Thursday 30 August 2012


I guess I'm really bad at keeping secrets. Two years ago, I blogged that I was doing funded work on alcohol.
Because I worry a lot about problems of one-sided skepticism, I wanted to ensure not only that any work I did would be entirely independent but also that it could be seen to be so. Consequently, this work is being undertaken as part of a consulting contract administered by the University of Canterbury, approved of by the powers that be here at the University. I drafted all of the provisions regarding academic freedom in the contract; they were happy with it. The contract guarantees our full academic independence, maintains our ownership of the intellectual property produced, guarantees our ability to publish the work as we like, and reserves to us the right to comment publicly on the work without restriction. I can't imagine what else we could have added to the contract to guarantee academic freedom. An honest application of standard economic method is what's most valuable to all parties concerned.
I presented the work first at the New Zealand Economics Association annual meetings in Wellington in 2011. The first footnote in that paper read:
This project was supported in part by the National Alcohol Beverage Industry Council (NABIC) through a grant administered by the Research & Innovation Office and the College of Business and Economics at the University of Canterbury.
I improved the paper subsequent to feedback at the NZAEs [errors in the older version corrected, though you can get the older version from the conference website to check I'm not lying; I don't want to link it and increase the pagerank of the deprecated version]. I was invited to present at the Australian Conference of Economists Policy Day later in 2011 on problems in economic analysis in health. We released the formal working paper (the improved version) while in town; my presentation at the ACE Policy Day noted that work along with other stuff like the MoH costings on tobacco. From the University of Canterbury's press release of September 2011:
His research, called “The Cost of Cost Studies”, found that at least 75 per cent of this figure would be dismissed if the 2008 researchers had used mainstream economic method.

Dr Crampton worked on the report with Matt Burgess from the New Zealand Institute for the Study of Competition and Regulation in Wellington, and Brad Taylor from the Australian National University. It was commissioned by the Australian National Alcohol Beverage Industries Council (NABIC) through a grant administered by UC’s Research & Innovation Office and the College of Business and Economics.
NABIC approached Dr Crampton to undertake the work after the organisation saw a similar analysis he and Mr Burgess did of the 2009 BERL (Business and Economics Research Limited) report, which was commissioned by New Zealand’s Ministry of Health and ACC to look into the social costs of drugs and alcohol in New Zealand.
And the first footnote in that proper working paper reads:
This project was supported in part by the National Alcohol Beverage Industry Council (NABIC) through a grant administered by the Research & Innovation Office and the College of Business and Economics at the University of Canterbury. We thank Nick Sander and Rachel Webb for excellent research assistance and participants at the 2011 New Zealand Economics Association meetings for comment. We also thank Seamus Hogan for careful review and comment; the standard disclaimer applies.  
We sent a summary of the working paper's findings, highlighting the difference in method between the Collins & Lapsley and BERL method and that found in more standard economics, to the NZ Med Journal. It came out last week. We there have two disclosures. First, in "competing interests":

Competing interests: The underlying study was funded by NABIC (National Alcohol Beverage
Industries Council) through a grant administered by the Research & Innovation Office and the College of Business and Economics at the University of Canterbury. This funding source is also disclosed in the first footnote of the paper. Very strict controls were employed to ensure academic freedom in the conduct of the study, and the only pressure we've been under has been to complete the paper subsequent to earthquake-induced delays.

 Then again in the first footnote when we point to the working paper:

Crampton E, Burgess M, Taylor, B. The Cost of Cost Studies. University of Canterbury
Department of Economics and Finance Working Paper. 2011. Available at (This study was commissioned
and funded by the National Alcohol Beverage Industries Council through a grant administered by the Research & Innovation Office and the College of Business and Economics at the
University of Canterbury.)
I this past week have been busy in Australia. The Australian Liquor Stores Association, one of the constituent members of NABIC, asked if I could come out to their conference as keynote to present my findings to their members; NABIC asked if I could spend a day in Canberra telling media and some Treasury / MoH / ANPHA folks about the work. I mentioned on the blog that I was in Oz for the conference, but I hadn't made a big deal about it; I talk at a few events that I don't wind up blogging about.*

In short, I'm horribly bad at keeping secret that my work on the Collins & Lapsley report (and only that work) has been commissioned and funded by industry via an arrangement that's as iron-clad in guaranteeing academic freedom as I can imagine is possible. When the Sydney Morning Herald's Mark Metherell asked me about independence, I told him what we'd set up to guarantee independence; he still seemed suspicious, so I asked him what else I could possibly have done that would have satisfied him. He couldn't think of anything else, or at least he didn't reveal that he could.

And so it's a bit amusing when the New Zealand Herald's Kurt Bayer writes things like this:

A university researcher who today claimed youth binge drinking has remained unchanged since the legal age was lowered to 18 has had his impartiality challenged after it was revealed his research was financed by the liquor industry.
University of Canterbury economics lecturer Eric Crampton has been slammed by academics [EC: note the plural] today after suggesting there was no strong evidence of increased problem drinking among young people since lowering the alcohol purchase age to 18 in 1999.
There are two very seriously misleading suggestions here, plus a quibble.

First, there is no secret to be revealed. You don't put out press releases about things you want to keep secret. Or at least it wouldn't be the most obvious strategy.

Second, absolutely no part of my having looked at the stats on youth drinking and concluded there to be little obvious basis for cracking down on kids was commissioned, funded, or requested by anybody. Actually, scratch that. The College's media person asked me if I could offer comment on the alcohol purchase age since I work in the area. Because I knew the stats and I knew the holes in the Law Commission's evidence on the topic, I pointed to some of them. And I pointed to more subsequently here on the blog.

And here's the quibble. Bayer uses the plural to claim I've been slammed by "academics". The only person he quotes is Doug Sellman. Doug Sellman is the basis for a headline saying academics are slamming me? Come on.

*  I haven't blogged yet about my talk for the ag industry group Strategic Link a couple months ago, partially because I'm not sure if Chatham House Rules allow me to say what I said. I'm not sure whether I blogged on my talk for the National Business Review's annual strategic retreat up at Waiheke back in the fall.

Alcohol purchase age

I caught a fun call from Christchurch Press reporter Joelle Dally yesterday afternoon. She noted that Doug Sellman disputed my figures and that Sellman claimed I was running a political campaign on the issue. Here are the stats I sent back to the reporter by email after a fire alarm on her side of the call cut the interview short. I must have gotten her email address wrong as none of them showed up in today's story. Anyway, here's what I'd sent:
Here are some of the stats to which I'd point. I'm sure that Professor Sellman would find reason to find a crisis in youth drinking in them, but I've a harder time seeing it.

First, I'll point to the figures from The Social Report that showed no increase in "potentially hazardous drinking" in the 15-24 age cohort from 1996/1997 to 2006/2007. Doug would be right that it could be the case that drinking within the "potentially hazardous" range could have changed, either becoming more or less intense, without showing any change in the proportion in that cohort. But it would seem odd if there were one cohort that were just getting worse and worse, without any changes in the proportion of people in that cohort. The Social Report also shows that there were some demographic shifts within the cohort of "potentially hazardous" drinkers: compared to 1996/1997, relatively more youths in that cohort are of European/Other background compared to Maori and Pacific.

Second, I'll point to ALAC's Youth Drinking Monitor. In 1998, before the purchase age change, about 25% of youths aged 14-18 were non-drinkers. That dropped to 14% shortly after the law change (2000).

But, if we look at more recent figures we find (Table 18) that 88% of 12-14 year olds are non-drinkers, 46% of 15-17 year olds are non-drinkers, and 11% of those aged 18-24 are non-drinkers.

Unfortunately, the age groupings make those a bit hard to compare. But look at those numbers and judge for yourself whether it seems plausible that we've had big increases in youth (age <18) access to alcohol since legalization.

The same pair of reports have 31% of youths 14-18 drinking heavily in 1998 (reported 5+ drinks at last occasion) and 30% of those aged 12-24 binge drinking now (4% of those 12-14, 27% of those 15-17, and 44% of those 18-24). Again, youth binge drinking is a problem. But it's harder to say that it's a problem that's worse now than it was prior to the change in the purchase age; the proportions seem pretty similar.

