Sunday 30 August 2020

Not enough worst-case thinking

A University College of London Vice-Provost and clinical academic lays out scenarios

I worry we're not worried enough about Scenario 4.

There are four likely scenarios for exiting this pandemic:
  1. Development of a vaccine. A vaccine that successfully blocks transmission would be the most ideal strategy. Over 170 companies are now developing vaccines with three in large scale, clinical trials. It is essential that vaccines show not only an immune response (which they all do) but that this is sufficient to protect against severe disease and against transmission of the virus. There are, however, concerns that any protection may not break the transmission cycle and that immunity will be transient, as is the case for natural immunity of people, infected with coronavirus. A new vaccine typically takes four years to develop. It is my view that a vaccine is unlikely to be successfully developed that leads us to exit the current crisis. The other concern is the recent report of the infection of an individual, who had already had coronavirus. Re-infection was with a mutated strain and this suggests that vaccines will need to deal with natural variation or target the invariable part of the virus.

  2. The SARS-CoV2 naturally loses infectivity, as may well have occurred during the Spanish influenza outbreak in 1918-20. However, coronaviruses do not mutate at the same rate as the influenza virus.

  3. We develop herd immunity: an R0 of 2.6 means that approx. 62% of the population would need to be immune, at any one time, in order to develop herd immunity. Herd immunity values are currently approximately 0.5-15% in different areas of the country. Any recurrent waves of infection will be required to acquire a 62% target and that would assume that immunity is long lasting, rather than transient.

  4. The virus becomes endemic: this would mean the virus circulating in the community with intermittent flare ups that are handled by local lockdowns and standard NHS treatment. Interpretation of the current data is that case numbers are going up but there are no significant numbers of admissions to hospitals, at least in London. This likely reflects the testing of younger, more mobile people who do not succumb to the symptoms of coronavirus whereas, in the past, the people, who were tested, were those who had been admitted to hospital.

Thinking hard about the worst case makes all options look worse. 

But before that - some cause for optimism. A vaccine may be more likely than scenario 1 suggests. Saloni Dattani reviews the evidence and the forecasts here.  

But let's go with a worst-case where either a vaccine isn't successful, or where vaccines need boosters so frequently and where new strains requiring new vaccines keep emerging, so even 100% vaccination implies new outbreaks in the absence of tight border controls against the entry of new strains. 

The rest of the world is not following an elimination strategy. The longer this goes, the less likely it seems that they will even try. Too many places haven't the political cohesion to pull it off, even if they wanted to. 

Whatever you think about Sweden, they aren't going to hit herd immunity in Scenario 4. Herd immunity doesn't exist. It's just recurring Covid seasons, just like the common cold, except it kills a lot of people and leaves a lot of people with long term disability. 

If you want to run this direction, expect to spend a pile of money increasing capacity in hospitals and ICUs. Disability benefits will go up as reduced lung capacity prevents employment; spending on retraining would have to go up to enable shifts into other sectors. There would, in NZ, be pressure to shift ACC to cover this kind of harm, and making ACC cover health more generally rather than just accidents would probably break it. 

Expect contraction in hospitality and similar sectors as those who are risk averse take their own measures to reduce risk - at least until self-testing becomes ubiquitous. Look at the app integration of the latest Abbott test, which will show your latest test result. That test doesn't get us all the way there yet because it isn't an at-home test. We will get to at-home tests that are reliable enough while people are infectious, and cheap enough that they can be used daily or close to it. Restaurants and bars that want customers will require a recent test result as condition of entry. Risker workplaces, where distancing is harder, will test staff frequently to avoid on-site outbreaks that stymie production. When tests cost about as much as a cup of fancy coffee, workplaces will just do it. We will all be poorer by the real resources consumed in that regular testing, as compared to the pre-Covid counterfactual, but it's better than not-testing. 

There will still be recurring waves of illness, with consequent loss of life and morbidity, because even with lots of cheap tests around, some folks you just can't reach short of compulsory measures. Unless the schools are willing to enforce a regular testing regimen and to exclude students while infectious, it would at minimum come through that route. Even if each test isn't that expensive, testing each kid regularly would add an awful lot to school operating costs. School budgets would have to go up to accommodate it. 

Expect also that aged care facilities will be under heavy restrictions for the foreseeable future, with strict testing requirements for entry. 

None of that sounds good.

But neither do any of the other options. The other options may be less bad. But all options are bad. Everything sucks in Scenario 4. What sucks least isn't obvious. 

New Zealand and a small set of East Asian countries could hold to a strict elimination strategy, defend the borders, have robust systems for monitoring and trusting each other, and free mobility within the Green Zone. Advances in testing and compromises on some civil liberties for those entering the Green Zone, including rigorous testing and check-in requirements. Periods in managed isolation (NZ) or monitored and enforced self-isolation (Taiwan) can be shortened if it's possible to track people as they leave isolation and require them to self-test regularly. That increases throughput at the border without increasing risk. 

But the border will remain a horrible and costly bottleneck. I don't just mean monetary costs, obviously. There are huge humanitarian problems in imposing massive costs on travel. Remember isochronic maps? The maps that showed all the places you could get to, from London, in the same amount of time? 

We wouldn't be going all the way back to 1914, when most of New Zealand was 30-40 days from London. But coming into NZ from anywhere outside of a green zone would be a day's travel plus, at best, a week in self-isolation. 

Here's 1914 as reminder:

And here's 2016.

If the Green Zone holds, and the rest of the world remains a red zone but without formal restrictions against travel, the 2021 maps would look rather different. Basically, the map as it is in 2016, but with a big wall around an East Asian Green Zone where crossing the boundary costs somewhere between a week and a fortnight. The effective distance between New Zealand and the rest of the world would increase substantially, and we will increasingly be left behind in areas that require frequent interaction for initiating new projects and ventures. This would also severely hit the 1.2 million Kiwis born abroad and the million Kiwis abroad, any of whom may have family emergencies that require rapid travel to New Zealand from abroad. Even in the best case of this worst case, these costs will be substantial.  

Recall that the average month in 2019 saw over 250,000 Kiwis returning home from business trips, foreign study, holidays, or visits with friends and family. Current capacity in managed isolation is about 14,000 per month. Having to stay in isolation will itself deter travel. But it isn't crazy to think we'd need something like five or even ten times as much capacity in managed isolation as we currently have, once we also remember that others may wish to come here as well. If Australia gets to elimination, a lot of the burden on managed isolation goes away - Trans-Tasman is a big part of Kiwis' travel. But I do not know that they are actually trying to get to elimination, or if they will get there. 

Unless we start thinking about more practicable options. 

What can increase effective capacity while maintaining an eradication strategy?
  • A more risk-responsive set of isolation requirements, combined with options for self-isolation.
    • People coming from places outside of the Green Zone where COVID is less prevalent could have shorter isolation periods, if other risk-management protocols were put in place like post-isolation testing and location tracking. Halving the duration of isolation doubles effective capacity.
    • People willing to wear location monitoring bracelets could be allowed to isolate at home, if that home were a safe place for self-isolation - not having other flatmates, for example. I have no doubt that the American Ambassador's self-isolation in his home will be safe, and that his being there instead of in a managed isolation facility frees up a space in managed isolation for someone else. But figuring out a regime to work it all won't be easy. It requires checking that the proposed venue doesn't have others who'd be mingling. It isn't crazy to think that the costs of vetting a venue could be a reasonable fraction of the cost of providing a room in managed isolation. And while putting a monitoring bracelet on people is relatively easy in principle, keeping others from popping round for a visit is harder. The default could be managed isolation unless the traveller can demonstrate a credible self-isolation plan, but all of that will prove hard. There will be muppets who decide to have a party while they're self-isolating, and unless there are good chances of being caught and penalised for doing so, it'll be hard to deter. 
  • The whole voucher system I'd suggested to allow more facilities to come into the MIQ system - but it is still really costly both in resource and time, even if a lot of people are contributing to the costs of their own stays.
Maintaining eradication will be easier as testing improves, as regular cheap testing of anyone who is anywhere near anything that might be a risk, so it becomes far less likely that shutdowns are required to deal with outbreaks. Instead, they'd be caught more quickly, and contacts could be tested on the spot. 

