Thursday 25 December 2014

A tale of the top 1%

Consider the two countries portrayed below. In both cases, the lines represent the share of national income earned by the top 1%.


In both countries, inequality fell from the early part of the 20th Century through until about the 1980s. From 1990, inequality started increasing in both, with a much sharper increase in the country with the higher line.

Draw some conclusions about the policies in the country represented in the upper line. Would you say that:
  1. The graph clearly shows that policy went wrong starting around 1994, allowing the rich to exploit the poor.
  2. We cannot really tell whether the graph demonstrates generalised good things or bad things; we need to know why the shares changed before assessing anything.
  3. The graph clearly shows that policy improvements in 1994 stopped the wage compression that was hurting economic growth; we should expect that overall outcomes have improved since then as well.
I, likely obviously, go for answer #2. Rising or falling income shares for the top 1% tell us nothing on their own. If it's a Wilt Chamberlain process, I'm cool with that, though Chamberlain isn't everything. If it's instead some crony deal where we tax everybody a dollar and give it all to one person, I'm not cool with that, unless the money goes to me.*

Now that everyone's had enough time to draw their conclusions, we can reveal that the top line belongs to South Africa. The end of apartheid brought a sharp increase in the income share of the top 1%, as pointed out in a new NBER working paper from Acemoglu and Robinson. The bottom line is Sweden.

Their abstract:
Thomas Piketty's (2014) book, Capital in the 21st Century, follows in the tradition of the great classical economists, like Marx and Ricardo, in formulating general laws of capitalism to diagnose and predict the dynamics of inequality. We argue that general economic laws are unhelpful as a guide to understand the past or predict the future, because they ignore the central role of political and economic institutions, as well as the endogenous evolution of technology, in shaping the distribution of resources in society. We use regression evidence to show that the main economic force emphasized in Piketty's book, the gap between the interest rate and the growth rate, does not appear to explain historical patterns of inequality (especially, the share of income accruing to the upper tail of the distribution). We then use the histories of inequality of South Africa and Sweden to illustrate that inequality dynamics cannot be understood without embedding economic factors in the context of economic and political institutions, and also that the focus on the share of top incomes can give a misleading characterization of the true nature of inequality.
On South Africa, they note:
No clear consensus has yet emerged on the causes of the post-apartheid increase in inequality, but one reason is related to the fact that after the end of apartheid, the artificially compressed income distribution of blacks started widening as some portion of the population started to benefit from new business opportunities, education, and aggressive affirmative action programs (Leibbrandt, Woolard, Finn, and Argent, 2010). Whatever the details of these explanations, it is hard to see the post-1994 rise in the top 1 percent share as representing the demise of a previously egalitarian South Africa.

And, across the OECD, skill-based technological change and increases in the size of the top firms explains much of the increase in earnings at the top. They caution that if a rising income share of the top 1% turns into entrenched political power, that can be disastrous; appropriate policy then should focus on institutional checks against power-grabs rather than wealth taxes.

* Gordon Tullock liked to give this example. The Tullock Tax would impose a small cost on each American but make Tullock about $300,000,000 richer. Despite that Tullock had a concentrated interest in such a tax, the tax never passed: the Logic of Collective Action isn't everything. 

Wednesday 24 December 2014

Treasury learnings

Things we didn't know before the latest set of Treasury working papers came out:

Tuesday 23 December 2014

The Status of Status Games

I don't worry too much about notions that consumption is driven by status-seeking. Not because it isn't - it would be surprising if at least some consumption weren't status-driven. Rather, because status-seeking affects just about everything, from consumption to not consuming to leisure to exercise and more. 

The richest entrepreneurs got that way by coming up with new goods and services to make the rest of us better off; in earlier eras, they'd have sought fame and renown through displays of prowess in killing people in battles. I prefer today's version.

Here's Hanson on the status of status-games.
But if you start to learn that many people you know are starting to see conspicuous authenticity as just another way that posers vie for status, then of course your community will come to not accept that as giving real status. No, you’ll start to see some new kinds of behavior as the sort of thing that people do who don’t care about status, but are just being “real”.
Then you’ll start to become aware that other people that you know agree with this new attitude of yours. You’ll get more comfortable with saying that you approve of these sorts of behavior in others, with hearing others say the same thing, and you’ll notice that you feel good when other people credit you with such behavior. You and your associates will all feel good about themselves, knowing they they are all good people who deserve respect because they do these behaviors, behaviors that they all know are not about status seeking.
At which point these new behaviors will have become your new status game. You see, status-seeking behavior must be a respected behavior that isn’t seen as overtly status seeking. Because we all agree that we don’t respect behavior that is done mainly to gain status. Even though we do, we do, we very much do.
Wellington has some major authenticity hipster beard-quality status-games going on in which I refuse to play a part.

In related news, a quarter of all Welshmen are descended from 20 men who won a particularly nasty status game fifteen-hundred years ago. Bill Gates has nothin' on them. [Update: Thomas Lumley, to whom I defer in such things, calls bogus on this particular stat.]

Previously:

Public service efficiency: Going postal

It takes less time for a badly addressed letter to get back to the US from New Zealand than it does to get back there from Canada. At least that's what Chong, La Porta, Lopez-de-Silanes and Shleifer find in a test of public sector efficiency [working paper version]

Here's a snip from Table 2, showing the proportion of test letters making it back to sender, the proportion doing so within 90 days, and the average number of days to get the letter back.


I note that New Zealand's postal service faces private sector competition, unlike some other countries' postal services.

NZ Post advertises that its cheapest international option takes 6-10 working days to ship to the US. The USPS gives no delivery estimates on its first class international mail offerings, but its Priority Mail option, more expensive, is 6-10 working days.

I'd expect that the letter to New Zealand spent a long time on a slow boat getting here.

More broadly, they find that measures of bureaucratic quality, red tape, education, ease of doing business, contract enforcement, infrastructure quality, and political accountability all affect mail efficiency in the expected directions.

 

HT: Bryce Wilkinson

Monday 22 December 2014

Three years too late

The Supreme Court got this one right. But, big picture, it is absolutely nuts that this was decided in December 2014 rather than back in 2011.

Recall that after the quakes, Christchurch Council hiked the building code requirements. Repairing earthquake damage then often meant bettering the condition of the building over its pre-quake condition. Insurers had not provided insurance against Councils mandating costly things and so did not want to be on the hook for it. And so off to court.

Finally, the Supreme Court has ruled, rightly, that insurers aren't liable for Council's changing the building code. But it's taken until now. The real problem here is that we've had over three years of legal uncertainty slowing the rebuild.

