Are infrastructure costs higher on the fringe?

Capital costs per dwelling of infrastructure provision - inner suburbs vs outer suburbs (from Trubka et al)

The exhibit above purports to show that the cost of infrastructure associated with building a new dwelling within 10 km of the CBD of a city like Melbourne is, on average, $50,503. In contrast, it costs $136,401 to provide infrastructure for an outer suburban dwelling i.e. located more than 40 km from the CBD. That’s a huge difference: $85,538 per dwelling.

The figures come from a 2007 report, Assessing the costs of alternative development paths in Australian cities, written by three Curtin University academics, Roman Trubka, Peter Newman and Darren Bilsborough. I’ve mentioned this report before, but that was primarily in the context of The Age and some public sector agencies tending to conflate economic costs with infrastructure outlays (they’re not the same!).

The figures above however are solely infrastructure outlays (not economic costs). Judging by the extent to which Trubka et al’s report is cited by government agencies, there appears to be strong demand for this type of information. It seems, however, that these are the only numbers on this topic around. That’s unfortunate because they have some very serious shortcomings as an indicator of the relative cost of providing infrastructure in inner and outer locations.

The key deficiencies are they’re old; they don’t relate to Melbourne; and they’re not transparent. Trubka et al sourced them from a 2001 report, Future Perth, prepared by the WA Planning Commission to assess infrastructure costs in Perth. Future Perth didn’t calculate its estimates from first principles but rather surveyed 22 earlier studies, some dating from as far back as 1972 and some relating to costs in the USA and Canada.

Future Perth is a working paper and hasn’t been published – hence the rigour of its methodology and those of the 22 studies it drew from hasn’t been tested. Unfortunately, Trubka et al provide scant explanation of their infrastructure estimates, relying instead on a reference to Future Perth.

I can’t say for sure the Trubka et al estimates are wrong, but I can say they’re unlikely to be right. I can also say they’re far too flaky to be relied upon to guide significant policy or investment decisions here in Melbourne. There’s clearly a demand for this sort of information so it would be sensible for the State Government to undertake its own rigorous and up-to-date assessment of the costs of metropolitan infrastructure provision.

Although not as decisive as the shortcomings discussed above, I also have some issues with how Trubka et al have set up their cost comparison. Actually, because the report doesn’t elaborate much on the various infrastructure items, I’ll treat these as questions, or areas that need clarification. Read the rest of this entry »


Is the Transport Department coming out on outcomes?

xkcd on the value of time (the corresponding figure in Australia is circa 4 minutes)

I don’t ordinarily read annual reports but this story in Melbourne’s Herald-Sun acknowledged its source was the Victorian Department of Transport’s 2010-11 annual report. The newspaper noted a 3.4% increase in reported crime on Victoria’s public transport system in 2009-10. Much to my surprise, the paper acknowledged up front that patronage had also increased over the period, by 1.6%. That’s a pertinent qualification but it surprised me that a News-owned tabloid bothered to make it. Well done to the Hun and reporter Peter Rolfe – a big tick!

Looking at the performance outcomes section of the annual report, I see the Department’s been clever – its performance indicator for reported crime against the person is effectively the change in offences minus the change in patronage growth. Bit hard to ignore that and it certainly helps to put the crime figure in perspective.

This got me wondering what progress the Department is making in dealing with crime on public transport. What, I wondered, is the trend? How does the latest figure compare with the previous year? This was when I discovered some serious deficiencies in the way the Department reports on outcomes.

A key shortcoming is the annual report fails to give the previous year’s outcome figure. So, for example, the report tells us that the total volume of freight carried by rail in Victoria increased by 8.31 billion tonne-kilometres in the most recent year for which data is available, but provides no indication of whether that is an improvement on the trend or a deterioration. What is the reader to make of this number?

Similarly, bicycle path use increased 2.8% and rail capacity utilisation in the morning peak grew by 9.7%, but in neither case is there a figure on the previous year’s outcome. This is in stark contrast with the financials, which use the convention of showing 2011’s numbers against the previous years. The examples I’ve cited aren’t isolated cases – none of the performance outcome results show the previous year.

