Wednesday, 1 March 2017

Swimming standards

I love Thomas Lumley's explanation of the new river water quality standards.

Suppose you wanted a target for river water quality so that people didn't get sick while swimming. River catchments vary but so too will water quality in any given river depending on whether there's runoff from recent rain.

The standard they're settling on is that the risk involved in swimming in a swimmable river should be no more than one chance in twenty of getting sick, 95% of the time that you might swim in that river.
Suppose we imagine a slightly implausible extreme sports facility that sends 100 backpackers on one-day swimming parties each day. On 95% of days (347 days per year), they’d expect fewer than 5 to get infected. On 5% of days (18 days per year) they’d expect more than 5 to get infected, but it couldn’t possibly be more than 100. So the total number of infections across the year is less than 5*347+100*18, or 10% of swimmers. That sounds bad, but it’s an extremely conservative upper bound. In fact, when the risk is less than 5% it’s often much less, and when it’s greater than 5% it’s usually nowhere near 100%. To say more, though, you’d need to know more about how the risk varies over time.

There are statistical models for all of this, and since everyone seems to be using the same models we can just stipulate that they’re reasonable. The detailed report is here (PDF), and Jonathan Marshall, who’s a statistician who knows about this sort of thing, has scripts to reproduce some calculations here.

Using those models, a `yellow’ river, with risk less than 1/20 95% of the time actually has risk less than 1/1000 about half the time, but occasionally has risks well over 10%. Our imaginary extreme sports facility will have about 3 infections per 100 customers, averaged over the year. About half these infections will happen on the worst 5% of days.

So, the 1/20 of 1/20 level doesn’t by itself guarantee anything better than 10% infection risk for people swimming on randomly chosen days, but combined with knowledge of the actual bacteria distribution in NZ rivers, seems to work out at about a 3% risk averaged over all days. Also, if you can detect and avoid the worst few days each year, your risk will be reduced quite a lot.
There's been the usual Twitter snark about how a restaurant that made one in twenty people sick would be shut down. Better to think of it this way: imagine that the restaurant runs every day, but sometimes the power goes out and some of the food in the fridge goes bad. One in twenty, over the course of the year, could be almost entirely on those bad days. Isn't it better then just to put a sign on the door saying "Sorry, the power was out, we're closed today"?

Tuesday, 28 February 2017

Reading Creedy

John Creedy is really good at using complicated maths to make simple points. I'll summarise the simple points in Creedy's working paper on sugar taxes, issued earlier this month.

Section 2.1 shows that, whenever people enjoy a bundle of goods of various healthiness, and whenever people are likely to shift from one good to another if prices change, any tax on a particular unhealthy good might reduce consumption of that good while increasing consumption of other less healthy goods. Consequently, you can't just say that a soda tax would improve health - you need to show what the effects are across a broader set of consumption goods. At minimum the you'll get less than you'd hoped for in terms of effects; it's even possible to wind up getting worse health outcomes.

Sections 2.2 and 2.3 work through the maths to get an equation specifying the necessary conditions for a tax on one calorie-heavy food to yield an increase in total calories consumed. That's equation 16.

Walk it through. Food 1 is taxed, Food 2 isn't taxed. An increase in the price of Food 1, presumably through a soda tax, will increase total caloric intake whenever the inequality in Equation 16 holds. It isn't all that likely to hold, but all of these things would also limit the effectiveness of any tax in reducing calorie intake even if we don't get a full reversal from the intended effect.

The first term in the equation is the cross-price elasticity of Food 2 with respect to Food 1. How much does consumption of Food 2 go up if the price of Food 1 increases? So imagine Food 1 is soda with a soda tax and Food 2 is chocolate bars. If the price of soda goes up by 20%*, by how much does consumption of chocolate bars go up? Suppose that chocolate consumption goes up by 5% with a 10% increase in the price of soda. That cross-price elasticity would then be 0.5.

On the other side of the inequality we have another set of terms.