Why oh Why didn't the government fund a few test cases on this in 2011 to get declaratory judgements that would allow works to proceed more quickly? This is the kind of thing that CERA should have been jumping to do.

Convention Centre follies

The Sky City convention centre deal, as I'd understood it last year and the year before, was about the least bad way of getting a convention centre if you're determined that the government should be in the "facilitating convention centres" business.

Sky's offer was to provide a new convention centre attached to a casino. I wrote last year:
We should think of this as two separate deals.

First, the government is auctioning off some gambling concessions. SkyCity has bought the right to have an additional 230 pokie machines, 40 gaming tables, assorted other gambling concessions, and, possibly most importantly, a guarantee that if some future government reneges on the deal by banning gambling or otherwise eroding the benefits provided to SkyCity under the deal, they'll be compensated. Now suppose that we opened that whole thing up to a general auction. People would then bid for those rights; the highest bid would approximate the expected flow of profits from having the concession.

Second, the government took bids for the right to build and operate a big convention centre. The high bidder, or rather the company willing to do it at the lowest subsidy, gets to build and run the convention centre.

In this case, SkyCity has to reckon that losses (if any) from building and running a convention centre are less than the gains from the gambling concession [NBR subscription, sorry]. And it isn't crazy to think that the bundle provides added value: convention centres near casinos tend to lose less money than those not so-situated; there are reasonable complementarities between the kind of facilities attractive to conventioneers and those that are in place in casinos.

Conditional on the government wishing that there be a big fancy convention centre in Auckland, this is likely the least bad way of doing it. I haven't gone through the accounting on it in any depth, but the bottom line has to be that SkyCity reckons it can make a go of it, since they're bearing the risk if they can't operate it profitably. And it isn't crazy to think that there could be some economic benefits from increased tourist traffic if we host more conventions. But whether those benefits are larger than the amount SkyCity might otherwise have bid in an open auction for the gambling concessions, where the revenues went into the general fund rather than into a big convention centre, that's rather less clear. It's possible, but it's far from certain.
As I'd understood things, Sky was to be on the hook for cost increases:
On the plus side, it's unlikely this arrangement yields another Claudelands. SkyCity's on the hook for costs. They've reasonable incentive to keep things running properly: the Centre would complement their existing operations. The more conventioneers, the more potential customers at the nearby casino.

...Bottom line: I do not believe the "broader benefits" case for building convention centres. Nothing stops hotels near a proposed centre from getting into sponsorship arrangements where they pay for listings in standard convention centre marketing materials to help fund the centre; I find it more likely that the benefits are too small to make it worthwhile than that public goods problems, easily resolved by assurance contract, stop things. But if we're going to have a new convention centre, this might be the least bad way of limiting fiscal risks. That's not a strong endorsement, because I don't particularly like the potential regulatory takings from the local pubs currently hosting pokies if the new machines are part of the sinking lid policy, and because there's a whiff of cronyism to the whole thing. But honestly, who else would be able to pull off the casino-convention centre combination in Auckland?
If it turns out that the government is on the hook for cost inflation, then much of the point of the whole deal is lost. Matthew Hooton's been livid about this on Twitter. He's not wrong.

A government committed to using PPP arrangements also has to be ready to play hardball with contractors who lowball initial cost estimates lest they encourage stupidity in each and every future contract.

Every nice thing I said about this deal is retracted if government is on the hook for cost-overruns or operating expenditures.

Saturday 20 December 2014

In praise of inflation targeting

The New York Times celebrates 25 years of the Kiwi Inflation Targeting Technology.
Sometimes, decisions that shape the world's economic future are made with great pomp and gain widespread attention. other times, they are made through a quick, unanimous vote by members of the New Zealand Parliament who were eager to get home for Christmas.
That is what happened 25 years ago this Sunday, when New Zealand became the first country to set a formal target for how much prices should rise each year - zero to 2 percent in its initial action. The practice was so successful in making the high inflation of the 1970s and '80s a thing of the past that all of the world's most advanced nations have emulated it in one form or another. A 2 percent inflation target is now the norm across much of the world, having become virtually an economic religion. 
The piece misses that New Zealand's since flipped to a 1-3% rule, on average over the medium term, and that the RBNZ let things run hot in the leadup to the '08 recession - interest rate hikes in early '08 would have been a mistake regardless of that they were above 3%. But it's a great read.

Thanks to Don Brash, David Caygill, and Roger Douglas, who made it happen.

I'd not seen this picture of David Caygill before. But the general theme is not unfamiliar.


Update: Here's the version from the 2025 Taskforce report.


Update 2: The Kiwi Inflation Targeting Technology: KITT.

HT: Ben Atkinson

Friday 19 December 2014

An incoherent argument for spying

The Timaru Herald gets this one pretty wrong.

Peter O'Neill's editorial there makes the following case:

  • Yahoo!, a private company, saw that one of its service's users was trafficking in child pornography.
  • Yahoo! alerted the American authorities, who got in touch with Internal Affairs in New Zealand, as the user was in New Zealand.
  • The Americans then asked Yahoo for a few more details to allow the Kiwis to find the guy.
  • The police here arrested him in Timaru.
That's all fine. Then the punchline:
  • Therefore opposition to TICS, the NZ legislation making it easier for the GCSB to spy on internet users, is great and all the civil libertarians were wrong.
I just don't see how the last part follows from the first. All of the first chunk could have happened with or without TICS.

Thursday 18 December 2014

Cost-benefit analysis and impartial umpires

In contrast to his fiscal rules, Osborne has had enormous success with his greatest fiscal innovation: the Office for Budget Responsibility (OBR). It was created in 2010 and has been successful in curbing the problems of over-optimistic forecasts that plagued the Treasury through the 2000s. Its ruthless transparency has lent credibility to the government’s plans and, in only a few years, it has grown in stature to the point that it can openly rebuke the Prime Minister and force changes in the Budget with its forecasts.

Expanding the role of the OBR, commensurate with its growing stature, would help overcome the fits and starts of UK fiscal policy. Where rules are fragile and inflexible, institutions grow and evolve in response to circumstances. As the OBR continues to perform effectively and its credibility rises, it can sustain a greater burden of responsibility for holding the government to account. Already, it has cross-party respect and support: last year, the shadow chancellor Ed Balls asked it to inspect the budgetary pledges of each major party ahead of the election. That is surely a good idea.

