Congestion charging is back on my list of good ideas
For a while there, I was influenced by the equity arguments against it. The lack of substitutes to travel, and the unique role of commuting to work in a person’s well-being tipped me against congestion charging. Good economic reform doesn’t throw out equity every time it can get an efficiency dividend, and I decided congestion charging equity problems made the policy unworkable.
But I’m swinging back to support for a simpler price signal. What has captured my attention is the following graph from a new Grattan Institute report. It shows the extent of congestion in Sydney. Amazingly, most people experience almost no congestion. Their commutes are swift.
What this tells me is the the impact of a congestion charge is actually not likely to be widespread. Serious congestion, of more than ten minutes in a trip, is the purview of a small subset of commuters.
That subset is likely to be going into the CBD, where congestion is real.
Remember that despite the importance of CBDs, most jobs are still in the suburbs. If we know one thing about CBD jobs – especially nine to five CBD jobs – it is that they tend to be the good kind.
City centres are where the business services jobs are. The specialised jobs that pay big coin, as opposed to the population-serving jobs (pharmacies, florists, bakers, doctors, schools) that are found disproportionately in the suburbs.
It looks like driving into the city in peak hour is an elite problem. No wonder it gets so much attention. The Grattan analysis makes it clear the congestion charging would really only have to be applied in a narrow area.
This fact also counters the argument that congestion charging can’t be introduced until better public transport happens. Melbourne and Sydney have radial public transport systems that provide terrific CBD access.
Traffic is bad. The absence of price signals on the use of existing infrastructure causes crowding and delays. You end up listening to way too much FM radio. But that might not be the most costly effect. The big downside is probably the pressure to build yet more infrastructure.
Daniel Andrews has green-lit the West Gate tunnel – a big freeway that will not only soak up $5.5 billion but also lock Victorians into a regimented tolling regime (not a congestion charge system) for decades.
Big freeway projects have a lot of side effects.
One is making the places that they travel through less pleasant. Place-making is a big theme in urban planning now and a lot of money is spent on making areas seem nice. This “tunnel” which is actually an elevated road for a good section of its length,is kind of the opposite of place making.
A second side effect is city-shaping. You can cut travel times to the city, but that encourages yet more sprawl and inefficient urban form. (Thanks, marchetti’s constant._
If you want a policy that is likely to be equitable, can potentially conserve scarce government funds for more valuable projects, and prevent the paving over of the inner city, then congestion charging is your horse.
–
To finish with here’s some data to make you go “huh!” – rain, apparently, has no effect on traffic:
Sometimes you can see a policy change coming a mile off. For about the last two decades, drug legalisation looked like such a case.
The positive results of decriminalisation in Portugal, and the examples of marijuana legalisation in Uruguay and various states of the US were becoming more widely known. The Penington report in 1996 argued for decriminalisation of marijuana and when Victorian Premier Jeff Kennett ignored its recommendations it was seen as a stance justified only by retail politics.
It seemed only a matter of time before expert recommendations on decriminalisation and legalisation were taken on board by Australia and nations across the world. An armistice was about to be announced in the increasingly stupid war on drugs. So it seemed.
Then the opiates crisis began. America is in the grip of a really shocking wave of premature mortality, caused by addiction to opiates. The scale of it is really awful – at 32,000 deaths a year, roughly equal to the numbers killed by firearms in that country.
(If you’d like your faith in journalism to be restored utterly while your heart is smashed into a million irrevocable pieces, I recommend this piece, Seven Days Of Heroin, from the Cincinatti Enquirer.)
The US opiates crisis has forced some hard thinking on the merits of legalisation (for drugs beyond talking about drugs beyond marijuana, mostly.)
Opiates are not only a gateway to heroin abuse but a problem in themselves. Legal opiates accounted for 20,101 overdose deaths in the USA in 2015 compared to 12,990 related to heroin. If a legal drug, tightly controlled by law and available only under prescription, can be abused in a way that spirals way out of control, what does that say about the prospects of ending prohibition of drugs?
TOO MUCH TRUST
With legalisation, nothing is going to end up as available as buying flour at the supermarket. There will always be controls – regulation, licensing, etc. Choosing them is critical. But there is one shortcut we tend to take.
We love to rely on doctors as one of those controls. “Only available via prescription” sounds like a big barrier to drug availability. We have a lot of trust in doctors at a personal care level and that transfers over to a policy level.
Meanwhile, even the best-intentioned doctors are at the mercy of a pharmaceutical system that is itself far from perfect.
