Why the “induced demand” argument against roads is no good.

I read a lot of forums and articles about transport policy where most of the people agree with each other. The splintering of discussion into groups that agree is a common and well-understood phenomenon that is amplified on the internet.

Listen up, Melbourne

As well as getting excited by optimal stop spacing, one thing people in these circles agree on is that roads are bad. The most common argument made against roads goes like this:

“Essentially, if you widen roads to reduce congestion, people who were avoiding the road because of congestion will find it more convenient and take more trips, thus increasing traffic again.

So what do you have then? A big expensive project to eliminate traffic, and more traffic.” Streetsblog

“New roads will create new drivers, resulting in the intensity of traffic staying the same.

Mann explains how this counterintuitive reality can possibly be true: “As it turns out, we humans love moving around. And if you expand people’s ability to travel, they will do it more, living farther away from where they work and therefore being forced to drive into town. Making driving easier also means that people take more trips in the car than they otherwise would.” Planetizen

“Research indicates that generated traffic often fills a significant portion of capacity added to congested urban road. Generated traffic has three implications for transport planning. First, it reduces the congestion reduction benefits of road capacity expansion. Second, it increases many external costs. Third, it provides relatively small user benefits because it consists of vehicle travel that consumers are most willing to forego when their costs increase.” Victoria Transport Policy Institute

While I support many of the ideas this argument is rolled out to support, I find the argument itself utterly unconvincing. If you want to argue against investing in roads, you’ll need a better argument. Here’s why.

1. Public Transport will behave exactly the same.

If you widen a road, it will encourage more people to drive on it, bringing congestion to an equilibrium level* on that road, and spilling congestion into other parts of the network.

Likewise, if you increase the frequency with which a public transport service runs, it will attract more people to ride on it, bringing congestion to an equilibrium level and spilling congestion into other parts of the network.

This means that induced demand operates as an argument not to invest in any popular/crowded transport where crowding levels may be deterring travel. I’m not sure that’s what proponents of the induced demand argument intend.

*The equilibrium level is not necessarily the same amount of congestion as before, it’s the level where congestion deters travel.


2. Induced demand is good.

Induced demand means previously people had latent demand for travel, but they were unable to satisfy it. Now they are able to access jobs, get to shops that sell items that match their needs better, visit friends and family more. Any argument about transport investment that objects to people doing more travel should be treated with suspicion, if not contempt.


3. Focusing on congestion is the wrong way to look at transport policy.

If your approach to solving a city’s transport problem is to seek out choke points and jams and try to untangle them, you may end up fighting unwinnable battles against geometric problems. You will become frustrated, growing ever more sure that transport is a zero sum game with no easy answers.

But the answer is not necessarily eliminating every queue – there are planners who believe congestion is perfectly acceptable, a sign of popularity and even of success. (1, 2, 3,)

Forget congestion. If you make your yardstick access, you focus on what people want: What places can I get to?

A rational approach to transport planning in a city would be to measure access: From each address, a sum of all the jobs, services and other addresses that can be accessed, averaged across a time period encompassing peak and off-peak.

That number would be aggregated across the entire city to create a total score. Then, you would provide incentives for bureaucrats to improve that number. These bureaucrats must have more than just transport levers at their disposal.

They should also be able to make decisions on zoning and land-use, and road pricing. Improving access can as easily mean moving jobs to people as people to jobs. If a particular investment is expected to have spill over effects that worsen transport times in a far-flung part of the network, the access measure should pick this up and permit that effect to be compared to its positive local effect.

Build a new school, access improves. Add traffic light priority for buses, access improves for some at the expense of others. Price a road, access improves for high-value trips at the expense of lower value trips. Add a new lane on a freeway, access improves locally and probably diminishes elsewhere.

Focusing on access allows trade-offs and comparisons to be made. But often, improved access will involve turning latent demand for travel into real trips. We shouldn’t object too loudly to that.

The point of a city is that people want access to other people, and services. Let them at them.

More on this topic:

Trevors in Traffic: a PR strategy for congestion charging.

Selling the street: A land use hypothetical…

New competition report tries to go hard on road pricing, but is naive

How to dodge a big payout on cancelling a road contract AND avoid creating sovereign risk

The government of the state where this blog is produced is in a pickle.

Prior to an election last November the then Opposition promised to cancel or defeat in court a contract for a big controversial road tunnel. The tunnel, worth perhaps $6 to $10 billion dollars, has not been built yet. Nothing beyond planning has commenced

Now the former Opposition are in power, they are finding that the old government left them a poison pill. 