The first of those is the absolute value of soda's own-price elasticity. Why absolute value? Because own-price elasticity is negative (consumption drops when price goes up).

The second is the ratio of the calories in soda as compared to chocolate.

The third is the ratio of budget shares spent on soda as compared to chocolate.

And, finally, the last is the price of chocolate divided by the price of soda.

So what's needed for the soda tax to actually increase calorie intake?

  • The cheaper chocolate is compared to soda, the more likely we get an increase in calorie intake;
  • The more people spend on chocolate as compared to soda at the outset, the more likely that the soda tax has perverse effects;
  • The more calories per unit in chocolate bars as compared to soda, the more likely we are to have perverse effects.
And even if you don't wind up in the case where the soda tax increases calorie intake, you still get less calorie reduction than you'd have expected whenever the three points above are more important. 

I love John's work, but it is pretty tough to read through it unless you know the maths. I don't think he deliberately writes these things to exclude people who can't understand the maths, but it's pretty hard for anybody who doesn't follow the maths to read. 

And so it was ... frankly bizarre ... to read this from the Public Health Blog people:
The Government has an action plan to tackle childhood obesity, but it lacks a tax on sugary drinks – a strategy for which there is good evidence.  A new Treasury Report on soft drink tax price elasticities has just emerged. It has the look of a strategically published document that if and when – during election year – certain politicians need to defend non-action on taxing sugary drinks, they can point to this Report and obfuscate.  Indeed, this New Zealand Treasury Report has already been used for this purpose in Australia. We critique this Report in this blog, with a view to preventing its misrepresentation and to encourage a more informed discussion on taxing sugary drinks. [emphasis added]
If the government wanted to push a strategic anti-sugar-tax report in an election year, it wouldn't have had John Creedy write it. Like, just look at this summary he provides of his illustrative example.

And do note that the text below the graph is about as layperson-friendly as John gets. Does this look like a political document? I wonder what has to be going on in your head if you think that John Creedy is part of some big government conspiracy to write heavily mathy things that maybe one in a thousand people can read.

The public health people then use John's equations to show that it isn't particularly likely that we get an increase in total calorie consumption with a soda tax, but that's not the main point in there. The main point in there is how the effectiveness of any soda tax drops with the combination of things listed in my three bullet point summary.

And as for any alternative theory that would have "Oh, they're writing this up so that people like Crampton can use it in arguing against sugar taxes", well, I didn't even see the point in blogging this piece until they mentioned it. It seemed way too impenetrable even to try blogging about. I'd be a bit surprised if anybody were even still reading this post by now after the maths above.

I did tweet about the report when it came out, mostly because I was exceptionally proud that The Initiative's Jenesa Jeram's work on soda taxes was cited by one of New Zealand's gods of empirical public finance (see footnote 2!).

The other fun bit in the public health people's blog post: they have a big summary of all the not-economists who agree with them about taxes, but continue to ignore Waikato University Professor of Economics John Gibson's work.

But that doesn't mean you have to ignore it. He's giving a talk in Auckland on it that I'd be attending if I were in Auckland. He shows how the estimates of the effects of soda taxes in Mexico are overestimated because consumers shifted to cheaper sodas and, consequently, reduced their soda consumption by maybe a third as much as is often suggested.

And note that John's work on soda taxes is funded through the Royal Society's Marsden Fund, so it's harder for the public health types to simply dismiss on the basis of its funding.

Update: On further thought, it seems a bit odd that Boyd Swinburn is one of the authors on here. The first paragraph implies that John Creedy's work was politically swayed. For a guy who is suing Whaleoil for defamation...

* Note: only an idiot would think that this does not apply to excise just because I used the percent sign and excise is done per unit of whatever rather than in percent terms. Elasticities are just done in terms of percentages. You can turn any excise into a percentage change around different price thresholds to get point elasticities. The effect of any per-unit excise depends on price elasticity. You can probably judge whether somebody gets to count as an economist based on whether they get this. Getting it isn't sufficient for being an economist, but it's necessary.