(If you’d like your faith in journalism to further cement while your faith in capitalism is smashed into a million irrevocable pieces, I recommend this piece, ‘You want a description of hell?’ OxyContin’s 12-hour problem, from the LA Times. It describes how a big Pharmaceutical companies lies about its products, got loads of people hooked on opiates and evaded a whole lot of systems designed to stop exactly that from happening.)
To some extent this is like the story on Elon Musk yesterday. It bothers me when too much trust is vested in an entity, person or institution that doesn’t deserve it. And nobody deserves as much trust as we invest in doctors without an panopticon of ombudsmans, review committees and inspectors.
I think we can move towards legalisation of drugs. But what is crucial in regulating anything is the fine details of the way they are controlled.
Hoops to jump through
I wrote about this in The Right Amount of Smoking. Finding the exact sweet spot for control and legalisation is hard. You can fiddle with public and private ownership of suppliers, taxation, occupational licensing, sales licensing and controls on consumption.
At this stage, we probably don’t have enough controls for gambling, and we have too many of the wrong kinds for most drugs.
Finding the right kinds of control is hard and requires ongoing adjustment of the policy settings. Trying to outsource the difficulty we have in solving that to doctors is an attractive shortcut, but not the answer.
Elon Musk gets on my nerves. Whenever I see him in a headline my teeth start grinding.
But why? I agree with all his goals. I love the idea of clean energy. I want better batteries. I’m excited by colonising the universe and digging cheaper tunnels. So why does his every pronouncement get me upset?
I’ve been dwelling on this recently, and can only conclude it’s because of the lack of public skepticism he encounters.
Fly to most places on Earth in under 30 mins and anywhere in under 60. Cost per seat should be… https://t.co/dGYDdGttYd
Whenever I think about the future, I like to consider it in probabilistic terms. So when I hear Elon Musk talk about using rockets to travel from New York to Sydney in an hour, I naturally try to imagine what the likelihood of this happening is. I generally come up with numbers awfully close to zero.
Apparently other people’s thinking goes off in different directions, wondering about comfort during take off:
I don’t find myself thinking about g-forces. I’m too busy puzzling over why he should be able to make a roof including solar panels for less than the price of a roof. What does he think roof manufacturers have been doing for all this time?
Musk is not short of ambition or afraid to make his life more complex. For example, the original Telsa plan had nothing in it about automation or self-driving. He just bolted that onto the plan, presumably expecting it would be doable if the engineers just tried hard enough.
I remain skeptical.
When people think about the progress of science, they have an awful tendency to be swayed by survivor bias. They think especially about progress in personal electronics – because that’s where the progress is. They infer that technology can utterly transform itself with a decade or two.
But when you take a broader sample, you see something different. For every iPhone that did get invented, a flying car failed to be. While we beat back AIDS, cures for dementia and multiple sclerosis languished. And not for want of effort. You can’t tell in advance which fields will yield to effort.
I was a big fan of an old website called Paleo Future, which goes back and looks at old predictions of the future. They’re mostly silly.
In fifty years, most Musk plans will seem as silly. But they’re being repeated across all forms of media. That credulity, and the adulation that goes with it, really rustles my jimmies.
The other relevant cognitive bias is the base rate fallacy. People ignore the fact that in a given domain (colonising space, say) background probabilities of success are very low. They prefer instead to focus on some other seemingly salient factor, like whether the person making the plan to do so is a genius. (And I’m perfectly willing to admit Musk is.)
Now, the charm of having so many cognitive biases running in your favour, is you can attract a lot of capital and hire a lot of good employees. You get to make a lot of bets at once. Take one 5 per cent chance, you’re set to fail. But take ten and you have a 40% chance of one of them coming good.
So I’d be surprised if everything Musk tries from hereon turns to poop. He can probably go down in history as a genius inventor. But at the moment he’s getting way ahead of himself.
STRATEGIC THINKER?
Musk’s strategic thinking has worked well so far for Tesla, but past performance is no guarantee of future performance. You only need to look back on his Tesla “Master plan Part Deux” from 2016 to get a sense of how iffy it can be. It contained a very peculiar section on taking the aisles out of buses to make room for more seats. Ignore for a moment that aisles are important to buses – the point is that that kind of fine detail has no place in a strategic plan. Shortly afterward, he walked back the whole section on buses anyway. The whole thing made me wonder if his success came because of or despite his strategic vision.
It is possible that long before he has a chance to be proven wrong on intercontinental travel, Mr Musk will have a reversal of fortune.