There should be a very strong burden of proof on those who would impose large costs on all kids aged 18-19 who drink responsibly - the gains of potential reductions in bad behaviour have to outweigh the costs we impose on those who aren't causing problems. Again, this can't rule out that maybe each and every one of those binge drinkers have gone from 5 drinks to 50 (or from 25 to 5).

And, as I'm sure Doug Sellman will have made insinuations about it anyway, I have done work for the Australian alcohol industry contrasting the methods used in estimating the social costs of alcohol with methods used in standard economic analysis. This work was conducted through a consultancy grant handled through the University of Canterbury's Research Office and subject to strict conditions around ethical conduct in research and around academic independence. Indeed, the University sent out press releases last year celebrating some of my work in this area; it's hardly been secret.

I have one other statement of pecuniary interest to make. If a split drinking age goes through, I'm going to make money on iPredict; if we go to either 18 or 20, I'm going to lose a lot. I'm a supporter of "Keep it 18", but I expect the split verdict to obtain unless they change the standing orders on how the vote proceeds.
Joelle didn't include those stats but did include a note from Sellman that the Law Commission had considered all this and still thought that there were problems in youth drinking. Look carefully at the Law Commission's review around youth drinking. Leave aside for now the general stuff about that it's bad that kids drink and look to how they treat evidence around the change in the alcohol purchase age in 1999. The Law Commission correctly notes that there was an increase in youth drinking in the year subsequent to the law change. But they don't say much about more recent trends.

At paragraph 16.15 they cite these bits of evidence on changes:

  • The year following the law change had an increase in ED presentations by drunk kids
    • But if you look at the reports I'm citing above, you find a blip upwards in bad stuff immediately following the law change, which subsequently reversed back to the status quo ex ante. So it wouldn't surprise me if there were an increase in ED presentations in the year subsequent. But if that has followed the same trend as youth binge drinking, it will have reversed.
  • There was an increase in youth drink driving related problems subsequent to the purchase age change.
  • "Our officers report bad stuff" reports from the police
  • Evidence of increased binge drinking by kids through 2000 or 2002
    • nothing about how that trend subsequently reversed back to status quo ex ante.
The rest of the LC's evidence on this question is around how bad youth drinking is in general, not about changes due to the purchase age. And while there's really good evidence from the United States that increasing the drinking age to 21 did a whole lot of good in reducing drink driving by kids, New Zealand seems to have achieved that goal with changes in the drink driving limit for kids. 

This paper analyzes the impact of increases in the minimum drinking age on the prevalence of alcohol and marijuana use among high school seniors. The empirical analysis is based on a large sample of students from 43 states over the years 1980 - 1989. We find that increases in the legal minimum drinking age did slightly reduce the prevalence of alcohol consumption. We also find, however, that increased legal minimum drinking ages had the unintended consequence of slightly increasing the prevalence of marijuana consumption. Estimates from a structural model suggest that this unintended consequence is attributable to standard substitution e ffects.
I stand by that there is not sufficient evidential base for imposing large costs on those moderate drinkers aged 18-20.

Wednesday 29 August 2012


I've been having a bit of fun across the ditch. Back in Christchurch late tomorrow. From today's Sydney Morning Herald:
The NZ economist Eric Crampton said yesterday that a widely cited Australian study had relied on incorrect economic arguments to support ''paternalistic'' policies to combat excessive drinking.

Dr Crampton, a senior lecturer in economics at the University of Canterbury, has had a paper published in the New Zealand Medical Journal stating that only a fifth of the social costs of alcohol abuse asserted by the Australian study could be ''plausibly counted''.

He said many of the costs totalling $15 billion in that study were inadmissable in a standard economic framework.

''My worry has been that while the $15 billion is economically meaningless, it is policy meaningful. If people expect this is a cost to … their back pocket because of other people's behaviour, that increases the demand for certain types of policy,'' Dr Crampton told a briefing in Canberra hosted by the National Alcohol Beverage Industry Council.

Questioned about his independence, given his research and visit was financed by the alcohol industry, Dr Crampton said he was subject to his university's strict controls to ensure academic freedom. [What, would SMH have paid to have me come over and talk to their journalist? Airfares aren't free!]

The chief executive of FARE, Michael Thorn, said the study Dr Crampton questioned was in accordance with the approach laid down by the World Health Organisation.

He challenged the alcohol industry and Dr Crampton to generate a new estimate of the cost of alcohol abuse.
Usually the burden of proof is on those wanting to change policy to come up with studies showing why they're right; I would have thought that simply showing the evidence in favour of further restrictions to be highly flawed would have been enough. But perhaps not.

Coming up with a new estimate of the social cost of alcohol use, using a sound economic method, falls pretty heavily into the "this is going to be a pretty big job" file.

Tuesday 28 August 2012

The Dismal Science

Continuing this jet-propelled, monkey-navigated blast down the alkali flats of economics blogging, I'm proud to announce the launch of a new New Zealand economics blog: The Dismal Science. Our launch post:
Thomas Carlyle called economics the “Dismal Science”, in contrast to what he termed the “Gay Sciences” of poetry and literature. The Dismal Science feed at SciBlogs brings top commentary from the New Zealand economics blogosphere to those more familiar with the Bench Sciences. Curated by the University of Canterbury’s Eric Crampton, who blogs at Offsetting Behaviour, the feed picks posts from our country’s top economics blogs, including Anti-Dismal (Paul Walker), Fair Play and Forward Passes (Sam Richardson),Groping Towards Bethlehem (Bill Kaye-Blake), Offsetting Behaviour (Eric Crampton and Seamus Hogan) and The Visible Hand in Economics (Matt Nolan, James Zucollo and co-bloggers).
The Masthead at The Dismal Science borrows Fleeman Jenkin’s illustration of barter as a rather beautiful dance, with lines of exchange among individuals.
Although people sometimes think that economics is just about money or business, it is much broader. Economics analyses individual choice under conditions of scarcity and how those individual choices aggregate into social outcomes – the dance Jenkin illustrates. Posts at The Dismal Science will reflect that rather broad domain of analysis.
We wear Carlyle’s epithet proudly: Carlyle deemed economics Dismal because of John Stuart Mill’s opposition to slavery. Economics insists on an approach based in methodological individualism – that individuals are the fundamental unit of analysis, that their individual subjective valuations as revealed by their choices forms the basis of economic notions of welfare, and that all individuals’ welfare get to count. The economic approach is antithetical to Carlyle’s insistence on a hierarchy of men where, by his assumption rather than revealed by their choice, one group are made better off by being enslaved.
We're still working out some back end issues to let me efficiently curate the different inbound feeds. When everything is working right, I'll see a morning dashboard with a list of new posts up at the source blogs that their authors deemed worthy, then schedule them for appearance at Dismal. I'd also like to be able to pull classic posts from our combined back archives when topics like capital gains taxes or stadiums become timely. Peter Griffin, the Editor at SciBlogs, is seeing what we can do to set up the system's back end. In the meantime, I today there highlight a worthy post from Groping Towards Bethlehem.

Alcohol and heart attacks

Another finding for the pile: among male heart attack survivors, all source mortality risk for moderate drinkers (1-3 standard drinks per day) was 66% of the risk for non-drinkers. Here's the latest from Pai, Rimm, & Mukamal:
Moderate alcohol consumption of up to two drinks per day was significantly inversely associated with both all-cause and cardiovascular mortality in basic and adjusted models. The multivariable-adjusted HRs for total mortality were 0.78 (95% CI: 0.62–0.97) for 0.1–9.9 g/day, 0.66 (95% CI: 0.51–0.86) for 10.0–29.9 g/day, and 0.87 (95% CI: 0.61–1.25) for ≥30 g/day (P for quadratic trend = 0.006) (Table 2). Similarly, for cardiovascular mortality, the adjusted HRs were 0.74 (95% CI: 0.54–1.02) for 0.1–9.9 g/day, 0.58 (95% CI: 0.39–0.84) for 10.0–29.9 g/day, and 0.98 (95% CI: 0.60–1.60) for ≥30 g/day (P for quadratic trend = 0.003). Additional adjustments for caloric intake, omega-3 fatty acids, ST-elevation and site of initial MI, and creatinine did not appreciably attenuate the risk estimates. Although HRs were similar across alcoholic beverage types, in analyses mutually adjusted for beer, wine, and liquor, there was no significant association between specific beverage type and mortality
Ok, what if the people who were sickest after their heart attacks quit drinking and confound the measure? They also check mortality against pre-MI drinking patterns and find pretty much the same result: the adjusted hazard ratio increases from 0.66 to 0.7.
Our results clearly support the hypothesis that long-term moderate alcohol consumption among individuals with prior MI may be beneficial for all-cause and cardiovascular mortality. This U-shaped association may be stronger among men with better long-term prognosis after MI and further examination is warranted to determine the suitability of moderate alcohol consumption among individuals with other severe cardiovascular conditions.
Wonder if the doctors here note this kind of advice...

