The high cost of free parking: urban intensification edition

Problem definition matters in policy.

I worry some people in Auckland are experiencing poor parking management by Auckland Council as being a problem of the removal of mandatory parking minimums. 

The problem isn't removing mandatory parking minimums. 

The problem is the failure to price or manage on-street parking in places subject to intensification. Expectations of being able to find free on-street parking reduces demand for on-site parking where on-site parking has high opportunity cost. 

So solve the underlying problem. It isn't that hard. Set resident parking passes for the street, keep some spaces back as paid parking (easily managed through apps), and let folks decide whether they want to pay for an on-street spot or forgo some living space on-site or not have a car. 

My column in yesterday's Post.

The case for considering fuel surcharges

Last week over in Newsroom, I laid out the case against subsidising fuel use during a supply shortage. 

That case is sufficiently obvious that you wouldn't think it would need to be made. But the New Zealand Taxpayers Union really disgraced itself by urging those kinds of subsidies via reductions in petrol excise. Petrol excise already is inadequate to cover the cost of the roading network. Taxpayers already subsidise road use. The NZTU wanted even larger subsidies from taxpayers to road users when it is manifestly unclear that we will be getting more fuel shipments. 

The piece also made the case for just letting fuel prices rise rather than contemplating various kinds of rationing. Letting prices rise automatically culls the lowest-valued uses of fuel. Governments concerned about equity implications can transfer money to poorer people. This point ought to be obvious to any economist who's done intermediate micro and the welfare theorems. Always surprised to note which bank economists might have missed the lecture on the welfare theorems. 

And it then made a point that I don't think I'd seen others make. I think it's also obvious. Real option value matters. 

But I also worry about the Commerce Commission’s insistence that prices should only reflect current supply costs.

Consider an analogous problem faced by hydroelectric generators.

The ‘cost’ of running water through a hydroelectric dam in February is that the generator cannot run that bit of water through the dam again in July. Rain is always uncertain. Holding storage back, in case inflows are lower than expected, has value.

If fuel storage tanks might not be replenished because import routes are disrupted or exports are restricted, the ‘cost’ of fuel is not whatever notional price fuel companies contract with overseas suppliers. A litre used now might mean a litre that cannot be used in June. It is possible that everything resolves itself before then. But it is also possible that things turn considerably worse. The Strait of Hormuz is less predictable than hydro lake inflows.

The commission’s messaging risks confusing today’s sourcing cost with the value of fuel in storage. If replenishment is uncertain, the relevant cost of selling a litre today is not just what it cost to source, but the value of keeping it in reserve. That is like telling hydro generators to run the sluices in February despite the risk of a dry winter.

And that gets us to the case for fuel surcharges. 

I do not think it's any kind of knock-down case. I certainly don't have the background figures to be able to tell. And if the government is confident that its agreement with Singapore and alternative sources of supply are secure enough, then none of it's needed. But it can be worth thinking through in advance, just in case.

Suppose the storage tanks are at their normal levels. Prices have increased, mainly reflecting increased price of fuel being loaded onto transport ships. But each ship now carries some risk of being diverted before it reaches New Zealand. That risk means each litre in storage carries some real option value: there is some chance that using a litre from storage today means that there will not be a litre to use in May or June. 

All going well, fuel companies price this in. If the transports stop arriving, fuel in storage is far more valuable than the current price. Every fuel company ought to be weighing the price it can get for a litre of fuel today against the potential price that litre could get in future if kept in storage, weighted by the likelihood that future prices are higher. 

It'll be complicated, and there aren't great numbers to throw in as probabilities. But it's in-principle doable. Fuel companies should have strong incentive to get that right. If a litre of diesel is worth $20+ in June if transports stop arriving, then selling today for $3 would be a mistake if the risk of the bad scenario is high enough.

And in the first best, government is in constant communication with the fuel companies about the negotiations it's having with foreign suppliers, and fuel companies are always weighing that up alongside their own market intelligence (and odds at Polymarket etc). 

