Wednesday 30 June 2021

Better testing

Last week, Pattrick Smellie tried to figure out what evidence backs up the Covid saliva test that the government has hired in for testing a small number of border workers. 

There doesn't seem to be much, or at least it isn't being disclosed. 

The firm’s saliva test, which the Ministry of Health had wanted in place by mid-April when it first sought bids for a contract it has indicated could be worth $50 million a year, is expected to become available in Christchurch in coming days with a national rollout occurring on a staged basis through July. 

APHG says it has been conducting its saliva test on a small number of border and managed isolation workers since the start of the year, but has never published any information publicly on the test. 

On Monday last week, APHG’s relationships manager Trevor English told BusinessDesk “our test is pretty much the same test as you would run for the nasopharyngeal test” – the nasal swab test currently treated by the ministry as the ‘gold standard’ and only officially accepted test for the presence of the virus. 

English said APHG had compared the saliva test it was using with two others that are in use internationally – developed by Yale and Illinois universities in the US – “and we know our test performs better in our hands than those two”. 

However, APHG is unwilling to provide transparency on that and other questions relating to a saliva testing regime the health ministry resisted until being instructed in February to implement one.

Meanwhile, we keep getting more news on the validation of Rako's test. From a Vic Uni press release:

Scientists at Te Herenga Waka—Victoria University of Wellington have used real-world samples to confirm the 98.7 percent accuracy of a saliva test for COVID-19. This is the first, and currently the only, COVID-19 saliva test to be diagnostically validated in Aotearoa New Zealand.

Associate Professor Janet Pitman from the University’s Te Kura Mātauranga Koiora—School of Biological Sciences led the study. “Despite a view that saliva tests aren’t as accurate as the standard nasopharyngeal test, our research shows this one is,” she says. A paper describing the study has been prepared and submitted for publication in a medical journal.

Initial testing was carried out on artificially infected saliva, where heat-inactivated SARS-CoV-2 virus, which causes COVID-19, was added to saliva in the lab. The now-completed diagnostic validation involved testing samples from actual positive COVID-19 patients.

Associate Professor Pitman’s team tested paired samples, sent from the United States, from 152 people. The paired samples were a sample of saliva and a sample from a nasal swab taken at the same time. Thirty-four of these people were positive for COVID-19.

In all but two instances, the results matched—in one instance the saliva sample tested positive when the nasopharyngeal sample didn’t, and in the other the reverse occurred. The discordance between the two samples is likely not due to a failing of either test. It is more likely due to differences in the timing of the virus’s presence during the early stages, and disappearance during the late stages, of the disease at the two biological sites (salivary glands and the nasopharyngeal region). Salivary glands contain large numbers of cells that replicate the virus and are then shed into the saliva, making saliva a great indicator of live virus production.

The researchers’ results showed the test is highly sensitive. “An infectious person has about 1,000 viral copies of SARS-CoV-2 per microlitre of saliva, equivalent to one thousandth of a millilitre,” says Associate Professor Pitman. “Our test can measure one copy in this volume. It is as sensitive as the standard nasopharyngeal test. We found this test is sensitive enough to measure asymptomatic people well before they become infectious.”

Associate Professor Pitman says there are many advantages to this method. “The great thing about saliva testing is not only can you use it for detection but also as an ongoing surveillance method to keep an eye on what’s in the community. You can test more regularly as it’s non-invasive.

“I’m not sure how long the public is going to accept the nasopharyngeal ‘brain scrape’ test. With our test, you just let the saliva pool in your mouth, then drool into a spoon, without creating bubbles, and tip it into a tube. However, people must not eat or drink for one hour before providing the saliva sample to ensure a good quality sample.”

There are also other advantages to saliva tests, she says—for example, they don’t require health care professionals to take the sample.

Associate Professor Pitman stresses that COVID tests are going to be part of our lives for years to come.

“This virus isn’t going away, and eventually we’re going to open our borders. What we need is a suite of highly accurate, diagnostically validated saliva tests for SARS-CoV-2. And for that to happen, the information surrounding saliva tests should be transparent and publicly available. We’re going to need these tests in our workplaces to give people peace of mind at work, at our borders to enable people to travel, and within our communities to gauge any outbreaks. The more tools we have in our toolbox to fight this and keep New Zealand safe, the better.”

The saliva test was developed by the University of Illinois in the US and is authorised for emergency use by the US Food and Drug Administration. The test is provided in New Zealand by Rako Science, which asked Associate Professor Pitman’s team to validate the test in New Zealand, for rollout here through the molecular diagnostic lab iGENZ. It is currently being used at Auckland airport, Air New Zealand and other large workplaces and will also service the New Zealand Olympic team.

Wouldn't it have been helpful if rapid saliva testing could have been deployed as part of the recent Wellington case? It can run at scale, quickly. It doesn't need nurses collecting samples, and shortages of nurses were cited as a reason that the government couldn't shift vaccination efforts from Auckland to Wellington. 

You'll have to go read the rest of the BusinessDesk piece on what may be going on here. 