Sunday 26 August 2012

Mitigating Risk

It looks like the local elementary school came to the same conclusion we did about local tsunami risk. South New Brighton elementary has bought life jackets for all the students; we still need to figure out where to source a family-pack.

Any large tsunami is likely to come with plenty of evacuation notice as it would reach us after a large quake off the coast of South America. But there also is some risk of a locally generated one that would come with less than 20 minutes notice. The Avon River separates Southshore, South Brighton, and New Brighton from any higher land except that which might be reached by a very long drive North. After the February earthquake, none of the bridges could be crossed by vehicles. So it could very easily be very impossible to evacuate after a local tsunami-causing earthquake.

Our emergency plan for anything big and local? Sit on the roof of our two-story house with a life-jacket. Our house is on a two meter rise from the street; the street is about two meters above sea level; the roof is about 4 meters from the ground. I can't see how we can do better other than by having a quick-inflate emergency life raft up there. But those aren't cheap. And a locally generated tsunami would likely to be small enough that a life jacket could reasonably reduce the risk.

Saturday 25 August 2012

Alcohol stats

I wrote this for the op-ed section of Friday's Christchurch Press. I'm currently in Oz where I'll be talking about alcohol policy for the hospitality industry association liquor retailers conference and with a few journalists. So I'm posting this Wednesday and hoping that the piece did come out in Christchurch on Friday. Anyway, enjoy. [Update: here it is!]


It would be pretty easy for diligent Christchurch Press readers to conclude that alcohol is the worst scourge of humanity. Doug Sellman featured several times over the last fortnight warning us all of alcohol’s dangers; health reporter Georgina Stylianou wrote two articles on a new estimate of alcohol’s cost to the Canterbury health budget. And the University of Otago’s Professor Jennie Connor told us that minimum prices are remarkably effective in reducing alcohol consumption, pointing to a Canadian study “that showed a 10 per cent increase in the minimum price of alcohol reduced consumption by 16 per cent relative to other drinks.” Connor was quoted in a Mainlander feature on Doug Sellman explaining how Professor Sellman is not in fact a wowser; presumably it’s only free-spirited libertines who write op-eds in the New Zealand Herald worrying about how Woodstock commercials encourage boys “to have sex with their best mates’ mothers” and that advertisements “encouraging middle-aged women to get their teenage clothes back on and flirt with their son’s best mates” do not contribute to a healthy society.  (30 October 2009)

I do not particularly care what the jury decides on who is or is not a wowser. But I work with and care about the numbers around alcohol policy. And the impression most readers would get from the latest reporting in the Press is a bit at odds with, well, reality.

Let us begin perhaps with Jennie Connor’s citing of “a Canadian study” on the effects of minimum pricing. Can a ten percent increase in the minimum price of alcohol really reduce total alcohol consumption by sixteen percent? No. The number is so far out of line with the vast consensusof the literature that I got in touch with Chris Auld, one of the authors of what had to have been the study Connor had read, a piece in the May 2012 issue of Addiction. And my read of his study was right. If you increase the price of beer by ten percent, you could get a reduction in beer consumption of sixteen percent, but that’s because beer drinkers move over to other alcoholic drinks. Across-the-board increases in the minimum price of alcohol have far smaller effects: a ten percent price increase reduces aggregate consumption by only about 3.4 percent, as is made reasonably clear in Auld’s paper.

That the sixteen percent figure is screamingly wrong should have been obvious to anybody who is familiar with the literature. The largest estimate of the responsiveness of alcohol consumption to price measures in Wagenaar’s 2009 survey of 112 different studies suggested that a 10% price hike reduced consumption by 8.4%; Wagenaar concluded that 4.4% is the best consensus estimate. And for heavy drinkers, it was 2.8%. Auld’s 3.4% is then right where a reasonable person would have expected an estimate on the effects of minimum prices might fall. Sixteen percent – that’s right out.

The Auld study does come out in favour of minimum prices, but in conjunction with a decrease in alcohol excise taxes: it’s then a way of increasing the cost facing heavier drinkers while doing less harm to moderate drinkers if heavier drinkers drink cheaper alcohol. I worry more that the policy does disproportionate harm to moderate-drinking poorer people.

I was a bit more surprised to read of the new commissioned BERL report on the health costs of alcohol in Canterbury. Their prior work in the field, a 2009 study that included a measure of costs to the health system among other costs of alcohol, seemed to put a pretty heavy thumb on the scales in estimating the costs borne by the health system. My article in today’s issue of the New Zealand Medical Journal lists a few of those problems [note: link may only be active after lunchtime Friday, sorry]. While alcohol greatly increases the burden of disorders like liver cirrhosis, it also reduces the costs of cardiovascular disease. BERL here replicated work done in Australia by Collins and Lapsley (2008). But where Collins and Lapsley added up all the costs imposed by those disorders where alcohol makes things worse and subtracted from that total all the cost savings from those disorders where alcohol reduces costs, BERL simply erased any beneficial effects of alcohol for disorders including ischaemic heart disease, cholelithiasis, heart failure, stroke and hypertension.

As I had warned BERL against this kind of method when I served as discussant on their paper at the 2009 New Zealand Economic Association Meetings, I was curious whether they’d revised things. But, the new BERL paper wasn’t available. Two separate stories in the Christchurch Press, with lots of reaction quotes from doctors, were based on a paper that the reporter did not have and was not yet available for public critique. I received the paper Wednesday courtesy of the CDHB. And BERL, at footnote 14, reports they’ve done the same thing again: “The Collins and Lapsley fractions indicate some alcohol use may be beneficial for some conditions. We concentrate on harmful drug use, and assume zero fractions for such conditions.” So their measure of the costs of alcohol to the Canterbury health system relies on an assumption that there can be no health benefits from alcohol – an assumption that runs contrary to the weight of international evidence. Assuming one’s conclusions is hardly proper method.

Unfortunately, alcohol policy is one of those areas where a lot of people’s convictions about the right answer put a pretty strong lens on how they assess the literature. How often do you read that problem drinking among 15-24 year olds was no different in 2006/2007 than in 1996/1997 before the change in the alcohol purchase age? Or that per capita alcohol consumption is down substantially since 1991? Or that light drinkers have about a 14% reduction in their chance of dying from any cause than people who never drink,correcting for the host of other health-related behaviours that are usually given as reasons for ignoring the health benefits of moderate drinking?

Be skeptical of the moral crisis around alcohol.

Friday 24 August 2012

A symposium, of sorts

The latest issue of the New Zealand Medical Journal features three papers on alcohol policy, including one from Matt Burgess, Brad Taylor and me, along with a commissioned editorial piece on the set. I have not yet had a chance to read the other two papers in the series but the editor of the journal kindly forwarded along a copy of the editorial piece late Thursday night.

Doug Sellman, lead author on the editorial piece, says about what I expected he would say about our work on alcohol.

Our piece in the NZMJ contrasts the standard economic approach to costs and benefits with that employed by BERL in their work. We worked through the different cost categories in the BERL report to illustrate the differences between a standard economic approach and the cost of illness method they used before spending a bit of time on the importance of marginal analysis rather than measuring total benefits and total costs of anything.