A scaled up safe border is a lot more important in this world. If what we're looking at isn't a system that has to just see us through for the next year, but rather one that has to work for many years, the fixed costs of establishing better systems become worth fronting. It has to be done not just to enable Kiwis to come home, it also has to be done to mitigate some of the economic harms of Covid. There will be lots of sectors where rolling waves of disruption due to Covid are particularly costly. Some of those sectors might be able to relocate to New Zealand, if the costs of being a week or two away from the rest of the world are lower than the costs of disrupted operation. There will be piles of people who have shifted to remote work who could work remotely from here, and pay taxes here, while being paid by their overseas-based employer. 

But there are lots of options between eradication and Sweden. Canada and much of Europe are trying to keep a lid on things. Canada allows travellers to self-isolate at home, without the kinds of monitoring that Taiwan would impose. 

In the worst-case world of Option 4, everything looks worse. There will be no herd immunity. There will be no effective vaccine. There will only be dealing with it, for years and years. 

I don't think proponents of any of the possible strategies for dealing with it have carefully thought through what the "for years and years" really means. 

Proponents of Swedish approaches downplay that there will be deaths and misery for years and years, and that individual actions to reduce the risks they face also result in drops in economic activity. 

Proponents of elimination downplay the consequences of putting New Zealand back to the travel world of 1914 when the rest of the world is in 2020, and haven't thought seriously enough about what scaled up border capabilities have to look like to deal with any of this. 

And middling options that try to wipe out cases and clusters as they emerge, but without eradicating the virus, will face some of the same border costs (Canada's self-isolation requirements, for example) and ongoing restrictions and risk for those at risk, but without the potential opportunities that could come of being a covid-free place. 

I also don't think it's obvious what the best approach is.

If there were a button I could push for the set of policies that would simultaneously invest heavily in scaling up border systems (with costs recouped on a user-pays basis for overseas visitors) AND became far more welcoming to overseas arrivals by making it really easy to get a "work from New Zealand" visa for those who'd come here to work remotely for their foreign employers, making sure that there's enough managed isolation capacity to deal with it, making it easier for foreign investors to set up shop here as the Covid-free place. If it looks like normal campus life would be impossible in Covid places, a scaled-up managed isolation system could make NZ very attractive as destination. 

But I worry that the xenophobia that's come through during the pandemic makes that button impossible to push, even if the systems are safe. 

Meanwhile, the American university system is learning. They have hundreds of universities, all trying different things for dealing with the crisis. And look at what the University of Arizona has pulled off. Wastewater testing at each of the campus dorms found an outbreak before it happened, and stopped it.  
Click through and read the thread. The University of Arizona is testing wastewater coming out of each dorm facility and coupling it with rapid antigen testing. Everyone was tested before they were allowed to move into the dorms. When one dorm showed Covid in wastewater, everyone in the dorm got a quick antigen test, two asymptomatics turned up positive and were put into isolation, contact tracing is in place, and they're dealing with it.

That's what one university came up with. 

There are plenty that are providing very good examples of what not to do. But students will shift to the places that have decent systems. It'll be part of their selling point, if this is here for the long haul. 

New Zealand has one creaky system that has a hard time adapting. There have been suggestions here around wastewater testing, but has anyone started looking at testing the water coming out of each of the MIQ facilities? If it shows up in wastewater pretty early, you could run regular wastewater screening and then test everyone in the facility if the wastewater shows a positive case. 

It is hard to pick what policy options suck least when you combine worst-case thinking about the virus becoming endemic with worst-case thinking about government capabilities. Maintaining elimination has a lot of option value. Treatment keeps improving. Testing keeps improving. Systems elsewhere for getting on top of outbreaks will keep improving, and learning from experiences elsewhere will make things here better. 

But it is also easy to imagine scenarios where things abroad improve substantially for treatment, for keeping on top of outbreaks and for preventing transmission into vulnerable communities. Lots of reliance on cheap testing, layered protections for aged care facilities, isolation of cases before transmission chains can get much established. Maintaining quarantine for those arriving from abroad may not make sense in that scenario: you'd rely instead on pre-travel testing, location monitoring requirements for recent arrivals, and requirements that new arrivals do daily self-testing. We are nowhere near that making sense now. But would we be able to switch to it when and if it does make sense, under worst-case thinking about politics?

Friday 28 August 2020

Burton on the bureaus

This is the sort of thing that should wind up on the syllabus in politics and public administration classes. Tony Burton on how government departments really work:

In the days after the Nazis surrendered in 1945 the allies had a lot to do, so there was a curious period when the senior bureaucrats of the old regime, holed up in a nice castle in north Germany, were left alone. So what did those bureaucrats do as their country lay in smouldering, starved ruins?

“… in a former school room still smelling of chalk, we solemnly met on the dot of ten every morning, sat down on brightly coloured straight chairs around a brightly painted square table and discussed the non-existent plans of a non-existent country.” (Albert Speer quoted in Sereny 1995)

One suspects that after the border control debacle there are New Zealand ministers wondering if something similar is happening in the streets around the Beehive. The government’s strategy had two parts. The hard part was largely achieved by the “team of five million” agreeing to stay at home at the risk of large scale unemployment. The second part, the seemingly simpler task for an island country four thousand kilometres of ocean away from its nearest neighbour, closing the borders. The first part of the plan has succeeded beyond expectations. The second part, not so much.

When I was part of the government machine I was struck by how little understanding even those receiving the eye-watering fees to teach “Masters in Public Policy” have of the way government operates. (If you want an example, look up “policy cycle” in a textbook on government where you will find a hamster wheel schematic and text describing how, apparently, government is run by hamster bureaucrats scuttling round it.)


To get from the Minister to a person providing a service, say a person at the border, takes anything between five to ten of these layers. That is, five to ten people, each with a range of jobs, one of which is to “cover off” a policy like the border controls to keep out Covid. Apparently this was not bureaucratic enough, so since last week’s Public Service Act came into force, the State Services Commissioner has been added as an additional layer between Ministers and the departments that deliver services.

To understand what happened at the border it helps to look at the writings of a neo-liberal. The term “neo-liberal” is now used so promiscuously it has become the left’s contribution to a dumb and dumber act with the right’s “political correct”. But at one point it described a mid-twentieth century approach to government whose proponents included a bureaucrat turned academic called Frederick Hayek. Contrary to myth he would have been a sceptic about the 1980s reforms in New Zealand, not the reduction in the size of government but any “rational choice“ thinking behind the new model for departments, and his views on the power of convention would probably have led him to predict the recent reversion back to a more British style civil service.

Unlike Keynes, and many others who theorise on the role of government, Hayek spent his time in the civil service working as an administrative cog in a typical silo of government, in his case the department that managed Austria’s government debt. This gave him an insight into how little relevant information is available for most administrative purposes and he applied this to his thinking on both government and private sectors. The key problem government organisations face, that private organisations do not, is lack of incentives to find and use the information they need to do their job properly.

Less abstractly, imagine you work on the front line of the New Zealand Customs Service and want to play your part in keeping New Zealand free of Covid-19. If you spotted gaps in the system, you would be required to “work through the line”, so your manager could pass the information up the managerial hierarchy.

