Friday, July 25, 2014

Touch your toes...

If lack of physical activity is really to blame for increased obesity, rather than increased caloric intake, do we switch from soda taxes to mandatory morning callisthenics?

From the American Journal of Medicine's press release on two new studies (here and here):
Sedentary lifestyle and not caloric intake may be to blame for increased obesity in the US, according to a new analysis of data from the National Health and Nutrition Examination Survey (NHANES). A study published in The American Journal of Medicine reveals that in the past 20 years there has been a sharp decrease in physical exercise and an increase in average body mass index (BMI), while caloric intake has remained steady. Investigators theorized that a nationwide drop in leisuretime physical activity, especially among young women, may be responsible for the upward trend in obesity rates.
By analyzing NHANES data from the last 20 years, researchers from Stanford University discovered that the number of US adult women who reported no physical activity jumped from 19.1% in 1994 to 51.7% in 2010. For men, the number increased from 11.4% in 1994 to 43.5% in 2010. During the period, average BMI has increased across the board, with the most dramatic rise found among young women ages 18-39.
I expect to do substantially more walking on moving to Wellington. Optimal policy, if obesity reduction were your goal, would need to consider the relative cost elasticities of caloric consumption and exercise. The jump from about 20% to about 50% of adult women reporting no physical activity does seem pretty large.

Hanley writes as commentary on the Ladabaum et al research:
What is missing from the Ladabaum et al paper is societal context. The finger-wagging Puritan in me wants sedentary folks to get up off the couch and exercise, but my public health background cautions me to go beyond the data tables and look at the lives of Americans today. 
Life, work, and leisure have changed dramatically in the US since 1988. With the blossoming of the Internet; the widespread use of computers and mobile devices at home and at work; and the increasing popularity of video games, our lives have been transformed. On an economic level, the prosperity of the 1990s dissipated after the September 11, 2001 terrorist attacks; years of war; the off-shoring of manufacturing jobs; and the Wall Street crash of 2008. Ubiquitous home foreclosures, lay-offs, and
continued unemployment/underemployment have fueled historic income inequality.
I note that none of the factors listed in her second paragraph really comes into the Ladabaum study, though poverty does correlate with obesity.

She points to poverty among single mothers as potential reason for little exercise (though the study does not give any breakdowns by single/couple household status); I'd wonder if it isn't just practicality. Who's going to get a babysitter to head out to the gym? We're hardly poor, and I couldn't imagine our shelling out for that. And while playing outside with the kids can be rather physically exerting (and is indeed my main form of physical activity), that's most easily done while somebody else is in the kitchen making dinner. In the summer, one of us would fairly regularly take the kids out to the beach or park after work and school while the other got dinner ready. You can't do that as easily in a single-parent household.

Further, while video games have indeed gotten more awesome, television is the main reported leisure-time activity of the poor and low educated. Walking around the block is not expensive; income alone should not be the main constraint except inasmuch as it means longer work hours. I'm reasonably sure, though, that recent data has higher income workers also putting in more total hours than lower income workers. Further, the time-rich unemployed have not been immune. Motivation could perhaps be a bigger issue.

The Australian Monthly [HT: @clairlemon] recently hit on some of the realities of morbid obesity - a very different problem than the smaller expansion in average waistlines.
I ask a young 200-kilo patient what he snacks on. “Nothing,” he says. I look him in the eye. Nothing? He nods. I ask him about his chronic skin infections, his diabetes. He tears up: “I eat hot chips and fried dim sims and drink three bottles of Coke every afternoon. The truth is I’m addicted to eating. I’m addicted.” He punches his thigh.
Addicted. The word is useless in my clinic, a mere barrier to any hope of self-determined change. My patient is not addicted; he’s a very lonely, unemployed young man who has gradually become socially isolated to the extent that the only thing available to him for comfort and entertainment is food. He has no friends, no money to buy other consumables, little education, no partner, no job. Some days he doesn’t leave his bed. The choice for him is to eat this food or experience no pleasure. The surgeon and I discuss his situation, concerned that he may overeat after the band has been fitted. We tell him that surgery may not be appropriate for him, given his situation. The patient is perturbed. “Well, what are you going to do for me if you won’t do the operation? Don’t you have some kind of ethical responsibility to help me lose weight?”
This is where the obesity-as-disease concept leads us – to a situation in which people demand that medicine shoulder the responsibility. What about the responsibility of the individual? And of society? My patient cries because the highlight of his day is returning from the supermarket with a plastic bag full of junk that he will eat and drink in his empty lounge room. What can I do for him? I can threaten him with his early demise, intensify his shame. I can offer him some evidence-based motivational lifestyle interventions – swap Coke for Diet Coke! Prescribe exercise? Walk for an hour at an average pace and you’ll only burn off the equivalent of one slice of bread. I could take the old-fashioned approach and wire his jaw shut. I have no hope of resolving his loneliness, his hopelessness, his lack of a job. I could, and do, refer him to a psychologist – if he’s lucky he may land one who is talented and sensitive and will try to get to the root of why this young man hates his own guts. More likely he’ll be offered a few sessions of behavioural therapy that will make everyone except him feel better.
It does not seem likely, at least to me, that folks like this would change their behaviour all that much in response to a soda tax. Better to try to address problems of long-term unemployment that give rise to this kind of despair.