MODEL 3
Tesla’s plan to ramp up production of Model 3s in a new facility looked risky to me from the start. Manufacturing is hard and Tesla is new to doing it at scale. Today we learned initial production of the smaller more affordable car has fallen short.
Expecting to move smoothly to mass production was pure hubris – big new projects regularly have huge cost overruns and delays. (I used to work on Defence projects so I’ve seen cost overruns and delays.) Furthermore, in making its new factory, Tesla skipped a step most manufacturers use, getting all its tooling made before they’d had a dry run. That will save time and cost only if all the systems fit together neatly and as expected.
I understand why they’re rushing. There’s two reasons Tesla must sprint to survive.
First, so much debt has been brought on that they need a lot of sales to pay the interest (with the share price so high I don’t understand why they wouldn’t just issue shares, which don’t need to be periodically refinanced).
So far Tesla burns cash just to stay running. Having big debt and negative cashflow is not sustainable. There’s not many times corporate finance is heart-in-your-mouth terrifying, but Tesla is making it like watching one of those guys in a wingsuit.
Second, the longer they delay the more competitors with proven manufacturing ability can catch up and steal the market. The Chevy Bolt is a proven success and we heard last week that even vacuum manufacturer Dyson is entering the electric car market. A Bloomberg article published today had a huge list of Tesla competitors. Fifty new electric vehicles are going to hit the market in the next five years, from companies with a strong history of making quality products.
I think the Tesla corporate structure needs careful steering to not end up on the rocks. The technology and brand could well be for sale within five years, and gleefully bought up by someone like General Motors or Google. That’d be awful for Tesla investors and employees but mostly fine for society, as the losses incurred in creating all this technological progress would be internalised by all the investors who’ve done their dough.
SHOULD I BE SO MAD?
So am I justified in being so cross at Elon Musk and all the people who believe in him?
One argument is I am not. To the extent that he is making great progress, I should shut up, and to the extent he is selling risky bets, his main victims are private investors who are welcome to include in their portfolios a few risky bets.
While money will be wasted, technology will also be created. If it has value, that technology will presumably be up for grabs if Tesla (or SpaceX, or Hyperloop) ever needs to make their creditors whole, and society will still be able to benefit from them. From this perspective, the cult of Elon Musk is just a big scheme to get private investors to take the risks of moving science forward. And it’d be awfully pig-headed to be mad at that.
From another perspective, investor money is finite, and we should be careful to steer it toward those schemes with the highest chance of success.
So tell me, dear reader. Am I being too much of a grouch toward Mr Musk?
I love government. But it is not a blind love. Government is not done as well as it could be. I’m very much for the idea of achieving collective goals to improve society, and very much open to reforming how we do that.
Here in Victoria we have two of the most potent and innovative public agencies: TAC and WorkSafe. The Transport Accident Commission works to reduce injuries and deaths from transport accidents. Worksafe works to reduce injuries and death at work.
They do a lot of good preventative work. Both have been very effective.
Victoria has the second lowest workplace accident rate in Australia, measured by fatalities (after the ACT) and by workcover claims (after the NT), as these next two figures show.
1.
2.
Victoria also has some of the best road safety performance in Australia. It has improved substantially over the last decade, as the next chart shows. (Of course, improvements in safety also come from improvements in cars themselves, but via programs like How Safe Is Your Car? the TAC encourages Victorians to buy safer cars, accelerating those positive changes.)
What makes Worksafe and the TAC effective?
Reason one is their independence. (The 2014-15 blip in Worksafe performance that you see in the above graph may be related to the state government meddling with its operations at that time. Independence matters.) They are statutory agencies free from direct ministerial control, giving them the ability to take extra risks.
The innovations that are most visible to the public are their communications – Worksafe sponsors a football team while the TAC pioneered using TV advertisements to change culture and reduce the road toll. This kind of communications strategy is rarer in government departments, where ministers face tough questions over spending.
The second, related reason is that these are not just policy agencies but insurers. Worksafe takes premiums from employers, and TAC from car registrations. They pay out when a worker is injured, or when a person is injured in a vehicle accident. This gives them not only a funding source independent of annual budget rounds, but also a clear financial incentive. (nb. Worksafe is also the workplace safety regulator and inspector, giving it further powers. TAC is not.)
These are both world-leading organisations which have had powerful positive effects on society, and organisations whose successes I have admired. So I was excited to see, earlier this year, the head of the Productivity Commission throw up a powerpoint slide with a dot point that argued we should “Address disease prevention as directly as we address workplace accidents.”