Those managers are expected to spend 40% of their time “managing up”, that is making those above them in the hierarchy happy. So, at each point that information on border deficiencies rises, it competes with agency and departmental priorities of far greater importance to people in that line, like team budgets, annual performance reviews and “strategic visions”. At the end what is left are abstracted summaries designed to make senior bureaucrats feel powerful and ministers feel in control.

Worse, it is rare for departments to spontaneously present bad news, particularly to ministers. No-one wants to be known for telling the government that its policies are not working. At its most extreme, a former Chief Executive of MSD commanded “no problems without solutions” so only problems that had already been solved could be presented to senior managers. More commonly, there are long delays as senior officials angst over how to present the information in a way that does not show the department in a bad light. Even if good information somehow gets to the upper reaches of the hierarchy, it will not necessarily reach those who need it in a coherent or timely form.

This works the other way too. Ministers very rarely talk to people at the front line. Their decisions are largely informed by meetings with people at the upper end of the hierarchy who are equally ignorant of what is happening where services are delivered. Decisions are and are then passed down that line of five to ten people. Anyone who has played Chinese whispers will know some of the problems with that. Unlike the children’s game, the whispers will be carefully “messaging” to promote executives, their teams and their department.

The miracle is that this ever works. We observed one reason over the last couple of weeks: if those administering the system go too far from the real world, the real world tends to tap them on the shoulder and remind them of its existence.

Thursday 27 August 2020

Better testing

More rapid COVID testing options keep opening up. This week, Abbott's Ag Card Test received FDA emergency authorisation. Antigen tests like this won't pick up Covid as early in an infection as a PCR test would, but are decent after symptom onset. 

And this can then play a role as part of a testing and management system. 

For example, rather than go in for a PCR test because you've started having symptoms, and wait around for a couple days or more to find out whether you're positive, you could take this kind of test. This test runs $5 USD and gives results in 15 minutes. It would be cheap and easy for everyone just to keep a couple at home, in case needed. 

Community testing when someone feels unwell would shift from being a potentially scary nose-swab, coupled with costly stay-home requirements when waiting on results, to something easier and cheaper than a home pregnancy test. 

Those leaving managed isolation could be given a few of them and asked to test themselves in case of any symptoms, and required to send through a test result a week after leaving isolation - just to be sure there wasn't any late-onset. Note that this test comes with some rather interesting app options:

NAVICA mobile app will help facilitate return to daily activities
Abbott is also offering a mobile app at no charge that will allow people to display their results obtained through a healthcare provider when entering facilities requiring proof of testing. The NAVICA app is optional and an easy-to-use tool that allows people to store, access and display their results with organizations that accept the results so people can move about with greater confidence. The app is supported by Apple and Android digital wallets and will be available from public app stores in the U.S.

"While BinaxNOW is the hardware that makes knowing your COVID-19 status possible, the NAVICA app is the digital network that allows people to share that information with those who need to know," said Ford. "We're taking our know-how from our digitally-connected medical devices and applying it to our diagnostics at a time when people expect their health information to be digital and readily accessible."

If test results are negative, the app will display a digital health pass via a QR code, similar to an airline boarding pass. If test results are positive, people receive a message to quarantine and talk to their doctor. As they're required to do for all COVID-19 tests, healthcare providers in all settings will be required to report positive results to the CDC and other public health authorities, regardless of whether they use the app. The digital health pass is stored in the app temporarily and expires after the time period specified by organizations that accept the app.

The app's user interface is supported by a back-end digital infrastructure that is cloud-based, scalable and secure. It's been designed to support a very large number of users and enable access from anywhere. The app is not for contact tracing and only collects a person's first and last name, email address, phone number, zip code, date of birth and test results.

All kinds of other options start opening up as well. 

Everyone working anywhere near the MIQ system could be tested daily. They could have a set of tests at home so that anyone in their family who gets a cough could self-test immediately. 

Pre-flight testing becomes dead simple. Adding $7.50 NZD to the cost of an airline ticket would get a test for every person boarding. You could use it for every flight during Level 2 or 3, adding only trivial cost, and potentially as a way of getting rid of seating restrictions - consult with your local epidemiologist on whether the risk reduction from testing makes that equivalent. If so, the test requirement could reduce the cost of plane tickets despite the cost of the test, because it would let planes run closer to capacity. 

Workers in sensitive facilities like cold-stores and meat packing plants could be tested daily during any outbreak, rather than facing either shutdowns or restrictions on operation. 

And all of it would mean that contact tracers would have an easier job because infections would be more likely to be caught quickly. 

All of this could make elimination much much cheaper to maintain. 

Update: To be real clear - it isn't available yet. The Emergency Authorisation in the US doesn't even yet allow at-home use. But more of these things keep being developed

For A Reverse Public Works Act

Earthquake strengthening is expensive enough on its own. Layering heritage restrictions on top of those costs can make it impossible. 

The Herald reports on the former Tramway Hotel in Mt Cook, Wellington. It has the lovely unreinforced brick masonry facades that look so very similar to the ones that fell over and killed people in Christchurch. It was supposed to have been strengthened by December 2013. It hasn't been yet. 


The building is basically useless. The owner would like to demolish it, but has to go through a consenting process for that. Council would like to take over the building temporarily, spend millions strengthening it, then hand the building and the bill back to its owner. The Herald says strengthening will cost more than $6 million, and the building would only be worth $2.6m after those works are done. 

If the heritage value of that building is really in excess of $3.4 million, which seems impossible to believe, then Council should be fronting that cost - not the owner. The heritage value clearly doesn't accrue to the owner. 

But Council can't afford to spend millions and millions on each of the hundreds of buildings it has decided are of substantial heritage value. Instead, it wants to load those costs onto the owners of piles of buildings of dubious heritage value. 

This is clearly a regulatory taking. Council is forcing these owners to provide a public amenity, at the owners' own cost. When we'd looked at this stuff back in 2016, nobody could explain to us the selection criteria for heritage listings. There certainly wasn't any cost-benefit assessment going on. The best explanation we were able to get was that some Council official likely drove around with a clip-board, ticking off buildings that looked to be nice a few decades ago. Back then, there wouldn't have been substantial cost associated with a listing. Now, a listing means it is impossible to undertake earthquake strengthening works, which are compulsory, without going through ridiculous layers of heritage consenting bureaucracy. 
In 2017 Parbhu sent WCC plans to build a bigger hotel but that was not progressed after what he considered to be unlikely support from heritage officers.

He told the Herald his prime goal was to preserve the building, but he needed to break even and that meant adding to the existing structure.

Parbhu was reluctant to comment any further while the demolition application was before council.

Another option he has considered is selling the building for land value, but that has also proved unsuccessful because of heritage restrictions.
Iona Pannett, on Wellington Council, heads up the Heritage portfolio:
But speaking generally, she said the council's approach was to work constructively with owners.

"Obviously owners do have property rights in this country, and we want good outcomes for the city as well, so our preference is not to spend ratepayers' money to come in with a big stick to take enforcement."

Pannett said there were more than 500 earthquake-prone buildings in the city and she acknowledged some owners were struggling to find an economic solution to pay for strengthening.

Safety, heritage, and owner's ability to pay was a difficult balance, she said.

"I don't think you can say that one thing takes precedence, we're trying to balance all those things in a reasonable manner but it's not easy."