Screen Shot 2015-02-06 at 1.07.08 pm

If the road is not built for any reason, the government must pay the company that would have built it $1.1 billion. This clause was added by the previous government. The companies might have done only $50 million of preparatory work – being generous here – but they get paid $1.1 billion simply for missing out on finishing the job.

Ignoring for a moment the morality of inserting such a clause into a contract (it’s vile, wasteful, ridiculous, and would in a better world result in a range of senior bureaucrats and politicians going to jail), we turn our minds to how the present government can deal with it.

There are three main options.

1. Avoid the payment and make the road. This would involve reneging on a major election promise, but you don’t waste the money.

2. Avoid the road and make the payment. This would gift a billion dollars from an indebted state government to a consortium of companies including Lend Lease Group, worth $9 billion, Acciona, worth €3.6 billion and Bouygues SA, worth €10 billion.  It would probably be politically convenient too.

3. Avoid both the road and the payment. The government has one big advantage. It makes laws. It can write legislation that annuls the offending contract. But the big risk in such a course of action is that it establishes an extremely unwelcome precedent that promised payments can be cancelled at whim by the government, and valid questions being raised about sovereign risk.

I want to look more closely at option three. Is there a way a law could be drafted that gets a just result and avoids sovereign risk? I think there might be.

Any law to cancel the payment provisions in the east-west link should:

1. Make it clear that this is a once-off by raising the hurdle for ever cancelling this kind of contract again.

For example, the Government could include a clause requiring that in future passing legislation that annuls any contract above a multi-billion dollar value threshold requires a supermajority in parliament, e.g. two-thirds of votes. The requirement for a super-majority should not apply to contracts where the cancellation provisions are substantially greater than the cost of the work done.

Sovereign risk only applies if a company can genuinely fear its contract provisions may be changed by legislative fiat. If they fear risk, they will raise prices.

Reducing the risk of such legislative action should attenuate the real costs of sovereign risk (although it won’t prevent the political costs of big companies mouthing off about it.)

2. Legislate against any future government ever introducing “poison pill” contract clauses into infrastructure contracts. (Part of me wonders if this law could apply retrospectively?)

3. Legislate that any large contract signed during the “caretaker period” in the lead-up to an election should be agreed upon by the leader of the Opposition as well as the Government, in order to prevent sneaky surprises. Part of the problem with the east-west link project was that it was never an election pledge, was controversial for 3.5 years, and with weeks before the election it looked set to lose, the government signed a contract.

We’re in a tangled mess.

Now. Could the state government of Victoria pass such legislation? It has a lower house majority, so it could pass it there, no problem. In the Upper house it holds just 14 of 40 seats. But The Greens have five, and they are likely to support such a plan. Then the government needs just a couple more, drawn from The Democratic Labour Party, the Sex Party, Shooters and Fishers, and Vote 1 Local Jobs. It might require some side promises, but it may be possible.

I welcome your thoughts and comments on this idea. Please leave a comment below, or hit me up on Twitter.

How is the peak oil movement dealing with the oil price crash?

Things are eerily quiet over at the Peak OIl subreddit. There’s been only a handful of posts in recent weeks. None of them seem to engage directly with the elephant that just parachuted into the room:


At oilprice.com, they’re furiously counting barrels produced and making bold predictions that this is the view we get from the summit. This is it! This is peak oil. There is scarcely a mention of price. Similar at PeakOilBarrel.

Over at peakoil.com, however, they’re addressing the issue, listing reasons that cheap oil doesn’t mean the end of peak oil. The key arguments basically insist that the oil price will go back up.

“(12) Chevron have released a presentation for their investors … which indicates an expectation that 40% of the “new oil” will come from deepwater fields, 20% from U.S. shale, 10% from increased tar-sands production, 25% from OPEC growth (Venezuelan extra-heavy oil?), and around 5% each from shale outside of the U.S. (Russia?) and “onshore and shallow offshore”.

(13) Chevron also stress that production from these sources will not come cheap, and will probably be of the order of $100 a barrel (“Breakeven price” or “marginal cost”).

(14) Hence at under $50 a barrel selling price, these projects will not go ahead, or they will be money-losers (cost more to produce the oil than it sells for). This year, $150 billion worth of new projects may face the axe, which are mainly from heavy-oil, deepwater, tar-sands and shale-oil.”

I think this analysis looks sound as far as it goes, but it doesn’t address two big issues.

1. What OPEC might do next.

Although oil prices had fallen around 30 percent before the seismic November meeting in Vienna, the cartel is the reason behind much of the oil price crash.  Will they regroup? Or is this the end of any and all caps on production, and might the oil price sink further.

2. Demand.

Talking about a market from only one side, as though demand was limitless, is as silly as assuming supply is limitless.