Tuesday 29 June 2021

Agriculture and the ETS

Interesting idea for getting some agricultural emissions into the covered sector: bring fertiliser into the ETS. 

No Right Turn points to the proposal:

Nitrous Oxide is a potent greenhouse gas, 265 times worse per ton than carbon dioxide. Its also long-lived, with a half-life of 121 years in the atmosphere. New Zealand emitted 8.36 million tons CO2-equivalent of it in 2019, almost 10% of our total emissions. And yet, we're doing nothing to limit that. Why not? Because its emitted by farmers, of course.

1.5 million tons CO2-equivalent of those emissions is caused by farmers' (over)use of fertiliser. Stuff's Olivia Wannan points out that we have an easy way to control this: bring it into the ETS:

In its new report, the Climate Change Commission noted fertilisers could be charged the same levy that is already put on every litre of petrol. By pricing fertiliser under the Emissions Trading Scheme (ETS), the country could start making in-roads on the greenhouse gas now.

However, the Government took this option off the table in its deal with the agricultural industry – a 2019 partnership called He Waka Eke Noa, set up to solve the thorny problem of how to measure and price greenhouse gases from sheep and cows.

That solution won’t take effect until 2025, which will give the country just 25 years to get nitrous oxide to net zero.

Which isn't good enough. This is the lowest-hanging of low-hanging fruit, and sticking a carbon price on it would incentivise farmers to use less. And it makes perfect sense to do it at the importer/producer level (as is done for oil, coal, and everything else), not least because its far easier to audit two fertiliser companies to ensure they're not cheating than thousands of farmers. As for how much it would cost, at the current carbon price of $43.35 a ton, we're talking an extra ~$31.60 per ton of fertiliser. According to this article, urea sells for $1290 a ton, so we're talking a roughly 2.5% increase - less than the extra they'll be paying on a new, dirty ute. And of course that cost can be completely avoided just by using less.

Our refusal to bring fertiliser into the ETS is effectively a $65 million a year subsidy to farmers, ot encourage them to pollute not just the global atmosphere, but also our rivers and streams (which is where the nitrogen ends up). Its also a subsidy for dead babiesbowel cancer, and poisoned drinking water. We wouldn't accept it if the government subsidised tobacco companies to cause cancer - but that's effectively what they're doing with farmers. Pretty obviously, I think that needs to stop. Farmers need to pay the full cost of their pollution, as well as facing environmental controls to prevent them from poisoning our waterways. And if the government refuses to do that, we need to get a government who will.

The last paragraph's rhetoric is over the top. But we do need better freshwater management, and to me that means cap-and-trade over multipollutant space in catchments big enough for it to make sense. 

I haven't checked the numbers here provided. But it's hard to see any obvious case against getting a CO2-equivalent price onto fertiliser, charged like petrol at point of manufacture or import, by bringing it into the ETS.

Subsidy incidence: EV edition

It's pretty standard drill in intermediate micro to work through some tax incidence questions. 

The general point is to remind students that it doesn't matter whether the government says some tax has to be paid by the business or by the customer, or whether it has to be paid by the employer or the employee. The burden of the tax depends on relative elasticities of supply and demand; the statutory incidence really doesn't much matter in determining the economic incidence.

The extension to subsidies is obvious - the same drill applies. 

And I guess that supply of used electric vehicles in Japan is relatively inelastic. NZ's announced EV subsidy scheme, yet to come into effect, is bidding up prices there. 

Turners' subscription fleet was made possible with a grant from the Government's green transport fund. But used EVs were becoming more difficult to source from Japan, Hunter said. "In terms of pricing, the vehicles in Japan have all gone up for the amount of the subsidy – $3450."

Robert Young, director of New Zealand's biggest used car importer Nichibo Japan, agreed prices had gone up, though not by quite as much. He said the market had been "perverted" by the subsidy, and New Zealanders would see that as the new EV and hybrid stock was shipped over from the Japan used car auction houses.

He estimated about half the $3450 subsidy would end up off-shore, benefiting the auction vendors in Japan and the UK as well as new car manufacturers. More would go to GST – meaning Kiwi EV buyers would pocket only about one-third of the subsidy.

"This policy is more about political run-scoring than achieving its objective," Young said. "It will drive up new and used car prices for New Zealanders and increase the age of the national fleet which reduces road safety. 

The funniest part is Minister Wood thinking he can do much of anything about a bidding up of auction prices of used cars in Japan.

But Transport Minister Michael Wood said the Government was keeping a close eye out for any attempts to take advantage of the subsidy.

“The new and imported used vehicle market is very competitive and I’m sure anyone attempting to distort market pricing will be called out," he said.

"I have asked my officials to keep a close eye on the market as the Clean Car Discount gains momentum. Over the lifetime of the policy, it will help make cleaner cars more accessible to Kiwis.”

Suppose that you were the official tasked with 'doing something' about higher used car auction prices in Japan. Where would you even start?  

Monday 28 June 2021

Reader mailbag: Bus contracting edition

A little while back, I'd suggested that the problem with Wellington's buses wasn't whether the operator was private or public, but rather the contracts under which they were operating. 