I was immediately captivated by the idea of using the Worksafe model to try to fight disease. The upside looks to be huge. Australia’s preventative spending on health is fairly terrible, as is well-documented, and was again confirmed in a paper by two Public Health academics in July this year.
“Treating chronic disease costs the Australian community an estimated $27 billion annually, accounting for more than a third of our national health budget.
“Yet Australia currently spends just over $2 billion on preventive health each year, or around $89 per person. At just 1.34 per cent of Australian healthcare expenditure, the amount is considerably less than OECD countries Canada, New Zealand and the United Kingdom, with Australia ranked 16th out of 31 OECD countries by per capita expenditure.”
Could we fix the preventative health spending deficit by setting up organisations akin to Worksafe and the TAC? Might it have the exact combination of novelty, innovation and actual prospects for success that could make politicians and public servants agree on it?
At a high level the opportunity seems to be there (for preventable diseases but not, so far as I can see, for non-preventable ones like MS). Does it persist when you dive down into the details?
THE DEVIL INSIDE
Imagine you were setting up a statutory insurer to fight against adult-onset diabetes. The insurer would collect premiums and pay for treatment after a person was diagnosed.
The big question is where the premiums would be levied. There is a key difference between this scenario and the workplace safety situation. Employers opt in to insuring employees when they hire them. Likewise, road users opt in to the TAC scheme when they register a vehicle.
The sort of population-wide coverage required by a diabetes insurance scheme means the beneficiaries could not be expected to cover their own premiums. (i.e. not without undermining the public nature of the health system! This fact may motivate some skepticism towards this idea from people who fear it is a Trojan horse for dismantling public healthcare. I don’t think it is.)
IT ALL COMES BACK TO CANBERRA
The only plausible premium-payer would be the federal government. That raises the question: How different, ultimately, would this be from Medicare? If the government is paying premiums into a public insurer and taking out the payouts to cover treatment costs, isn’t this just replicating an existing system?
The answer is it might be a replication. But if there is something in the culture, funding or control systems of the Health Department that makes it less than optimally effective, then there is a chance of improving outcomes by making a new organisational structure.
Following the Worksafe and TAC models, a good insurer would be focused on a single disease or group of diseases that we have at least some ideas how to prevent (lung cancers, diabetes, heart diseases). It would work on culture change and system changes to try to find the most cost-effective ways of reducing the incidence of that disease. (As an example of systemic changes, the organisation paying the premium might be more inclined to levy a sugar tax if it knew that would reduce its premiums for diabetes insurance). If it had success, the premium it would have to charge to the government for coverage would fall.
NDIS-ESQUE?
Any such systems would be different from the NDIS. The NDIS is far more about organising and coordinating the care for people who have disabilities. It is premised on helping people after the fact and is a vital service.
It does not seem, so far as my reading has shown, to have a focus on identifying avoidable disabilities and investing to avoid them. ( I am sure there are exceptions down in the details but at a high level NDIS is more about service-delivery than prevention.)
Sometimes puppers are not relevant to the post so much as reward for reading so far.
ISSUES
I’ve painted a picture above that I find promising, but I’m not going to expire in the proverbial drainage channel for this idea. I can see its weaknesses.
For starters, this looks like a case of bower-bird problem solving. You spot a shiny thing (TAC, Worksafe) and take it back to your nest. Then you’re seeking out a good way to use it. That is different to taking a first-principles approach to figuring out how best to optimise preventative spending. There may be better ways. And it could be that the azure sheen of TAC and Worksafe blinds one to the inherent unsuitability of the model in other environments.
Secondly, I may have mis-identified the effectiveness of those two organisations. I know their existence correlates with big improvements in the outcomes they’re targeting, and I know they are well-regarded but I can’t show causation. Other jurisdictions have similar organisations that are not as effective.
Third, the advantages might all fly out the window when you have the federal government paying the premiums rather than other customers. That’s a powerful customer and it might be hard to fight for justified premium hikes in tough fiscal situations, in which case the independence of the agencies becomes blurry.
Having said all that, I’d love to see a superlong .pdf getting into the guts of this idea and figuring out whether there is promise in it. If you work for a thinktank or a department and you’ve already written such a thing, please let me know!
Anyone else, please leave a comment sharing any insights or aspects you think are relevant!
Remember that despite the importance of CBDs, most jobs are still in the suburbs. If we know one thing about CBD jobs – especially nine to five CBD jobs – it is that they tend to be the good kind.
2.
Thomas the Think Engine 