Peak oil appeals as a theory because it is self-evidently true. Oil really is non-renewable, and we use a lot of it. There will be a single day on which more non-renewable hydrocarbons are extracted from the earth than any day before or after. That day is the peak of production.

But – Peak oil is a hot topic, and despite debates about the accuracy of reserve estimation techniques, it’s not because of the physics of reserves. Peak oil matters because of the assumption it is going to be a problem.

Really, peak oil is not about oil. It’s about whether we can adapt. It’s about how we manage demand. That’s a question about the operation of markets, about the functioning of governments, about technology.  About action guided by markets and action guided by collective planning. And that, in my opinion, is far more interesting.

The main assumption regarding demand is that oil consumption rises as the world grows richer.

This is the unstated fact behind the implication that peak oil is a catastrophe. Rising demand has been true for a long time. But need it always be true? Sometimes long-standing trends defy extrapolation.

There may be a few reasons to start thinking about peak oil from a demand perspective. Here are three.

1. It’s well-known that the energy intensity of a country’s GDP declines as that country goes from being involved in heavy industries to producing more services. China burns 250mL of oil to make a dollar of GDP. The UK uses more like 80mL. As places like China develop, their economies will become more fuel efficient.

But the relationship is not simple across countries, with very poor countries using very little energy and some rich countries using a lot

energy efficiency

It’s possible to see the above curve as a Kuznets curve lying on its side. (I’ve written about Kuznets curves before in a very different context).  Very poor countries and very advanced ones are more efficient. Middle income countries (and freezing cold Canada) are worst. If the most populous countries continue to develop, oil demand could shift.

2. Technology is crucial. Obviously, high prices drive the adoption of energy efficient practices, which is one reason why the Toyota Prius sold so well in the last ten years of high oil prices, and cars that get 13.3L/100km, like the one below, did not.

the car of the future

If oil prices stay low, that will be because those technologies – electric cars, solar power, biofuels –  are having their effect. But oil prices will probably go back up ( in fact they already have bounced a little). If that happens, households and companies will regain motivation to adopt technologies to conserve usage, and we will probably be having these debates about when peak oil will occur in another hundred years.

3. Climate change. We may see peak oil happen long before we’ve used up half the world’s known reserves. Moving carbon from beneath the earth’s crust to its atmosphere produces externalities that could, any moment now, become so obvious that they drive a political consensus to leave the carbon beneath the surface, where it is relatively safe.

What country flies most?

I was wandering up to the entrance at Melbourne airport late last year when three members of the Rebels motorcycle gang arrived at the terminal.

Approaching the sliding doors at the exact same time as these three large, black-clad, sunburned men, I realised I would either have to slow down or speed up to avoid bumping shoulders.

I went to speed up, then, remembering the news about bikies killing someone at an airport a couple of years ago, I suddenly reversed my decision. I might just keep my eye on this lot, I thought, and let them through first.

What happened next?

They passed through security and I later saw them sitting near the gate. Like any other Australian, a bikie sometimes needs to travel by air.

Flying is a regular part of life for most of us. I doubt there is anyone reading this blog who hasn’t held a boarding pass in their hand in the last year. It’s a convenience, but it’s also a pain. All that waiting, the tiny seats, the way food tray is so crowded with things and hard to manage. It’s easy to forget that most people in the world have still never flown in a plane.

That fact made me wondered if I could quantify Australia’s tendency to fly.

It turns out Australia is in the top 10 countries globally for passenger movement, according to World Bank data. (The World Bank uses domestic and international aircraft passengers of air carriers registered in the country, which is a fair proxy in the case of some nations, and not for some others).

When you look at the data, you could easily conclude that Australians fly more than the citizens of any other country.

This chart uses that World Bank data to show domestic and international flights for the most flying-oriented countries, plus a few others for comparison. Nine countries score higher than Australia on this metric

Screen Shot 2015-02-04 at 2.09.39 pmWhen you look at the chart, it makes sense. The countries with the most flights per capita are islands. They are also often small countries where there’s nowhere else to go except to fly out. Several of them are tourist destinations or aviation hubs. (nb. Ireland and Antigua probably top the list because of their tax status (and RyanAir), not their proclivity for flying. The Irish statistical office claims only 22 million international flights) .

The countries above may also top the flights per capita listing because they are small. Australia is more than double the size of any country above it on the list (UAE, home to Emirates airlines, has 9 million inhabitants.)Screen Shot 2015-02-04 at 2.08.55 pm

Given all this, I am prepared to say Australians probably fly more than citizens of any other nation.