The reasonable way to balance these things is for Council to set a heritage budget, to pay the owners of these buildings an annual easement reflecting the annual value the community presumably enjoys from the existence of the building as it is, and to remove all other encumbrances. If the owner decides to forgo the annual easement by demolishing the building, or to take a reduced easement because strengthening processes detracted from the heritage value, that provides the balance between safety, heritage, and the owner's ability to pay. We went through this in a report in 2016

Another option would be a Reverse Public Works Act. 

Under the Public Works Act, the Government can compel a property owner to sell property to the government because it is needed for some public purpose. 

Under a Reverse Public Works Act, the owner of a property encumbered with a heritage designation would be able to compel Council to purchase the property at its rateable value, or at the value of the underlying land, whichever is higher. Council could avoid that obligation only by de-listing the building. In that case, if the real value of the heritage amenity were high enough, Council would buy and strengthen the building in heritage-sensitive ways, and then either sell or lease out the building. If Council instead viewed the heritage value as being less than the cost of buying the thing and strengthening it, it would come to the decision it should have come to in the first place and de-listed the building so that it could be bowled or strengthened in more cost-effective ways. 

In the alternative, if Council insists on sticking with this process, the heritage people on council should be personally liable if the earthquake comes and fatalities inevitably result. The owner's trying to make the building safe by demolishing it; those on council standing in the way really should be culpable here. 

Wednesday 26 August 2020

Behind the veil

Michael Huemer points out the numerous problems in John Rawls' maximin criteria.

Rawls argued that, behind a veil of ignorance where you can't tell your final position in society, people would agree on a rule that maximised the position of the worst off.

Economists have generally argued this is nonsense - it implies a degree of risk aversion behind the veil that is never seen anywhere else. 

Huemer goes through the problems. A snippet [note that OP means the Original Position, behind the veil]:

II. Diminishing Marginal Utility

Wealth has sharply diminishing marginal utility, above a certain minimum level. So there’s not much reason for the parties in the OP to try for greater wealth, above the amount they would get under the Difference Principle.

a) That’s an argument for not maximizing expected wealth. It’s not an argument for not maximizing expected utility. The diminishing marginal utility of wealth is already taken into account by utilitarianism (obviously), so it can’t be an argument against utilitarianism.
b) If there is really a minimum amount of wealth that you need for a decent life, and additions above that amount give little or no improvement in your welfare, that still wouldn’t support the difference principle. Rather, it would support the principle, “maximize the number of people who are above the minimum.”
c) Rawls hasn’t supported those factual assumptions anyway. It’s plausible that there is a minimum amount of wealth needed to have a decent life, and there is also an amount above which you get no noticeable benefit from more wealth. But those two amounts are obviously not the same amount.


Tuesday 25 August 2020

Distancing rules

This is a bit of an odd one.

On Monday, the government announced that masks on public transport will be compulsory under Covid Alert Level 2. Hooray! The new rule will take effect from next week - they still have to draft the order to give it effect. 

So far so good.

But from 11.59 pm this past Saturday night, the COVID-19 Public Health Response (Alert Levels 3 and 2) Order (No 2) 2020 came into effect. Clause 25 of that Order removes physical distancing requirements on public transport services.  

Adjusted alert level 2 requirements for public transport services

(1)  This clause adjusts the alert level 2 requirements for public transport services.
(2)  Public transport services need not maintain compliance with either the 1-metre physical distancing rule or the 2-metre physical distancing rule for people using those services.

And public transport services are defined earlier:

public transport service has the same meaning as in paragraph (a) of the definition of public transport service in section 5(1) of the Land Transport Management Act 2003 but—
(a) includes air transport that is available to the public generally; and
(b) excludes small passenger service vehicles (as defined in section 5(1) of that Act)
So you don't need to be physically distanced on busses or airplanes, unless I'm misreading this.

Now I can see how masks could be a substitute for distancing on busses. Masks and distance would be safest, but if you want more people to be able to fit on a bus without increasing risk, you could combine it with a mask order. That all kinda makes sense.

But what is weird is that the distancing order came into effect more than a week before the mask requirement is coming in.

And, more importantly, and as best I can tell, the Air Transport rules haven't been updated to reflect the change in the physical distancing rules - and this was updated yesterday.

Physical distancing (previously approved seating configurations) in aircraft applies. 

But the Public Health Order says distancing on planes is no longer required at Level 2. And the coming mask order will mitigate the risk of lack of distancing. 

Imagine trying to run an airline on this basis. One set of rules from Transport sets out seating configurations that make it harder to run any flights. Level 2 looks like it could be in place for a while. Jetstar dropped out rather than run under those rules. But the Public Health Order says distancing is no longer required. 

If the government's intention was to allow airplanes to run with masks in place of distancing, that would be consistent with the Public Health Order - but the Air Transport Rules would need to be updated to reflect it. And who knows what the legal basis is for "approved seating configurations" under Level 2 if the Public Health Order no longer requires distancing. 

If the government's intention was not to allow airplanes to run with masks in place of distancing, and to book out the middle seat under L2 (or the partner seats in the smaller planes with two seats on each side), why did they abolish distancing requirements for domestic transport in the Public Health Order (there's an exclusion for persons subject to the Air Border Order, so international will still have to be distanced)?

And the Prime Minister keeps emphasising the need to maintain distancing - despite the Order saying it's no longer required on public transport under Level 2. 

What is an airline actually supposed to do under these circumstances? If they guess wrong in one direction, they're throwing away money and offering fewer flights. If they guess wrong in the other direction, they risk pissing off a Minister or Ministry for not complying with what the intention had been regardless of what was written in the Order - and potentially a hasty Order leading to cancelled tickets. 

Stuff falls between the cracks between Ministries all the time; this seems a potentially messy one. 


Under that interpretation, the Order wasn't meant to supercede anything from Transport about seating but just stuff that might happen during boarding and the like. But I wonder what the legal basis for Transport directions around seating might be when a Public Health Order says there's no requirement to distance.

Protecting the Canadian Dairy Cartel

Canada might not be playing fair under the CPTPP. 

Here's Farmer's Weekly:

In the two years since the agreement came into force low-tariff and tariff-free quota created to open their markets to more imports of dairy products have gone largely unfilled.

Canada has been the worst offender, with just 12% of quota for dairy imports from CPTPP countries filled last year and just 4% so far this year.

In Japan just 40% of dairy quotas have been filled while fill rates have also disappointed in Mexico.

Dairy Companies Association of NZ executive director Kimberly Crewther said administrative foot-dragging by the three countries’ governments, along with President Donald Trump’s withdrawal of the United States from the agreement in 2017, meant the $96m of annual gains predicted for the industry once CPTPP was fully implemented now looked unachievable.

Not a particular surprise that the Canadian government would pull whatever tricks it could to support the dairy cartel.

The Ministry of Foreign Affairs and Trade was aware of the problem and raised it with Canadian, Mexican and Japanese counterparts and was due to do so again at an online meeting of CPTPP officials last week.

Crewther said it was disappointing but not unusual for countries to use quota administration to protect local industries from the increase in imported competition that followed free trade agreements.

“The agreement of new access in FTAs is great but it is the implementation around that access that is really crucial to whether it is usable or not.”

In the case of Canada up to 85% of quota was allocated to local processors who sourced subsidised raw milk from the country’s farmers and tended not to import dearer foreign dairy products.

Only 10% was for importers, who were also prevented from getting their hands on quota allocated to local processors but not used.

Canada’s chronic underutilisation of its CPTPP import quotas was reinforced by its Milk Class 7 milk payments system, which subsidised prices paid to local farmers and undercut imports and has long been a bugbear of the NZ industry and other exporters.

Monday 24 August 2020

Covid and co-morbidities

A tallying of the costs of road accidents that included fatalities but ignored disabilities would result in too little investment in road improvements that might reduce accident rates.

