
This is a valid comment and I did acknowledge that the manual cycling counts are highly susceptible to variation, particularly due to weather and human error. If we were seeing one year with a crazily high count and then another year with a very low one I’d be sceptical.
Instead, what the cycle counts suggest is that there has been a gentle increase over time from “8417 cyclists in 2007, 8783 in 2008, 8427 in 2009, 10659 in 2010, and 10917 in 2011″. These don’t look to me like numbers that are being skewed by crazy variations in counting techniques.
Instead they look quite similar to what you might expect given that the Household Travel Survey also suggests there has been a gentle increase in cycling rates as a proportion of total mode share over the same time period.
Lucy,
“But I’d love to see you present evidence that there has been a per capita decline in cycling rates in either New Zealand or Australia since 2006-7″
Given that no one is claiming there has been a known recent decline, or that a recent boom is impossible, why should anyone present this evidence? The burden of proof is on those who made the positive claim that there has been a cycling boom (and the absolute numbers they have given so far, as the author explains, do not meet the burden). Failing this, we can reasonably reject the new and contrary claim of a cycling boom.
I have repeated this argument in various forms already, but if you would like to continue to insist that the burden of proof lies elsewhere, or that anyone else is asserting an unsupported claim, then so be it; I will refrain from a futile debate.
“the same author also talked a lot about helmets in the associated press release that came out with this article.”
It is wholly unfair to judge the author’s original academic work by what an unattributed press release selectively quotes him as saying. We know from the published study how rigorous (or not) his research is, and from his op-ed what his opinions and priorities are on this particular subject of a cycling boom. The press release takes a very different angle (possibly because it is written or edited by someone else?) and actually covers other studies at some length. So all that should not detract from what we know is the author’s own work on the topic at hand.
“I still haven’t seen any conclusive evidence that helmet laws lead to a dramatic decline in cycling rates or, indeed, that removing them would lead to a dramatic increase in the absence of any improvements to cycling infrastructure.”
This is a different issue from the question of evidence regarding a cycling boom. Even the author is not claiming in any of the three relevant publications that helmet laws are certainly a cause of the decline. Generally, he is merely not ruling out its potential contribution — i.e. after infrastructure and investment issues. It is appropriate, given such confounding factors.
Of course, he may well believe it is a cause of the decline, with or without any statistical basis. But it is beyond the scope of the discussion of a cycling boom. Only in the press release is he quoted as saying “[t]he most likely deterrent to more people cycling” (note: not the same thing as “the most likely cause of the decline in cycling”) — but that is still in reference to a previous study he did, not the one in question here.
So he has conducted other studies and written further op-eds on the subject of helmets. You may be interested in examining those separately, if you care.
“Find me some and I’ll be convinced”
I suppose you will accept evidence even if it is not properly calculated!
Seriously, I am sure you know there are plenty of conflicting studies about this. I am not interested in trying to convince you of it one way or another. The only reason I mentioned the Finch study was as an example of the sensitivity of such small numbers to random factors — especially when sampled in a similar way to the regional cycle monitoring programme you subsequently quoted from. (Essentially, their optimism based on that data point was unfounded, whatever the cause of the long-term trend.)
And those Auckland regional figures are fascinating in light of the original discussion:
The percentage increases reported are year-on-year unadjusted figures. Accounting for Auckland’s estimated resident population growth in roughly the same period, the per capita adjusted increase in cycling movement counts for 2007/2011 is more like 23%, not 30%. Also, the most recent year 2010/2011 is given as a 2% positive change, but it is actually closer to 0.77% after adjusting for an estimated 1.65% population growth in roughly the same period. (That is less than a 1% change for a transport mode whose share is itself on the order of 1%.)
These numbers show anything but a “gentle increase”. In relative terms, there is an abrupt leap at 2009/2010, which is itself slightly greater than the overall per capita adjusted growth from 2007 to 2011, and so it accounts for most of the recorded positive change. It even follows a significant drop in the previous year (-5.5% adjusted).
I find it hard to believe that a “cycling boom” would suddenly appear and last for just 12 months at best before flattening again. If I was studying these figures rigorously, I would explore a range of plausible correlations, and perhaps methodological changes near that period, instead of eagerly resorting to a miracle explanation.
Although this counting and projection approach has many limitations, it is a decent enough method. So putting aside human error etc, I would also investigate the possibility that individual cyclists are coincidentally funneling through multiple counting sites on legitimate trips (gives sites measure discrete movements, not unique cyclists). Perhaps the apparent growth reflects favourable changes in trip generators or trip distribution or routing around the observation sites, rather than a real uptake in mode choice?