Sellman's piece, or at least the part of it criticising us rather than lauding the other two papers, begins with* a critique of our focus on external costs rather than costs borne by drinkers themselves; he pitches this as being responsible for the $4 billion difference in measured cost. I would note here that about a billion and a half of that $4 billion difference comes from BERL having double-counted intangible costs of statistical lives lost with forgone earnings. Even if your focus were entirely on total costs, there's a big chunk of that total figure that's just wrong.

I'm also not sure if he's amplifying for rhetorical effect, didn't understand the point, or whether we didn't phrase it clearly enough. But he writes:
The narrowness of the Crampton group’s approach is exemplified in the statement that “the only policy-relevant costs [of a fatal drink driver accident] are those imposed on emergency services in responding to the accident”. Equally the lost productivity through mortality reveals the dismal view from the world of this particular economic approach: early death is not seen as a cost because the economic unit (the deceased person) can simply be replaced.
In the case of drink driving accidents, we were very clear that deaths to those outside of the drinker's vehicle impose very large costs that are definitely deemed external and policy relevant. But if you were considering a single vehicle accident with a solo driver where nobody else was hurt and only the driver's own car was wrecked, then we would only count the costs to emergency services. There costs of deaths associated with that kind of crash remain very real, but if you want to count them, you have to also count the benefits enjoyed by those who took similar risk and didn't have an accident.

Similarly, Doug doesn't quite get why we discounted costs of forgone output. BERL used a measure of GDP per capita as forgone output rather than using the deceased worker's prior wages; that addition requires very strong assumptions around worker irreplaceability and capital-labour complementarity. But the main reasons we discounted lost output were first that it's double-counted with the intangible costs of lives lost, so it should be reduced even if you want to count private costs, and second because it's largely a cost borne by the drinker himself.

But here's my very favourite paragraph in the whole thing.
From a non-economist standpoint Crampton et al would appear to be the fundamentalists in this debate and their conclusions need to also be considered in the context of their receiving funds from an alcohol industry, which benefits from their conclusions. Further, it is important to note that “standard (neoclassical) economic approaches” are coming under increasing scrutiny and criticism, particularly following the recent global financial crisis, which was not predicted by “standard” economic models. New economic models based on better science and more rigorous mathematics are now progressing.5 [cite to Keen!]

I'm pretty sure that these can't all be true at once:
  1. I'm a fundamentalist
  2. I'm bought out by industry, saying whatever they want
  3. We should throw out neoclassical economics because it yields conclusions like mine and the financial crisis too.
I suppose 1 and 3 could go together if I'm a fundamentalist economist. But the thing is, you can't really buy out a fundamentalist. A true zealot can be encouraged to spend more time on one area of work rather than another [and a consultancy contract that needs to get finished, earthquakes or not, does sharpen incentives on that front], but you can't buy a fundamentalist's conclusions. It's perhaps worth noting, again, that the net effect of industry funding on conclusions, when it comes to my analysis of the BERL study, was to substantially increase the measured external costs of alcohol use: thoroughly understanding the model on which the BERL study was based led us to find an error in our prior unfunded work - one not noted either by BERL or by Marsden Jacob in its $60k hatchet job on that prior piece.

* Actually, it begins with Doug making a trivial but risky error: he says that BERL produced its report for the Law Commission's review of the alcohol literature. I understand that the Law Commission, back in 2009, sent out a couple of threatening letters when they were characterised as having commissioned the BERL report. I doubt it'll happen here, because LC isn't being criticized (incorrectly) for having commissioned a shonky report. They didn't commission a shonky report; the Ministry of Health and ACC did. The Law Commission relied on a shonky report. That's different.

Thursday 23 August 2012

Tobacco excise incidence

The latest paper by Callison and Kaestner* makes me a bit more worried about the incidence of Tariana Turia's proposed excise increases. Recall that the MoH is modelling things using a fairly high estimate of price elasticity of demand. At paragraph 55 here, MoH says:
the price elasticity of demand (ie. the extent to which each percentage increase in the tobacco price causes consumption to fall) - current estimates are that each 10% rise in tobacco prices will engender about a 4-5% drop in tobacco consumption, but based on recent Treasury revenue returns and behavioural research by Auckland University, there is emerging evidence that price responsiveness may be increasing.
Presumably the -0.4 to -0.5 elasticity estimate is what they had NZIER use in their analysis; MoH seems to be suggesting this is a lower bound.

Callison and Kaestner start by surveying the existing literature.  My clipping from their survey:**
  • Lewit and Coate (1982): adult smoking participation elasticity (age 35+) of -0.15; younger smokers more elastic.
  • Evans and Farrelly (1998); Farrelly et al (2001): same dataset as above; younger smokers respond to price changes; adults aged 40+ not responsive.
  • Wasserman (1991): Adult participation elasticity -0.17 in 1985.
  • Ohsfeldt (1998): Participation tax elasticity for teen and young adult males of -0.15 to -0.22, but only -0.07 for males over age 45.
  • Tauras (2006): Same dataset as Ohsfeldt but more recent years' coverage: participation elasticity among adults 18+ of -0.12.
  • DeCicca and McLeod (2008): Post-2001 data. Participation elasticity of -0.3 for 45-59 year olds and -0.2 for 45-64 year olds. 
  • Gallet and List (2003): A survey of existing elasticity estimates giving a price elasticity of demand (I read this as combination of participation and intensity elasticity) of -0.32 for adults aged over 24.
When O'Dea ran his analysis, he used a price elasticity of demand of -0.5 and a participation elasticity of -0.2. 

Callison and Kaestner use a difference-in-difference method comparing smoking prevalence in states with large excise increases with those in states that didn't increase taxes. Among adults, they found participation and price elasticities of demand on the order of -0.02 to -0.05: a 10% tax increase reduces consumption by -0.5% or less, not 2% and not 5%. They write:
Considering all the evidence, we conclude that there is insufficient justification for the widespread belief that raising cigarette taxes will significantly reduce cigarette consumption among adults, even young adults. Our evidence suggests that, at best, increases in cigarette taxes will be associated with a small decrease in cigarette consumption and that it will take very sizeable tax increases, on the order of 100%, to decrease smoking by as much as 5%.
Maybe that's why the Australian Treasury modelled tobacco excise revenues on the assumption that excise revenues will be increasing.

Recall that, on an assumption of a -0.2 participation elasticity, O'Dea reckoned that, for a 20% price increase, 129,200 households in the bottom four deciles that continued to smoke would suffer losses of $396 per household; these were balanced by gains to 5,400 quitting households of $2,988 per household in cash savings (no accounting for reduced consumption benefits for those who enjoyed smoking) plus 2 Quality Adjusted Life-Years in each household. They value a QALY at $50k. So losses to non-quitting households for the 20% tax increase total $51.2 million while gains to quitting households total $16.1 million plus $540 million in QALY benefits. 

Callison and Kaestner say that the participation elasticity isn't -0.2, it's -0.05 to -0.02. What happens if you simply halve O'Dea's participation elasticity estimate to -0.1? Then half as many households quit: we have 2700 quitting households (in the bottom 4 deciles) enjoying aggregate cost savings of $8 million  plus $270m in QALY against 131,900 non-quitting households paying an extra $52.2 million in tax. If we go all the way to the -0.02 estimate, then a tenth as many households quit: 540 quitting households gain a total of $55.6 million (including QALY) and 134,060 non-quitting households suffer aggregate losses of $53.1 million.

The O'Dea numbers are calibrated around a 20% price increase from 2005 prices and prevalence levels that then obtained; MoH has been talking about excise increases well in excess of that. I'd be nervous about applying the Callison and Kaestner estimates to New Zealand as total prices here are well in excess of anything they'd have in the US data. On the other hand, non difference-in-difference estimates from the US aren't crazily out of the ballpark on participation elasticities: O'Dea was using -0.2, which is well in line with the US studies that used somewhat less sophisticated techniques. 

Note as well that if reduced effective income hurts QALY measures, then we perhaps need to add QALY costs on the non-quitting side as the poorest households effectively get poorer. 