Shortly after, I received a helpful pointer in the reader mailbag, but have only now had a chance to hoist it up here.

Bus services in the main centres used to be run by municipalities – plus Railways Road Services in some areas. The industry was regulated: you couldn’t run in opposition to the licence holders even if you wanted to. When the government deregulated urban bus services, it also forced the municipalities to sell their fleets and to contract for the services they wanted, with the contracts awarded basically on price. 

The effect of the first contract round was to reduce costs by around 30%, which in most cases was translated into improved services. The contracting model adopted enabled operators to register services commercially, then the public transport authority would go out to tender for any additional services it wanted. Contracts were net, which meant that the operator provided the service and collected the revenue and was paid a fixed sum per month to cover the difference. If an operator missed a service or the service was unreliable, the operator suffered the loss in revenue. Since the operator kept the revenue, there was a strong incentive to try to attract passengers and indeed patronage grew significantly. Operators registered services commercially if they could as it gave them control of the service – contracted services might pay better, but when the contract ended you might lose the contract to someone else. 

But the bureaucrats didn’t like them at all. Particularly in Auckland where they were seen as inhibiting the introduction of connecting buses for the new rail services and making integrated ticketing difficult. So they persuaded the government to change the operating model to give the commercial services to the councils. The other thing the bureaucrats didn’t like was that under net contracts, the operators had profited from the increase in patronage that had accompanied privatisation. 

So the new public transport operating model (PTOM) changed to gross contracts, where all the revenue went to the transport authority and the operators were just paid a price per kilometre for running the buses. Under PTOM, the operator doesn’t have to worry about the passenger any more unless the contract contains penalties or incentives. Getting the penalties and incentives right is difficult. The buses that connect with the trains at Paraparaumu didn’t wait for trains that were running late as the bus operator would cop a late running penalty. And as you say, the penalty for failing to operate is insufficient to warrant the operator fixing the problem. 

To my mind the best solution is to revert to the original net contract model.

Afternoon roundup

A very long overdue closing of the browser tabs brings these worthies:

Supply and demand denialism on the right

Normally, it's the left that makes a hash of basic supply-and-demand thinking - at least in New Zealand. 

That wasn't the case in last week's Wellington Council debates on the proposed spatial plan. 

Councilors broadly of the left voted in favour of reducing or eliminating heritage restrictions on development, and in favour of allowing building to greater heights downtown. 

Councilors broadly of the right voted against all of those, but did support allowing new subdivisions - unfortunately, without support from the pro-housing councilors on the left.

And Iona Pannett ... well, she's on her own anti-housing axis that isn't quite left-right but that seems to view building housing as neoliberal and consequently bad. 

But this from Nicola Young was disappointing.

“The character suburbs are literally what gives us our character. Bowling them is not going to help affordable housing, as it’s the most expensive land in the city,” Young said.

Wellington housing is musical chairs with a lot more people than chairs. If you allow people to subdivide the larger nicer chairs so that a lot more people can find a seat, fewer people are left scrambling when the music stops. That means more room for everyone, and less bidding-up of prices. 

I'd expect townhouses or apartments in Mt Vic to be high-end, given the land prices. There I'd agree with Young. But the people moving into them are then putting less pressure on other existing properties. They're freeing up other chairs in Wellington's horrible game of musical chairs, where we currently wind up with too many people crowded onto cracked, broken, moldy seats. 

Anything that enables adding chairs helps. 

Another case for Cat Bonds

This week's column in the Dom draws on the joint RBNZ-Treasury workshop on post-Covid macroeconomic policy that preceded last week's Covid-truncated NZ Association of Economists conference. 

A snippet:

Overall, the workshop felt designed to warm the economic policy community to higher public debt levels for a longer period. The risks of the approach were noted: interest rates can rise, and there will be problems if they do.

And the approach only makes sense if projects funded by that debt really do pass cost-benefit assessment. That conventional cost-benefit assessment processes ensuring value for money seem out of fashion was not noted as any substantial constraint.

Higher levels of government debt bring risk not only in case of interest rate increases, but also in case of natural disaster. Maintaining headroom to take on a lot of debt in a crisis has been important. If public debt is higher for longer, and global credit conditions become less friendly, the Alpine Fault becomes even riskier.

If the public sector is determined to encourage politicians’ imprudent pursuit of higher debt levels, it should encourage that some of that debt be funded more prudently: through catastrophe bonds.

Catastrophe bonds pay investors more during normal times but void most or all of the bond if a triggering event happens. If an earthquake required substantial government funding, existing catastrophe bonds would void and would provide some necessary headroom.

They may be a more prudent approach in imprudent times.

Nightmare scenarios do still exist

There were other interesting bits on the day. One presentation went through some simulations of different paths for fiscal consolidation (getting debt back down); the least costly approach, which also yielded long-run benefits, was through increased consumption tax - GST - and/or reduced transfer spending. The worst approaches were increased taxes on capital, and/or reduced government investment spending (on the assumption that that investment spending is on stuff with positive BCRs, which is a bit heroic). 