Assessment of opposition and government policies is a role already performed by similar bodies overseas, such as the Dutch CPB. The CPB is nearly 70 years old and has gradually taken on a central role in analysing the implications of election manifestos and Budget promises. In America, the Congressional Budget Office has also expanded its purview over time. Both provide indispensable analysis of the implications of the government’s policies and proposals, describing the trade-offs and estimating the costs.
Bryce Wilkinson and Khyaati Acharya here at the Initiative made the case for a New Zealand version.

I'd love such a body to act as clearinghouse for cost-benefit analyses. Regulatory Impact Statements vary in the quality of cost-benefit analysis; there are always ways of getting the number over the line if the Ministry's particularly keen - just consider the Australian East-West rail road link. You need an impartial arbiter to make sure that the cost-benefit analyses are being undertaken to a common standard across Ministries.

Further, since few Ministries have in-house capability to conduct these analyses, they often have to outsource it to private consultancies. At least one of these consultancies has a reputation for providing the number that the client wants rather than one that's sound. But Ministries can't pay the piper unless they know the tune, or unless somebody who can read sheet-music has a look over things afterwards.

Here's a simple rule that could work. Submitted Regulatory Impact Statements would be required to include a cost-benefit analysis for any rules with substantial effect. That cost-benefit analysis must be vetted by Treasury, or by a new Fiscal Council, with enough of the workings provided by the Ministry or the consultants to allow for replication and sensitivity tests. Treasury is working up new guidelines for cost-benefit assessment; that would be the benchmark. If the cost-benefit analysis fails to pass muster, it's sent back to the Ministry. And consultancies that produce cost-benefit assessments that fail to meet the standard more than, say, one time for every ten reports produced, are put on a naughty sheet barring Ministries, government agencies, SOEs, local governments, regional governments, or any other part of the government I've missed here, from engaging their services for a few years.

Further, for a cost-recovery fee, that same agency should be able to vet analyses produced for private sector clients. A failed report there could get the tagline "Had this been produced for the public sector, the consultancy producing it would be barred from producing further analyses for any government agency for two years." And that tagline could follow it in all public discussion of that report.

I think the naughty sheet would do a lot to improve the standard of cost-benefit analyses in New Zealand.

Wednesday 17 December 2014

The gender wage gap and hours worked

From my piece with Bryce Wilkinson in last week's National Business Review:
Suppose a management consultant pulled you aside and said, “Hey, have I got a deal for you. With this one simple trick, I can lower your labour costs by 14% and it won’t cost you a dime in productivity. Trust me, I know what I’m doing.” You’d probably be a little sceptical about the claim, or at least I would.

Statistics released in the last week by the State Services Commission reportedly found that the pay gap between men and women in public sector management roles averaged 14%.

...
Sadly, there are few one-simple-trick paths to success. More rigorous analysis of gender pay gaps that accounts for differences in work experience, training, time outside of the workforce, choice of industry, and hours worked generally wipes out the bulk of the headline pay differences.

The recently released New Zealand Income Survey data also makes things very clear. Hourly earnings for full time male employees, at $23.97/hour, is only about 6% higher than women’s $22.54; women in part-time work, averaging $17.26/hour, beat men’s $16/hour. Much of the headline pay gap comes from that more men are in full-time work rather than part-time work.

Treasury released analysis a fortnight ago showing that Working for Families (WFF) has reduced the number of hours worked by married women. Prior to WFF, among married men in employment, about 10% worked part-time, 59% worked a 40-hour week, and 32% worked a 50-hour week. For women, 55% were in part-time work, 30% worked a 40-hour week, and 15% worked a 50-hour week. None of those figures changed greatly with WFF, but women in part-time work shifted to working fewer hours, and about 9000 married women dropped out of the workforce entirely.

In a lot of professions, wages start climbing substantially for employees willing to put in the longer work-weeks, as economist Claudia Goldin pointed out in her 2014 Presidential Address to the American Economic Association. Much of the measured pay gap comes down not to discrimination, but to choices.
Update: NBR subscribers can catch it here.

A few relevant links:

Tuesday 16 December 2014

Kreskin Cosh

Well, Colby Cosh called this one before it happened, didn't he?

Cosh from last week:
The point is not that Bismarck [subject of many assassination attempts] was particularly hated, although he was. The point is that this period of European (and American) history was crawling with young, often solitary male terrorists, most of whom showed signs of mental disorder when caught and tried, and most of whom were attached to some prevailing utopian cause. They tended to be anarchists, nationalists or socialists, but the distinctions are not always clear, and were not thought particularly important. The 19th-century mind identified these young men as congenital conspirators. It emphasized what they had in common: social maladjustment, mania, an overwhelming sense of mission and, usually, a prior record of minor crimes.
From the Sydney Morning Herald on yesterday's hostage incident:
Manny Conditsis, a Sydney lawyer who represented Monis last year when he was charged with being accessory to the murder of ex-wife Noleen Hayson Pal, told ABC News that Monis was an isolated figure and "damaged goods".

"His ideology is just so strong and so powerful that it clouds his vision for common sense and objectiveness," Mr Conditsis said.

...
"Knowing he was on bail for very serious offences, knowing that while he was in custody some terrible things happened to him, I thought he may consider that he's got nothing to lose," he said.

"Hence participating in something as desperate and outrageous as this."

Monis had an extensive criminal history, which included being charged with 50 allegations of indecent and sexual assault. He had also been engaged in a protracted battle to overturn his conviction for sending offensive letters to the families of dead Australian soldiers between 2007 and 2009. 
BK Drinkwater also claims, ex post, to have gotten it right:
One potential lesson from the whole thing?
I could support shifting funds from spying over to mental health support. I doubt it would have helped in this case, as Monis looks to be somebody who really should have stayed in prison for a very long time. But it does seem a better general-purpose technology. This image sticks with me:


Friday 12 December 2014

Bah humbug

I led the affirmative charge at the GEN debate on the following proposition: "This house believes that the Christmas Extravaganza is a waste of time and money." We didn't get to pick our sides, but I didn't mind partnering up with Patrick Nolan from the Productivity Commission.

As Bronwyn Croxson from Ministry of Health was leading the charge for the opposition, I attempted a pre-emption of her most likely argument. Even though her side wound up taking more votes at the end, I'm pretty happy with my mind-reading. But I did not expect that she would give gifts of chocolate almonds to the audience to sway the votes. Well played, Bronwyn. Anne-Marie Brook from Treasury was second for the opposition.