Harder than it needs to be

The Niskanen Centre shows that immigrants in the United States make less use of government-funded welfare programmes.

It's a nice piece that quickly debunks a lot of popular myths about immigrants.

But it makes me weep how much harder that work is to do here than it is in the USA.

If I want to know, in the US, about whether foreign-born people have higher welfare uptake rates than domestic, all I need to do is go to IPUMS, go to the Current Population Survey, select the variables I want and the years I want extracted, hit the extract request button, and wait for the email saying that the subsample has been extracted for me.

Or, even more easily, I can use their online data analysis tool. And so I did. I'd registered previously with IPUMS. I had to fill in another form to get access to CPS data, which I hadn't used there before. The form auto-populated with the stuff they had on me already. I told them why I wanted the data, then clicked over into the SDA interface for the CPS. Here's what the interface looks like: it's the basic SDA interface Berkeley has that you'll be familiar with from the US GSS - which anybody can use without any login, straight from the website (here you need multi-week approval processes not just to get the data, but for any subsequent use of the data which is different from the purpose which you specified when you got the data in the first place).

If you care about open data, head over to IPUMS to see what is possible. Or even just navigate the SDA interface for the American General Social Survey.

That same work in New Zealand wouldn't be easy. The closest you could get is income source by ethnicity in the 2013 Census publicly available files. I think. Linking that to country of birth would be a specialised data request to Stats, or a trip to the data lab. There is a Confidentialised Unit Record File for the Census - similar to the PUMS available in the US on ACS data. But where anybody with a browser can download the American 5% or 1% samples, you have to make application through the data lab for access to the confidentialised unit record files.

Today StatsNZ put up some new work showing that the wage gap between childless men and women is tiny, while the wage gap between fathers and mothers is larger. It's nice that they've thought to check this, but the damned problem with StatsNZ is that their whole setup requires a stats analyst to think "Ok, maybe people would also be interested in this way of cross-tabulating the data." If you want to know something they haven't added, it's a multi-week application process for access to New Zealand's General Social Survey, for example. And running the HLFS data to adjust for whether the respondent has had a child - that would likely be a trip to the data lab. Meanwhile, American data is just a mouse-click away.

Monday, 27 February 2017

A stupid Newshub beat-up [updated]

Newshub today helped make Kiwis just a little bit stupider. But Ministers not knowing the underlying stats didn't help. [See update below though!]

To recap. The government was put on the spot about whether they're rorting tourism numbers. MSD will sometimes put people in hotels or motels as temporary emergency accommodation. Whether that happens too often relative to an ideal is a different question we'll leave to the side for now. Question at hand is whether that's inflating the tourism numbers. 

Tourism Minister Paula Bennett was asked whether the tourism numbers were wrong because of this. 

The correct answer is "MSD clients are a tiny fraction of overall hotel nights, so it really cannot affect the figures either way." 

Minister Bennett clearly didn't know what's going on in the underlying stats because she said that they aren't included because they're not tourists. Hotels don't know why guests are spending the night. They just report up to Stats how many nights they've provided. [Update - see below] Other non-tourists included in the figures:
  • A couple getting a room for a discreet encounter, who aren't tourists;
  • Someone who realises he is in no shape to drive home and would rather spend the night in the hotel rather than go home drunk in a cab;
  • Someone taking a night at a hotel after a row at home;
  • Someone renting a room as a meeting space;
  • Someone staying in a hotel room during some renovations, or before taking possession of a place they've just bought.
None of it matters. Why? There are almost 22 million domestic guest-nights per year in New Zealand hotels, and over 15 million international guest-nights. How do we know this? The tourism satellite accounts. Here's Table 8.