There is no case for surcharges in that first-best world. Everyone's already weighing up real option values and they're priced in.

Here are a few ways that can go wrong. 

1. Suppose companies think that they will be punished by the government for increasing prices today to reflect real option value, ahead of the risk materialising. They are not crazy to think that, given the kind of messaging that our populist Commerce Commission provided. In that case, they will choose some price lower than would otherwise be optimal. Their optimisation would be 'what is the highest price we can charge, below the actual optimal price, without drawing ire from the government?' 
2. Suppose companies think that they will be punished by the government for increasing prices to very high levels in future if transport stops. In that case, a company holding fuel back against that future risk (and potential future returns) isn't able to realise those future returns. The social value of holding fuel back for future use would be high, but the private return would be expropriated through government seizure of stored supplies or other forms of de facto expropriation. 
3. Even if the government were sensible, consumers may not be. Suppose one fuel company runs real-option-value pricing and another does not. The one charging higher prices today will be damned today for charging higher prices. The one that does not, will not. The one that charges higher prices today will be further damned if the bad state of the world obtains and they charge very high prices in June. The one that runs out will not be blamed for running out. How could they be blamed for running out? The Straits are closed. It isn't their fault. The one that really deserves blame is the one that 'profiteers' in June. Running real option pricing when others aren't means taking a hit to revenues today in expectation of higher returns in future if the bad outcome obtains. But the brand damage could be far higher than can be recouped, because approximately nobody is willing to think in real option value terms. This underlying mechanism is what drives populist government responses too.

I do not know what fuel prices reflecting real option value should be. But $3/litre diesel feels low when thinking about those risks. 

And that gets to the case for surcharges. If companies fear backlash for using real option pricing, government could set a surcharge reflecting the difference between actual prices and prices that incorporate that real option value. 

I'm not convinced by that case. And I have little confidence that government could get to the right number on it, even with smart people making their best efforts. But it seems worth thinking about. 

At a minimum, the government really shouldn't be saying anything that could be read as discouraging fuel companies from increasing prices. 

And the Commerce Commission's public messaging ought to be more around real option pricing. Saying the words 'real option pricing' would be a mistake. But I think enough people understand the hydro lakes to get the point? The generators hold water back when the lakes are high, resulting in higher electricity prices at that time of year, if they worry that the lakes will otherwise be too low mid-winter. If they didn't do that, we'd risk winter blackouts. And that's worse than higher prices when the lakes are high. 

Addendum/Postscript 31 March: Storage is a constraint. Real option value of fuel held back now is consequently lower, unless it were cost-effective to pay an arriving tanker to just stay here. And I doubt that currently stacks up.

I am very glad that the government here has not discounted petrol excise or jumped to other populist stupidities that tempt the Australians. 

Recursive technology, non-recursive adoption, and Schumpeterian displacement

Tyler pointed to this piece, from Citadel Securities, making the point Tyler's made before: AI is growing more powerful by leaps and bounds, but people and institutions are slow to adopt new things. 

Diffusion and bottlenecks within institutions will prevent an AI-powered growth explosion and rapid displacement. 

Recursive Technology ≠ Recursive Adoption

The current debate around artificial intelligence conflates the recursive potential of the technology with expectations of recursive economic deployment. In other words, because AI systems can improve themselves or accelerate their own capabilities, commentators are extrapolating a future in which automation and productivity compound indefinitely at exponential rates. Technological diffusion has historically followed an S-curve. Early adoption is slow and expensive. Growth accelerates as costs fall, and complementary infrastructure develops. Eventually, saturation sets in, and the marginal adopter is less productive or less profitable which causes growth to decelerate.

Despite this – markets often extrapolate the acceleration phase linearly but history implies pace of adoption plateaus as organizational integration is costly, regulation emerges and diminishing marginal returns exist in economic deployment. The risk of displacement declines with a slower pace of adoption.

