Wednesday, 29 March 2017

Publication bias and the file drawer

Economists have known about file drawer problems for rather a while; DeLong and Lang was on the syllabus when I was in grad school in the late '90s. 

My Canterbury colleague Andrea Menclova's wished to do something about it. If papers with insignificant results get put in file drawers for want of suitable publication outlets, why not start a journal of insignificant results? She's been thinking about this one for a while; it was a not infrequent lunchroom conversation topic.

She's now blogged on it, which I'm taking as an optimistic sign that her proposed journal might actually get going. She writes:
Series of Unsurprising Results in Economics (SURE)Is the topic of your paper interesting, your analysis carefully done, but your results are not “sexy”? If so, please consider submitting your paper to SURE. An e-journal of high-quality research with “unsurprising” findings.
How does it work:
  • We accept papers from all fields of Economics…
  • Which have been rejected at a journal indexed in EconLit…
  • With the ONLY important reason being that their results are statistically insignificant or otherwise “unsurprising”.
To document that your paper meets the above eligibility criteria, please send us all referee reports and letters from the editor from the journal where your paper has been rejected.  Two independent referees will read these reports along with your paper and evaluate whether they indicate that:
  1. the paper is of high quality and
  2. the only important reason for rejection was the insignificant/unsurprising nature of the result
Submission implies that you (the authors) give permission to the SURE editor to contact the editor of the rejecting journal regarding your manuscript.
SURE benefits writers by:
  • Providing an outlet for interesting, high-quality, but “risky” (in terms of uncertain results) research projects;
  • Decreasing incentives to data-mine, change theories and hypotheses ex post, exclusively focus on provocative topics.
SURE benefits readers by:
  • Mitigating the publication bias and thus complementing other journals in an effort to provide a complete account of the state of affairs;
  • Serving as a repository of potential (and tentative) “dead ends” in Economics research.
Feedback is definitely invited! Please submit your comments here or email me at

Monday, 27 March 2017

Water pricing

And here we go for another edition of "Because you've misspecified the problem, your proposed solution might make things worse."

The past couple of weeks have had renewed anger about water bottling plants. If you buy land with water drawing rights attached to it, you don't get charged for drawing that water. The value of the water is baked into the selling price of the land. And so whoever owns the land with the drawing rights gets to appropriate any unexpected increase in the value of water drawing rights, and anybody buying land at market with those rights would, expectationally, only wind up earning a normal rate of return on that investment.

So far so good. But when people buy land to irrigate paddocks, run cows, milk the cows, dry the water out of the milk, and export the milk powder to China, nobody gets too upset. When they just bottle the water directly and export it to China, nationalists get mad and anti-corporate people see it as a big giveaway.

Labour's proposing charging water exporters. I expect it's feasible to do it, or at least for new plants. But the underlying problems are all still there: nobody really knows whether the highest valued use of water is in spraying it onto paddocks, or in bottling it for export, or leaving it in the river. If you put a charge on one type of water drawing, but not on the others, and if it turns out that the water really is more valuable being bottled directly, then you could be doing harm.

So the underlying problem is getting the right allocation of water across potential users, within constraints imposed by existing use rights and likely iwi ownership claims if drawing rights ever turned into more formal property rights.

Charging water exporters for drawing solves a political problem, but otherwise doesn't do much good. Land with water drawing rights already attracts a premium reflecting the value of the water. Any charge on bottling plants then requires that bottlers put a higher value on water than do irrigators: the bottler not only has to outbid an irrigator for the land, but also has to pay some premium on top. Since there aren't a ton of these things showing up, the value of water to bottlers can't be *that* much higher than the value to irrigators.

What's a potential solution then?

Fritz Raffensperger proposed a pretty innovative water trading system for the Canterbury plains, where the underlying hydrology was mapped well enough. It recognised the flow-on effects of drawing from a pile of different node-points on a pile of other node-points, and set node-pricing accordingly. Then, holders of drawing rights could trade: buying the right to draw more, or selling back their existing allocations, with water flowing to its highest valued use.

But what about the environment and rivers? His system allowed a minimum flow constraint that would automatically scale back drawing rights in dry years to maintain a minimum flow rate.