I'm not offering any policy solutions to obesity. If the health and productivity costs of obesity were terrifically large, employers would start running morning callisthenics programmes to reduce their insurance premiums.* It would be great if cities fixed their regs to allow higher density so more folks could choose to be in more walkable neighbourhoods, but I'd hardly want to force everybody to live in those either.

Further, it may well be that optimal body mass has increased with the increasing opportunity cost of exercise-related leisure activity. It takes some value judgements to say that individuals' observed obesity outcomes are suboptimal where information about the benefits of exercise and about nutrition is readily available. Sure, most folks surveyed will claim that they wish they could be thinner. But whether that's notional or effective demand where half of those surveyed are undertaking no physical activity, well...

In related news, Brand-Miller and Barclay have been cleared of allegations that they'd falsified data in their study showing declining sugar consumption in Australia. Overall sugar consumption is down, though it may be up in some population subgroups. The review notes that increased obesity among those consuming less sugar may come down to reduced exercise.

* Of course, offsetting behaviour can kick in with these things. Follow the link....

Thursday, July 24, 2014

Paternalism as insult

This article sketches a third conception of paternalism — one that locates its normative significance in neither coercion nor motives. This approach maintains that paternalism involves expressive content. Paternalism expresses the idea that the actor knows better than the person acted upon; it implies that the other party is not capable of making good judgments for him or her self. The normative significance of paternalism derives from the typical impermissibility of making such an expression. That is, paternalism is wrong in the same way that an insult is wrong. This understanding of paternalism’s normative significance provides the tools to make intelligible the charge of paternalism leveled against some policies and conversely to explain why other paternalist policies are permissible.
I've argued that paternalistic interventions tend to hit on those vices which enjoy moral disapprobation and which are enjoyed by those looked down upon by the political elites: there are strong classist underpinnings to the interventions.

Cornell notes the implied insult:
The objectionable feature of paternalism is not the nature of the interference (more or less coercive), nor the intention behind the interference, but rather what is expressed by that interference. It is, of course, true that more forceful intervention will frequently be a statement of even greater superiority in judging one’s welfare. So for this reason (and the fact that the potential for injury is greater), forceful coercion will often seem more objectionable than the gentler nudging. But it is not simply how coercive an act of paternalism is that dictates whether it is objectionable, but rather how much respect it shows to the subject. 
...In particular, the ability to make choices may be valuable as an expression of the respect others have for me. It is this value that the expressive theory says is harmed in cases of paternalism. And whether the symbolic value of choice is undermined does not necessarily correspond with the degree of coercion involved. 
Paternalistic interventions on an individual level can be justifiable, depending on context. It can be ok to arrange with your overworked partner's office that his or her meetings be rescheduled so that you can both head out for a long weekend, but nudge-interventions, like hiding her smartphone, aren't so nice.
On the basis of this humility, we should be willing to accept some implicit questioning or criticisms of our judgment. If my friend gives me a CD of music that I profess to dislike and tells me that I will enjoy it if I try, I should be willing to think that maybe my judgment was clouded. If my wife starts forcing me to eat healthier foods now and then, I need not immediately object that I know what’s best for me.
Although in both cases the action calls into question my judgment about my own well-being in a particular matter, in neither case does the implication seem offensive. It would, of course, be totally different if a stranger approached me on the street and informed me that I would look so much better if I refrained from wearing pleated pants. Or if a stranger next to me at the cafĂ© were to suggest that I eat healthier foods. 
Cornell notes that paternalistic interventions are less insulting where they are general rather than specific: seatbelt laws affect everyone, for example, and are far less insulting than putting in higher fines for ethnic groups that are less likely to wear seatbelts (were there ethnic differences). But I think he misses here that the new types of paternalism being suggested, like fat or soda taxes, or bans on smoking in cars, or restrictions on things like RTDs rather than wine, are strongly targeted in application despite being written in general terms. Poorer people are more likely to smoke; obesity is more concentrated among the poor; there are strong class differences in types of alcohol beverages consumed. Sodas and chips get targeted for taxes while double-mocha frappucinos don't. But he does conclude that you need a fair bit of background context to judge whether a policy is objectionably or unobjectionably paternalistic.