The importance of context is shown by the Herald-Sun’s report. Peter Rolfe did some legwork and found reported crime dropped 10% in the previous year, while patronage grew 6%. That’s much better than the latest year’s 3.4% vs 1.6%, suggesting a marked deterioration in performance. Although he failed to acknowledge the difference between the two years might in part be due to increased policing (a cross!), he nevertheless illustrates the value of having something against which to compare performance measures.

Unfortunately, the shortcomings of the Department’s annual report get worse. While each indicator has a target “increase” or “reduction”, most of the “results” actually provide a snapshot or balance statistic – they don’t show the change. For example, one outcome indicator purports to show the increase in the proportion of trips starting or ending in Central Activities Areas (CAA). But what we’re told is that 3% of trips started or ended in CAAs in 2009-10. That’s valuable information in its own right, but it’s not telling us if the Department is succeeding on this indicator. It’s not telling us if there was an increase. It’s not telling us what the indicator says it tells us!

Another example is the target to reduce fuel consumption of petrol vehicles – the result is given as 11.4 litres per vehicle kilometre, but again there’s no indication whether things are getting better or getting worse. Likewise, the result for the car occupancy rate indicator is 1.2 persons (it’s meant to reduce); the greenhouse emissions of the vehicle fleet is 384 g/km (it’s meant to reduce); the average weight carried per freight vehicles is 4.19 tonnes (it’s meant to increase); and so on – but do these represent an improvement or a deterioration on the trend?

This is not a minor issue. Of the 34 indicators that claim to show an “increase” in good things or a “reduction” in bad things, 20 do not even attempt to show how they’ve changed. They’re all snapshots. Experts will in many cases know the context but annual reports are prepared for parliamentarians and for the people. Read the rest of this entry »


Why do outer suburban streets look so bland?

Manna Gum Drive, Ventnor, Phillip Island

Critics are gunning for Victoria’s Planning Minister, Matthew Guy, following his decision to rezone 5.7 Ha of farmland at Ventnor, Phillip Island, for residential use despite the opposition of Bass Coast Shire Council. The rights and wrongs of the Minister’s decision is no doubt a fascinating topic, however my present interest is in the way this land is likely to be developed.

I got to thinking about that after reading a letter in the paper on Saturday from the owner of a beach house at Ventnor, expressing the “hope that this natural wonderland does not become transformed into a home of little courts, high fences, and narrow streets filled with the McMansions of suburbia”.

I wouldn’t be holding my breath if I were him – there’s got to be a much better than even chance that any new residential development will end up looking more like nearby Manna Gum Drive (see exhibit) than the more traditional ‘beach houses’ of Ventnor. In fact even that seems optimistic – given the enormous decline in average lot sizes in recent years, a more probable scenario could be this development in Melton.

He can probably rest easy about his fear of McMansions though. Two storey behemoths are likely to be too expensive for most Ventnor newcomers – it’ll probably be single storey brick veneers with low tile roofs and two car garages.

Being near the beach doesn’t faze the standard suburban form – it’s ubiquitous. Drive 100 metres back from the beach in large parts of the sub-tropical Sunshine Coast or Gold Coast, close your eyes, and you could as easily be in the bland streets of Melton or Campbelltown. You’ll even have the same experience in tropical Cairns.

I expect they all look much the same because the economics of land development and cottage building produces the same solution everywhere. Affordable lots are 500–700 sq metres with high fences for privacy. The houses look more or less the same because the home building industry is pretty efficient at churning out economically priced detached houses in low-maintenance brick veneer.

As with most mass produced items built to a price, the scope for differentiation isn’t high, often just a tweak of the front facade. Buyers can have something markedly different if they want – they might, for example, commission an architect – but they’ll have to pay a lot more for the advantages of a bespoke design. But that’s just not an option for the vast bulk of buyers in areas like Melton and, I daresay, Ventnor.