So how do you manage the politics around all of it and incipient iwi claims? I don't know if the sketch below would work, but I think it's a starting point.

First, define the minimum flow of a river for it still to really be that river. Maybe the flow rate that's at, say, the 25th percentile of unencumbered flow in a typical year. I don't know what the right number is there, but there will be one. That can be the minimum flow rate in the Raffensperger trading setup.

Next declare that the river owns itself, but that the local iwi is its guardian and acts on its behalf. Grant the iwi rights to another margin on top of the river's minimum flow rate. They can leave it in the river, or they can sell it to other users; they could also buy back drawing rights from other users to achieve higher river flow. And so too could folks like Fish & Game for rivers where they'd want higher flow for trout fishing.

Finally, existing drawing rights have to be respected. If the catchment is already over-allocated relative to sustainable extraction rates, then you have to scale things. We already have that kind of setup in the fishing quota management system where quota landings are a fraction of the year's declared total allowable catch, with the TAC varying as sustainable harvest rates vary. The government might have to buy back drawing rights if the catchment is over-allocated once the minimum river flow and iwi allotments are sorted. On the other side, though, the government would earn money through selling drawing rights in under-allocated catchments.

Then run the Raffensperger water trading model. There would need to be a bit of fiddling because irrigation has some water flow back into the rivers/aquifer and bottling plants don't, but it could be sorted. The government would buy back drawing rights through that setup to get over-allocated catchments into line.

That's just a first cut sketch at a solution: it would need a fair bit of working up yet. And there are plausibly other better solutions. But this one at least is trying to get at the real underlying problems. And of course it would all be better if partnered with a market in effluent built on the kind of model Taupo's running for nutrient management.

  • Respects existing use rights;
  • New drawing requires buying the right to do so;
  • Water flows to its most highly valued use.
Disadvantages and things to be worked out:
  • Setting initial minimum flow constraints would be very contentious and risk locking in too high a level (note that it's easier to buy water back into the river if the flow constraint is too low than it is to get agreement to ease a too-high constraint later).
  • The politics could still be a disaster. I've specified here what seems to me a reasonable outcome of a negotiation process, but iwi would likely open by claiming that all water is theirs and the Crown claiming none of it is and farmers objecting to minimum flow constraints in dry years. And assigning ownership of residual drawing rights in under-allocated catchments to Council, iwi, or Crown would be contentious. As an economist, I just invoke Coase and note that I don't care who owns it so long as they can then trade it. Maybe they could each get some of the unallocated parts. 
  • How to handle dry years is still a problem. If you set the drawing rights like fishing quota, as part-shares of the available water that year, then farmers needing set water volumes for stock on hand would face very high costs in buying that water - while the river maintained normal flow. If it's instead a quantity right that doesn't scale, the government would need to buy back allocation to maintain minimum river flow - and that would be expensive. Ultimately the thing likely needs to scale the minimum flow constraint to the shadow price that drops out of the optimisation. 

Wednesday, 22 March 2017

Economic Impact Assessments and the value of fishing

The New Zealand Marine Research Foundation’s commissioned economic impact assessment of recreational fishing in New Zealand says that fishing generates $1.7b in economic activity. Legasea argues “We now have a handle on the value of recreational fishing” and that the results provide “enough evidence to support a recalibration” from commercial fishing to high-value recreational fishing.

But there are two problems. First, even if the economic impact assessment figures were correct, that still provides no basis for deciding whether the next fish caught would be more valuable in a commercial boat or on a recreational hook. And, second, the numbers come from an economic impact assessment. Let’s deal with the second one first.
Enjoy! I also got to quote from one of my favourite Stephen Gordon tweets.

Our Research Fellow Dr Randall Bess provided the numbers I used at the end of the piece. Keep an eye out for his coming report on recreational fishing....

Age-restricted fast foods?

Doug Sellman's latest prescription for reducing obesity:
Does the fast food industry need to be regulated?

Prof Sellman believes only New Zealand's law makers can forcibly change and control the booming fast food industry.

"Government regulation is the answer. It is the tobacco and alcohol industry story all over again.

"The price of freedom from government regulation is enslavement of large numbers of the population to these profit, rather than health-driven industries."

But Prof Sellman isn't optimistic the Government will do anything to effect meaningful change in the fast food industry.