Wednesday, July 23, 2014

Coroner Recommends ... revisited

Kiwiblog reports on an Otago study looking at uptake of coronial recommendations. It's gated at the NZMJ, but Stuff and the Herald have coverage.

Jennifer Moore went through all the coronial recommendations from mid-2007 through mid-2012: 1644 of them. She makes a few reasonable recommendations about things the Coroner's Office could do to increase uptake of their recommendations: vaguely directed recommendations are unlikely to hit anybody's radar.

But she doesn't note one potentially important reason that at least some coronial recommendations don't lead to policy changes. The Coroners Act 2006 asks coroners to make recommendations of any changes that could reduce the likelihood of future similar deaths, but with no reference to any kind of cost constraint. So while the coroners will put out lots of really helpful and cost-effective recommendations, they'll also recommend things that seem extraordinarily unlikely to pass any kind of reasonable cost-benefit assessment. Mandatory high-visibility clothing (not just vests) for cyclists, mandatory helmets for riders of motorised skateboards, and mandatory fencing-off of farmhouses seem reasonable on a "if it saves only one life it's worth it" standard, but fortunately policy isn't based on that kind of standard. It's then hardly surprising that not all coronial recommendations turn into policy changes.

As I'd noted last year:
The problem seems to be in the Act. Pretty much anything that could reduce the chances of particular forms of death can be recommended; there's no consideration anywhere of costs. It's fine to say that that's Parliament's job. But Coronial recommendations carry some weight - people take them as being something more than "This is something that could save lives, but I have no clue whether it's worth it because I have zero training in policy assessment and cost-benefit analysis, so somebody else better figure out whether we'd be wasting a whole ton of resources in enacting it; moreover, the Act specifically asks me to just name any darned thing that might help even if it would cost a trillion dollars and save a life every fifty years."
It would be interesting to go back through Moore's list to see whether some rough ballparking of cost-benefit ratios predict which coronial recommendations are taken up.

Monday, July 21, 2014

Internet Party Policies

NZ's Internet-Mana Party announced a few policies this weekend. I wish they'd stuck to internet and surveillance issues. Here goes.
Internet Party leader Laila Harre said it would stop the controversial Trans-Pacific Partnership trade agreement, give students free tertiary education and instigate a royal commission on any spying on New Zealand residents by the Government Communications and Security Bureau (GCSB).
Let's take these in turn.
  • I'm a bit of a TPP-sceptic, but I'd sure not want to stop it until the final text had been hammered out. There's some chance yet that it could be well worth having. We just have to be ready to turn it down if it isn't. Getting better dairy access to the US would be great, but so too would avoiding having US copyright laws imposed on us. Until we see the deal, we won't know what the trade-offs are.
  • Free tertiary education seems a rather poor idea. After Labour put in interest-free student loans, a whole pile of other sectoral distortions followed. Why? Because every additional admitted domestic student meant a potentially substantial liability for the Crown. And so we started getting different funding arrangements with capped student intake. Either this policy would be extraordinarily expensive, or it would wind up reducing the number of places at Uni, or it would require substantial per-student cost-cutting at the Universities. 
  • I could be sympathetic to the last bit, but I'm not sure what the point would be. We kinda know what they did, and the government has made it all legal. We don't need a Royal Commission. We need a redrafting of the legislation with input from the tech sector at ground level. [Update: Chris Yong, Internet Party candidate for Te Atatu, notes that the Commission enquiry would be into the handling of the DotCom case; the total review of the GCSB Act would be separate.]
Back to their policies:
Mana Party Waiariki candidate Annette Sykes said it would close prisons, while co-vice president John Minto promised to abolish GST but introduce a tax on financial speculation, and bring back the inheritance tax ditched by National.
  • I'd really want to see what policies they're proposing to reduce the prison muster before commenting on closing prisons. If they can do it by legalising possession, cultivation and distribution of marijuana while releasing anybody jailed for only those offences, and if that were enough to do away with a prison, great. 
  • Replacing GST with a speculation tax is an extraordinarily bad idea. The GST is a nice clean tax. Financial speculation taxes are rather harder to implement and, worse, don't make a lot of sense. The best case for them is that they might reduce excess volatility in currency markets, but that doesn't apply well to the kinds of long term cycles we get in the NZ dollar. 
  • Inheritance taxes aren't uncommon around the world, but they do involve substantially more distortion than does the GST. Whenever you put in inheritance taxes, you have to put in a big apparatus around transfers from parents to children to avoid folks' just pre-gifting the inheritance. But when you do that, you also encumber a lot of ways that parents are already helping their children for non-tax-avoidance reasons. If an inheritance tax kicks in at $500,000, that will hit a pretty decent fraction of estates where the deceased owned a home. Do you then require that the grieving kids sell off the family home to pay the taxes? Or do you exempt the family home, and consequently induce even greater investment in housing? It's messy.