A key reason streets in fringe suburbs look so boring and nondescript to sophisticated eyes is their relative youth – trees planted in the nature strip haven’t had time to take off and residents haven’t yet established front gardens. Many streets in established suburbs were bland once too. The streets of Eaglemont and Ivanhoe doubtless looked pretty insipid at first with their small brick and tile houses, but generations of zealous gardeners cultivating their front yards and nature strips have created, in effect, a completely new streetscape.

Yet there are many streets in established suburbs like this one in Keysborough which are still pretty uninspiring despite the advantages of maturity. This relatively young street in Melton looks like it’s lost some street trees already and parts of the nature strip have become a parking lot. Here’s another newish one in Melton where the front yards don’t even pretend to be gardens – they’re all driveway and low maintenance ground cover. And most of the houses on this street on the Sunshine Coast were built at least 30 years ago (in fact some have been redeveloped) yet apart from a few desultory palms, trees with scale aren’t very common. Read the rest of this entry »


Can selection bias shoot down an argument?

Manhattan in Motion

Urban policy is rich in opportunities for fallacious thinking – for example, surveys that purport to show huge latent demand for a particular mode of transport, but sample only users of that mode. So I’m always interested in new examples of where we can so easily go wrong.

Alex Tabarrok at Marginal Revolution recently provided a pointer to this great historical example by John D. Cook of the importance of selection bias. It isn’t specifically related to urban policy, but nevertheless provides a valuable lesson. Cook says:

During WWII, statistician Abraham Wald was asked to help the British decide where to add armour to their bombers. After analysing the records, he recommended adding more armour to the places where there was no damage!

According to Cook, while it seems backward at first, Wald realised his data came from bombers that hadn’t been shot down:

That is, the British were only able to analyse the bombers that returned to England; those that were shot down over enemy territory were not part of their sample. These (surviving) bombers’ wounds showed where they could afford to be hit. Said another way, the undamaged areas on the survivors showed where the lost planes must have been hit because the planes hit in those areas did not return from their missions.

Wald assumed that the bullets were fired randomly, that no one could accurately aim for a particular part of the bomber. Instead they aimed in the general direction of the plane and sometimes got lucky. So, for example, if Wald saw that more bombers in his sample had bullet holes in the middle of the wings, he did not conclude that Nazis liked to aim for the middle of wings. He assumed that there must have been about as many bombers with bullet holes in every other part of the plane but that those with holes elsewhere were not part of his sample because they had been shot down.

Here’s another example via Marc Gawley, who noted a BBC story implying that three quarters of those committing crimes during the London riots had previous convictions or cautions. Is that really true, he wonders, or:

is it that those with previous convictions have their details on a police database and it was therefore possible to identify (and find) those people based on images from photographs and recordings made last month? Read the rest of this entry »


Is Melbourne pushing the boundary (too far)?

Melbourne in 1988 - a part of the 2010 image, showing the effects of the drought, is visible on the RHS

Geosciences Australia has released a new series of satellite images comparing the extent of development in Melbourne in 1988 with 2010. Click on the image above to go to Geosciences Australia’s web site where you can do a “swipe” comparison of the two images i.e. move the cursor across the image to progressively reveal the second image underneath.

According to a report in The Age, the images show a “massive” increase in urban sprawl, with Melbourne’s urban footprint surging to the north, west and south-east. Melbourne has “marched into surrounding rural landscapes” and its “unofficial boundary is now more than 150 km east to west”.

Melbourne has certainly expanded at the fringe over the last 22 years, there’s no doubt about that. However I’ve argued before that both the extent of sprawl and its downsides are routinely exaggerated, so I want to have a closer look at these images and at The Age’s interpretation.

One thing that struck me straight up is that any comparison of the two satellite images can be misleading unless the viewer appreciates Melbourne was still suffering the effects of a long drought in 2010 (compare the size of the dams in each year). So areas that appear to have gone from green to brown between 1988 and 2010 might reflect lack of rain, rather than an increase in development.