Tallying Covid's morbidity costs is an awful lot harder than tallying the morbidity costs of road accidents. We have years of data on road accidents and anything in New Zealand that comes consequent to a road accident runs through our ACC system - it's then not all that hard to get a handle on costs.

Covid is a lot harder. We all hear horror stories about the ongoing consequences for some who catch it. If those stories represent one in a million cases, they wouldn't have much effect on policy decisions. If they represent one in ten cases, then they're a big deal. But it is darned hard to find anything that summarises the numbers.

Dr Jin Russell points to an article published in Nature back in July that summarises the various conditions that can be consequent to Covid. 

Here's the abstract:

Although COVID-19 is most well known for causing substantial respiratory pathology, it can also result in several extrapulmonary manifestations. These conditions include thrombotic complications, myocardial dysfunction and arrhythmia, acute coronary syndromes, acute kidney injury, gastrointestinal symptoms, hepatocellular injury, hyperglycemia and ketosis, neurologic illnesses, ocular symptoms, and dermatologic complications. Given that ACE2, the entry receptor for the causative coronavirus SARS-CoV-2, is expressed in multiple extrapulmonary tissues, direct viral tissue damage is a plausible mechanism of injury. In addition, endothelial damage and thromboinflammation, dysregulation of immune responses, and maladaptation of ACE2-related pathways might all contribute to these extrapulmonary manifestations of COVID-19. Here we review the extrapulmonary organ-specific pathophysiology, presentations and management considerations for patients with COVID-19 to aid clinicians and scientists in recognizing and monitoring the spectrum of manifestations, and in developing research priorities and therapeutic strategies for all organ systems involved.

The piece is written for clinicians, telling them things potentially to watch for. Turning it into something that could be used in cost-benefit assessment would be a pretty big job.  

There are rather a few potential consequences.

The article will note things like "in a cohort of 107 patients admitted to a single-center ICU with COVID-19, their rates of pulmonary emboli were notably higher than those of patients admitted to the same ICU during the same interval in 2019 (20.6% versus 6.1%, respectively)."

But using things like that in any CBA work would require knowing what proportion of COVID cases in that area wind up in ICU. You need to know what proportion of infections wind up with that outcome to be able to say anything . You can make simplifying assumptions about the proportion of cases that wind up in ICU. 

Other conditions are reported as fractions of hospitalised patients: 17% of hospitalised patients in Wuhan wound up with cardiac arrhythmias. But what fraction of infections wound up in hospital? And were those conditions presenting only while in hospital during infection, or did they persist? Acute kidney injuries occurred in 37% of patients hospitalised in New York, with 14% requiring dialysis. But again, what fraction wind up in hospital, and is this something from which people recover, or does it continue?

I really wish we had a better picture of the state of play.

There are a bit over 1500 people in New Zealand who have recovered from Covid. They would not be representative of the population overall, should New Zealand ever have a broader infection. Those here who have recovered are going to reflect the cases we have had: mostly those who have returned from overseas, who are younger and healthier. 

But despite that, surely it would be worth knowing the outcomes for that cohort. It would be cheap in the grand scheme of things. The government could pay GPs for a series of follow-up visits with those who have recovered, and pay those who have recovered for their time. It feels like it would be hard to spend $10 million on this. $5k split between GP and patient would only get you to $7.5m. 

The series of follow-up visits would check for any ongoing health effects of Covid. I'd also really like to get some measures on just how bad the whole experience was for those who have had it. I'm not a big fan of numbers expressed in survey values as compared to revealed preference, but I'd expect we'd learn something useful out of it here. 

Get the big table of conditions that have existing DALY or QALY figures attached to them. Ask each recovered patient "So, which would be worse? Getting Covid again exactly as you had it, but with no longer term consequences, or breaking your femur?" Running that over a series of conditions would get a distribution of valuations, among the cohort here who have recovered from it. You'd have to be rather careful in extrapolating that up to the population, because that cohort will have far fewer pre-existing conditions that worsen things. But at least it would be something of a start. 

I don't get why this kind of follow-up work isn't being done. Somebody would have to fund it, and it would likely be more expensive than a standard grant. Could it just be that nobody thought to put in a funding line for this kind of thing?

Friday 14 August 2020

Public Sector Reasons

There's a standard list of things that the public service will point to when it just doesn't want to do something. Maybe there are privacy considerations. Maybe it's complicated. Maybe there are difficult intersections with employment law. They'll pick whatever seems most plausible from the list and patiently explain why something that seems an obvious good idea just can't be done.

We heard a fair bit of that around making COVID testing a job requirement for those at the border and in MIQ. It's too hard. You can't change people's employment contracts. You can't compel medical treatment - what, are you going to hold them down and force a stick up their nose?

But there's a problem with all of the excuse making.

The problem is simple. Section 11 of the COVID-19 Public Health Response Act 2020, assented to 13 May 2020, at Subsection 1, details the orders the Minister or the Director-General may make. The DG may make these orders if they are urgently needed and are the most appropriate way of addressing these matters. (Section 10). 

Subsection 1(a)(viii) allows the Director General of Health to require persons to:

report for and undergo a medical examination or testing of any kind, and at any place or time, specified and in any specified way or specified circumstances:

That means the DG Health, Ashley Bloomfield, at any point could have issued an order requiring testing of MIQ and border staff if he thought it warranted. 

It would have been most obviously included in the COVID-19 Public Health Response (Air Border) Order 2020 at Section 7

Section 7 includes a set of Obligations for arrivals that requires each person arriving in NZ by air other than those excluded to (at 1(a)) 

"report for, and submit to, medical examination and testing, as soon as practicable after their arrival, at the security designated aerodrome at which they arrive"

Don't you love how they call it an aerodrome? It's cute. 

Section (c) requires those arriving to report for and submit medical examination and testing as required while in isolation or quarantine. 

It would have been pretty easy to add in a couple lines requiring every person dealing with anyone who is arriving into New Zealand, whether they are airline staff, airport staff, customs officials, contractors of any of those, bus drivers, or employees at managed isolation and quarantine facilities, also to report for and submit to testing at least once a week. 

Whatever arguments people want to make about it being hard to add into employment contracts or whatever, it's just public sector bullshit. Everything can be overridden by a Public Health Order. And it wasn't done.

The government chose to have this outbreak. 

But it isn't just a failure of the political side of government. 

If Bloomfield had felt it was warranted, as DG Health, he could have issued the order anyway, unless I'm really misreading Section 10. There might be a bit of finagling where Section 10 allows orders that apply only within a single territorial authority district, but requiring it even only in Auckland would have been a substantial improvement.

It'll be an interesting standup at 1 PM. 

Update: Obviously, I'm not in favour of holding staff down and putting a stick up their nose. I am very very very much in favour of firing* each and every MIQ or border staffer who expressed the slightest reluctance about testing, because muppets who do not understand risk and who are unwilling to experience even the slightest inconvenience as part of reducing risk are the very last people you want anywhere near this particular risk. They will cut other corners the second you are not looking. 

* or redeploying elsewhere in the system, or shooting out of a very large cannon pointed very far away from New Zealand. 

Thursday 13 August 2020

Better border defences

Ages back, Tyler Cowen posted about the contrast between America's world-leading higher education system, and, well:

The United States circa 2012 is one of the most productive economies of all time, arguably the most productive if you take into account size and diversification (rules out Norway, etc.). Internationally speaking, in the richest and most productive global economy of all time, which is our most competitive sector?

Hollywood? Maybe, but it could well be higher education. Students from all over the world want to go to U.S. higher education. If we had nicer immigration authorities, this advantage would be all the more pronounced.

In other words, I work in what is perhaps the most competitive and successful sector in the most competitive and successful economy of all time.