Another cool bit of the Callison and Kaestner paper: they use only MSA data and correct for distance to nearest lower-tax jurisdiction to try to correct for leakage from smuggled product - the find little effect on their estimates from cross-border purchases. But if it is the case that smuggling rings are a bit more organized and folks with semi-loads of smokes from lower-tax jurisdictions don't much care whether they're driving 50 miles or 250 miles, then smuggling might still confound their measure: it probably well captures individuals' decreased likelihood to drive out of state to get smokes, but it might not catch organized groups handling the transport. As each state has its own particular tax stamp for cigarette packets, such opportunities may be limited. 

If New Zealand goes ahead with very large tobacco excise increases, the government might consider some offsetting income tax cuts for lower decile groups. Callison and Kaestner didn't split their sample by income decile rather than just by age group to see whether price responsiveness varied by income; price elasticities split by income could be important in assessing overall incidence of the NZ changes. But as smoking is fairly concentrated among lower decile groups, overall estimates likely aren't far from estimates restricted to lower decile groups. 

It could also be fun to have iPredict run markets on aggregate tobacco excise revenues over the next few years. 

* Previously noted here.

** Note that the participation elasticity is the percent change in the proportion of the sample who report smoking given a percent change in the price. So if a 10% price increase is associated with 5% of the sample quitting, the participation elasticity would be -0.5. MoH above is using a total consumption elasticity that combines those quitting with reductions in smoking among those who don't quit entirely. Callison and Kaestner also list some of the smoking intensity elasticities: the percentage change in the amount that you smoke (conditional on your smoking) given a percentage change in price. These values are usually about half of reported participation elasticities.

Wednesday 22 August 2012

Coercion everywhere: organs edition

I had an awfully fun time a few weeks back guest-lecturing in an honours health ethics course in  Canterbury's Health Sciences programme. They wanted to know how an economist approached ethical issues around organ donation. It was pretty clear pretty quickly that the grad students there hadn't been exposed to economic arguments before; they were, on the whole, remarkably receptive. The lecturer, a bit less so. And so it was great fun.

I started by explaining how economists go about estimating the value of a statistical life, the importance of such estimates in drawing the appropriate balance between policies that mitigate risk and those that enhance the quality of life, and the importance of individual risk assessments in deriving those estimate. I moved then to compare the risk of workplace death in various industries with the risk of death from voluntary live kidney donation - live kidney donation is pretty safe. If we're prepared to let people accept cash for risky things like working on a fishing boat, why aren't we prepared to let people accept cash for risky things like donating organs?

The lecturer worried a lot about coercion. I yesterday noted the distinction between voluntary and euvoluntary exchanges as highlighted by Mike Munger. I noted it there too. Sales of organs seem pretty likely to fail the euvoluntary test - income pressures could drive a lot of donation decisions. But that's also true of decisions to work on a fishing boat or to take a lot of other unpleasant and risky jobs. What makes the "coercion of being poor" so much worse for the decision to sell a kidney than for the decision to take a job on a fishing boat if the risks of death or other adverse health consequence from the two decisions are roughly comparable?

The students seemed pretty happy with the notion that if we let poor people trade risk for income on fishing boats, it's a bit odd to ban them from taking roughly the same level of risk for income from selling a kidney, especially as the latter can save lives. The lecturer wasn't as keen on the idea. But I really couldn't pin down just what made the two kinds of decisions different except for that one involved organs, despite rather a bit of pressing. At one point she was backed into the (to me) clearly untenable position that workers on commercial fishing boats take the job not for the income but because they love fishing, then denying people were taking money in exchange for risk. 

I'm still puzzled about why it's obvious to many people that transactions around organs are inherently coercive but those involving very comparable levels of risk or unpleasantness that don't involve organs aren't. Bans on trade based around inarticulable squeemishness concerns do have effects. The National Business Review reports that New Zealand has imported human tissues from RTI Biologics' subsidiary Tutogen [paid link - get a subscription!]; Tutogen has gotten into a bit of trouble about how it sources its parts. Caleb Allison at NBR points to the ICIJ report that shows how pretty much everybody in the system, and especially the companies trading in tissues, are able to profit from individuals' uncompensated donation decisions. Banning trade in such things here, where we could be pretty confident that harvested tissues could be registered and traced back to source, just pushes the problem to places where the regulatory regime may well be less sound. We would have fewer problems in sourcing organs and tissues domestically were we able to compensate donors. 

I'd love to hear a comprehensible reason why we ban compensation in this area but allow it for taking risky jobs in mining, logging, and fishing.

Update: do read EuvoluntaryExchange - the pdf above-linked has the in-one-place version, but the blog is where the argument lives

Tuesday 21 August 2012

Craft Beer is great...but...

I love New Zealand craft beer. Well, all kinds of craft beer, really. But I love keeping partial and total correlation straight even more. And I'm not sure Richard Florida does a great job of that here. He writes at the Atlantic:
Only four states come in the top ten in both both lists (total and per population): Washington, Colorado, Oregon, and Wisconsin. [Total number of craft breweries and number of craft breweries per 100,000 population]
With the help of my colleague Charlotta Mellander, I took a quick look at how the concentration of craft breweries per capita correlates with key state demographic and economic characteristics. As usual, I point out that correlation points only to associations between variables and does not imply causation. Other factors may come into play. Still, this analysis points to a number of interesting patterns.
Given how much craft beer costs, you might think income would play a role, with craft breweries more concentrated in more affluent states. But we found no statistical association between craft breweries and income, wages, or per capita economic output.
Education does factor in: Craft brewing is more concentrated in more highly educated states, with a modest correlation to the share of adults that are college grads (0.32).  
And craft brewing is more closely associated with higher levels of happiness and well-being (0.47). 
On the flip side, craft brewing less likely in conservative states, with a modest negative correlation (-0.3) to 2008 John McCain votes (there was no statistically-significant association to Barack Obama votes).
Craft brewing was far less likely in religious states — the correlation between religion and craft breweries was the strongest of any variables (a whopping -0.75).
Curiously, there was a negative connection between craft breweries and two other unhealthy behaviors or "sins" — smoking (-0.28) and even more so with obesity (-0.54). 
Surely the reasonable correlation between education and craft brewing concentration is likely driving the negative correlation between craft brewing and either smoking or obesity. The kinds of places likely to favour craft beer are likely the kinds of places that are just filled with awesome people doing awesome things, and that kind of underlying heterogeneity is likely driving the other results.

How does NZ do? Florida reports states range in craft brewery concentration up to 3.68 per 100,000 people. New Zealand has about 4.4 million people; we have about 66 breweries that would count as craft by production volume measures used in the US.* So we have about 1.5 craft breweries per 100,000 people: that would likely put us in the Top 10 among US states as only 13 have more than 1 craft brewery per 100,000 people.

NZ's beer awards came out last week. Christchurch did very well, with Harrington's picking up the Top Brewery award. I will endeavour to try all the winners I haven't before sampled.

* 6 million barrels or less per year; I would guess that only Lion and DB here have any chance of exceeding that threshold. NZ counts 40,000 litres or less per year as "craft" - 30 breweries meet that standard. 

NBER roundup [updated]

The late-night NZ-time twitter feed occasionally brings American morning delights. Tonight, it's the new NBER working papers. In the queue for when I'm back on campus and can read the NBER subscription papers:*
  • Callison and Kaestner find tobacco consumption less price sensitive than previously thought; they reckon it would take a 100% tax increase to get a 5% drop in consumption. This would be estimated around American tax levels, which are rather below NZ ones; I'd need to back that out into price elasticity estimates to translate it into effects of NZ tax increases. From the abstract:
    ...we focus on recent, large tax changes, which provide the best opportunity to empirically observe a response in cigarette consumption, and employ a novel paired difference-in-differences technique to estimate the association between tax increases and cigarette consumption. Estimates indicate that, for adults, the association between cigarette taxes and either smoking participation or smoking intensity is negative, small and not usually statistically significant. Our evidence suggests that increases in cigarette taxes are associated with small decreases in cigarette consumption and that it will take sizable tax increases, on the order of 100%, to decrease adult smoking by as much as 5%.
    If that's right, Turia's tax increases are more regressive than we'd expected.