Saturday, July 19, 2014

Testing bans

How far back should any ban on animal-tested products go?

Labour this week promised that, in the unlikely event of its election, it will ban the importation of overseas cosmetics that had been tested on animals. Suppose you agree with the intention of the policy: to reduce and eventually eliminate the testing of cosmetics on animals. Deciding which cosmetics to ban remains tricky.

At minimum, you would want to ban the importing of any products whose development involved animal testing where that testing happened after the ban went into force. We can only really ever affect behaviour in the future, not behaviour in the past, so banning future products that had been tested on animals would be the obvious start.

Suppose you went further back and decided to ban anything that involved testing since Wednesday, when Labour's policy was announced. Perhaps some manufacturers would have heard that this policy could eventuate in New Zealand and might decide to change their development process as consequence. You'd want to have been clear that the ban would apply from today-developed products onwards, but you can still make the case for it.

Going farther back, you start conferring rents on manufacturers who had chosen to veer away from animal testing and to impose losses on those who hadn't. It's possible that these rents could fuel further product development from the no-testing companies, and maybe it's desirable to confer those rents even if it doesn't. But the farther back you go, the more losses you impose on NZ consumers for no plausible gain for today's animals. If it turned out that Chanel #5 involved some kind of animal cruelty when developed in 1920, what possible benefit would come of a ban?

I don't know what the right answer here would be from the perspective of someone against hurting animals, but it seems unlikely that any ban should extend backwards in time all that far. I am curious how Labour would seek to apply this though. Their PDF says any products or ingredients that have been tested on animals would be banned. But it doesn't say anything about whether that applies only to current testing, or to past testing.

Friday, July 18, 2014

Zone wars

Adrien de Croy, who lives with his family in a home zoned for EGGS and also within the proposed One Tree Hill College zone, understood there was significant pressure on the rolls at AGS and EGGS.
"They can't really reduce the zones unless there's an alternative in place, and [the proposed One Tree Hill College zone] basically gives them the opportunity to reduce their zone. We see it as the first step to removing us from the Auckland Grammar and EGGS zones."
Mr de Croy, whose eldest child is 7, said at this stage his main concern was for the value of their property. A real estate agent had told him a typical premium someone would pay to get into the "double Grammar zone" was about 20 per cent.
Last year the Herald reported one Mt Eden home just 750m outside the area went for $516,000 less than a house up the road, valued the same but situated 250m within the zone.
One Tree Hill College principal Nick Coughlan said he understood such concerns, but they were unfounded.
The zone, and any overlaps, was informed by the need to not divide areas and homes around contributing schools. There was no intention to realign zones in the future, Mr Coughlan said.
We can do better than this, though, to gauge the effects of school zoning. For that, we turn to Waikato's John Gibson and Geua Boe-Gibson. They've estimated the effects of school boundaries in Christchurch pre-quake. From their abstract:
School attendance boundaries are a contentious issue in New Zealand, and have been relaxed and re-imposed depending upon political sentiment. Critics contend that a supposedly egalitarian state school system becomes one of selection by mortgage, with the value of ‘free’ schools capitalized into property prices. Attendance boundaries restrict the schooling opportunity set facing a student, who typically is unable to study at nearby high-performing schools if they live outside their boundary. We relate schooling opportunity sets to sales prices of over 8000 houses in Christchurch, controlling for dwelling attributes, neighborhood characteristics and geographic accessibility to a wide range of services. Our model explains over three-quarters of the variation in prices and we use this model to predict property prices if there were no attendance boundaries. Abolishing boundaries expands most schooling opportunity sets and predicted house prices generally rise. But prices would fall in some higher income neighborhoods with highly educated residents, who are likely to oppose reform of school attendance boundaries.
Gibson and Boe-Gibson use a year's worth of house sales in Christchurch, October '04 through October '05, to check the effects of school zones on prices after accounting for land and building area, building age, materials, parking, garage, and whether there was a deck, slope, or view. I hadn't known that QV data included information on the latter three. Importantly, they link in Census meshblock data on neighbourhood ethnicity, immigrant status, education, and employment, and meshblock crime. Some of the work on school zone effects will confound "good school"  with "seen-as-desirable (ie no rednecks) neighbours".