Had I had a chance to provide a closing statement before the vote, it would have been this:
I agree entirely with the opposing team on the value that can come from gifts. If you believe, as I do, that people can manage to do good things for each other without the quasi-coercive Christmas to force the issue, then you should vote for the motion. If you're a pessimist about human nature and think the only way that these will be provided is by forcing it through Christmas, and that the gains from that outweigh all the losses I've talked about, you should vote for the opposition. 
Had I been more confident in my mind-reading, I'd have had it at the end of my opening statement. I should have been more confident. Bronwyn made exactly the move I'd have made in her place, as I expected.

Here's the notes I'd made for myself for the debate. I varied from them as always, but here they are anyway.
The Christmas Extravaganza is a waste of time and money. What a waste of a season. Christmas has a lot of faults. Economists have known for at least two decades that the most prominent part of Christmas – the gift exchange – has some serious problems. But that’s just the start of it. Aside from the wastefulness of the gift-exchange, which I’ll touch on in a bit more depth and which Patrick will go through a bit more thoroughly, we also have substantial problems arising from bunching and clustering of holidays, to the detriment of both work and vacations.
 Because the wastefulness of the Christmas gift exchange is so very well established, I thought I might try to make the best case I can against it. Again, Patrick will hit it in a bit more depth with more recent data, but the standard argument runs as follows. The person best placed to decide how to spend his money is the person himself. Consequently, it would be remarkable if any gift-giver could provide a gift that the recipient would value more than cash. Cash can be transformed into anything the recipient really wanted. Bad gifts, not so much. And surveys show that gift recipients put less value on the gift than its cost. If I had been given the other side of this argument, what would I have said? I’ve always been a bit of a Christmas-sceptic; this has given me a chance to check my intuitions. So here’s the best argument I can give in favour of Christmas gift-giving. I’m not trying to set up a strawman here; I’m sure the other side will have come up with better arguments for Christmas. But here’s the only argument I currently think stands against the “Inefficient gifts” critique. 
Even if survey measures show that recipients put less value on the gift than the gift cost the giver, the fact that millions of people around the world choose to come together to give each other gifts at Christmas is prima facie evidence of that the whole process provides some kind of value to givers and receivers. Otherwise, why would they be doing it? The big point that the gift-inefficiency argument misses is that the giver gets enjoyment from giving the gift. The receiver’s enjoyment is only part of it: there’s a selfish aspect to giving too. It doesn’t have to be as selfish as the time that Homer Simpson gave Marge a bowling ball that he expected her to let him use, but there can be selfish joy in seeing somebody else enjoy something you’ve given them. The twentieth century’s greatest economist, Gordon Tullock, noted the super-efficiency of charity. If you give a dollar to charity, you must get at least a dollar’s worth of enjoyment out of the gift – or you wouldn’t have given it. The receiver also gets a dollar’s value out of it because it’s a dollar. So the anti-Christmas economists say that the recipient might only get 70 cents out of it rather than a dollar: you still likely have well over a dollar’s value, all up. So what’s the problem? 
There are a couple of rather substantial problems with that argument though, and they both stem from the fact that Christmas isn’t really entirely voluntary. Sure, there’s no law forcing you to participate, but not all coercive social arrangements are enforced by law. We can show this pretty easily by counterexample. Imagine going home, after this debate, and telling your partner,
“You know what? Those economists, they convinced me of something really important today. We shouldn’t give each other presents this year. You and I both know that we’d do a better job spending the money on ourselves, so instead of sending oblique hints to each other about what sorts of gifts might be appreciated, let’s just call the whole thing off.” 
If you think that’ll work, you might want to get the number for a marriage counsellor before you try it. Unless your partner’s also an economist, it’s not likely to go over really well. Even harder could be suggesting doing away with expansive gift-giving with the extended family. If two-person bargaining can be difficult, adding in more people doesn’t do a lot to make it easier. Whether or not you believe the case against Christmas, even broaching the subject with your family is dangerous. It could be taken as a signal of disloyalty, or even that you might not love them as much as they’d thought.
 And, the stakes are high. Divorce filings are most common right after Christmas. The owner of one website for divorcees says “I see a huge increase in pageviews and searches the day after Christmas. People start looking for information before the New Year starts, but they can’t do much until the attorneys are back in the office.”
 Ultimately, I think the whole gift-exchange process is best thought of as a costly signalling exercise where failure to signal is very costly. In a better world, we could show our loved ones how well we understand them and how much we love them by providing thoughtful gifts and gestures whenever we encounter a good potential one. Christmas forces the issue not only for those who haven’t provided such a signal, and perhaps should, but also for everyone else. And if you blew your one great thoughtful-gift idea earlier in the year, well, Christmas could be painful. In that kind of world, for a lot of gift-givers, we shouldn’t be worrying about having missed their joy of giving in the Calculus of Christmas, we should instead be worried about having forgotten to add in the brain-wracking expense of trying to come up with a gift that will meet the threshold – and cash sure isn’t allowed.
 And if you find the right gift at the wrong time, do you buy and hope you don't lose it before Christmas? Even if you can find it again at Christmastime for wrapping, the recipient still could have enjoyed it for months before Christmas if you could have just given a thoughtful gift at the time that you found it. But you couldn’t, because you’d then still have to figure out something for Christmas, and serendipity might not strike twice. But wait, there’s more! Even a giftless Christmas is destructive because of the whole "silly season" effect. The month before Christmas is a mad office rush of getting all the projects done before all the critical people disappear - AT THE SAME TIME - and holiday parties that bite into the time needed for the project-finishing rush. Loud brass bands on Lambton Quay make you wish for the agnostic's missile from the Monty Python sketch. Outside the office, when you've little option but to take holidays between Christmas and New Year's, everybody else is taking theirs at the same time - even if the weather is terrible and they'd have preferred holidaying in February. And so we have wrecked the economics of a lot of holiday destinations: huge peak load issues where smoothing over a broader season could work better. The roads are a nightmare. Everywhere's booked up, inducing ever-earlier pre-booking of venues in an arms race. And nobody's holiday is as enjoyable as it could be. Are the coordination gains from having a day or two off for extended family things really substantial enough to merit all this?
 The superior alternative? A Festivus for the Rest Of Us. One day. One pole. No tinsel.

My slides ended with Frank Costanza's words of wisdom:



I really should have given our Festivus stories as example of that you can really do all this without the Christmas part.

Previously:

Thursday 11 December 2014

OECD on inequality

Whenever a small country gets mentioned specifically in an international report, that report there gets noticed.