"Unfortunately, we live in intense neo-liberal economic times when public health is given less value by governments to the GDP contribution of big business, while the harms are relatively discounted."

What can be done to combat the booming fast food industry?

Prof Sellman suggests the following measures:
  • Dismantle the marketing
  • Increase the price
  • Reduce accessibility (density of outlets and hours of sale)
  • Increase the age of purchase
  • More incentives for people to leave their cars at home
Yup. Total neoliberal conspiracy that we don't have minimum purchase ages for food.

But remember, there are no slippery slopes from tobacco regulation to every other damned thing that Doug Sellman doesn't like.

Tuesday, 21 March 2017

Myopia and Discounting

Gotta love any technical paper that opens with a Böhm Bawerk cite. 

Gabaix and Laibson have a new framework up in which patient Bayesian imperfectly informed agents display behaviour observationally equivalent to hyperbolic discounting.

The intuition of the model is pretty simple. Agents get noisy signals about the future state of the world, and so there's option value in deferring some decisions until you get more certainty. You consequently get things that look like preference reversals, but they're really just information updating by patient agents.

Their summary here is rather nice:
We provide an illustrative example of our framework in Section 2, where we study a binary choice problem: an actor chooses between an early reward and a mutually exclusive later reward. We show that when the variance of forecasting noise rises linearly with the event horizon, Bayesian agents will act as if they are hyperbolic discounters, even though their deep time preferences are perfectly patient.

In Section 3, we describe the broader implications of our framework, and identify predictions that distinguish our framework from time preference models. First, we show that our (perfectly patient) agents exhibit preference reversals of the same kind that are exhibited by agents with hyperbolic discount functions. However, these preference reversals do not reáect a self-control problem. The preference reversals arise because the agents obtain less noisy information with the passage of time. Accordingly, our agents do not wish to commit themselves; they act as-if they are naive hyperbolic discounters (Strotz 1957, Akerlof 1992, OíDonoghue and Rabin 1999) rather than sophisticated ones (Laibson 1997).

In the cross-section, our framework implies that agents with greater intelligence exhibit less as-if discounting - their superior forecasting ability enables them to make choices that are more responsive to future utility flows.

In addition, our agents exhibit as-if discounting that is domain specific. They exhibit less as-if discounting (i) when they have more overall life experience, (ii) when they are more experienced in the specific choice domain, (iii) when they have more time to think about an intertemporal choice (e.g., Imas, Khun, and Mironova, 2016), and (iv) when they have more cognitive bandwidth to think about their choice (e.g., Benjamin and Shapiro, 2015).

In Section 4, we generalize our example by making the action set continuous. We provide sufficient conditions that imply that perfectly patient agents who are imperfect forecasters will act as if they are naive hyperbolic discounters.
And the Böhm Bawerk quip:
Diminishing sensitivity to future utils is also explained by imperfect information. For example, Böhm-Bawerk (1889) wrote that "we possess inadequate power to imagine and to abstract, or that we are not willing to put forth the necessary effort, but in any event we limn a more or less incomplete picture of our future wants and especially of the remotely distant ones. And then, there are all of those wants that never come to mind at all."
Hyperbolicy behaviour is a reasonably common justification for behavioural economics type interventions. Where the problem is information rather than self-control, providing information may well be the better solution.

Getting out of transitional gains traps

The only way I'd ever seen of getting out of transitional gains traps efficiently was by compensating the losers by taxing the winners.

Is that what's going on in Australia with Uber?

Let's recap.

Recall that Tullock's transitional gains trap obtains whenever the excess profit from a regulatory rent gets capitalised into the price of the asset in regulatory fixed supply. In the taxicab case, that's the medallion that gives you the right to drive a cab in supply-regulated markets. Once that capitalisation happens, the owners of the asset will have enjoyed a capital gain, but future buyers only should earn normal rates of return on investment. And there's the trap: the rule no longer conveys excess profits to anybody, but any changes will be fought because they'd impose capital losses.

Tullock thought that Pareto solutions were impossible in that kind of case, but I proposed something close to one anyway. Buy out existing licence holders at the value of their permits and abolish the permit regime (obviously use a price from before wind of the announcement got out). Issue bonds to cover the buy-out cost. Implement a tax on taxicab usage that pays off the bonds, and retire the tax when the bonds are paid off.