In fact 1988 was an unusual year. There was a La Nina event in 1988-89 – the first since 1973-76 – and hence there were wetter than normal conditions. Have a look at this older CSIRO comparison of satellite images of Melbourne in 1972 and 1988, and note the CSIRO cautions that 1972 was much drier than 1988.

It should also be borne in mind that 22 years is a reasonably long period in urban development terms. We shouldn’t be surprised to see substantial change when a metropolitan area is growing. Have a look at these images to see how spectacularly some other growing cities have changed in the course of 20 years.

The level of growth also needs to be considered. Melbourne’s population grew from circa 3.1 million to around 4 million between 1988 and 2010. That’s an increase of about 30%, which is considerably more than the apparent increase in the size of the urbanised area. That’s to be expected, as a large proportion of population growth – currently approaching half – is accommodated within the existing urban fabric via redevelopment.

And then there’s The Age’s claim that Melbourne’s urban footprint spreads “more than 150 km east to west”. It doesn’t. Using GIS, I measure at most 75 km from the western edge of Wyndham to Lilydale in the east and 85 km to Pakenham in the south-east. If I measure instead from Melton (putting aside that it’s separated by 9 km of green wedge from the continuously urbanised area), I still get less than 80 km to Lilydale and less than 100 km to Pakenham. Some might think that’s still too big, but it’s a lot less than 150 km plus. Read the rest of this entry »


What should be done with myki?

Road pricing in London, 18th century

The Age reckons myki is “failing at maths”, but I wonder if the key failing is actually with the way it’s managed rather than with any technical shortcomings.

This report in Tuesday’s paper said “hundreds of travellers (who use myki) are paying too much”. It follows an earlier report by The Age, back in April, when it was claimed that “one third of myki bills are inaccurate”.

Both these stories refer to MykiLeaks, a web site set up by Monash University student Jonathan Mullins. It enables travellers to review (some aspects of) the accuracy of myki statements on-line.

The Age’s latest report indicates a big improvement since April – the proportion of inaccurate statements is down from 33% to 15%. But if valid, that’s still a very high error rate even if, as The Age’s reporter says, the total dollars involved aren’t significant (the combined overcharge across circa 300 faulty statements was $1700, with the biggest error being a traveller charged $18 instead of $6).

But there are reasons to be careful about how much weight to put on the MykiLeaks results. One is that we don’t know how accurate the MykiLeaks algorithm is. Another is that only 2,000 statements have been submitted to the site since it started last December. That’s a very small proportion of the million plus myki cards on issue.

Such a small sample might not be problematic if it were randomly selected, but it seems unlikely the kind of people who use MykiLeaks are representative of the whole body of myki users. They could have good reason to be wary if, for example, they are in the small group who make the kind of complex, multi modal trips where myki appears to be weakest.

I don’t know for sure how accurate myki actually is or even what most of the errors are, but I don’t put a lot of store by MykiLeak’s findings. But that doesn’t really matter all that much because the key issue in my view is the poor public perception of myki – it’s pretty clearly a tainted brand. I’m the first to be wary of on-line polls, but surely it’s not without significance that a staggering 91% of the 5,657 readers who voted in The Age’s poll answered ‘No’ to the question: “Do you trust myki to charge you the correct amount?”.

Moreover, read through the comments on the article and it’s evident there are other issues too. Quite a few people have difficulty understanding their account statements. Others are annoyed by the considerable delay between topping up their account on-line via credit card, and when the funds become available for use. Many resent having to identify errors themselves and then contact a call centre to have it fixed up.

If The Age’s report is a fair account, the Transport Ticketing Authority (TTA) seems to be taking a defensive posture. That’s not good for business and it’s not smart politics either.

The best counter the Chief Executive, Bernie Carolan, could manage was to warn “myki users against giving their personal information to the (MykiLeaks) website”.  He also said he couldn’t confirm MykiLeak’s claims because the Government doesn’t have access to the site. Just as he did back in April, Mr Carolan again provided this reassuring, customer-friendly advice: “If a customer is concerned they have paid more for their fare than required, they should contact the myki call centre on 13 69 54”.