A new OECD working paper claims that income inequality hurts economic growth, and particularly hurt New Zealand growth. Note that this is a working paper rather than OECD position.

Let's walk through their method a bit before discussing.

They use a new OECD panel to estimate effects. Growth, and everything else, is measured at five year intervals from 1970 to 2010.

Rather than use standard fixed-effect OLS modelling, they use a System Generalised Method of Moments approach. I've not used this approach before but here rely on their description: it combines first differenced equations with a set of lagged first-differences of the explanatory variables as instruments.

They find that net inequality (after tax-and-transfer) hurts economic growth, that gross inequality (pre tax-and-transfer) doesn't hurt growth, that changes in human capital (education) do not affect growth one way or another - there's a slightly negative effect of education on growth in the set of specifications, but it's not significant; and, investment doesn't affect growth one way or another.

The set of results is then a little surprising. We usually expect investment to matter a lot for growth - both in physical plant and equipment (investment) and in people (education). They find that neither does anything and that the only thing that matters is inequality.

Further, when they break things down a little, what seems to matter most is the difference between 4th decile income and average income rather than incomes at the top. Incomes in the 9th and 10th decile relative to average income do nothing; differences between the fourth decile and the average matter hugely.

And now we start getting into the plausibility checks. Does this set of results really make sense?

By what mechanism does a sharper gradient between income at the 40th percentile and average income translate into worse growth? Imagine that we took this as policy conclusion: increase the tax on average earners to give money to people slightly poorer than them. Does that seem reasonable? They argue that the effect runs through reduced investments in education in the lower decile cohorts when income inequality is higher, but they found no effect of education on growth. Further, the countries examined, like New Zealand, went through rather a few changes to tertiary education over the period - from free tuition to tuition to student loans. All of these would affect lower-tier access to education, and none are accounted for.

The paper goes on to argue that its results on education must be wrong because everybody knows that education matters, then makes a strong case not for income transfers, but for increased spending on education. And they hang that case on other papers finding a strong effect of education on growth - but that also find that inequality increases growth!

I've a specific concern also about the use of New Zealand in this time series. The potted history of New Zealand inequality and growth. New Zealand growth rates tanked from the late 1970s through the early 1990s as first Muldoonism then necessary restructuring put a pretty high cost on the economy. The Muldoon stuff was nonsense. The economic restructuring set the groundwork for strong growth in the 90s and through the 2000s, but was really really painful. It was painful for laid off workers, and it was painful for a whole pile of firms, both large and small, that had to reinvent themselves for an open and free market. Them ships don't turn on a dime, and so growth tanked.

At the same time as NZ growth rates tanked due to restructuring, incomes at the top jumped - in part due to changes in tax and accounting that brought some of that onto the books where it previously had been hidden, and in part due to that folks with the skills to adjust to the new environment started being compensated for it. That rise in inequality happened almost entirely from 1985 through about 1992, after which it wobbled around but didn't have systematic trends.

If we look then at a long run data series, we get a big increase in measured inequality in New Zealand at the exact same time as economic growth takes a nosedive. Once the growth in inequality stops around 1992: blammo! Growth starts again.

Is it any wonder, then, that a regression approach based on reduced form fixed effect estimation with no dummying out of the reform period or other adjustment for it would find huge effects of inequality on growth in New Zealand? It's stuff like this that's meant that more recent academic work, unlike OECD working papers, has been shifting to use of microdata within countries to try to figure out what's causing what. I don't think the OECD papers get us there.

I note that I have profited from conversations with Matt Nolan on this, though I'm to blame for all errors here.

Update: Here's me and Tim Hazeldine agreeing about the report's merits over at Morning Report.

Tuesday 9 December 2014

The Asylum creeps in: health and safety edition

A reader emails me that the revised health and safety regime, which brings criminal penalties for company directors, will also apply to voluntary organisations. He writes:
I know you write periodically about crazy laws and being inside the asylum. As you might know, I'm a scout leader. A volunteer. One of things that I have enjoyed in scouting is the ability to let children and youth take risks. You know, tackle bullrush, climbing trees, crossing rivers, hiking, making and playing with gun powder, making their own bows and arrows (one shot an arrow clean through a window without breaking it :) ) and so on.

They keep making changes to health and safety laws which potentially undermine the ability to do this. The first was throwing in volunteers in the Health and Safety in Employment Act back several years ago. That added to our paperwork and probably did restrict some activities at the margin (but not hugely), which was a pain at the margin but that seemed to be all. But they are currently making a couple of other changes which increases costs to volunteer organisations and I'd predict the benefits are negative once they take into account lost consumer surplus from eventual changes and reductions in volunteer activities. You can find one mentioned at 
http://www.stuff.co.nz/manawatu-standard/news/63962701/New-safety-reforms-threaten-volunteers  

and another at
 http://www.scoop.co.nz/stories/BU1412/S00202/new-sentencing-act-exposes-businesses-and-individuals.htm 

They might not have any impact, but.....! Since volunteer organisations fall under the jurisdiction of the health and safety in employment act you can see why the sentencing act changes are a potential problem for them. It also seems to me to undermine the no-fault basis of the ACC scheme. I think ACC has some problems, but the no-fault provision is incredibly useful in not having to worry about frivolous lawsuits to extract payments to avoid the costs of the suits going to court, which then result in noticeable restrictions on people's activities, such as children playing where there is any tiny risk of an injury!
The link to the Manawatu Standard article has since died; I don't know whether they pulled it because the government has issued a very recent clarification, or if some other error has occurred. Can anybody confirm how the changes to legislation will be affecting volunteer organisations?

I also found a press release from an insurance company warning outfits to get liability insurance because ACC no longer shields against lawsuit:
Ms Cross says that under the previous legislation even the most serious accidents would rarely result in reparation awards over $100,000, “but with this new law, the figures are likely to be significantly higher.”
Ms Cross says that businesses should seek advice from their broker whether their statutory liability policy has the required level of cover, as this amendment will have potential impact when there is a breach of the Health and Safety legislation.
She pointed out that companies would not be able to insure themselves against any penalties but could get insurance that would cover reparation and legal costs.
Ms Cross says she will be working closely with her clients at Crombie Lockwood to prepare for this law change.
“At this stage it is still a bit unclear how the courts are going to use this new tool, but you wouldn’t want your company to be the guinea pig.”
I'd thought that the basic deal with ACC was that we gave up our right to sue in exchange for a mandatory kinda-ok-but-really-kinda-not insurance programme, with the whole thing making sense because U.S.-style tort excesses are worse. I suppose that potential damages are here limited: if the most that a party can be liable for is 20% of the total ACC claim, and if ACC doesn't really pay that much, then potential liability stays below that in the States.