Advantage: the winners compensate the losers, and we get to move to the more efficient state of the world. It is not a Pareto move, because plenty of owners would prefer not to have had their permits taken at yesterday's price, but it's not far.

Now what does all of that miss? Technological change always affects equilibrium regulatory outcomes. That's the standard Peltzman work on regulation. We haven't had disruptive technological change in Canadian dairy as yet (another great transitional gains trap), but we have in taxicabs.

Uber makes maintaining the taxicab cartel more expensive. Cartel enforcement requires political will, and Uber makes the costs of the cartel more obvious to voters. And Uber also enables cartel driver defection: they can drive Uber on the side. Equilibrium stringency then should fall.

And we've seen that in New York. Medallion prices are way down. I used to lecture on this stuff, and noted the successes of Medallion Financial, a specialised lender that provided capital for folks to buy taxicab medallions. Their website once bragged* about how medallions provided above market returns for decades and were highly secure. Here's their stock ticker:

Meanwhile, the price of medallions has dropped from about a million dollars in 2014 down to $250,000 again.

So what does this have to do with paying off the losers? The main point of paying off the losers to get out of a transitional gains trap is to enable the switch to the more efficient outcome. If you're going to get to the more efficient outcome regardless, then paying the medallion owners is just a transfer that might have potential equity justification. 

And so we come to Australia, where they're looking to tax Uber riders to compensate owners of cab licences. 

If the change were going to happen anyway, should the existing cab owners be paid? I hadn't before caught this excellent piece by Richard Holden

He works out which licence holders already earned back the price of their licences on a normal rate of return and argues there's no need to compensate those who've already earned back the value of their licenses. Most require no bailout on those grounds alone. Excess returns on these things are due to their inherent political risk anyway. Licence-holders in New South Wales who bought prior to 2012 had already earned back their investment, as had those in Victoria who bought before 2006. 

For those who haven't yet earned a fair return on their investment, because they bought their licenses too recently, there could be hardship grounds for providing assistance, but he argues that it's little different than other cases where risky investments have not paid out. Do we bail out everyone who invested in IPOs of companies that fizzle? People who bought late bought knowing that the tech was changing. And if the case for payment is on hardship grounds, it should be means-tested and funded out of general government revenues rather than by cab riders. 

Holdin also points out pernicious incentive effects if innovators have to pay off affected industries all the time. I've seen compensation as a last resort way out of horrible political equilbria, like Canadian dairy - not as something that should be the default. 

* This was on their website circa 2000. I saved it and used it in my lecture notes on rent-seeking and transitional gains traps. 
My grandfather got to this country from Europe, via Argentina, in 1923. He had $150 in his pocket, and soon after he got here, New York City issued 11,787 taxi medallions - the same number as there were until just a few years ago. Back then, they cost $10 apiece. My grandfather bought one, and he started driving a cab. In his mind, it was one of the few jobs in which success depended only on how hard you worked.

Soon, he had saved up enough to buy his second medallion, and by the 1960's, he had 150. They were terrific investments - better than stocks. We recently figured out that since the 1930's, the Dow Jones industrial average has gone up 11 percent a year on average, and taxi medallions 17 percent. Today, they sell for $250,000 each. In the 1970's, my father started selling off some of the medallions to diversify. But no bank would lend money to the buyers - immigrants from the Soviet Union, Haiti and India - because they didn't have any bank statements. So he started a loan business. We've lent over $1 billion and never had to possess a single loan even though the interest rates are higher than on bank loans.

Monday, 20 March 2017

A regulatory flexibility act?

The law focuses on “small entities” -- not only small businesses but also small nonprofits and small governmental units such as towns and school districts. It recognizes that small entities often bear no responsibility for health and safety problems that give rise to regulation; that regulation deters potential entrepreneurs from innovating; that treating small entities the same as large ones impairs productivity; that regulation is often a barrier to entry; and that it can be easy for big companies, and tough for small ones, to comply with expensive federal mandates.