I'm happy to agree with all that. 

And I think it was Josh Gans who, a few months ago, pointed to very wide within-industry dispersion of firm-level productivity as suggestive. Highly productive and far-less-productive firms coexist in single industries.

But if the productivity gap is large enough between firms that really get AI and those that don't, wouldn't we expect the laggards to be winnowed out? If old institutions are slow to move, but new upstarts can leapfrog them, you could have stats showing relatively small proportions of workers relying heavily on AI, as Citadel currently shows. But weighting that by market share could be different. 

I dunno. When Aella can quickly roll-out dataviz with Claude that beats anything Stats NZ has ever put up, it feels like there's potential for tiny firms to quickly eat large lunches. 

I find it difficult to ignore the role of stupidity in human affairs

Adam Thierer points to his favourite Ronald Coase quote:

my favorite Ronald Coase quote: "I find it difficult to ignore the role of stupidity in human affairs."https://t.co/VkrTfmmppu pic.twitter.com/UjHgK2Eu75

— Adam Thierer (@AdamThierer) February 18, 2026

With that in mind, a couple of things have been bugging me. 

Item the first. 

The government of New Zealand is exercised about supermarkets and supermarket competition. It simultaneously wants supermarkets to pay higher prices to their suppliers while reducing costs to consumers. It has established an entire regulatory agency whose job is to force supermarkets to pay more for inputs while reducing prices on outputs. At the same time another part of the Commerce Commission is trying to prevent Foodstuffs North Island from merging with Foodstuffs South Island on a very dubious theory of harm: the back-end efficiency benefits of merging two supermarket cooperatives that operate on different islands would be less than the harm to competition in the supplier market, which ComCom claims would wind up having longer term harm to consumers. 

It is not coherent.

But it gets worse.

One coherent part of the government's approach is a desire to make it easier to import food products. It would help reduce costs to consumers directly, and indirectly by easing a barrier to entry. I support that initiative. 

New Zealand has signed onto FSANZ, a joint approach to food product regulation with Australia.

FSANZ has proposed making bespoke Australia-NZ 'food-star' healthy food ratings compulsory. They are currently voluntary. 

Let's think this through. 

A store wanting to sell grocery products from outside of Australia-NZ would have two options for each product.

It could put the product up for FSANZ assessment and put a sticker on each and every packet before putting it on the shelf.

Or it could ask suppliers to provide production runs specific to the requirements of the New Zealand market. 

The former would have fairly high per-unit cost: a sticker has to get added to every product. 

The latter would also be expensive. It would mean limited production runs at higher cost that also limit flexibility to shift product across different markets in response to changing circumstances. 

It would disproportionately burden stores that stock foreign-sourced products - like the Asian grocers that provide fringe competition to the main supermarket chains. 

And it would increase the cost of entry for any overseas supermarket that wanted to bring its supply chain with it. 

One part of government wants to improve competition in supermarkets and bring costs down; another part of government wants to push up the cost of groceries while impeding competition.

Item the Second

Earlier this week, I received an email. 

I'd put $200 in there mainly to test their trading interface. I love prediction markets. I missed them. New Zealand had been truly world-leading in this area, then FMA/DIA/Justice/Simon Bridges killed the sector here. 

My submission to the CFTC on Kalshi from 2022 is here. I'd there concluded:

What experience we had with iPredict suggests CFTC really doesn't have anything substantial to worry about in allowing contracts on political events. If anything, they heightened voter engagement. The CE of iPredict even featured on the nightly news during the election, giving the latest on election market prices. And for that brief period, whenever blowhard partisans insisted that some outcome was going to happen, people could just point to the iPredict price on the event and ask them why they thought that price was wrong, and whether they'd actually put their money where their mouth was. It was a remarkable era. iPredict inflation forecasts (they also had markets on inflation going out several years - it was so very good) wound up being noted in our Reserve Bank's Monetary Policy Statements. I desperately miss it. I envy the opportunities Americans could have if CFTC takes a sensible approach to regulation.