Monday 8 December 2014

The voice of presumption

In an editorial co-written for the Australian and New Zealand Journal of Public Health, published yesterday, Dr Boyd Swinburn says scientists who represent the voice of the public in policy-making are being shouted down by large commercial interests and their views suppressed by attack campaigns.
In what sense do public health researchers, as earnest and honest as they may be, in any way represent "the voice of the public"?

At their best, public health researchers provide useful scientific data that can help individuals in assessing trade-offs, and that can also help inform policy as one of many considerations in the policy process.

But "the voice of the public"?

I suppose if that self-perception is common on the public health side, it explains a bit of how they come to the policy debate.

I'm far from convinced that either elections or polls can provide "the voice of the public", or that such a thing can even exist. People's views are too varied - thankfully. And there's that whole set of impossibility theorems on aggregating social preferences into any kind of "voice of the public".

As for being shouted down, it would be an interesting exercise to compare newspaper column inches of "large commercial interest attack campaigns" versus column inches of public health campaigners complaining about WhaleOil blog posts.

Swinburn continues:
In the article, co-authored with Australian professor Michael Moore, Swinburn said: "A blanket of suppression is insidiously descending on the voices for public health."
First, there were the interests of transnational corporationsand then there was the Government, which wanted to control public health information and messaging, Swinburn said.
While working in Australia's Deakin University, Swinburn said he experienced efforts to have reports cancelled or watered down and funding pulled, and he said that had started happening in New Zealand.
"There needs to be voices that are based in science standing up to that and speaking on behalf of the public," he said.
"There are plenty of voices on behalf of the commercial interest. But the number of voices on behalf of the public are getting fewer and weaker."
This is actually a tough one.

Suppose the government honestly contracts for some piece of research, the researcher produces and honest and true answer, and the government then suppresses it because politics. That by all accounts is bad.

Now imagine a contrary case where the government contracts for some piece of research, but the researcher produces a report that is, frankly, embarrassing: it would fail to withstand any rigorous scrutiny. Worse, the government's having funded the work could give it imprimateur if it were released. If the contracting agency then demands changes or wishes that the report not be released, that's different from efforts to water things down.

From the outside, a report's suppression in the two cases looks identical. The researcher in the second case would claim that the first case obtained, and the government in the first case would claim that the second case obtained. I suppose requiring rigorous truly external peer review on reports - and not just rubber stamps - is one way of preventing the latter but not the former.

Friday 5 December 2014

Alcohol marketing and evidence

Two contrasting stories in the weekly mailbag.

Item the first: Sally Casswell takes one look at the evidence in a viewpoint article in the NZMJ and concludes:
Stepping back and looking at the harm alcohol does in our society, the evidence on marketing’s effect on young people’s consumption and the popular support evinced for change, the only reasons to maintain alcohol marketing in its current largely unrestricted state are to: first, protect the profits of the transnational corporate producers by allowing them to appeal to new cohorts of young people with marketing which recruits them as consumers as early as possible and encourages drinking of larger amounts and second, to protect the financial interests of the advertising and media industries.
It is hard to avoid drawing a conclusion that government’s failure to act was based on a decision to protect the interests of these large corporations at the expense of protecting the health and wellbeing of future generations of New Zealanders.
A restriction on alcohol marketing similar to that adopted more than 20 years ago in relation to tobacco (Smoke-free Environments Act, 1990) or specific to alcohol similar to that in France (LoiEvin, 1991) will not impact in any meaningful way on adult consumers’ knowledge of the availability of alcohol.
Significant restrictions on alcohol marketing will, however, likely effect the normalisation of alcohol. Normalisation, the acceptance of ubiquitous and perception of unproblematic use, makes it more difficult for health promotion and social marketing to affect consumption among heavy drinking social networks or for family and whanau to place limitations on access to alcohol by vulnerable young people.
...
The ideal is a complete ban on alcohol marketing. This is feasible and a useful model is available in the Smoke-free Environments Act 1990. It has the advantage of thorough coverage of all marketing and sponsorship and includes internet marketing. 
Casswell's piece argues that the evidence on the harms of alcohol marketing to youths is so obvious, lack of government action means the government must have been subverted by evil evil industry.

Item the Second: Cochrane has been running systematic reviews of all kinds of things. Here's Cochrane on alcohol advertising.
OBJECTIVES: To evaluate the benefits, harms and costs of restricting or banning the advertising of alcohol, via any format, compared with no restrictions or counter-advertising, on alcohol consumption in adults and adolescents.

AUTHORS' CONCLUSIONS: There is a lack of robust evidence for or against recommending the implementation of alcohol advertising restrictions. Advertising restrictions should be implemented within a high-quality, well-monitored research programme to ensure the evaluation over time of all relevant outcomes in order to build the evidence base.
And here's what I'd said at the Ministerial Forum on Alcohol Advertising and Sponsorship back in the fall.
Bans on advertising, in a free society, are only justifiable on solid evidence of substantial net harm reduction. The existing evidence shows only weak associations between exposure to advertising and consumption behaviours.

...

No further restrictions to alcohol advertising are justifiable on current evidence. However, if the government is determined to impose such restrictions, it could at least design them in such a way as to allow for programme evaluation. If you decide to ban billboards, set up a rolling phase-in design across a set of towns to assess whether the ban has had any effect; if it has not, then remove the ban. If you decide to ban alcohol advertising in print, use a similar rolling phase-in design to allow for evaluation.
So Cochrane concluded there was no evidence sufficient for restricting alcohol advertising and said any measures should be done within a research design allowing for evaluation. I said that there was no evidence sufficient for further restricting alcohol advertising in New Zealand and that any measures undertaken should be set up to allow for evaluation.

But Casswell's piece argues that the evidence is so strong that only industry subversion can explain lack of action.

It remains astonishing that a government so deeply in Big Alcohol's pockets somehow keeps giving grants to Casswell's research outfit.

Thursday 4 December 2014

Cost-benefit analysis and the new drink-driving limit

NewstalkZB asked me Monday morning to comment on the new .05 drink driving limit.

I noted that it is likely to prevent some crashes, but that I did worry about whether it was worth the cost. The cost-benefit analysis that formed the background to the RIS wasn't bad, but did have a few things that gave cause for concern.

Here's the table of behavioural assumptions that fed into the cost-benefit analysis.