To reduce the problem, the Regulatory Flexibility Act directs federal agencies to identify and reassess existing rules that have “a significant economic impact upon a substantial number of small entities.” The legal requirement is simple: Ten years after rules are finalized, agencies have to determine whether they are having such an impact, and must decide whether they should be amended or rescinded.
Sunstein notes that the law hasn't had much effect as yet, but provides some recommendations for President Trump in order to make it more effective:
With a memorandum issued in 2011, President Barack Obama explicitly drew attention to the requirements of the law, directing agencies to offer flexibility to small entities unless they justified their failure to do so.

That was a start, but the Trump administration could go further. For example, it could make explicit provision for public outreach to small entities whenever it appears that they will be adversely affected by an expensive regulation. It might require agencies to respond, in writing, to serious objections from the Office of Advocacy (and thus give greater power to that occasionally important office). It might state that the Office of Information and Regulatory Affairs will not approve significant rules unless the most adverse effects on small entities have been eliminated, reduced or justified.

To be sure, small entities are not entitled to automatic exemptions from regulations. Whether large or small, companies should not be allowed to impose serious health risks on their workers. But it always makes sense to ask whether the arguments that justify regulation – for example, those in favor of increased energy efficiency – really apply to small companies, or whether the costs of burdening them outweigh the benefits.
It's fun to imagine consequences of this kind of rule in New Zealand. Maybe somebody in the system would have been forced to stop and think about the compliance costs they were layering on tiny iPredict in pursuit of money launderers.

Friday, 17 March 2017

Unintended consequences: bees edition

In this morning's mailbag:
I just had to share an observation that a colleague made when I casually raised the issue of beehive theft becoming an issue.

He said “You know why?  The Department of Corrections began providing bee-keeping courses to inmates who quickly ascertained the money to be made.”  Evidently the Beekeepers association had raised this concern explicitly with the Department but gained no traction.

He then said “You know what is next?  Agricultural spraying equipment.  Evidently the latest course to be introduced by Corrections is Agricultural Management.”

The courses are no doubt empowering the inmates but perhaps they should start with ethics.
From Bevan Wallace.


Wednesday, 15 March 2017

Like Uber, but for finding a realtor

I didn't particularly like the process of finding a realtor when we sold our house in Christchurch. We were happy with the ones we found, but the process wasn't all that great.

And so it's interesting to see this from Andreas Heuser and Henry Milne. Their startup, Agent Auction, promises to make things a bit easier, with realtors submitting marketing plans and fees for a vendor to choose among.

We've no plans on selling our place; if you're selling and wind up using Agent Auction, let us know how it goes. Andreas is with the Law & Economics Association, and will have thought carefully through the incentives on both sides of this market.

Monday, 13 March 2017

Tax cuts

I had a chat with Radio Live this afternoon on whether the government can afford tax cuts. I like to make a few notes for myself for these things ahead of time; here they are for you.

Over the medium to longer term, the government should aim for balanced books – or to run modest surpluses to get net debt down, providing headroom in case of future emergency need. Part of preparedness for future uncertainty is having fiscal room to take on debt in case of disasters like earthquakes.

Treasury projections have surpluses increasing through 2021. Those projections already incorporate core Crown expenditures increasing from about $74 billion currently to about $88 billion in 2021: core tax revenue increases more quickly than that, from about $76b to about $97b. So the expected path is for more government revenue and more government spending. Treasury also notes that its expected surpluses are structural rather than cyclical: this isn’t just a temporary thing that will wash out. On average over the medium to longer term, we see surpluses. By 2021, that surplus is projected to be $8.5b.

Part of the reason for surpluses is fiscal drag. Inflation is very low, but still present. And as incomes rise, more people jump into higher tax brackets. It’s a small overall part of the expected surpluses, but it still will give the government about $1.4 billion more revenue in total in the period to 2021 (about $300 million per year).