And yet what I see around me is a total, total mess. And I believe my school to be considerably above average in terms of how well it is run.

I've felt like this about New Zealand's success in keeping Covid out, and the daily stories about the latest bit of dysfunction at the border. 

And the inevitable obtained. We have an outbreak.

Not just an outbreak though. An outbreak where the government has no clue, as yet, where it came from.

That isn't good.

Under a "Gee, wouldn't you think they'd run this seriously?" assumption, it would be impossible, right? Everyone working in the Managed Isolation and Quarantine system, whether on the planes getting people here, in the airports, on the busses shuttling them through to facilities, or staffing those facilities, would be tested for Covid so often that it would be impossible for one of them to do more than pass it on to their own immediate contacts before it was noticed. 


Testing everyone weekly would mean it would be really unlikely that anyone could go more than 7 days without being caught. In the worst case, your test day is the last possible second before your viral load is high enough to be caught, and you're infectious one or two days later, and you never develop noticeable symptoms. Even in that case, you maybe have had 6 days to pass it on. In that period, it is not terribly likely that anyone you've infected has become infectious. It could all be handled quickly and easily through contact tracing. 

But we could do even better than that, right? In a system that cared about it, and that wanted to avoid $1.5 billion-dollar-a-month lockdowns of Auckland, you'd think they'd also be running those cheap-as saliva tests. Sure, they're not as good as PCR. But they're real fast, and they're real cheap, and if they were being used daily, it's real likely that anyone who was contagious would be caught. 

But no.

They were doing none of that, were they. 

There's some small chance that the current outbreak was caused by the virus having come in in refrigerated transport. Maybe. But I do know that under a better system, the government could have said "We've double-checked our audit trail. Every single person working in the MIQ system has returned a negative COVID test in the past week, so it had to have come from somewhere else." Or, they would have caught it long before it took an unknown number of hops to reach the positive case that was found. 

BusinessDesk asked me yesterday for a short piece on the mess. I had a bit of a rant. It's paywalled; here are some snips. 

It has been obvious for a while that border practices have a Dad’s Army flavour to them.

Last week, the government received a report suggesting that testing frontline border and managed isolation staff only once every two weeks might not be enough. When pressed by journalists, the report’s author, Professor Shaun Hendy, conceded people “shouldn’t be forced to take weekly swabs but strongly encouraged to do so” – to use a journalist’s paraphrase.

This is astonishing. In any sane system, agreeing to be tested regularly would be a condition of employment in the managed isolation and quarantine system.

The government could require everyone who leaves managed isolation to show up for one further test a week after leaving managed isolation. The government could require everyone who leaves managed isolation to turn on Google Maps location tracking and to share their movement history with contact tracing teams if required.

Those together would vastly simplify the task facing contact tracing teams.

The post-isolation test would mean that anything that did get out would be caught earlier.


Every traveller leaving managed isolation presents a small risk. That risk can be reduced through post-isolation testing and by helping the contact tracing teams. But there is also a risk with every worker in the managed isolation and quarantine system. Every worker on the airplanes delivering visitors to Auckland. Every bus driver shuttling visitors from the airport to their isolation facility. Every worker within those facilities who may be in contact with the isolated visitors.

Even with the best of personal protective equipment and facilities, safety protocols can still fail. Workers are human. The protocols must be checked to make sure they’re being followed and tightened up where necessary. Is every incoming visitor masked when getting on the plane all the way through to arriving in their managed isolation room? If not, why not?

But those protocols must also be augmented with frequent compulsory testing of workers in the system. New Zealand’s PCR tests are highly accurate, yet they are costly and can take time to return results. Saliva-based strip tests are cheap and fast, but only accurate for those with lots of the virus in their bodies. Daily checks with cheap and fast strip tests coupled with less frequent PCR tests could be an effective combination for managed isolation workers. That frequent testing would not stop a worker from passing the virus to a close contact. But they would catch it quickly.


If a month in level 3 costs Auckland $1.5 billion in financial terms, plus added misery, but each strip test costs about $10, each Aucklander could be tested a hundred times before hitting the financial cost of that month in level 3. And only workers in the managed isolation system need regular testing.

A full-bore cost-benefit assessment isn’t needed to start doing this. It is almost incomprehensible that the system considers regular testing of staff an “aspirational” goal. The prime minister said in yesterday's press briefing that asking every worker to undergo a daily swab test is unreasonable. But inadequate testing that leads to a new level 3 lockdown is even more unreasonable. And there are options sitting between irregular and daily swab testing.

Finally, the government must create an incentive for the workers to be tested.

The government could compensate employers for providing extra days of sick leave for self-isolating workers awaiting test results. For the past month, officials have urged people to be tested if they develop symptoms and worried about the declining number of people presenting for testing. But they have not looked to the financial disincentives that bar many workers from being tested, even when the tests are free. This must be remedied.
I live in what is perhaps the most successful country in dealing with Covid - barring Taiwan.

And yet what I see around me is a total, total mess. 

Tuesday 11 August 2020

Borders and Bubbles

There's no Covid on the Cook Islands. But isolation-free travel into New Zealand from the Cooks looks to be a long way off. 

The Prime Minister confirmed on Monday that Cabinet had considered draft text being worked on via officials which will become the basis of an agreement for quarantine-free travel between the Cook Islands and New Zealand.

"That draft text is near conclusion," Ardern told reporters. "The next phase will be the verification phase; that is where we have officials on both sides who undertake work on the ground to assure ourselves that we're meeting the expectations on both sides."

Ardern said the Government's expectation is that there will be travel between officials undertaking that verification work within about the next 10 days.

She said the third phase will be the finalising of details. The Government will get advice form the Director-General of Health Ashley Bloomfield and the Government will then confirm with the Cook Islands administration if travel can go ahead.

"Whilst we haven't put a timeframe around the reopening at this stage, our expectation is that it would be in place before the end of the year," Ardern said.

"We haven't wanted to give dates beyond that while we're still undertaking that verification work and that's on both sides - both for the Cook Islands and for New Zealand."

It's important that we don't send Covid into the Cooks. And it's important that New Zealand help the Cooks ensure that their testing is up to scratch and that nothing's getting in via their ports.

But I don't quite get why NZ couldn't start admitting visitors from the Cooks now without isolation, with an explicit warning to all travellers that restrictions could be imposed at short notice should an outbreak happen in either place. 

Here's what MFAT has to say about the Cook Islands in general:

Cook Islands is located between French Polynesia and American Samoa. It is self-governing in ‘free association’ with New Zealand. That means that while it administers its own affairs, Cook Islanders are New Zealand citizens who are free to live and work here. More than 80,000 Cook Island Māori live in New Zealand.

There is no Covid in the Cooks. Those who live there are New Zealand citizens. If it's going to be months before travel normalisation can happen, how long might it be before a travel bubble with Taiwan is possible?

Way back in March, Balaji Srinivasan started talking about red zone / green zone options. Areas that have stamped out the virus could start linking up and re-normalising travel arrangements within that green zone, while implementing strict controls at borders to the red zone to keep it out. 

Green zones would never be set in stone but contingent: if situations changed, controls could be reimposed. If the Cooks started getting very sloppy at their maritime ports, New Zealand could help them tighten things up and could reimpose travel restrictions if it looked like cases could have gotten in. 

The green zone would expand to include other Covid-free places where we could trust the stats and trust in that border controls would keep things out. We could start admitting visitors from Taiwan on direct flights tomorrow without quarantine - the risk imposed by a traveller arriving in Auckland from Taiwan is about the same as the risk imposed by a traveller from Invercargill arriving in Auckland. 