(The data from our own statistical agencies show the World Bank figures may underestimate the amount of flying. According to BITRE, Australia has 57 million domestic passengers a year and 23 million people, making 2.5 domestic flights per capita per year. Add in 33 million international passenger movements and the data suggest more than three flights per person per year – closer to 3.5. New Zealand may also be a contender for the flying-est nation, but its numbers are swollen by high inbound tourism, while Australia sees more departures than arrivals and a higher share of domestic travel.)

So Australians probably fly the most! We head to the airport as often as every three months. Woo-hoo?

I didn’t think so.

Meanwhile, at the other end of the scale, Indian nationals rack up on average one flight every 20 years. That’s less than Kenyans, but slightly more than the Cambodians.

What would happen if Indians started flying more? Such a densely populated country will probably never fly as much as Aussies do, but what if they flew at the same rate as the Americans, 2.35 trips a year?

That would mean another 3 billion people taking to the sky every year. That’s a doubling of current global passenger numbers, which, according to ICAO, topped 3 billion for the first time in 2012.

It’s safe to say this would be bad for carbon emissions. The contribution of aviation would jump from 2 per cent of the global total to more like 4 per cent. Sorry, penguins. Sorry polar bears.

Furthermore, I estimate the world would need an extra 8,000 planes. (Based on a generous 370,000 passengers per plane per year). For context, Airbus currently makes around 50 planes a month.

The conseuquences for the aviation fuel market could be significant. There are already problems with supply leading to rationing.

Those extra planes are going to need somewhere to land as well. What cities in the world have space for more runways?

China has already increased its appetite for flying tenfold. From 35 million flights a year in 1994, it has increased to 350 million by 2014. The lesson is this – economic growth sends people to the airports, for business and for leisure.

India’s economy grew at 5 per cent per year. At that rate it won’t be a “rich” nation for perhaps a century. But in 2010 it grew at 10 per cent, and if –  like China –  its growth rate can be held high, then it can become very rich very fast.

If everybody in India can afford to fly – even the bikies – its worth asking the question of whether the world is ready for the consequences.

Are the interest rate cut and the foreign investment crackdown linked?

In 2014, the RBA was loath to cut interest rates. But in 2015 it came out swinging, beating their previous record by knocking the official cash rate to 2.25 per cent!

Did the big bank decide:

a) the downside of a housing bubble isn’t that bad; or

b) the economy is in such dire shape that we need an interest rate cut even if it inflates house prices into a giant puffy monster; or

Screen Shot 2015-02-03 at 5.59.25 pm

c) the risks of a housing bubble can be contained?

Now, option a) seems unlikely. The US housing crisis is still very fresh in the minds of the globe’s central bankers. Just last year the Governor was warning in – for him – very strong terms about the risks flowing from housing. 

Option b) is not the answer either. This excerpt from Tuesday’s decision is not penned by an RBA frothing with fear:

“Overall, the Bank’s assessment is that output growth will probably remain a little below trend for somewhat longer, and the rate of unemployment peak a little higher, than earlier expected.”

So could option c) be the answer? They’ve found a way to contain housing risks?

It was only yesterday that Prime Minister Tony Abbott made a very public announcement about a crackdown on foreign investment in real estate.

“I am a friend of foreign investment but it has to come on our terms and for our benefit. The government will shortly put in place better scrutiny and reporting of foreign purchases of agricultural land and better enforcement of the rules against foreign purchases of existing homes so that young people are not priced out of the market.”

It was hard to process that announcement at the time, being bereft of context and detail. There had already been a crackdown announced. Was this new approach to have a louder crack, or push further down? It was a mystery.

Until the RBA rate cut.

I hadn’t though of a linkage here until I saw the question posed on Reddit. The person who asked the question in that forum sees causation running the other way, with an interest rate cut necessary to accommodate the deflationary effect of the crackdown.

But we know that the order in which things are revealed does not necessary accord with causation. Assuming so is to fall victim to the simplest kind of fallacy: post hoc ergo propter hoc.

The RBA may be betting that keeping Chinese money out of the market will help keep a lid on things.

Will that work? It might. Foreign investment represents only a tiny share of our market. But introductory economics textbooks tell you, economists think at the margin. A few extra bidders for a house can be what pushes the price of that house through the roof.

Think about it like this: if there 100 seat in lifeboats and the ship is sinking, the price of lifeboats will be zero if there are 99 passengers, and start to rise very rapidly if there are 101. The marginal bidder is important.

But that’s just the theory. The reality is that foreign buyers are scooping up very little real estate indeed.

which country bought most value held gorwing faster than value sold

pie chart

The RBA faces a market where investors are doing a lot of the lifting. They accounted for around 40 per cent of buyers late last year, a record. This interest rate cut could see auctions turn into a frenzy.