A balanced policy response to the situation we face would include:
  1. A third of the projected surpluses being devoted to tax cuts (so about $2.8b by 2021).
    • First, undo the effects of fiscal drag. We really should just index the tax thresholds to inflation so that this didn’t happen any more. But governments like the automatic tax increases that come with inflation because it lets them announce tax cuts regularly (or spending increases). 
    • Remaining tax cuts should be focused, in my view, on rates reductions on lower-income earners, at least in part because that helps bring down effective marginal tax rates that can be very high for poorer cohorts when income taxes get added onto clawbacks of various income-contingent benefits like Working for Families. 
    • Treasury notes that a percentage point drop in the each of the personal tax rates would cost about $1.3 billion currently. Dropping each by two points, and cutting the 33% rate only to 32%, would cost about $2.3b. So there’s room to move on tax cuts, even if only a third of future surpluses went to tax cuts. And these could be scheduled to phase in as expected surpluses grew.
  2. A third of the projected surpluses being devoted to spending increases:
    • This should be targeted at infrastructure improvements to deal with the housing crisis. Recall that Treasury projections already bake in some spending increases, including superannuation spending and non-NZS welfare spending.
    • One mechanism that we at the Initiative have liked is rebating to Councils at least some of the GST associated with new housing construction in their areas for a period over the next few years in order to encourage councils to consent and permit a lot more activity a lot more quickly. Councils could use that funding to roll out infrastructure more quickly, or to otherwise grease the wheels that need greasing to let new development happen. We need a lot more housing, quickly, to get through the current crisis.
    • The Productivity Commission is likely this week to recommend substantial redesign of the urban planning framework, but those changes would take a long time to implement.
  3. A third of the projected surpluses being devoted to paying down government debt:
    • Government debt is already on track to reduce as a fraction of GDP;
    • Very low Crown debt means that if a large earthquake hits Wellington, or somewhere else, it is easier for government to sell bonds to fund reconstruction. If the government goes into a crisis with higher debt, it is more expensive to raise the funds needed for rebuilding. Pushing debt down more quickly is a prudent part of preparedness.
Government should resist calls to simply increase spending in response to surpluses. The government has been pushing hard, over the past few years, to encourage a greater focus on the value that government delivers for its expenditure as the measure of its success rather than just what it spends. If it identifies areas where spending delivers strong value for money, by all means increase spending in those areas. But that has to come with a commitment to pare back spending in areas where spending doesn’t really achieve much. Blunt calls to increase overall expenditure miss that the government still has a lot of work to do in identifying areas where spending should be reduced because it is ineffective.

Thursday, 9 March 2017

The costs of advice

Recall that the government chose to default everyone into Kiwisaver. It subsequently put a big regime around provisions of financial advice, with all kinds of rules about who's allowed to say what about which and all kinds of penalties for getting things wrong.

Those two came together in interesting ways:
The Financial Markets Authority has softened its stance on KiwiSaver advice, after it found providers were not providing any, in fear of breaking the law.
New Zealand is baking stupid levels of risk aversion into business decisions. We see it in health & safety rules; we see it too in financial advice.

And, surprise surprise, the fault apparently lies with those subject to the rules for being too risk averse in applying them when big penalties apply for getting things wrong:
“With the revised guidelines, we’ve tried to help providers understand what they can give in the way of class financial advice,” says FMA director of regulation Liam Mason.

“What we’re saying is there’s a lot more help that can be given without fear of crossing a regulatory line, than was thought before… You can give help to people making KiwiSaver decisions without worrying about straying into personalised advice.”
It's good that they've put in revised guidelines that, presumably, companies can point to if they're ever pulled up on this stuff. Eventually regulators might update so that they're not constantly surprised that the threat of big big penalties results in strong risk aversion while people learn what the boundaries are.

The piece also notes that the FMA wants providers to have improved the amount of advice given to customers. More boxes to tick before the providers' next meeting with the FMA....

HT: Bryce Wilkinson

Monday, 6 March 2017

Reducing the alcohol purchase age

Reducing the alcohol purchase age from 20 to 18 did not increase the number of road accidents. If anything, risky driving dropped for a while.

Stefan Boes and Steve Stillman have updated their earlier work (noted here) on New Zealand's alcohol purchase age to bring in some more recent accident data to allow for longer term trends. It's now up as an IZA working paper.

They make an important methodological point - one that applies to a lot of work on minimum legal drinking ages. It is really easy to set up a regression discontinuity design using the birthday as the point of the discontinuity. The 17 year old, before the birthday, is the same person a week later, except with access to alcohol. Sure. But the RDD method can't tell you whether you've identified a birthday effect that will wash out over time, or a persistent negative effect resulting from a change in access to alcohol.