Where it would go from there would have to depend on what the epidemiologists would say. Maybe travel to Tasmania could be reinstated - they have zero active cases.

Every country or region that could safely join the green zone would mean one small but important bit of freedom being restored. Border controls like New Zealand has now aren't good in their own right. They're terrible. They're only defensible because they avoid something even worse: a renewed outbreak. 

But when we're talking about travel between places that are safe, by people who have not been outside of the green zone, travel restrictions only do harm. Managed isolation requirements at best in those cases only increase the cost of travel. More likely, the scarcity of spaces in Managed Isolation and Quarantine means travel is effectively forbidden - despite that that travel is not risky. 

As I put it in this week's column in the Dom Post:

The system obviously is not working.

It has too little capacity and cannot find any reasonable way of reconciling the conflicting demands of tens of thousands of people desperate to get here.

It puts people into queues for managed isolation who, because they come from places without Covid like Taiwan or the Cook Islands, have no business taking up scarce spaces in those facilities. Isolation requirements for those travellers only do harm.

And safety protocols have been slipshod.

Managed isolation of those entering New Zealand from risky places is obviously critically important in keeping the country safe until effective treatment or a vaccine is available. Strict safety standards are essential.

But none of that requires the Government to be handling all of the bookings, or decide who should get which scarce spaces.

This isn't some dumb health vs economy thing. Preventing another outbreak is good for both. But there are no health benefits in preventing people from Taiwan from coming to New Zealand. What good is done by this?

I don't know whether there are any particular economic gains from re-establishing normal travel arrangements with Taiwan. But I do know that a traveller coming here on a direct flight from Taiwan imposes no harm, and makes at least one person better off.

Over at BusinessDesk, Rebecca Stevenson covers the broader picture on managing managed isolation. Her piece includes a few bits from me, as well as noting the support of former Prime Ministers Clark and Key for improvements. As for Rob Campbell's quoted comments on the topic in that piece, well, I think SkyCity's track record on Covid speaks for itself.

Bottom lines:
  • Regardless of the economic benefits of any "green zone" among safe places, there is a pressing humanitarian case to stop limiting travel by people who do not impose risk. 
  • Safe protocols at the border for those coming from risky places are really important. The government's system for allocating scarce spaces will not be adequate for the long haul ahead.

Monday 10 August 2020

Education departments are weird

So our Joel Hernandez has completed some more work on what's all going on in New Zealand's school system and an Auckland Uni education prof is mad about it

Oh well. 

Joel's long term project has been to look at differences in outcomes across students and schools, using the administrative data held in the StatsNZ data lab to adjust for a rather broad assortment of things that students bring with them into the classroom. 

Naive league tables will credit, or damn, schools for outcomes that are largely due to differences in the communities that those schools serve. Getting better measures on outcomes, adjusting for the differences across families that we can see in the data lab, helps. 

Our measures don't tell you what's going on in any particular school, but they do let you know whether a school is doing about as well as expected in the current system given the kids it teaches, or whether it's a place that the Education Review Office might want to go visit to see what's going on. It could be that better-than-expected performance in one school has nothing to do with that school's practices but instead has everything to do with an after-school tutoring club the parents set up - for example. 

Earlier, Joel looked at differences across schools to show that most of the difference in public school performance, by decile, disappears when you account for the differences in the families those schools serve. Piles of low-decile schools showed up as top performers, if you run the stats properly. 

The broader project, which will take some time because lab work is onerous and we only have Joel doing this work for us, will extend to a much broader set of outcomes going beyond NCEA. The Ministry of Ed has like 3000 staff; we have a bit over a dozen staff and we only have Joel in the lab. 

I'm keen to know how different schools vary in stuff like Not In Education, Employment or Training (NEET) status in years following high school completion; tertiary completion; salaries a few years after completing high school; crime rates; benefit uptake - there's a lot that can be looked at. But it'll take a while. We start with one set of data matches and build outward from there, adding things on as we go. 

Anyway, Joel's most recent project looks at whether there are differences in outcomes across state, state-integrated, and private schools. Not many kids go to private school in New Zealand, but private and integrated schools dominate the league tables for achieving University Entrance. 

Because of the cost of private schools, they're mostly going to be attended by kids from richer families. We're still looking at outcomes observable in school data available in the data lab. School data includes every student's performance on every NCEA standard they've sat, and whether they've achieved University Entrance. But it doesn't include data on whether they took up options available in some private schools to attend International Baccalaureate classes instead, or to take the Cambridge exams instead of NCEA. So Joel looked at UE as basis for comparison. 

And remember that the broader project will eventually get to a lot more outcomes. Those take time, and we have one econometrician on the job. 

Joel found that state-integrated schools outperformed state and private schools on University Entrance, adjusting for all the family background characteristics observable in the lab. You can't adjust for everything in the lab, but the stuff you can adjust for, like parents' education, will also be correlated with some of the things you can't observe. 

So, for example, the weight and value parents put on education can matter a lot, but you can't observe that in the data lab. If parents who put the highest value on education will both push their kids harder at home, helping them through, and be more likely to select out of public schools and choose an integrated or private school, then you could be unfairly crediting private schools for effects that come from family background. But, at the same time, if, on average, the parents who put the highest value on education also have high levels of education themselves, then you'll have mopped up some of the effects of "parents value education" by controlling for parents' own education. It isn't perfect, but so long as the unobservables correlate positively with the observables, then you've handled some of that selection issue. 

Or, at least, you've somewhat bounded it. Take a very different area: the persistent arguments about whether unobserved confounds drive the J-curve in alcohol and health. If adjusting for all of the observable health behaviours you can find doesn't do much to reduce the J-curve, and those observable health behaviours are real likely to be correlated with unobservable health behaviours, then it isn't plausible that unobserved confounds are driving the rest. Here, adjusting for the kitchen sink of family background reduced the coefficient on state-integrated schools but hardly got rid of it. You'd need the effects of the unobservables that aren't already mopped up by the observables to be as big as the effects of the observables, and to have driven the selection of private over state schools, to knock out the effect of private schools - and you'd need huge effects of unobservables to take out the effect of state-integrated schools. 

Anyway, here's Auckland Uni Prof of Ed Peter O'Connor on it all:

However University of Auckland Professor Peter O'Connor argues that the focus on UE results alone "reduces the complexity of learning to such a narrow construct that it becomes meaningless".

"As a professor of education, I'd fail my master's students on that. It's a false science," he said.

"There is nothing to suggest that going to a private school means you will be happier, lead a more purposeful life, contribute more to the world, have better relationships with your partner or your children. That in fact your life matters for having been lived.

"You might have better connections, even more money, but that isn't much in the grand scheme of life."

It's kinda funny. We never said anything about happiness, leading a more purposeful life, or any of that. We were just looking at the average effects of integrated and private schools as compared to state schools on university entrance - a metric a lot of people still do care about, and the one that it's possible to check in the lab. We will be broadening to more outcomes in future. 

But I guess if you're a student contemplating doing grad work in education that has any kind of econometrics in it, you probably shouldn't pick O'Connor as supervisor. If he doesn't like whatever numbers you get, he might fail you because he didn't understand the study or the methods. Michael Johnston over at Vic's education department would be a way better choice if you wanted to do it in an Ed department. 

Or, perhaps even better, do it in economics. 

Either way, you can even start with all the code Joel's used to run the data matches - we've got it all up in there freely available for anyone else to build on. There are years and years worth of studies to be done in there, and we've only got Joel. It's basically the best administrative dataset in the world for linking high school students' grades, their family backgrounds, and their later life outcomes. Drop us a line if you're considering picking up on any of this in your thesis work - we're always happy to provide a bit of advice. 