One option put up by that paper's discussant, which hadn't come up in the paper, was to use migration settings. You can drive down net debt to GDP by increasing population size - though you'd have to be careful on how the necessary infrastructure were financed.  

Michael Reddell has a good run-down on the macro session

Friday 18 June 2021

Contracting for buses

The Ministry of Transport is consulting on the Public Transport Operating Model.

Sounds boring as all heck, right?

You can submit on it via a survey, and boy are there some worrying questions in there. 

Let's step back a minute and consider the problem first. 

Wellington in particular screwed up its bus service. I have not seen the RFP or contracts that Regional Council put out for the bus service, but it's easy to diagnose the problem from the symptoms.

Here are the symptoms:

  • Bus services that are frequently cancelled for want of drivers;
  • Complaints about driver pay;
  • Concerns about there not being enough drivers. 
Any doctor can tell you the most likely cause of those symptoms. 

If the contract for service imposes low penalties for missed services or poor standard of service, relative to what the bus company would have to pay in higher staff costs to avoid having missed services or poor standards of service, then the bus company will optimise by running a lean staffing model. 

Bus companies compete to provide the service. Given the terms of the contract, the one able to deliver the desired service, and desired here means "follow the terms of the contract and respond to the incentives it provides in ways that are utterly predictable to the person writing the RFP if the person writing the RFP isn't an idiot", will be the one doing so at lowest cost. 

This isn't a problem of greed, or of the bus company putting profits over people, or of the bus being privately run rather than publicly owned - it's none of that. And it isn't a problem either of choosing the lowest-cost bidder either. 

The problem the contracts had to have set very low penalties for missed services. Just think about it for a minute. If it were cheaper for the bus companies to hire on more staff so there'd always be someone ready to take on a shift if a driver came up sick, that's what they'd be doing, right? Because they care about their bottom line. If it's cheapest to eat the fines and pay less, that's what they'll do. If the fines were higher, they'd eat the higher staffing cost instead. 

Now an important consequence of that higher-fine model would be that the bus company would have to make more on the route - whether through higher fares or higher council subsidy for running the route. How do we know that? We know that because the cost to the bus company would go up. 

So my diagnosis is that council was contracting on the cheap, not wanting to wear the higher costs either in fares or in route subsidies required for levels of service consistent with community expectations, and then foists blame on the contracting model, on the bus company, or on neoliberalism. 

Given that diagnosis, the potential solutions offered up in the MoT work are a bit wanting. 

They're looking at going back to public sector ownership, living wage mandates and the like. 

It misses the point. 

A living wage mandate might be a nth-best solution to the problem in that the bus company would be able to hire as many staff as the company would want at that wage, but it wouldn't solve the problem of the bus company not wanting to roster staff in case of staff not showing up. It would make that problem worse, unless the fines for dropping a service went up. Why? Because the cost of keeping extra staff around in case relievers are needed would be higher. So you might have fewer no-shows, because staff would be more worried about having to shift to some other job that pays less well, but you'd have fewer relievers ready to deal with no-shows, and there will still be an optimisation on how many relievers to keep on in case of sickness days. 

And public sector ownership doesn't solve anything either if the underlying problem is Council not wanting to stump up enough for a reliable service. 

Then there's a bunch of other stuff in the document about making it mandatory to run electric buses, despite transport being in the ETS, and about whether Councils should own and provide the bus depots to facilitate electrification. 

If this thing follows the fashionable direction of travel, we're going to wind up higher cost public transport without the desired improvements in service. 

Wednesday 16 June 2021

Afternoon roundup

The afternoon's closing of the browser tabs brings a few worthies:

Assorted updates

Blogging has been light. These are the bits I've missed telling you about.

Self-recommending. Meaning, I'm in them, and I'm recommending them. Likely not in the better way that Tyler Cowen uses the term. 

Tuesday 15 June 2021

Smokefree 2025

The Government's proposed approach for achieving SmokeFree 2025 is a bit over-the-top. 

The proposals would restrict tobacco sales to a smaller number of to-be-licenced R18 outlets, which could then be subject to a sinking lid; impose an annual one-year increase in the purchase age for tobacco until full prohibition were achieved; restrict nicotine content in cigarettes to very low levels; prohibit filters in cigarettes; impose minimum cigarette pricing; and further restrict flavourings.

In short, the only way to get a proper cigarette would be through the black market. The Ministry's betting on folks shifting more heavily to vaping or heated tobacco. I'd expect instead that imposing all this stuff would have smokers flip to black-market excise-free cigarettes, and that smokers would be less likely to switch from those to vaping. I also wonder whether some smokers might try soaking loose tobacco in nicotine e-liquid to get the nicotine levels up, and I don't know what smoking that stuff winds up doing. Heating is different than combustion. 

Our submission on the consultation document went through a couple of weeks ago; have been a bit pressed and hadn't gotten around to blogging it. 

But it has been fun watching more stories of black market tobacco coming though. Sometimes they're sold for organised crime groups; sometimes they're sold as church fundraisers. Loose tobacco is fungible like that. 