New Zealand makes for a good case study because the change in the minimum legal alcohol purchase age is relatively recent. So we can check what happens around birthdays before and after the change, set appropriate counterfactual trends, and look at overall effects on accident rates for those in the age group that switches status.

They conclude:
Overall, we find no evidence that changing the drinking age from 20 to 18 led to more vehicular accidents or alcohol-related accidents among teens. This is true both in the short-run following the law change and when examining cumulative accidents for the affected cohorts. We find that accidents do increase after one’s 18th birthday, but this appears to be a short-run phenomenon. Finally, our parametric regression models suggest that reducing the drinking age led to a decline in risky driving by youth who were already 15 at the time of the law change but had no longer-run impacts on youth risky driving among younger cohorts. We speculate that this occurred because of the extensive public discussion about the drinking age change that took place and because teens are likely to be particularly focused on the near future. We also present supportive evidence from infrequent health surveys showing a similar pattern for alcohol consumption among different youth cohorts. Our results support the argument that the legal drinking age can be lowered without leading to increases in detrimental outcomes for youth and call into question previous studies that have made policy recommendations by extrapolating from results identified using age-based RDDs.

Overall, our results support the argument being made by groups like Amethyst Initiative and Choose Responsibility (see that the legal drinking age can be lowered without leading to detrimental outcomes for youth. The current age limit of 21 in the US is higher than in Canada, Mexico and most western European countries. The arguments against lowering the drinking age typically include the idea that, even if a new steady-state with a lower drinking age might be beneficial, the transition to that new steady-state might be very costly. The evidence in our paper from a country with drinking habits very similar to the US suggests that this does not have to be the case.

Saturday, 4 March 2017

Why do they need this?

I'm a big fan of letting NGOs benchmark their effectiveness using government data. But I don't quite get why the government needs NGOs to collect some of this information on their behalf.

Here goes.

The government holds a huge amount of linked administrative data on all of us in the Integrated Data Infrastructure. All kinds of stuff can be linked up in the back end, under some pretty tight access controls.

The government's now demanding client information from NGOs who are the contracted service delivery arms for the Ministry of Social Development.
The Privacy Commissioner is investigating the Government's demand for client information in exchange for funds, after non-profit groups raised concerns.

The Ministry of Social Development's (MSD) new requirement for non-governmental organisations (NGOs) to supply identifying client information has blindsided non-profit groups.

Every person accessing MSD-funded services must agree to share their personal details – including their name, address, gender, date of birth, primary ethnicity, iwi, as well as dependents' names, dates of birth and relationships to client – for the funding to be released.

NGOs are adamant this requirement was not made obvious in the tender process.
If the NGO provides the government with their clients' names and unique identifiers (health identification number, for example), then all the other stuff should be able to be pulled in at the back end without bothering clients about it. Gender, date of birth, ethnicity and dependants should all be linked up via census records, for example. Why ask for stuff that the government already has?

At the same time, I can see a great case for making it easier for these kinds of NGOs to have their own work benchmarked. All they should have to do is go to the Social Investment Unit, give SIU the unique identifiers for the clients they're helping, and the outcomes they're targeting. SIU would tell them whether that outcome data is in IDI or not. If it is, it should be pretty simple to build synthetic control groups matched on the ex ante characteristics of the NGO's clients and then see whether the treated group shows improvement as compared to the synthetic control.

Imagine if charities could include, in their donation appeal, that kind of credible indication of their effectiveness in improving outcomes for the people they work with. It would help enable effective altruism.

Friday, 3 March 2017

A couple of talks

I presented on the Initiative's work on inequality at the ACT Party's 2017 conference in Auckland this past weekend. I don't think the presentations were recorded; if one turns up, I'll post it here.

I hate going to presentations where people just repeat what's on the darned slides, and so mine tend to just have the key bits to move the story along and the graphs that make the point. But I've covered related issues here at Offsetting, and I drew on our reports on inequality and education.

On Wednesday morning, I presented at the New Zealand Property Council. They wanted me to cover the current state of the world and how New Zealand should be thinking about Brexit and Trump.

I argued that we should be trying to bring over people working in America on H1-B visas who'd be great for the tech sector here and who won't be feeling particularly welcome there, and that we should be forging ahead with Pacific area free-trade agreements. As America gets more unreliable and talks about setting up non-tariff barriers, it's more important for everybody else to have strong working relations with each other.