They expect that 94% of drivers who are currently drinking to less than the .05 level will have no change in behaviour, 2.4% will take alternative transport, 2.4% will switch to drinking at home, and 1.2% will flip to drinking to less than a .03 level. 

While there's been a lot of media play on how much you can drink while staying under the .08 limit, and that's up to 8 standard drinks for an 85 kg male on a 2-hour window, that's not the safe limit. Rather, it's how much you can drink and have a 50/50 chance of being nailed for drunk driving. See table 3, here

A half litre of Renaissance Stonecutter Scotch Ale is 2.8 standard drinks. Three half-pints in a two-hour window, 4.77 standard drinks, is safe for an 85 kg male at the .08 limit but not safe at the .05 limit. At the .05 limit, the "safe" level is just over one pint per two hours. For a 55 kg male, even a half-litre is unsafe because the safe threshold is 2.5 standard drinks.

It consequently seems pretty implausible to me that we wouldn't have more substantial changes in behaviour than those here assumed. While that might have some consequent improvement in accident reduction, the effects on consumer surplus will be rather more substantial. The risk-averse will be targeting .03 to avoid hitting .05 and consequent penalties.

Their mid-range estimate has the 10-year aggregate loss in consumer surplus at $900,000 for the whole country. That's around $130,000 per year if they applied an 8% discount rate, or $2500 in lost consumer surplus, per weekend, for the whole country, assuming that all drinking is done on the weekend. It just doesn't seem plausible. I'd be willing to pay $100 per year to be able to stick with .08 instead of .05, which would let me keep aiming at .05 rather than having to target .03. If there are at least 1300 people like me, across the whole country, then they've messed up the consumer surplus calculations substantially. In a country of 4 million and change, well...

It'll be interesting to see how this plays out, not only in the accident stats, but also in patronage at after-work functions and at rural pubs.

Wednesday 3 December 2014

Air Gini

In today's glass-half-empty news, inequality in airline space has increased.

While it used to be the case that the top 7% of passengers got about 15% of the room, transatlantic configurations give proportionately more space to first-class passengers: the top 21% get about 40% of the plane. So the Gini index increases from 8(ish) to 25. Epopp at OrgTheory provides the illustrations. The top one shows an earlier equal plane; the bottom one shows a modern configuration.

plane 1plane 3

Isn't it terrible, eh? All those fat cats in first class taking up space that could be used by the proletariat in economy. Horrible stuff.

Except for that, in the earlier period, a whole pile of those economy passengers might have been taking the bus instead. The cost-saving innovations that allowed cheap economy-flight transport increased in-flight inequality and allowed more poor people to fly. The correct diagram would have a couple of Greyhound buses running alongside the spacious early airplane for domestic travel, and no other real options for international. But it's easy to forget that and to focus on the horrible horrible oppression experienced by those flying economy class who, in an earlier time, couldn't have afforded any air travel.

I haven't ready access to the American price series, but here's the New Zealand real cost series from 1981 through 2010. Domestic travel has gotten a little bit cheaper, in real terms, but has otherwise just tracked inflation; international travel has halved in cost.

Graph, Domestic and international air transport real series

The divergence of international and domestic prices is interesting. I wonder what portion is due to greater cost-saving innovation in long-haul than in short-haul transport, what portion is due to relatively greater international fare competition, and what portion is due to greater elasticity of passenger destination choices among internationally competing airports than in domestic service. If you're flying domestically to Auckland or Christchurch, you likely need to get to that place. If you're flying in from overseas and don't much care whether you start in Auckland or Christchurch, airport landing fees start getting more competitive.

Tuesday 2 December 2014

Information failure?

Last week's column for Insights argued against Obama's new regulations mandating calorie counts on standard menu items at large chains. A snippet:
The costs of mandatory calorie counts are pretty obvious. In addition to the usual labelling costs, they also make menus a lot more rigid. For places like McDonald’s, it’s pretty easy.

Standard menu items across thousands of locations make it easy to spread the fixed costs of assessing calorie counts and changing the menus. But what about smaller chains that like to encourage their chefs in a bit of experimentation? Trying out new dishes gets a lot harder if you then have to run a nutritional analysis on each one before serving.

And it isn’t just restaurants: a scoop of ice cream at the dairy counts too, if the dairy is part of a chain of 20 or more; movie-theatre popcorn also counts.

It’s always possible that, despite the costs, the information is sufficiently valuable to consumers that the policy makes sense. But there are two pretty big problems.

First, if customers really wanted calorie counts, they’d already have them. Restaurants are highly competitive: they run on thin margins and have to work really hard to draw customers. Most restaurants offer well-labelled options catering to vegetarian or gluten-sensitive customers, helping them to make choices. And some restaurants, like Subway, provide calorie counts.

Second, if customers really found the information valuable, it would show up in changes in the choices they make when presented with calorie information. Instead, field experiments show consumers don’t change their behaviour.

I hope that when people predictably ignore the calorie counts, the government won’t move from labelling nudges to harsher shoves; I also hope New Zealand remains outside of this particular asylum.
A correspondent suggests that I've missed the information market failure that could have motivated the regulations, and suggests that, since OIRA requires cost-benefit assessment, the policy is likely fine. Since we've had a bit of back and forth on it, I thought I'd post my reply here in case others might wish to see it.
There can be an information-failure case for public education campaigns explaining how calories work, what recommended daily calorie limits are, and how extra calories turn into fat. So generalised ad campaigns like “Did you know that eating 10% more than the daily recommended calorie intake will make you x kg heavier by the end of the year?” – I can see a case for that. And if that then has consumers asking restaurants for more calorie information, that’s all fine.

But I remain entirely unconvinced that there exists any real market failure in restaurant information provision to customers. The FDA’s case here is really weak.

In the FDA’s cost-benefit assessment, they note that restaurants are highly competitive, they have to respond to consumer needs, and that if customers really valued nutritional information, restaurants would provide it. I agree entirely. Then this:

“Notwithstanding this point, and although many of the usual market failures that justify regulatory action, such as the existence of market power or public goods, cannot be found here, the primary support for government intervention is an absence of sufficient nutritional information, produced by an inadequate incentive for restaurants to produce that information on their own. An absence of adequate information is of course a standard market failure, justifying disclosure requirements or provision of information in many contexts.”
This isn’t a market failure case; it’s a merit-goods argument! People aren’t demanding as much information as FDA would like (as restaurants would be providing it if customers did want it), therefore market failure? That just doesn’t hold.