All the talk about prohibition reminds me that I have forgotten to tell you something else important. The new season of Cocaine & Rhinestones is up. It's Tyler Mahan Coe's podcast of the history of country music, and it's superb. 

Episode 4 is on prohibition and the leadup to George Jones's White Lightning

He goes through the rather evil history, not forgetting the part where the federal government poisoned industrial alcohol and murdered a pile of people who'd previously been filtering the bad tasting stuff out of industrial alcohol. But also the history of the whisky rebels before that. Great stuff. Recommended.

In other words, the U.S. government empowered a bunch of thugs to enforce organized crime’s monopoly on illegal alcohol distribution in most major markets of the nation. This is how Chicago came under the thumb of Al Capone, who was targeted by Elliot Ness and his Untouchables, yes… But this unit of officers were called “untouchable” because of their surprising ability to resist the near-universal corruption laid bare by Prohibition.

Beats me why the New Zealand government expects better results out of tobacco prohibition.  

Monday 14 June 2021

Turns out you can pay to get vaccines faster

Alex Tabarrok and coauthors have argued, convincingly, for substantial investment in vaccine manufacturing capacity by richer countries. 

Paying a lot up front to get vaccines faster doesn't have to be about shifting vaccines from one place to another. It can instead be about getting more resource to those companies to scale up production more quickly. And that matters because the lines can then just keep running, pumping out more vaccine more quickly for everybody. 

So, for rather some time, I've argued that New Zealand should have at least tried paying a lot more to boost capacity and to get vaccines here faster. Our border walls are expensive and less secure than they could be; getting everyone vaccinated faster reduces risk and would restore more normal travel arrangements more quickly. Getting that capacity in place matters a lot too. 

And I've often been told that that would be impossible because things don't work that way. 

Here's CTV news on how Canada got their vaccines faster by paying more.

Canada paid a premium to get more than 250,000 doses of the Pfizer-BioNTech vaccine delivered last December, weeks earlier than planned.

The detail is contained in heavily redacted contracts released to the House of Commons health committee late Friday, but any specifics on what price was paid or how the delivery schedule was amended were deleted before the contract was published.

Canada reached a deal with Pfizer in July 2020 to buy at least 20 million doses of the COVID-19 vaccine it was developing with German-based BioNTech. The first contract was signed on Oct. 26.

Pfizer Canada CEO Cole Pinnow told The Canadian Press in February that Canada's negotiations were based on an expectation the first vaccines wouldn't be authorized for use until late January at the earliest, and deliveries were planned to start after that.

But within a month of the contract being signed, Health Canada's chief medical adviser, Dr. Supriya Sharma, was signalling that her department was about two weeks away from giving the vaccine the green light.

That appears to have sent Canada racing back to Pfizer to see if its contract could be amended to get some doses delivered early. On Dec. 4, Canada and Pfizer signed an amendment allowing for that, but at a cost.

"Whereas (the) purchaser has requested, and Pfizer has agreed .... to amend the delivery schedule so that a certain number of contracted doses are delivered prior to Jan. 1, 2021 and in consideration thereof the parties have agreed to increase the price contracted doses which are delivered prior to Jan. 1, 2021," the contract says.

Every detail in the contract related to the price paid for dose was deleted before the documents were made public.

The contract with Pfizer stipulated nothing would be shipped until Health Canada had authorized the vaccine, which happened on Dec. 9. The first shipment of 30,000 doses was on a plane to Canada within days and the first Canadian was vaccinated on Dec. 14.

Ultimately, Pfizer shipped 255,450 doses before Jan. 1. It has since delivered more than 22.5 million shots, and its vaccine has become the main component of Canada's effort to get all 38 million residents immunized against COVID-19.

Canada expects to have the whole country fully vaccinated by the end of September. 

Update: No rush on vaccination though really. No Covid here, no chance the Delta variant will make it through the border defences...

Film subsidies and election campaigns

I'm a bit skeptical about campaign finance restrictions in general. Money probably does less to buy votes than is generally expected, especially when we consider that candidates with a better shot of winning probably draw more money in the first place. And restrictions in one place can often just make things bulge out in another place

But we have a fairly extensive set of rules, as well as prohibitions on a lot of speech activities that might influence people during election campaigns. Third parties need to be registered and their spending is tracked. Billboards and hoardings have to come down on election day, and you aren't supposed to even tweet about your preferred party on election day. Both rules seem increasingly anachronistic where more people vote at advance polls, but the rules are still there.

So things could get really really weird if the proposed film about the Prime Minister's response to the terrorist Christchurch mosque attacks goes ahead. 

The Electoral Commission doesn't consider books about candidates to be election ads or otherwise subject to regulation, so long as the books are sold at commercial rates. And they've suggested applying similar logic to a film about a candidate. 

You could then wind up in a spot where, prohibited from seeing tweets about candidates, all hoardings pulled down, you could nevertheless hear ads for a movie about the Prime Minister while driving past billboards for the movie about the Prime Minister on your way to a movie about the Prime Minister before heading out to the polling booth, where everyone would be banned from wearing political party regalia that might unduly influence you.