Thursday, 2 March 2017

Sweet storable substitutes

A couple more important points on the Mexican soda case.

First from the comments on Tuesday's post:
Mexicans also love to drink uncarbonated sugary drinks, like horchata, and drink more of those now carbonated beverages are more dear. Much of that market doesn't go through regular retailers, being sold by street vendors and cafes, so I doubt they have any idea how much is being drunk.

The Public Health people are also very naughty about their moving target. They say "sugary drinks", but they can't ever mean that, since they would have to tax a whole bunch of fruit juices. Sometimes they move to "soda" or carbonated, but I'm guessing that they won't include any wines in that.

What they want to tax is just the cheap soft drink that the naughty fat poor people drink. By the time they divide the market down to that sector the effect will be trivial, even should it work. The weirdness of taxing soda drinks but not taxing lollies that are almost pure sugar seems to escape them.

If they were honest they would want to slap an excise on all sugar, everywhere. The effect would be to just raise the cost of food and piss people off. But if they actually want to cut down the amount of sugar consumed, that would be the only logical step.
When the Mexican tax came in, I asked whether anybody was keeping track of this.
Other fun questions I've not elsewhere seen asked:
  • The Mexican tax seems to be per litre of sweetened beverage, regardless of sugar content. If I lived in that environment, in places where customers have access to clean water, I'd sell a concentrated syrup for dilution at home rather than a ready-to-drink product. I wonder whether anybody there is doing that. Alternatively, I'd sell a bottled sugar-free product with a sugar sachet attached to it with a rubber band. Mix your own!
  • Even if things aren't turning into syrups, you should expect an increase in the sugar concentration in sugary beverages: taxed for a teaspoon, taxed for a cup.
  • I don't know whether the tax applies to just a bag of sugar at the supermarket. There is an 8% ad valorem tax on "a defined list of non-essential highly energy dense foods". A bag of sugar is pretty cheap. Even if the 8% tax applies to it, an 8% tax on a kilo bag of sugar has got to be cheaper than a peso-per-litre of sugar-sweetened beverage. 
  • Enterprising children should surely be setting up lemonade stands and dodging the tax man. Here's a recipe for Mexican lemonade: about a quarter cup of sugar per litre. A kilo of sugar is 5 cups, so 20 litres per kilo. That would draw 20 pesos in excise tax alone, were it taxed as a beverage. I suspect that the kilo of sugar itself would retail for less than that, with or without an 8% ad valorem tax.*** Sweet sweet tax-dodging lemonade stands.
  • One also wonders about whether any of the purchased bottled water to which folks are substituting is being used in untaxed home production of lemonade. The folks who have the Nielson household sales data should be able to tell whether purchases of sugar have changed. Why wouldn't you check that and report on it?
There are more consistent advocates out there who've pushed for comprehensive sugar taxes, but I expect those advocating for soda-only are aiming for salami-slices. Put in a low and politically palatable soda tax, find out (surprise surprise) that it doesn't do anything, then broaden the base and hike the rate by salami slices until you're at the more comprehensive setup that always made more sense, given the objective, but would not be politically saleable.

Second, do remember that soda can keep for a while in the cupboard. If you buy lots of it when it's on sale for consumption over a longer period, then your measured price elasticity of demand will overstate your true price elasticity of demand. The University of Massachusetts, Amherst's Emily Yucai Wang covered this in the RAND journal in 2015 rather well; here's an ungated version.
The typical analysis on the effectiveness of soda taxes relies on price elasticity estimates from static demand models, which ignores consumers' inventory behaviors and their persistent tastes. This article provides estimates of the relevant price elasticities based on a dynamic demand model that better addresses potential inter-temporal substitution and unobservable persistent heterogeneous tastes. It finds that static analyses overestimate the long-run own-price elasticity of regular soda by 60.8%, leading to overestimated consumption reduction of sugar-sweetened soft drinks by up to 57.9% in some cases. Results indicate that soda taxes will raise revenue but are unlikely to substantially impact soda consumption. 
She also finds the taxes to be regressive not only because poorer households consume more soda, but also because their demand is less price elastic.

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"?