    Update: Oh wow. Read this bit from the full paper:
    Using this method, we found that for adult smokers ages 18 to 74, a 10% tax increase is associated with between a 0.3% to a 0.6% decrease in smoking participation and a 0.3% to a 0.4% decrease in smoking intensity. More surprisingly, given past research suggesting that youth smoking is more sensitive to taxes and prices, we find very little difference by age in the association between cigarette taxes and cigarette consumption. A 10% increase in state cigarette tax is associated with: between a 0.3% to a 0.7% decrease in smoking participation for those ages 18 to 34; between a 0.2% to a 0.4% decrease in smoking participation for those ages 35 to 54; and between a 0.3% to a 0.6% decrease in smoking participation for those ages 55 to 74. Similarly a 10% increase in state cigarette tax is associated with: between a 0.3% and a 0.5% decrease in smoking intensity for those ages 18 to 34; a 0.3% decrease in smoking intensity for those ages 35 to 54; and between a 0.3% and a 0.4% decrease in smoking intensity for those ages 55 to 74. Finally, standard errors of estimates are of a magnitude that rule out cigarette tax elasticities with respect to smoking participation (intensity) among adults greater (more negative) than -0.12 (-0.13).
    It is notable that estimates in Table 3 provide no evidence to support the hypothesis that smoking behavior is more responsive to taxes (prices) among younger persons than older persons.
    There may be effects in encouraging kids younger than 18 to avoid starting smoking; the analysis here is restricted to adults. But if this is right, it means that whatever benefits come from Turia's Tax will be over a very long time horizon while the costs on low decile households through reduced net-of-smoking disposable income will be very large for a rather long time. I probably ought to pull this up to being its own post. There are rather a few careful controls in here that need more discussion.

  • Reyes gives more evidence that banning leaded gasoline, and other restrictions on environmental lead, was a very good idea. From the abstract:
    The paper finds that elevated levels of blood lead in early childhood adversely impact standardized test performance, even when controlling for community and school characteristics. The results imply that public health policy that reduced childhood lead levels in the 1990s was responsible for modest but statistically significant improvements in test performance in the 2000s, lowering the share of children scoring unsatisfactory on standardized tests by 1 to 2 percentage points. Public health policy targeting lead thus has clear potential to improve academic performance, with particular promise for children in low income communities.
    Reyes previously estimated that reductions in environmental lead can account for a 56% reduction in violent crime in the 1990s.

    Update: The paper gives some nice benchmarking of the effects of lead reduction: the improvement in test scores that came of the reduction in the proportion of low income kids with high blood lead concentrations would be comparable to the improvement in test scores you'd expect if per capita incomes improved by 15% in low income communities - a rather substantial effect.

  • Hastings et al provide more evidence that kids winning lotteries allowing them to attend the school of their choice enjoy better outcomes

  • Fergusson, Robinson, Torvik and Vargas set up a model testing an Orwellian idea: that leaders whose power is augmented by warmaking have little incentive to let the war end. They test against Colombian data. From the abstract:
    We find that after the three largest victories against the FARC rebel group, the government reduced its efforts to eliminate the group and did so differentially in politically salient municipalities. Our results therefore support the notion that such politicians need enemies to maintain their political advantage and act so as to keep the enemy alive.
    War is the health of the state...

  • And, finally, Lacetera et al on compensation for marrow and organ donation. The abstract:
    In an attempt to alleviate the shortfall in organs and bone marrow available for transplants, many U.S. states passed legislation providing leave to organ and bone marrow donors and/or tax benefits for live and deceased organ and bone marrow donations and to employers of donors. We exploit cross-state variation in the timing and passage of such legislation to analyze its impact on organ donations by living and deceased persons, on measures of the quality of the organs transplanted, and on the number of bone marrow donations. We find that these provisions did not have a significant impact on the quantity of organs donated. The leave legislation, however, did have a positive impact on bone marrow donations. We also find some evidence of a positive impact on the quality of organ transplants, measured by post-transplant survival rates. Our results suggest that these types of legislation work for moderately invasive procedures such as bone marrow donation, but may be too low for organ donation, which is riskier and more burdensome to the donor.
    Becker and Elias reckoned it would take about $15k in compensation to encourage kidney donation. If the tax benefits added up to less than that, it would have been surprising if there had been large effects on live donation rates. The data appendices (free access; the article is gated) shows no state provided more than $10k as tax deduction. Note that a $10k tax deduction isn't $10k in hand: it's $10k that you get to remove from your taxable income total. So it's only worth $10k times your marginal tax rate - in other words, very unlikely to motivate donation from the cohorts more likely otherwise there to be price sensitive.

    Update: A few neat bits on seeing the paper rather than just the abstract:
    • Where some worry that paying for organs worsens quality, the authors found instead weak evidence of quality improvement.
    • The authors seem to have reached the same conclusion: the payment levels via tax deductions are likely below the reservation price for live kidney donation.
* I've only caught the abstracts of these thus far; if there are grievous errors in method that aren't obvious from the abstracts, my apologies.

Monday 20 August 2012

Socialising private costs

What's a good way to make sure that smokers cost other people money? Make it illegal for private health insurers to charge smokers more. Seems a particularly stupid policy, but Australia's going for it.
Private health insurers have failed in a push to charge smokers higher premiums. While the federal government triumphed over big tobacco this week, the Minister for Health, Tanya Plibersek, said smokers should not be penalised with higher premiums - and people should pay the same regardless of the state of their health. She ruled out allowing the health funds to charge smokers more.

Private insurers - led by NIB - are interested in the idea and argue it is the next step the government could take to reduce rates of smoking.

''We should be able to offer a discount for good healthy behaviour like not smoking and exercise,'' the chief executive of NIB, Mark Fitzgibbon, said.
The Australians have a pretty extensive regulatory apparatus around "private" health insurance; see Robson et al for a recent analysis. I'd suggested in the introduction to that issue of Agenda that whatever goals are sought by community rating could better be achieved by providing transfers to individuals based on the exogenous parts of their individual risks and letting insurers set actuarially fair premiums: poorer and less healthy people are given money to help them buy insurance, but incentives to make efficient investments in your own health are maintained.

Once you've banned insurers from charging differential fees for smokers, all kinds of anti-smoking policies can have second-best cases supporting them; first best would just let insurers set fair premiums. Government creates the "market failure", then supplies the interventions that partially offset the failure. And everybody gets to shout about how markets are terrible because absent intervention, everybody else has to share the costs imposed by smokers.

Coercion everywhere: welfare edition

It's hard to draw the line between coercion and choice. Some people see coercion in normal market transactions between consenting adults where relative wealth differences are large - prostitution markets are often banned as somehow coercive; most countries would ban me from selling you one of my kidneys for fear that the money offered had coerced me. I don't see any of those as being coercive; others do.

Mike Munger defines exchanges as being Euvoluntary if they satisfy a few requirements, the last of which likely motivates bans on organ sales, prostitution and price gouging. I'm going to highlight the two that deal with coercion:
(4) neither party is coerced, in the sense of being forced to exchange by threat, and (5) neither party is coerced in the alternative sense of being harmed by failing to exchange.
Munger doesn't argue for bans on voluntary transactions that fail to meet the more stringent conditions for Euvoluntary exchanges; indeed, he says they remain just and useful. I'd argue transactions failing requirement (4) ought to be banned because we have pretty strong expectations that one party is left worse off by the transaction, but that those failing requirement (5) should generally be allowed.

National announced changes around welfare policy a few months ago.* One proposed change would provide free contraception for young women receiving benefits and for the teenaged children of families on benefits. Most people, at least by wholly unreliable web-polls on news-sites, think this an excellent idea. With over 12,000 votes cast, 85% say it's great. I also think it's a pretty good idea.

Some of the Kiwi Twitteratti worry whether there's a slippery slope towards stronger forms of encouragement for long term contraception use for women on benefits. They could be right. But, would that kind of requirement be coercive?