They simulate the effects of a standard deviation increase in NCEA Level 1-3 pass rates on median house prices and find that, all else equal, having access to a school with a standard deviation better NCEA pass rate is worth between $14,300 and $19,900 for the median house. This gives a few implications.

First, the market value of policy innovations that improve school quality is very high: a policy that improved NCEA pass rates by a standard deviation is worth about $42 billion.

Second, locking poorer people into poorer schools seems a pretty bad policy. Gibson calls it "selection by mortgage", and worries it can reduce social mobility especially among minority groups.

Finally, while abolishing school zones would increase the total value of the housing stock because gains to those getting access to better schools exceed losses to those currently sitting on regulatory rents, it's unlikely to happen because those earning the rents are more effective at protecting turf. While the average goes up in value by about $25,000, houses in preferred zones drop in value by about $20,000. I expect this is an upper-bound estimate as other forms of rationing would have to come in for the better schools in the absence of mechanisms allowing them to grow, and as I'd expect that those with current access would find ways to maintain such access.

Rich people can afford to pick their preferred public schools; poor people get locked into whichever schools service poorer neighbourhoods. The problem is worse in much of North America, where schools are funded from local property taxes rather than from general revenues, ensuring that poor places can't afford good schools; New Zealand's decile funding system works to provide equitable funding across schools. But zoning still causes problems.

Gibson and Boe-Gibson conclude:
...the property market becomes the main schooling selection mechanism for New Zealand parents who are ambitious for their children. Even though schools may nominally be ‘free’, students from poorer households face more restricted schooling opportunities than do wealthier students, being constrained through the housing market.
Almost two decades have passed since New Zealand’s brief experiment with relaxing school attendance boundaries in the 1990s. The frequency of reselling houses makes it likely that most home-owners have paid a price for their dwelling that includes the expected value of access (or exclusion) from particular schools. Consequently there will be windfall gains and losses if future policy reform allows a weakening of attendance boundaries and an opening up of school enrolments. Nevertheless, the wide variation in school performance and the contribution of attendance boundaries to reducing social mobility suggests even difficult reform is worthwhile.
I would love to see a replication of this work in Wellington. In particular, I'd love to know the relative magnitudes of school zone and all-source earthquake risk on property values. Is there a bigger difference between moving from Wellington College zone to out-of-zone than from moving from a low-medium quake-risk property to one that will fall off the side of a cliff in an earthquake? I suspect so, but it would be nice to know.

An Uber experiment

Reason asks an excellent question: is Uber helping to cut drink driving rates in the US? When it's cheaper and easier to get a cab, maybe more people will do it instead of chancing a drive home when they shouldn't.

Reason points to some preliminary work on the topic done by Uber, looking at Uber's entry into Seattle with San Francisco as control. The work's suggestive, but hardly conclusive - especially when there are dozens of cities that could have been chosen as treatment or control.

So, here's the Masters thesis for somebody. Get a city-level panel of DUI rates and of taxi fares. As a first step, just run fixed effects with Uber entry and exit dates. Then run a few more complicated versions, like matching cities by probability of Uber entry based on city characteristics and taxi fares (comparing those of similar ex ante probabilities with different ex post resolution). Or exploit the city-by-city variation in pre- and post-Uber prices. There's loads of potential here, if city panel data on DUI arrests is available. There's loads of wonderful not-related-to-DUI variation in whether cities allow Uber or not making for something close to a natural experiment, though that will be less true if Uber's started using DUI-effects in its lobbying.

In other news, I had my first ride in an Uber cab in Auckland a couple of weeks ago. The cabbie was very enthusiastic about telling me all about it: he's on an hourly rate, but flips to commission when it's busy enough (says he gets 80% of the take). It makes sense that Uber would put cabbies in new markets on hourly to make sure that there's enough supply there when new customers come in.

If you sign up with Uber on code ericc294, you get $10 off your first ride and I get $10 in credit too.