Movies, best I'm aware, tend to have rather more substantial marketing budgets than books. Pretty common to see billboards and ads for movies. A lot less common to see them for books. 

And then we start thinking about the film subsidy regime.

If the movie about the Prime Minister were considered an international film production, it would get 20% of its cost back as a rebate from taxpayers - with an option for an extra 5%. If it were considered domestic, it would get 40% up to $6 million, and an option for additional grants beyond that.

Total campaign donations and loans to campaigns, in 2020, were about $7.5 million. A party's election spending limit was just under $1.2 million.

So you could easily wind up in a spot where taxpayer subsidies for a film about the Prime Minister, released during an election campaign, advertised heavily during the campaign, and shown on election day exceeded the total combined campaign expenditure of all political parties. That seems, well, really big relative to the scale of NZ election campaigning. 

I don't know if there's a fix to the campaign spending rules to deal with this kind of mess, were it ever to eventuate. But there's a simple fix to the film subsidy rules. Some projects are already ruled out. There's a line excluding film grants for pornography, for example. 

Add in a line excluding grants for films about current MPs or those standing for office. 

I covered this in this week's Dom Post column.

I'd tweeted suggesting this back on Friday morning when the film was announced. The usual film-subsidy fans and Labour supporters got mad at me, until the emerging consensus was that the film was bad because they hadn't actually gone and talked with the survivors about centering the Prime Minister in a film about the attacks. 

But to be perfectly clear:

  • I'm not saying Labour or the PM had anything to do with the film, or knew about it, or encouraged it, or anything like that.
  • I'd have said the same thing about a hagiography of John Key a decade ago, or about an attack film on Ardern today, or about some "Air New Zealand: The Chris Luxon Story" film today, or about a movie about a Green MP's refugee childhood in New Zealand, or about a Green MP's human rights work serving as lawyer for an accused war criminal. Too easy to imagine too many MPs being able to be cast as heroes or as villains depending on the slant of the team making the film. Taxpayers shouldn't be funding that.
  • I'm not saying that they'd definitely get the grant or that they even applied for it yet, but the standing rules make the project look eligible and that would be part of any sensible marketing campaign for investors in the film at Cannes.
  • I'm not saying that the film will definitely go ahead; lots of films will get pitched at Cannes and not find backing. All the more reason to get the rules changed before folks start signing contracts based on expectations of coming subsidies. 

Wednesday 9 June 2021

Bindingness and the ETS

The Commission's final report came out today. It has a series of recommendations that cannot do much to affect net emissions when they run up against the binding cap on net emissions in the Emissions Trading Scheme.

So the Commission decides to claim that the cap is really non-binding. 

Concern was raised by several submitters that the NZ ETS has a 'neutralising effect' on emissions reductions achieved by other policies. They cited that in an ETS with a fixed emissions cap (limit on total emissions), every tonne not emitted by one party will be available for someone else to emit.

The NZ ETS, however, does not have a fixed cap. This is partly the legacy of how the NZ ETS was run in the past, which has led to over 130 million units banked in participant accounts. This represents significant oversupply beyond what is likely needed for annual demand and hedging purposes.

The lack of a fixed cap is partly by design - recent reforms have implemented price measures to either withhold or release units, to put a brake on the emissions price from going above or below certain levels. These reforms reflect the political context in which the NZ ETS operates, where policy makers are concerned not just about efficiency but also about where costs fall. 

This is not unusual. Every functioning ETS in the world today contains market stability mechanisms that alter the number of units available depending on the market price or other factors. This means that ETSs are hybrid instruments, with safety valves to manage price or adjust the cap in response to economic changes - given the inherent uncertainty in setting a cap based on forecast emissions.

The recent NZ ETS reforms also implemented a flexible, five-year rolling cap. Emissions reductions that are expected to be achieved through other policies can be factored in when the cap is set. 

The cap can be adjusted over time to reflect actual emissions reductions achieved through other measures, or reductions to emissions not covered by the NZ ETS. This is important in Aotearoa, as significant emissions are not covered by the NZ ETS, such as agricultural emissions and emissions and removals by some forests.

The combination of oversupply, price measures and a flexible cap in the NZ ETS mean that it will not necessarily guarantee a specific emissions outcome. It also means that the NZ ETS can be managed in conjunction with other policies so that emissions reductions or removals from other policies are not a wasted effort (see Chapter 19: The direction of policy for Aotearoa of the 2021 Supporting Evidence for more information). 

The Commission is wrong about the implications of legacy units, and wrong about how the ability to re-jig future limits makes it desirable to pursue complementary policies.

Let's take them in order.

The Commission is right that people stockpiled carbon units when the price was cheap against a time when prices were expected to be higher. And the Commission is right that more units would be released if ETS prices hit the cap. 

But the setting of the cap is sensitive to those. Here's how the Ministry for the Environment explains it. The website is dated April 2021, so I doubt it's changed since. 

Calculating NZUs available at NZ ETS auctions

The overall limits are reached by following a series of steps to determine the total number of NZUs that will be available for sale at NZ ETS auctions.