They try to hang the whole thing on consumer time preferences being too high at point of sale (discounting future health costs) and argue that presenting the information would increase its salience and promote informed choice.

In the absence of market failure, it’s a lot harder to run the cost-benefit assessment. We can’t presume that consumers find the information valuable, either in gross or net terms. It isn’t at all inconceivable that some consumers in some circumstances have a preference for not being presented with calorie information, like when going out to a fancy dessert place for a celebration. Neither can we conclude from consequent changes in behaviour that the information was valuable, though I think we can conclude that, in the absence of any change, the information likely wasn’t valuable.

When the FDA runs the numbers, the cost-benefit assessment proceeds as though there are nothing but benefits from consumers’ having changed behaviour consequent to seeing the information: otherwise, they’d count some lost consumer surplus against health benefits to get a net. The only tallied costs are costs falling on firms for compliance; there’s no estimation of reduced consumer surplus from menu items that cease to exist because of labelling hassles. On the benefits side, they don’t even try to calculate it – they just say that the policy pays for itself easily if a small fraction of the obese cut their consumption by 100 calories per week. There is no real cost-benefit analysis anywhere in here. Meanwhile, the grocery council estimates that costs on them alone (since they also fall under the regs) would be a billion up front – double FDA’s estimate for the entire regulation. Sure, the grocery council will have had incentive to overestimate compliance costs, but the FDA’s assessment of $1,100 per covered establishment seems really low relative to the constantly-changing menus at grocery delis.

This isn’t so much an ideological point as a methodological one, though I suppose you could say the two link up. When we overturn voluntary consumer decisions in the absence of real market failures other than assumed “they’re not weighting the future enough” and without the ability to see into consumers’ utility functions, any cost-benefit assessment is fraught with difficulty. We can’t just assume information-based market failures where consumers don’t demand the information, and where there are plenty of existing ways for consumers to get the information if they want it. Heck, Samsung provides a free calorie-estimator app as part of the S5’s bundle, and plenty of other free ones already exist.

I really need a convincing story here explaining the market failure and justifying the mandatory provision of information that consumers might not find to be worth its cost. FDA really just points to the potential reductions in obesity and says that any trivial reduction would be sufficient to cover their estimate of the costs of the regulation; where those benefits are borne by individuals who could, right now, choose either to run a calorie-counting app or to frequent restaurants providing calorie information, and who are not doing so, I’m pretty sceptical. If there were some complementary market failure in the restaurant sector – say high barriers to entry and lack of competition, I could start seeing how it works. But that isn’t the restaurant sector. The story requires some fundamental underlying assumption of consumer stupidity or irrationality: I think it’s the kind of policy suitable for those inside the asylum, not those outside of it.
For those interested, here are the FDA's analysis and the OIRA statement.

Monday 1 December 2014

Black Hole Research

Last year, Steven Stillman showed that regression discontinuity design doesn't work very well in identifying the effect of changes in the drinking age on other outcomes.

Suppose you wanted to know whether changing the legal drinking age, or alcohol purchase age, would affect outcomes among the cohort subject to the change. A lot of folks will reach for a Regression Discontinuity Design to compare outcomes for those who've just reached the minimum drinking age with those just under it (who are otherwise pretty similar other than a small age difference), and then claim that increasing the drinking age would reduce harms by the magnitude of that discontinuity extended over the interval, or that reducing the drinking age would similarly increase harms.

The problem with that approach is that you're identifying not just on reaching the legal age, but also on having a big birthday milestone. And you could also be picking up effects of inexperienced drinkers. Pretty hard to tell whether the RDD is really catching a drinking age effect or a mix of birthday and experience effects. Hard, that is, unless you actually have a country that changed the purchase age. Stillman and Boes showed that despite an effect showing up in RDD with the New Zealand change from 20 to 18, that effect doesn't extend to a longer term difference-in-difference analysis that tracks actual changes for the affected cohort.

I'd summarised last year:
They also make an important point on method, and I thank Steven for having explained this one to me slowly; hopefully I've understood it properly. Now recall that they found no effect of the law change on traffic accidents: changing the alcohol purchase age from 20 to 18 did not increase the accident rate among youths. Nevertheless, results from a regression discontinuity design comparing accident rates among kids just under the alcohol purchase age with those just over the alcohol purchase age shows an increase in accidents on reaching the alcohol purchase age after the law change. Why could there be an effect in RDD but not overall? Either the RDD is picking up the effect of the alcohol purchase age on the cohort of inexperienced drinkers who change their behaviour on reaching 18 and experience worsened outcomes, or it's picking up a particular changed behaviour around the time of the birthday. We would have overestimated the effects on traffic accidents for 18-19 year olds by extrapolating from the discontinuity around the 18th birthday.

But, we only know this because New Zealand actually had the policy change so we can compare actual outcomes with those estimated by the RDD. Now, suppose you're in a country that hasn't reduced its drinking age and you want to estimate what the effect of decreasing the drinking age might be. If you ran an RDD looking at the effect of reaching the age of majority on traffic accidents, took the break around the birthday as being the effect of being able to drink, then extrapolated that effect back across the cohort of younger drinkers who would be eligible to drink under a different minimum legal drinking age, you could pretty easily be overestimating the likely effect of a real change in the alcohol purchase age. The RDD picks up the effect of reaching the legal drinking age rather than the effect that would obtain by lowering the legal drinking age.
So - we should be wary of RDD estimates of the effects of the alcohol purchase age.

Well, unless you have preferences over outcomes and want it to look like there's a big effect. In that case, RDD's great.

Toumbourou, Kypri, Jones and Hicki survey the literature other than Stillman's work and conclude that the drinking age should jump. Lindo and Siminski reply in a letter to the Medical Journal of Australia:
Toumbourou and colleagues argue that the Australian legal age for buying alcohol should be increased. However, they overstate their case by only citing research that supports their position, giving an impression of scientific consensus on several key issues when there is strong contrary evidence.
They note the Boes and Stillman work ignored by Toumbourou et al.

Toumbourou, Kypri, Jones and Hicki's reply? That the Boes and Stillman paper can be ignored because it's not yet published.

Second case? A piece in the American Journal of Preventative Medicine, again using RDD to claim that Quebec should hike its drinking age, with zero reference to the Boes and Stillman critique.

If only econ journals published as quickly as do some other fields' journals, or, better, that some other fields took the time to get expert referees in to provide comment on technique. Hopefully this will change when the Boes and Stillman work finally comes out. I'm not all that optimistic, but hopeful.