Receipt of various benefits already comes with a laundry list of conditions. If you're getting the accommodation supplement, you have to report family income. That gives some people incentive either to hide that they're living with their partner or, worse, to have one partner leave. Here's one list of the requirements around disclosing to Work and Income whether you're "in a relationship for income assistance purposes". Is it coercive that, if you care about being truthful, you may be forced (in the (5) sense) to avoid entering into a relationship? Whatever the costs of coercion in those cases, they seem outweighed by the ability better to target benefits to those most in need.

I have a hard time seeing how adding contraception as condition of receipt of benefit is different in kind from the other forms of coercion that already surround receipt of welfare payments. People still choose whether to accept the bundle of restrictions and payments. The exchange fails to be Euvoluntary as (5) is definitely violated. But (5) is pretty likely to be violated if any conditions are attached to welfare receipt.

Lew Stoddart thought I was engaging in moral gymnastics in hating state coercion except when it comes to welfare. But most interactions with the State involve coercion (4): threat of force. On my anarchist days, I hate all of these and would side with Peter Cresswell on getting the state out of welfare. But on other days, I take the tradeoff: some coercion in exchange for some public goods, albeit in a total bundle that's rather larger than I'd like. All of the potential rights violations can be put into the utilitarian calculus and weighed against the benefits that come either directly from the coercion or indirectly through what the coercion enables. We already do that when we weigh the costs of pointy relationship questions asked of beneficiaries against the benefits of greater targeting.

What are the tradeoffs for contraception mandates? There's a coercion (5) cost imposed on people on benefits. There's the cost imposed on children who fail to be born but would have valued their lives. Some of this will just be a transfer: when people come off benefits, they may wind up having other children who will only exist because some other child wasn't born previously. So we'd need to take the net decrease as a cost. On the benefit side, there's less type (4) coercion of those whose taxes would otherwise go to supporting children born to families on welfare. There will be some children of higher income families who get to come into existence if their parents keep more of their income. And there will be some relatively small number of children who fail to come into existence who, behind the veil, would prefer staying behind the veil.

I don't know which way things would fall on a full accounting. But ruling it out as eugenics misses that there are already pretty strong eugenic effects built into policy. Ex ante welfare policy subsidised childbearing by those who might otherwise chosen smaller families. Proposed changes to welfare policy imposing work requirements more quickly after the birth of additional children to those on benefits also have expected effects on fertility. So does free contraception for those on benefits.

I'd also disagree with Lew on one other point. While I agree with him that there is strong bipartisan consensus and majority agreement with taxpayer provision of support to the poor, that agreement is part of an implicit social contract. If you're down on your luck and trying your best, or if you're disabled and incapable of working, there seems pretty strong support for relief. But does that same level of broad consensus exist for supporting those deciding, while on benefit, either to have more children or to fail to take measures to avoid having additional children? I'd love to see some polling data on it.

Meanwhile, Kearney and Levine survey the evidence on teen childbearing in the United States. They find it pretty hard to find significant predictors of panel variation (why some states decline more quickly than others). Most targeted policies didn't have any effect. But two policies - expanded family planning assistance to poor women through Medicaid and reduced welfare benefits - did seem to matter and accounted for about twelve percent of the drop in teen childbearing. They frame this as being a fairly small effect, but it's still the only policy combination they found that had any effect. And, it's also the policy combination that National's looking at running: tighter work requirements for those with young children and enhanced access to contraception.

More recently, National's been suggesting drug testing for those on welfare, though the latest versions of it are rather weaker than the trial balloons first floated. The best argument in favour of this kind of policy comes from the Beaulier-Caplan take on the behavioural economics literature [see here and here]. While some American evidence suggests few there tested come up as being on drugs and so the policy isn't worthwhile, we'd really need better evidence to draw that conclusion if the existence of the policy reduces drug use; low rates of positive tests can be viewed as a success of the policy rather than a failure. It may here be worth waiting a few years to see what happens as different US states adopt mandatory drug testing. I'd be betting against that the policy winds up being found to be desirable, but why not free-ride on the Americans' experimentation with this one?

* I'd initially had this post queued up for 5/9/2012 when I'd meant to have it queued for 9/5/2012. Oops. Updated and now posted.

Saturday 18 August 2012

International influence

It can't just be coincidence, can it? A big UK lobby group puts out an ad about tobacco plain packaging and a New Zealand one just happens to put out an almost identical Kiwi variant soon afterwards? How long are we going to tolerate these big multinational efforts to influence New Zealand legislation, and especially when they're using kids to do it?

Dick Puddlecote points out the very strong similarities between the ad put up by government-funded lobby groups funding plain packaging advocacy in the UK and the one that's now up here in New Zealand. You watch and spot the difference. And hit Puddlecote's blog for a lot more context.

Here's the British one, courtesy of Cancer Research UK. The concluding quote:
Unbranding cigarette packs won't stop everyone from smoking, but it will give millions of kids one less reason to start.
Here are the New Zealand variants. The English version is embedded below; the Maori one here-linked. They're courtesy of Plain Packs New Zealand. Who are they? Their supporters listed at the bottom of the website include a great big list of agencies receiving cash from the New Zealand Government for provision of various health services. Someday, somebody's going to have to do a thorough audit of how much money the New Zealand Government gives these outfits to lobby the government on policy.

The quote mid-video (0:29):
Unbranding cigarette packs won't stop everyone from smoking. But it will give our kids one less reason to start.
Maybe we need a national inquiry into the international conspiracy to influence New Zealand legislation here. Ok, not really. But where the anti-tobacco folks express shock and indignation that tobacco companies might coordinate efforts across countries, it's worth noting that they're hardly the only ones.

I'm more serious about the need for an audit to find out how much tax money goes to pay anti-tobacco advocacy groups to engage in policy advocacy.

Thursday 16 August 2012

Evidence and minimum alcohol pricing

Otago's Jennie Connor and Alcohol Action NZ's Doug Sellman are angry again. This time, because Justice Minister Judith Collins cited some Masters' research done at Massey AUT showing that students surveyed said they'd not change their binge drinking habits if prices increased; they instead want the government to rely on international peer-reviewed evidence based on actual consumption patterns rather than on surveys.

And fair enough.

Fortunately, we have some evidence ready at hand. Byrnes et al, 2012, Drug and Alcohol Review. They use Australian household surveys from 2001, 2004 and 2007 to see how changes in alcohol prices affect the number of reported days of no, low, moderate, and high alcohol consumption; they find that while price increases do reduce consumption, they tend to reduce the number of days of low consumption while not changing the number of days of moderate and high alcohol consumption. This would be consistent with binge drinkers dropping the occasional beer or wine with dinner to save up for the big nights out. If policy is more worried about binge drinking than about light drinking, this might matter.

It's also mildly amusing that an Otago healthists complains about policy being based on surveys. I wonder what Connor would make of her Otago colleagues' call for banning smoking outside of bars on the basis of a survey of thirteen youths recruited in part via Facebook; the youths reported in focus groups that they'd be less likely to smoke if they couldn't smoke outside of bars. Maybe that one's ok because it's published in a journal rather than being a Masters Thesis.

Connor also notes that heavy drinkers tend to purchase cheaper alcohol relative to moderate drinkers.
“It has been established that hazardous drinkers spend less per unit of alcohol than others, and drinkers compensate for price increases by shifting to cheaper drinks. In the United States, the heaviest 10% of drinkers spend approximately $0.78 per drink compared with $4.75 per drink for the lightest 50% of drinkers.”
This only is a relevant comparison if the heaviest drinkers and lightest drinkers are drawn from the same parts of the income distribution. Suppose for sake of argument that heavier drinkers are more likely to be drawn from poorer parts of the income distribution and lighter drinkers from higher income parts of the distribution. If that's the case, we would want to compare the price paid by heavy and light drinkers correcting for any differences in income. If people with demographic characteristics similar to the heaviest 10% of drinkers but who are light drinkers spend $1 per drink (just a guess here), then minimum prices pushing above that hit both heavy drinkers and light drinkers of modest income. I know that consumption benefits from alcohol count for zero in the healthist world, but they ought to matter for policy.

At least they're not today claiming that a 10% increase in minimum prices reduce consumption by 16%...