Neither the international guest nights nor the growth in international guest nights is likely to have been affected at all by MSD clients; they wouldn't have been reported as domestic visitors. It is unlikely that MSD clients have any material effect on the overall domestic guest nights either - it would be like thinking the water volume of Lake Taupo is overstated because nobody netted out the mass of fish in the lake. Yeah, there's fish, but it won't make much difference to the overall figures.

How much effect could it have had? The Newshub story reports 8,860 emergency housing grants in the last quarter of last year at a cost of $7.7 million. Let's say that those are all hotel room nights. Since they're emergency nights, they're not going to be getting "book ahead and save" rates. And they're also potentially riskier for the motellier. Let's say that the room rate is $100 per night but I'd think I'm erring on the low side there. That's (top end) then about 77,000 nights in that quarter. If the room rate is $200/night, then it's 38,500 nights. 

If that had persisted for the whole year, the total number of guest nights would still have rounded to 22 million - but the measured growth rate would have been a bit lower. But, again, would it matter? The government crows about international tourist numbers and guest-nights. Domestic doesn't get noticed as much. 

Prime Minister English noted "if they're counting them as tourists, they shouldn't be." It maybe wouldn't be that hard for MSD to tell Stats how many nights they've purchased and then have those netted from the tourism satellite accounts, but it's stupid hassle for no particularly good reason. And unless they do it all the years back, they're going to break the continuous data definition. 

So, some bottom lines:
  • There is a housing crisis;
  • The government is not fudging the tourism stats by including MSD clients in the tourism satellite accounts, and neither is Stats NZ;
  • It is stupid, and damaging, and unethical, to undermine trust in official statistics in this kind of Gotcha! attack on Ministers who cannot reasonably be expected to know what's in the definition of particular stats - and especially where it is inconceivable that whether or not it is included it would make a whit of difference to the measured tourist night numbers. 
  • I hope that the Statistics Minister, on advice from Stats NZ, would also have told Newshub that the 22 million nights context means that this would just be rounding-error stuff anyway. If he did, and Newshub didn't report that part, that would be worse for them. 
  • If we ever get to the point where MSD emergency grants could materially affect the domestic accommodation guest-night figures, we're going to need a bigger word than crisis to describe what's going on in New Zealand's housing situation.
UPDATE: MBIE's tourism estimates, like the monthly regional tourism estimates, don't use the accommodation survey figures anyway. So if they did use them, it wouldn't matter because the numbers are tiny. But they don't. In this evening's reader mailbag (haven't seen a source link yet):

The Accommodation Survey is produced by Statistics New Zealand monthly to provide information on short-term commercial accommodation activity at a regional and national level. This includes all people staying in commercial accommodation, not only tourists.
Domestic tourism is currently measured by visitor spending, which is not informed by the Accommodation Survey.
MBIE does not use the Accommodation Survey to produce key tourism data products, such as:
  • Monthly Regional Tourism Estimates
  • International Visitor Survey
  • New Zealand Tourism Forecasts.
People with emergency housing grants are not included in these statistics.
However, while it does not inform these products and measures, the Accommodation Survey is part of a suite of statistics that we use to understand the tourism market, both domestic and international. People with emergency housing grants make up a tiny percentage of the approximately 38 million visitor nights recorded annually in the Accommodation Survey.

Wednesday, 22 February 2017

Diversification as insurance: bolthole edition

Idealog asked me last week whether it made sense for billionaires to see New Zealand as a bolt-hole against apocalypse. I had a lot of fun taking a punt at an answer.

New Zealand would do better than other places in some doomsday scenarios, but hardly in all of them.
But Dr Eric Crampton, an economist and the head of research at the New Zealand Initiative, says the billionaires prepping for doomsday may have it wrong.

“For many end-of-the-world scenarios, the best bolthole isn’t the most isolated place in the world, but the wealthiest place in the world,” Crampton says.

“Scenarios leading to collapse in international trade are worse for small countries that are trade dependent than for very large countries that have large internal free-trading areas.”