We start with the total volume of the emissions budget, then:

  1. Forecast emissions from outside of the scheme are removed to calculate the NZ ETS overall limt.
  2. Any required technical or forestry-related adjustments are made.
  3. An agreed volume of units is removed to drive stockpile reduction.
  4. The number of units projected to be freely allocated, or provided through negotiated greenhouse gas agreements are removed.
  5. A limit is set on international units
  6. The remaining volume is available to auction.

Point 3 says how they deal with the accumulated stockpile. They reduce the number of units created for sale into the system by the government so that more units are sold from those stockpiles. The expected sales from the stockpile are in the cap. If, in any given year, the number of sales from stockpile are higher than expected, they must consequently and necessarily be lower than otherwise expected in some other year. And future iterations of the cap would then come with revised expectations about flow from that stockpile. It does not plausibly affect Net Zero 2050 unless those with the units expect to do best by holding those units until 2050 and selling lots of them at that point. 

And if government is worried about that threat, they could presumably legislate a use-by time fuse on the stockpiled units such that they would no longer be accepted for meeting surrender obligations after 2040. 

The right comparison for the stockpiled unit volume isn't annual emissions, but total net emissions over the path to 2050. That path has to include release of fewer units by the Crown, offsetting the emissions that come from the stockpile. That's it. The cap continues to bind. Figure 1 from MfE makes it really abundantly clear. Look at the picture. It's split into two halves. One half is the uncovered sector, that still needs to face emission prices - agriculture. The other half is covered by the cap. On which side of the line is the grey pie slice titled, helpfully, "stockpile reduction volume"?

Again, this is MfE's explainer on how the cap is set, dated April 2021. Free industrial allocations? They're in the cap. Stockpile reductions? They're in the cap too. Neither of them makes the cap non-binding. Because they're part of the cap. 

Next, the price cap. Yes, the ETS has a price cap. But that does not affect the bindingness of the ETS's cap on net emissions. Rather, it introduces fiscal risk for the Crown. 


The cost containment reserve
The cost containment reserve is a reserve of NZUs which are available for sale only if a trigger price is reached in the auction. Making these extra NZUs available at auction eases demand for emissions units.

The effectiveness of the cost containment reserve in dampening the overall NZU auction price is dependent on the volume of units available and the demand for them.

Cost containment reserve units need to be backed by the Government
If any NZUs are sold from the cost containment reserve, causing the emissions budget to be exceeded, then the NZUs need to be backed by the Government. This means the Government would need to procure equivalent emissions reductions, for example by purchasing international units or funding activities that reduce emissions domestically.

So if we hit the price cap and the government releases one more NZU, the government must simultaneously find a way of avoiding a tonne of emissions. Allowing one more unit while reducing one more unit means net zero relative to the existing cap. 

It does not affect the bindingness of the cap. It cannot, unless you think that the government will renege on its obligation to back units. 

If buying credits in Europe is the most tractable way of backing an NZU, and European prices are higher than the NZ ETS price, then the Crown has potentially substantial liability. That liability increases the political risk that a future government might renege. 

But there's a simple solution. 

Update the cap so that instead of tracking $50 plus 2% per year, it just anchors in the European price. Whatever the European price happens to be at time of auction is the price cap at that auction. If the price cap gets hit, the government could just use the money it gets by selling those extra units to buy European ones and shred them to back the NZU released. Simple. Effective. Obvious. 

Finally, what about the Commission's view that the effects of complementary measures can be considered when setting the subsequent years' caps? 

It's true, but misses the point. 

The government can reduce the cap more quickly, with or without implementing complementary measures. 

If the government is *very* lucky, its regulatory solution is one that wouldn't have been found by ETS participants - perhaps there was an underlying real market failure that was getting in the way. In that case, the regulatory measure could be easily justified on a standard CBA laying out the intervention logic and the costs and benefits. 

If the government is *only* lucky, its regulatory measures will replicate the least-cost solutions that would have been found by simply reducing the cap without taking those regulatory measures.

But if the government is *not* lucky, it will find interventions that cost more than the going ETS price. It could have done better by reducing the cap more quickly - without also implementing measures like EV subsidies or petrol vehicle bans. 

So the problem isn't then that the the government can't offset things by lowering the cap more quickly, it's that the measures taken have a high hurdle to pass if we want to know that we're doing the most good possible. Because if it costs the country a billion dollars to reduce emissions through some regulatory measure costing twice the current ETS price, the government could instead have cut the cap by (roughly) twice as much for no greater cost to the country as a whole. Don't we want to be able to cut the cap faster rather than slower?

The way that caps are set isn't secret. The Climate Commission has to have known that stockpiled units don't make the cap nonbinding except in the most meaningless of ways - in any particular year, you'll get overs and unders around the path to net zero. It simply doesn't matter. And it also doesn't make complementary policies more desirable. 

TwitterNZ, tax, and transfer pricing

BusinessDesk reports that Twitter is setting up a local affiliate. So we can expect, a year from now, performative outrage about local tax paid. 