Tuesday 14 August 2012

School heterogeneity

New Zealand schools have not been particularly pleased with the government's plans to release data on student performance that would allow for the construction of league tables; they've argued that heterogeneity in incoming student characteristics produce unreliable results.

A new NBER working paper by Ellison and Swanson suggests that, controlling for these kinds of incoming demographic characteristics, US schools are highly heterogeneous in producing top students. They set up production functions for top math students: those scoring very well on the AMC 12 exam. As expected, there are pretty strong demographic correlates of high-scoring schools: zip codes with more highly educated parents and with more Asian-American students produce higher AMC-12 results. But within sets of schools with similar demographic characteristics, there's still strong heterogeneity in production of top students. 
Our results suggest that there is a lot of variation among seemingly similar schools. The most notable feature of this variation is a thick upper tail of schools in which students are many times more likely to reach high achievement levels than are students in the typical school with similar demographics. This thick upper tail is present in all of our analyses of students with high AMC scores, but is not present when we look at where students with high SAT scores are coming from. This contrast suggests that that the thick tail is not because of the self-selection of high-ability students into a particular subset of schools. We suggest that a potential explanation is that almost all schools see it as their responsibility to provide English and math courses that cover material necessary to do well on the SATs, whereas there is much less uniformity in whether schools encourage gifted students to develop more advanced problem solving skills and reach the higher level of mastery of high school mathematics needed to do well on the AMC.
Relative to the literature on the gender gap in mathematics, our comparison of school effects relevant to girls suggests that schools are perhaps even more important for girls: we estimate that the 99th percentile high school in our sample is producing high-scoring girls at more than ten times the rate of an average school with comparable demographics. We also note that there are many low-performing schools that will only very rarely have girls reach the AMC performance levels we have studied. [emphasis added]
Our finding that there appears to be a lot of variation across schools with similar demographics could be seen as hopeful: the number of high-achieving students would increase substantially if low-achieving schools could be brought up to average; and upper-tail schools might have programs that could be emulated to produce even larger improvements. 
If it's impossible to tell which schools are high- or low-achieving given the demographic characteristics of their incoming cohort of students, it's harder to encourage the low-performers to improve. 

Monday 13 August 2012

Unquestionable costs

Back in 2009, BERL told us that harmful alcohol use cost New Zealand's health system $286 million dollars for the 2005/06 year. There were a few problems with their method. It's worth revisiting these now that a new BERL number on alcohol health costs is out. I'd critique their current numbers except, well, I'll come back to that one.

The BERL report was modelled on an Australian study by Collins and Lapsley. Collins and Lapsley used a table of alcohol aetiological fractions to assign health costs. So if alcohol is responsible for 100% of alcoholic liver cirrhosis, 100% of the costs of alcoholic liver cirrhosis are attributed to alcohol. For some disorders, alcohol consumption reduces costs to the health system. In particular, cardiovascular disease is reduced by alcohol consumption. So while someone consuming 8 standard drinks per day is likely to blow his liver out (and impose costs on the "treating bad livers" part of the health system), he'll likely have pretty clean arteries (and reduce costs on the "fixing plugged arteries" part of the health system). To get a measure of net costs you take all the fractions, some positive and some negative, and multiply by the costs of each to the health system. This will remain an imperfect measure of aggregate health costs: if moderate drinkers live longer, they'll impose greater total costs on the health system by living longer than teetotalers; if heavy drinkers die early, they'll cost the health system money early on but save the system money in the longer term. But the method gives you a first cut.

So, what did BERL do? They took Collins and Lapsley's table and surgically removed any disorders for which alcohol provided protective effects. Why? Footnote 15:
A related, but separate, issue is that of the beneficial consequences of drug consumption.  This report concentrates specifically on the social costs of harmful use.  It does not analyse the impacts from non-harmful use, such as any protective health effects of alcohol consumption.  That is, beneficial impacts of alcohol use are not included as cost offsets.
They do a bunch of handwaving about how health benefits are contentious. But their source document, Collins and Lapsley, included those benefits. And it's not a great out to say that there are no medical benefits once you get past 4 standard drinks per day. Sure, that's about where aggregate mortality risk goes past the baseline for teetotallers. But that curve is drawn for net effects: assuming away the benefits side puts a pretty big thumb on the scale in measuring total costs to the health system, especially when metastudies like Corrao's find strong cardioprotective effects well into the range where total mortality effects are pretty negative.

So after assuming away any possible health benefits from alcohol use, BERL found that alcohol imposed a large burden on the New Zealand health system. We didn't have the resources to reverse-engineer their pretty shonky zeroing-out of potential health benefits when we critiqued their study but simply noted it as something that made their figures rather higher than they should have been.

Anyway, BERL has a new figure out on costs of alcohol to Canterbury's health system. Or so reports the Christchurch Press. Their reporter, Georgina Stylianou, seems to be running off the press release here. She interviews a bunch of the usual suspects about just how awful alcohol is. Any fact checking on the figure? No. Why? The paper isn't yet released.
The report will be released in full this week and will show the range of conditions in which alcohol is a contributing factor.
Since I can't critique their new figures, I'll remind folks about what was wrong with their figures last go-round. I wonder if they've fixed anything.

It's not encouraging that the Christchurch Press seems to be playing into the media strategy that BERL and CDHB here are running: get a story on the press release, maybe another one when the report is made available; get the figures into public debate but don't let anybody have a chance at critiquing them. I'd hoped for better.

Saturday 11 August 2012

If this isn't a wowser, I don't know what is

The Press has a puff piece letting Doug Sellman explain how he's not a wowser.
In short, 'wowser' suggested Christian women and other busybodies who wanted to get between a man and his drink. And not just drinking but many forms of pleasure. When Australian artist Norman Lindsay was attacked over his nude paintings in the 1930s, he said, "I am sick and tired of this wowseristic country". To this day, Lindsay's authorised online biography has "irate wowsers" in North America burning his paintings.
Closer to home and in our time, Doug Sellman has had "wowser" thrown at him a few times over the years. Along with other, less polite words. All because he has talked back to the alcohol industry.
Ok. This from the guy who got mad about the Woodstock mixed bourbon and cola drinks not just because of the booze, but also because they encouraged teenagers to have sex with their friends' mothers. Here's Sellman in an oped from the Herald not too long ago:
Mr Kerr is saying these young people and women need to be more responsible in the face of aggressive targeted marketing. But the industry he is defending is cynically trying to convert these people into heavy drinkers. Is that socially responsible?
Alcohol advertising on TV has recently reached a point that one wonders whether it can sink any lower. The latest Woodstock advert[*] aiming to get boys to drink bourbon sweetened by Coca-Cola in a product called Woodstock, associates a can of Woodstock, a "woody", with a penile erection and not unsubtly is encouraging them to have sex with their best mates' mothers.
The ad also seems to be encouraging middle-aged women to get their teenage clothes back on and flirt with their son's best mates. I presume some would defend these adverts as responsible business practice and contributing to a healthy society.
So Sellman isn't leading an old time Christian Temperance Union crusade against fun.  As the billboard says, Yeah Right.

Here's Sellman warning that alcohol is a Group 1 carcinogen, comparing it to plutonium, and ignoring that birth control pills are in the same darned category.

Here's Sellman pushing for taxes on soda.

Here's Sellman saying that muesli is addictive.

Further in the Press article, Sellman says it's all about the science:
As wowser and related insults show, alcohol reform is an emotive field where much is at stake - revenue streams included. Sellman wants to emphasise the science.
"It perhaps needs to be stated my colleagues and I are not basing our suggestions for alcohol law reform on our own opinions or life experience," he concludes. "They are based on the best international scientific evidence available at the current time."
The Press piece had earlier quoted the press release from Doug Sellman and Alcohol Action NZ colleague and U Otago Prof Jennie Connor that grievously exaggerated the findings from Chris Auld's paper on the effects of minimum pricing in Canada. Maybe that press release was "based" on the best international scientific evidence in the same way that a lot of stuff on TV is "based on a true story". Because it sure misrepresents what Auld actually found. Why should we trust anything he or Connor say about alcohol?

* I'd posted the ad here when Sellman first posted this op-ed about the moral decay of today's youths and their mothers.