And now they've taken Kalshi from me too. 

Newsroom has the detail on what happened, along with some pretty misleading nonsense from the Department of Internal Affairs.

Before we hit the Newsroom / DIA stuff, here's the defensible case. Or, well, defensible given the constraints.

The government auctioned off a monopoly to run sports betting in New Zealand. It is consequently obliged to protect that monopoly on behalf of the licensee. Remember that cartels and monopolies are only bad if they aren't government-authorised. In that case they are wonderful things. 

Kalshi is a CFTC-regulated exchange. It trades event contracts in the same way that iPredict did. It runs AML/KYC; I had to go through it to trade there. It's safe. It's fine. But some of its contracts are on sporting events, which in the US are treated as futures derivatives rather than sports betting. That provides Kalshi with a few advantages that Matt Levine has had an awful lot of fun discussing. 

And, at the same time, New Zealand Members of Parliament have become very exercised about online gambling full-stop. Their proposed solution is not very good. They want to license online gambling operators, which can be okay, but they also want to set tight limits on the number of operators. That's neither here nor there in this context, but it's part of the moral panic in the general area. 

Now. If you were the regulator charged with defending the monopoly provided to the sports betting licensee and if you viewed Kalshi's sports contracts as sports betting, the obvious thing to do would be to ask them to segment their markets such that NZ-based traders could not trade on the sports contracts but could trade on other contracts. 

It is entirely possible that Kalshi would prefer to close off access than so-segment its markets. But that would be a defensible position for DIA given its role as monopoly-enforcer for the company to which the New Zealand government auctioned the right to run sports betting. 

Instead we get this:

Polymarket, Kalshi and similar prediction markets are illegal under New Zealand’s gambling laws, the nation’s gambling regulator has decided.

Polymarket and Kalshi are online markets where users can place bets on future outcomes, ranging from New Zealand provincial cricket results to what phrases Donald Trump will use next month.

Whether the Reserve Bank of New Zealand will increase, decrease or hold the official cash rate when it meets on Wednesday has had over $127,000 placed on it.

...

Department of Internal Affairs gambling director Vicki Scott tells Newsroom it believes they are illegal gambling platforms.

“We consider platforms such as Kalshi and Polymarket to be gambling under New Zealand law,” Scott says.

“Since they aren’t authorised operators, they are prohibited from offering their gambling products to people in this country.”

“To the extent these platforms are taking bets from New Zealand customers, they are breaching the law here and can expect to hear from us.”

The event contracts on things like the OCR are very obviously financial derivatives. iPredict offered them as an exempt futures exchange in NZ a decade ago. DIA always hated iPredict and viewed it as treading on their turf - maybe they want the whole stock exchange to come under their authority as a gambling operation. 

This is not the 'online casino gambling' that is subject to the NZ gambling licensing regime. It's a CFTC-regulated futures exchange. 

And this part seems particularly off:

While these websites are still technically accessible, Scott says using unlicensed sites is risky because there are no guarantees to ensure they pay out as promised or take action to minimise gambling harm.

Does DIA have any credible reason to believe that Kalshi and Polymarket welch on bets? 

In a better world, DIA would have asked Kalshi to exclude Kiwis from its sports contracts (which I think is stupid, but that stupidity is imposed by Parliament) and if Kalshi either could not or would not, they would have put out a statement blocking Kalshi on that basis without insinuating that they're dodgy online casinos. 

Item the Third

Sam Bowman writes in the Washington Post:

By some measures, teenagers’ mental health does seem to have gotten worse over the past 10 years, and this does coincide with widespread adoption of smartphones. But that is where any clear correlation between the two ends. Multiple studies have either shown that smartphone and social media use among teens has minimal effects on their mental health or none at all. As a 2024 review published by an American Psychological Association journal put it: “There is no evidence that time spent on social media is correlated with adolescent mental health problems.”