Friday 7 August 2020

Mapping Covid in NZ: Genome says?

Things I learned from what looks to be a superb study on Covid in New Zealand using genetic analysis of 56% of all confirmed cases:
  1. Only 19% of cases that came into New Zealand resulted in more than one additional person being infected while 24% led to a single additional infection - presumably policy substantially reduced transmission;
  2. Lockdown reduced R-naught of our biggest cluster from 7 to 0.2 within a week;
  3. The 649 cases analysed showed 277 separate introductions of the virus into New Zealand;
  4. There is no evidence of the virus circulating before the first reported case on 26 February;
  5. The largest cluster came from the US, and North America provided most of the cases resulting in transmission linkages.
I was very worried in February that it had already gotten here. Somehow, it hadn't. We've been very lucky. 

Oh, and obviously, lockdown worked. 

Things you wouldn't think need explaining, but somehow still do

The world's a puzzling place.

Maybe cognitive constraints bind a lot more tightly than I'd ever thought. 

The government runs New Zealand's managed isolation system for arrivals at the border. The Ministry of Health was making an awful mess of things, so the military took over parts of it. 

This week we learned that the government hasn't really been testing frontline isolation staff for Covid. They have an aspirational target of testing staff every two weeks, and do have more regular health checks for fever and the like. 

This seems like one of those things that anyone who's been paying the least bit of attention to the whole Covid thing might have already known, without a report. 

Just look at this. 
The Ministry of Health has seen the report and asked the authors to provide more details about the difference between a test every two weeks compared to once a week.

“You definitely get a lot of extra benefit from the weekly test as opposed to two weeks,” Hendy said. “The Ministry is certainly keen to understand the risks and how to manage it.”

Currently, there are tests available for workers who develop symptoms. Hendy concedes people shouldn’t be forced to take weekly swabs but strongly encouraged to do so.

“If those weekly tests are available then that drastically cuts the risk of them passing the disease undetected onto other people such as family members or others in the community,” Hendy said.

“It would mean we caught it early enough before it got passed on more widely.”
Just amazing. 

It's amazing that the Ministry has to be told that testing is a good idea.

It's amazing that the Ministry would assign ANY staff into these roles who would not be willing to undergo regular tests. It's stupid because the testing matters, and it's stupid because the kinds of people who would refuse to be tested are the last people you want anywhere near a freaking managed isolation system. What other corners would those Covidiots be happy to cut? Don't hire muppets in these roles! It's too important!

This thing is going to be around for a long time. Getting the border processes right matters. It's amazing that we haven't had an outbreak yet despite all this. 

Tuesday 4 August 2020

Innovative island nations

An intriguing proposal from an innovative island nation offering safer respite from the pandemic:
The government of a Caribbean island has a tantalizing suggestion for quarantine-weary Canadians: Working from home is a lot more palatable when you're doing it remotely from a tropical paradise.

The island nation of Barbados has launched something it's calling a Barbados Welcome Stamp, a one-year remote working visa that gives foreigners the right to live and work remotely in Barbados while they ride out the COVID-19 pandemic.

Starting now, applicants can send in their personal information at a portal website. The application will be processed within 72 hours, at which point they may be approved to come live and work remotely in Barbados.

There are a few stipulations, namely that you have to make $50,000 US a year and there's a non-refundable fee of $2,000 US for an individual and $3,000 US for families, but once that's paid, a successful applicant is all set.

"You don't need to work in Europe, or the U.S or Latin America if you can come here and work for a couple months at a time, go back and come back," Barbados Prime Minister Mia Amor Mottley said when she first suggested the idea earlier this month.
There would be a lot of interest in this kind of option in New Zealand as well, with arrivals covering their own costs of testing and of managed isolation. Because I've been rather active in this space, I get emails from folks abroad who'd be eager to join us. This one came in last week: 
Greetings Dr. Crampton,
I noticed your tweet last week, quoted here:
"It is really hard to overstate the potential gains if NZ can sort out scaling up managed isolation to enable some of these workers to bring their jobs with them to work remotely from here." -- @EricCrampton
My wife and I are former XXXXXX engineering Directors who just left XXXXXX to start a start-up. We're US-born, living in Northern California. We realized it would be great to work in NZ for a year or a few years, not just because of covid, but because the US political situation is not great. In the time of pre-Covid NZ immigration, it looks like this was easy, and we probably would already be in NZ on a short-term visa, and would be applying for an entrepreneur's visa, with the idea of staffing up starting in NZ.
But of course there is Covid. My best understanding of the NZ gov web sites is that there are no channels open to us now,
Can you share any advice about how we should best proceed?  Of course, if there's someone better for us to talk to or work with (whether in the domain of gov, org, or commercial expediters), please let me know.
I advised my correspondent that nothing here is likely to change this side of the election, so they might either wait, or try a more innovative island nation like Barbados instead. They've said they're waiting. 

Safely scaling up managed isolation matters. 

I cover this stuff in this week's column over at Newsroom - currently gated, but usually comes ungated later in the week. A snippet:
Effective capacity in managed isolation has increased to just over 14,000 arrivals per month. While that sounds like a lot, the average month in 2019 saw over 250,000 Kiwis returning home after business trips, foreign study, holidays, or visits with friends and family. Non-resident Kiwis returning home from abroad for the longer term added about another 1,750 per month.

There will not be a lot of Kiwis keen on travelling to the Covid-ridden parts of the world, but the longer the pandemic lasts, the harder it will be to continue to defer travel. Even if Kiwis cut their travel to a quarter of what it was before Covid, they would still take up more than four times as much room as is available in the managed isolation system. Add to that tally the Kiwis abroad who would also like to come home, as well as the overseas specialists necessary in a wide range of business and infrastructure projects, and the need to safely scale up managed isolation becomes rather obvious.

If the Government expected vaccines or effective treatment to be just around the corner, maintaining the system as it is could be defensible. People can usually defer travel for a few months, barring emergency cases. Holidays to visit family and friends abroad can be delayed. Big trips abroad are once-in-a-lifetime events for a lot of us and putting them off for a year might not matter so much – and especially when going abroad is particularly unappealing. Business trips can be delayed, with video chats taking their place in the short term.

But the longer this all lasts, the harder it is to defer travel.

The odds of family emergencies abroad get higher over longer periods. 1.2 million Kiwis were born overseas. If ten percent of them have a family emergency in any given year requiring a trip abroad, that’s 10,000 spaces in managed isolation per month as they return home.

The costs of forgoing business travel increase as opportunities deferred become lost contacts and contracts – over 32,000 Kiwis returned from business trips abroad every month, before Covid. And companies here needing foreign experts can only defer those arrivals for so long before costs start rapidly escalating.

None of this is any argument for prioritising ‘the economy’ or business over health. Any outbreak here resulting in another lockdown would be economically devastating. Rather, it is an argument for building the systems and infrastructure necessary to be able to safely accommodate far more travellers than the system can currently handle.

By these numbers, scaling up is critical even if we consider only the needs of Kiwis. If we allow ourselves to think a bit more broadly, it becomes even more important.

New Zealand’s success in managing Covid makes the country a very attractive proposition. Students who would have studied in America, but who do not like the prospect of lectures via Zoom, could find studying here to be a very attractive alternative.

And many abroad, working remotely due to the pandemic, could bring their jobs with them to work remotely from here instead. As they would continue to be paid by their overseas employers, their work in New Zealand would count as the export of a service while they spent their earnings, and paid taxes, here. Other countries rightly see this opportunity: last week, Barbados began offering a one-year remote working visa encouraging people to bring their jobs with them to their island in the Caribbean.

The system has to change.