When I taught public econ at Vic, we had a short bit on company tax and international considerations. I invited the students to imagine that there were no Google New Zealand. Instead, New Zealand companies would be invited to bid for the right to manage NZ-based advertising sales for Google, getting a cut of the revenue from the adwords auctions. In the limit, where NZ is a competitive marketplace, how much do Kiwi companies bid for that right? The price of the right to be the contracted local affiliate should eliminate any local economic rents. The rents instead accrue to the IP, developed, run and taxed in the United States. 

Under that setup, the local affiliate doesn't pay much tax in NZ. It collects a lot in revenue, which it pays as its license fee to the US. Company tax is based on revenue less cost, and that license fee is part of the cost. It's revenue to Google US, and then gets taxed there.

It's obvious when you run it as separate companies. When instead Google NZ pays Google US to get the advertising product it onsells here, people argue about the transfer pricing arrangements and whether they're set to replicate the numbers that you'd get in the separate-companies kind of scenario, or whether they're set to shift taxable income to places where taxes are low. There are OECD transfer pricing guidelines on this stuff, and I understand that Google works with IRD to make sure that their practice is sound. 

Like, imagine that people were outraged that, while Honda makes a lot of money selling cars here, they write off as expense the cost of buying the cars in Japan and shipping them here, and called for taxing Honda based on gross revenues. Is it that different from what Jonathan Milne says here, from today's Newsroom Pro newsletter?

Facebook doesn't disclose its earnings in New Zealand, but the G7 agreement coincides with Google NZ filing its annual accounts with the Companies Office. Its press statement highlights the $3.3m tax it is paying on a profit of $10.19m. It fails to mention the $517m that Google NZ has paid its US parent company in service fees – an expense that it books against its gross earnings in order to report a laughably small taxable revenue in this company.

Anyway, maybe there's some advantage in Twitter bringing its NZ sales in-house and having local staff. But expect outrage next year when similar amounts of money go back to head office to cover the IP, overall tax paid doesn't much change, but the optics of it do because it'll be a transfer pricing arrangement. 

Update: Pattrick Smellie at BusinessDesk is worth reading. You should subscribe. 

However, the actions of Facebook and Google in NZ also indicate how differently the big tech companies can respond to such pressure. 

It is clear Google NZ is inclined to make nice, choosing in the last two financial years to declare advertising revenue booked by its NZ office paying tax on that. 

It presents an activist face in its government relations and is trying to stay ahead of a government showing increasingly regulatory instincts. It will placate local news media by implementing its Google News Showcase product, which will pay news producers something for republication and is being hailed in Australia as reviving news media companies' fortunes. 

In short, Google is pursuing a self-interested, carefully calibrated exercise in good corporate behaviour to maintain its political and social licence to operate in NZ. 

Facebook apparently doesn’t give a flying one. 

The company has a few employees in NZ, but the only time this reporter has ever clapped eyes on a Facebooker in an official capacity was before the 2017 election when executives from Sydney hosted a breakfast to explain how journalists could do a better job of providing free content for Facebook to monetise. 

At the same time, the company has simply stopped disclosing anything about the scale of its commercial activities here. 

Where Google pursues what it hopes is enlightened self-interest, Facebook’s approach is cynical - and arrogant to the extent that it has been flouting its filing obligations for five years.

Friday 4 June 2021

Bridge Maths

My excellent colleague Matt Burgess has done some fun back-of-the-envelope work on the carbon benefits of the proposed $685,000,000 bicycle bridge across Auckland Harbour. Some people have strongly favoured the bridge, noting its potential effects on reducing greenhouse gas emissions. 

Here's Matt's back-of-the-envelope reckoning. The strikeouts are because a first iteration of the work hadn't adjusted car emissions for average vehicle loadings; the updated version is so-corrected. But still back-of-the-envelope, order of magnitude stuff. 

The new Auckland crossing costs $685M. That is a daily cost of $112,500 at 6%. The minister justifies the bridge based partly on emissions.

Assume 10,000 one-way crossings/day and an average commute profile between Takapuna and Sky City = 8.7km.

Depending on whether each bridge user would otherwise have done their commute by bus or car, the bridge will avoid 8-23 8-14 tonnes of emissions per day.

So that is $4,800$7,800-$14,700/tonne abated.

To be competitive with the ETS at $37 would require 1.3 2.1-4.0 million crossings per day.

If bridge users would otherwise use the ferry, which may be the most likely alternative, then the bridge avoids just 0.5 tonnes of emissions per day at a cost of $237,000 per tonne.

He also notes that these kinds of cost-per-tonne figures would have serious consequences for New Zealand's ability to get to net zero, if we did it all via MAMIL-approved bike bridges (and the like) rather than by relying on the ETS.

The cost of getting long-lived emissions down to net zero is:

  • At $7,800/tonne: 94% of GDP
  • At $14,700/tonne: 176% of GDP
  • At $237,000/tonne: 2600% of GDP
The bike bridge is (roughly) 200-6300 times more expensive, as carbon mitigation initiative, than just buying credits in the ETS and running them through the shredder.

And so I fully expect that Rod Carr will strongly endorse the bike bridge in next week's Climate Change Commission report.