You can also catch The Studies Show on the topic. If there are effects, they are very small.

Overseas jurisdictions implementing social media age limits are, predictably, now looking to regulate access to VPNs as complement. 

And so the government of New Zealand is rushing to have legislation on it through before the election. As it is going through as a member's bill rather than a government bill, there won't be much accompanying assessment. The Prime Minister is 'deeply supportive'. 

Sigh.

There was another good line in that old Coase piece, quoting Axel Oxentierna, a 17th Century Swedish Chancellor. "You do not know, my son, with how little wisdom the world is governed."

Surely a case for outcome-based contracting

Outcome-based contracting would have solved this:

A stop smoking provider says the decision to cut its contract on the basis it has refused to hand out vaping kits as part of its cessation programme was not a part of the contract when it was signed.

Takiri Mai te Ata Trust regional manager Catherine Manning said the Wellington-based trust received an email this week stating it had to either start supplying the vape kits or have its contract terminated, despite signing a new contract earlier that year.

Vape kits are effective in stopping smoking - and more effective than patches.  

If a stop-smoking provider didn't want to distribute vape kits, flipping that provider to outcome-based payments could make sense. If whatever they've come up with is as effective as vaping, then they'd be paid for that outcome. And if they were just handing out ineffective patches to no reduction in smoking rates, they'd wind up having to close down the service. 

Outcome monitoring's a good idea regardless. But outcome-based payments can help the state be agnostic across methods. 

The state of the books

StatsNZ has put up its year-end accounts for the government, split out across functional areas. 

Their data goes back to 2009 in the main table; I'm sure earlier data's available somewhere in Infoshare.

But sticking with the Excel sheet they've provided, we can lob in June-year population statistics and June quarter consumer price index values to get per capita real measures on operating expenditures net of finance cost and on spend in different functional areas. 

Here's per capita real central government operating expenditure net of interest expenses. Netting out interest expenditures is helpful if you want a handle on core government operating expenditure. Government can decide to take on less debt; it cannot decide to stop making its interest payments. 

You may have heard a lot of stories about austerity. Consider that both the government and the opposition may want to convey the impression that it has happened, despite it very much not having happened. 

Throughout the 2010s (barring #eqnz), per capita real operating expenditure net of interest expenses ranged from $17,143 to $18,653 - with 2019's jump to $18,653 being well out of line with the prior track. Labour substantially increased spending under its wellbeing focus - as was its prerogative. 

Per capita real operating expenditure net of finance cost has been above $21,000 since then; the provisional figure for 2025 is $21,648.

Here's the breakdown by functional categories. Let's start with the bigger-ticket items. 

The largest-spend category here by far is social protection: benefits and superannuation. 

Spending on those has risen over the period and is now close to the peak that they hit during the GFC and Christchurch Earthquake. There was always going to be a slow rise in these with an aging population in the absence of reform to NZ Superannuation. 

Health is next. It peaked during Covid. Vaccines cost money. Again, an aging population will get you an increasing track on this one - but we were averaging about $3700 per capita before Covid and we're now around $4600. 

Education had a large Covid spike, then retrenchment to a new level a bit above the old per capita level. 

But remember that this is per capita, not per capita under the age of 25. And the proportion of under-25s has been dropping at the same time as the proportion of over-65s has been increasing. 

Any giant shedding of government staff will show up in General Public Services.

The austerity really stands out in this picture. Can't you see it too? 

'Economic affairs' spiked with the wage subsidy scheme but has maintained a substantial ratchet. 

The category's spend, pre-Covid, peaked in 2019 at just over $1,200 per capita. Now it's just over $1,700 per capita. 

Public order and safety has plateaued just over $1,200 per capita. The increase began under the Ardern government. National has maintained their increase in resourcing to police. 