I will not even discuss AT’s monthly automatic counts at length. If there are any shoddy statistics in this discussion, that would be them. There is no acknowledgement of error — statistical, methodological, systematic, operational or otherwise — in their reports, and no accounting for variance other than self-evident seasonal effects. There is no adjustment for population growth over a year, and the numbers are not normalised to something sensible (I would suggest “mean movements per day per month”, so they can usefully be charted together). There is no clarity around which sensors were counting when, and whether the efficacy of each sensor remained constant. And so on. It is a promising initiative, and it could be a source of great data given time and further enhancement, but I cannot take the statistics at face value currently.
Still, given the overall mood of the numbers, and frankly even my own subjective experience on the ground, I am tempted to believe there is some faltering per capita increase in cycling in Auckland. This is a legitimate claim to make based on the data, but sadly it is often overestimated and under-qualified when presented as an unabashed cycling boom. We agree that more and better data would help; but I would prefer caution over exaltation in characterising what little it reveals for now.
Geez, cheer up people, it just isn’t that bad!
I’ve just been looking at the regional annual cycle counts http://www.aucklandtransport.govt.nz/about-us/publications/PlansProgrammes/Pages/Regional-Cycle-Monitoring-Plans.aspx
they show a region wide 6% drop, possibly a reflection of the poor summer not encouraging people back into riding
I don’t think I’ve ever said there has been a cycling boom anywhere on this blog ever. I’ve said that I think there has been a gentle increase which at 30% is faster than the rate the population is growing at 10%. So we pretty much agree. You also have come to this conclusion – you worked out actually exactly how much it would be – 23% – great work
I would also distrust the sudden increase in 2009-2010 if it was not supported by a very similar figure in 2010-2011 and as it happens 2011-2012.
You discredit the monthly sensors based on the fact that they are not adjusted for population but that is how all traffic and public transport patronage counts I have ever seen work. The point is that you compare them to the same month in the year before and you wait 2 or 3 years to be sure you are seeing an ongoing trend. They always report data from the same counters – the ones they have had running now since the start, and the counters run all the time (hence why they don’t say WHEN they are counting). That’s why the automatic counters are mainly good for seeing a trend. A point I think I’ve also made on thsi blog many many times.
I am sorry you think these figures are so inadequate. Interestingly, Portland’s claims of a significant increase in cycling rates seem to be mainly also based on annual manual counts – at least based on their website. So perhaps there simply is no better way of doing it?
“Given that no one is claiming there has been a known recent decline, or that a recent boom is impossible, why should anyone ”
Well, actually, I think if you go back to the article that is my point. AT the beginning of the article the author speaks dismissively about claims of an increase in cycling based on figures from the late 1990s to early 2000s. He then goes on to present figures from 1986 to 2008 and say that these prove there is no cycling boom. So I think the way his article is worded (and his quotes in the subsequent press release which I also believe are relevant, just as I think that the views that Chris de Freitas gives on climate change outside his academic work are relevant to his research because they suggest that he has a very strong bias towards proving a certain point) does suggest that he is trying to discredit the claims of a recent cycling boom.
Anyway, you’re obviously not going to go and actually look up any statistics about recent cycling rates in Australia or find me any more convincing helmet research than I’ve previously seen so let’s just call this discussion quits eh?
Just to be absolutely crystal clear. Here is the blog post I wrote about the Household Travel Survey for example. Which does show a moderate increase in cycling mode share. You may note that I gave this blog the title of “The most boring graph in the world”
http://cyclingauckland.co.nz/general/2012/05/the-most-boring-graph-in-the-world/
Here is anoyther blog post titled “May cycling report shows modest 5.6% increase”
I then go on to say “If this is a sign that more people are beginning to cycle to work (faster than people are taking up recreational cycling) this is great news. Overall, it’s probably too soon to say that for sure.”
I don’t think that suggests that I have been running around saying that cycling rates in Auckland are absolutely soaring out of control.
And here is a blog post where I talk about the idfferent counts and here is what I say about the automatic counters
“Well, the manual counts and automatic counters probably provide a good indication of the trends in cycle rates – are they increasing or decreasing. In particular, the automatic counters will help to give an idea of seasonal and monthly variation.
But they probably can’t give you such a good idea of what amount (or proportion) of total trips in Auckland are made by bike.”
Personally, I think you’re going to achieve more for your cause (which is presumably to make Auckland a more cycle and pedestrian friendly city) by being positive BUT realistic about the numbers, rather than always looking for a negative explanation.
But up to you! If you think your approach will generate great results then more power to you.
I’m on a shitty computer so those links don’t work but I’m going to stop this now and not look at this blog for 3 months. Maybe you could write some blog posts while I’m away about cycle counts. You seem to have a lot to say
@lucyJH yes, unfortunately “let’s just call this discussion quits” is how the dangerwear camp like to park this particular elephant in the cycleway.
But this ‘question’ ain’t going away, and cycle advocates had better get their house in order if they want to be seen as credible, coherent and authoritative – particularly in relation to the international best practice – and if we desire anything more than crumbs at the table of the transport czars.