This sounds to me like an organisation that’s a little out of touch with its customers. People don’t want to hang on the phone to a call centre to correct a mistake they didn’t make. Moreover they expect the TTA will identify its own overcharging errors, not leave it up to customers and wait till they complain. Perhaps most of all, they want to be reassured – they want the TTA to tell them honestly and plainly if they’re being over-charged or not. Mr Carolan didn’t answer that question – he responded with bluster. Read the rest of this entry »


9/11: Should America move on?

South Pool, 9/11 memorial, Ground Zero

It’s a very brave or very insensitive American who would publicly question his nation’s entire 9/11 memorial project on the day of the tenth anniversary. Yet just a few days ago, on the morning of 11 September, US economist Robin Hanson began his internationally popular blog, Overcoming Bias, by complaining that half the comics in his Sunday paper “are not-funny 9/11 memorials”.

I did my own mawkish 9/11 ‘memorialising’ last time (the video On the transmigration of souls), so I too was guilty of dishing out largely uninformative and non-analytical content on this occasion. But Professor Hanson has a much bigger message (the links are his):

In the decade since 9/11 over half a billion people have died worldwide. A great many choices could have delayed such deaths, including personal choices to smoke less or exercise more, and collective choices like allowing more immigration….

Yet, to show solidarity with these three thousand victims, we have pissed away three trillion dollars ($1 billion per victim), and trashed long-standing legal principles. And now we’ll waste a day remembering them, instead of thinking seriously about how to save billions of others. I would rather we just forgot 9/11.

Do I sound insensitive? If so, good — 9/11 deaths were less than one part in a hundred thousand of deaths since then, and don’t deserve to be sensed much more than that fraction. If your feelings say otherwise, that just shows how full fricking far your mind has gone.

Just to really drive home his point, Professor Hanson updated his post with a link to news agency Aljazeera, which featured an article Let’s forget 9/11 by American Tom Engelhardt, author of  The American way of war: how Bush’s wars became Obama’s and The end of victory culture.

Not surprisingly, given this was 11 September, some comments from readers were pretty direct:

This will, sadly, be the last time I read your blog.

You, Mr. Hanson, are an idiot. You have no conception of mythos or national dignity. Good bye, sir.

Robin, you’re “fricking” out of your mind.

A co-worker used to have a little sign in her office: Most people know how to remain silent. Few people know when.

Some other comments took issue with his basic argument:

Insensitivity will not stop people from being affected by some tragedies more than others.

Human nature is to attach greater significance to things that are close to us and have evoked personal feelings, rather than basing it on magnitude alone. And there is nothing wrong with that. Your moralizing is no better than the ridiculous moralizing that has gone on in the past ten years in reaction to 9/11. Everyone has their story to tell; yours involves looking down on the natural human reaction to a dramatic event. That is all.

Defence against terrorists, not solidarity with victims, explains the “pissing away” of three trillion dollars.

You cannot reasonably expect this sort of brazen and demented contrarianism to induce readers in search of a new moral framework to think, “hrm, maybe I’ll give this utilitarianism thing a whirl and see where it takes me!”

Might as well forget about the Holocaust too. In the bigger scheme of things, what were only 6M Jews?

But quite a few comments were supportive of Professor Hanson’s view:

I think what Robin is saying is that the response to 9/11 didn’t do any of the things that the “leaders” who pushed those responses said those responses were going to do. What those responses did do was piss away $3 Trillion (and counting), kill a bunch of US soldiers, and squander the “good will” of the rest of the world while committing great evils; the killing of many Iraqi civilians, the strengthening of al Qaeda and the bankrupting of the US economy.

If only one 1/100 of the social capital spent on 9/11 remembrance was spent on organ markets… I would expect more than 100x as many lives would have been saved compared to any anti-terrorist measures implemented because of 9/11. Read the rest of this entry »