Housing and Community Amenities jumped from close to nil to just over $400 per capita under the prior Labour government, and has stayed there under National.

Defence has been more frugal - but remember that this is just operating expenditures, not capital expenditures. That's around $400 per capita throughout. Minor decline in real per capita operational expenditures since 2009. 

Recreation, culture, and religion rose and has retrenched somewhat. I wonder what the heck is in the 'religion' category. But the apparently arts and culture hating National-led coalition is spending more in this category, in per capita real dollars, than Ardern ever did pre-Covid. 

Environmental protection has been maintained at a slightly higher level than pre-Covid. Under the Key government environmental protection was around $160 per capita; pre-Covid Ardern had it around $175. Just look at that slash and burn from the environment-haters. It's right there. Can't you see it? That little downward tick in 2025?

We continue to be in substantial structural deficit. 

Increasing revenue to match spending is always an option, but one that would entrench the increased spending in just about every category since Covid - and still leave a demographic problem to be dealt with as superannuation entitlements keep rising along with health spending. 

And I've not included the costs of financing an increasing government debt. 

Alas, it was not to be

It would have been just one bad part of an overarching very silly policy. Exempting it from the policy regime I suppose makes the policy a bit more tractable. But it also makes it a lot less potentially funny.

New Zealand's government supports the creation of cultural content by paying for it through various grants. TV stations and streaming services can then run it, or not, as they want. 

Canada does things the dumber way. I'm sure they also have direct subsidies. But they also have Canadian Content regulations that prescribe the proportion of each day's broadcasting that must be Canadian content.

It was bad enough in the linear TV era. The ridiculousness of it all had the excellent SCTV pad out the extra couple of minutes of the Canadian version of the show (fewer ads than on the US side) with a very explicitly Canadian segment: the most over-the-top CanCon possible. Bob and Dough MacKenzie - the hosers.

The first segment including them had a lengthy scroll after the segment explaining how the segment meets official Canadian guidelines for what counts as Canadian content and was almost as funny as the MacKenzie brothers.

That was fifty years ago now - or thereabouts.

Times change. 

A decade ago, Canada decided that its regulatory reach extended to the entire internet if the internet could be viewed from Canada. If you wanted to stream to Canada, you'd have to meet CanCon rules. Quite how to make that work when people choose what they want to watch and plenty of potential platforms might not really care what Canada things about anything - well, they've been taking a while figuring out how to apply the principles. 

And they've finally decided that, despite or perhaps because of the uniquely Canadian content that might be created, to great hilarity, to meet the rules, the CanCon rules will not apply to pornography streamed in Canada. 

This has long been one of the more onerous demands of the CRTC, given the relative dearth of erotic media that would meet their terms as “Canadian content.”

Article content

Under the CRTC’s definition of the term, it’s not enough to have a Canadian performer or a Canadian setting.

Article content

Rather, it’s determined via an elaborate “points” system that, among other things, requires the producer and at least one of the lead performers to be able to prove Canadian citizenship.

Article content

At least three quarters of the financing must also come from “Canadians or Canadian companies.”

Article content

In extreme cases, this means that a video of a Canadian couple having sex in Canada and directed by another Canadian would not qualify as Canadian content if only 74 per cent of the financing was provably Canadian.

Whenever one despairs about policy in New Zealand, Canada and the UK provide superb reminders that the rest of the world generally remains even worse. 

Meanwhile, Australia's looking to impose Australian-content mandates on streaming services. 

The federal government has put laws requiring streaming services to produce Australian content back on the table after postponing them due to concerns about how they would interact with Australia's trade agreement with the United States.

The government has confirmed it will introduce legislation this week to mandate that any streaming services with more than 1 million Australian subscribers must produce Australian drama, children's, documentary, arts or educational programs.

I wonder whether there are enough subscribers to any single platform for Australia to run into Canada's difficulties here. It would be very funny if there were.