But he can understand why US citizens would prefer to seek refuge on the other side of the world, particularly in the current political climate.

Crampton is a Canadian that moved to New Zealand from the States in 2003 to take a job at Canterbury University, and he’s stuck around ever since.

“Arriving here, it felt like the outside of the asylum. America was growing increasingly mad. Airports were armed camps. The police seemed to have a very hostile relationship with the public and were heavily armed,” he says.
There's a lot more at the link, including zombies. Enjoy! Bottom line, best billionaire strategy likely has a portfolio of boltholes against a range of potential disasters, and New Zealand could be a nice part of that bundle.

The costs of golf

Auckland Council owned thirteen golf courses as of last year. The NBR talks with Julie Anne Genter and Jacinda Ardern, the two candidates for the Mount Albert by-election, about whether Council should turn the Mount Albert golf course into housing.

I don't know, but strongly suspect, that that land would be of much higher value in housing. Julie Genter is almost certainly correct that the place should be in housing.

But here's how you'd find out.

  1. Zone the land for the highest density of housing that's possible given the infrastructure around the place (or higher, if development could fund upgrading the infrastructure).
  2. Put in an SHA allowing fast development for any future owner. 
  3. Sell the land to the highest bidder, letting the owner keep it as a golf course or develop it.
  4. Tax the land on its market value, whether it's used as a golf course or as housing.
If the golf course can turn a bigger profit from greens fees from paying customers than a developer could earn by turning it into housing, then it should be a golf course. Otherwise, it shouldn't. 

The NBR quotes the golf club's treasurer on that the club currently turns a profit. I wonder what Council charges itself as rent on land that would be worth a fair bit in alternative uses. 

Tuesday, 21 February 2017

Picking zones and picking winners

The push for more localist approaches to policy problems in New Zealand continues to gather steam.

Earlier this month, the McGuinness Institute argued for what they're calling Demarcation Zones for policy trials. Their formulation differs a bit from what we at the Initiative proposed in 2015, but the core idea is similar: let local communities take on additional devolved powers and see what policy variants seem to work better in which places.

I go through some of the differences in last week's print-edition NBR column. Here's a snippet of the piece:
Trialling policy at a local level can make a lot of sense. Not only does it let policy be more sensitive to local conditions, it also helps central government figure out what kinds of policies might work in a broader rollout. But it will always be tempting for central government to use local zones as a way of funnelling benefits to electorally sensitive areas instead. Just imagine what could have been in a Northland special economic zone proposal at the time of 2015’s by-election.

We worried about this problem in developing the Initiative’s Special Economic Zones proposal in 2015. The worst thing that can happen with a policy proposal is not that it is ignored but that it turns into a nightmarish hall-of-mirrors distortion of the original vision.

Dangerous temptations

Even leaving aside crass electorally driven measures, being able to regionally target policies could provide other dangerous temptations to a micromanaging central government. Imagine a tax concession zone for a three-block area in downtown Wellington to boost the IT industry. Or a zone with special depreciation rules to help ensure an irrigation project makes it over the line. Good policy provides a general framework that lets winners emerge, rather than picking them at the outset.

But there is a way of telling whether a proposed zone is a policy trial or something less laudable. Policy trials, if they are successful, should be able to be rolled more broadly. But projects that channel benefits to electorally sensitive areas or that try to pick winners cannot be.

And so we argued that policy measures shown to be successful in any area should be available to any other region wanting the same treatment. An inefficient tax concession zone to boost employment in one region would not put that big a dent in the government’s purse. Requiring that the same concession be available to all regions would make the policy too expensive to contemplate – unless it really were potentially to the broader good.

Localist approaches hold a lot of promise. Central government should welcome proposals that have the demonstrated support of local communities, which have taken programme evaluation seriously, and that could be taken up by other towns. Baking this kind of accountability into proposals helps give central government, and the rest of us, confidence that local government is ready to take up the challenge.
You should subscribe to the NBR. But we do have an ungated version here.