Monday, September 18, 2006

Addiction and inequality

I am in Adelaide attending an NCETA (National Centre on Training and Addiction) Summer School ‘A Fair Go for All? Policy Responses to Alcohol, Drug and Gambling Issues’. The intent is to look at links between social inequality and various addictions.

A lot of the people present have come from the field of population health. Much of their analyses take the form of large numbers of bivariate graphs between a target variable Y (e.g. extent of addiction) and some explanatory variable X (e.g. some index of social inequality). If any type of relationship exists geometrically then the language used suggests that X causes Y.

There is no economic analysis at all. Some type of addiction (e.g. addiction to gambling) is identified as a problem of concern and measures are then discussed to eliminate the problem. No benefits are attached to the activity at all. It is a far more illiberal approach than even I could ever accept.

On the inequality issue the discussion has been very disappointing. Even if it is accepted that activities such as smoking, drinking alcohol and gambling do involve social costs, a difficulty is that measures to deal with such issues – hefty taxes on cigarettes and gambling or restrictions on the supply of gambling outlets will have highly inequitable effects on the socially disadvantaged. This leads to some peculiar arguments – for example that the poor should give up smoking because it is an expensive drain on their budget. That it is expensive because of hefty taxes is not even mentioned.

The population health school needs to learn some statistical technique and some economics.

Have I got value from the discussions? Yes I have. The main benefit for me is that it forces me to reassess my own somewhat illiberal views on the types of demerit goods this conference is on about. You can always learn and if you are honest you should be prepared to adjust your views.

In addition there were two very good papers – descriptive papers by Wayne Hall on alcohol and inequality and by Michael Gossop on nicotine dependence and inequality. I also liked a paper by the ANU’s Richard Eckersley on social determinants on health. He linked addiction problems to findings in happiness research and to individualist ethics in a way I found interesting. (The effort was dulled a bit at the end of the day by a loquacious, Marxist, dope-smoking public health enthusiast who stretched this idea to a ludicrous exaggeration).

The good papers are all apparently available at the NECTA website.

But the best part of my day was a great Italian meal at Parlamento Bistro on North Terrace. The seafood Zuppa soup was close to divine! Despite the great food and the few good papers today I won’t wait around for the first session tomorrow – the value-added is not enough.


mlesich said...

It's a pity I'm not there or I'd have a complete field day picking apart the papers at leisure. First point that one could make is that the authors haven't really shown that there is enough evidence to conclude that there is a significant relationship. Don't forget that ommitted variable bias can also skew the results

I wish I was in SA as well, after all, I hear that the Barossa Valley is lovely this time of the year

conrad said...

I think you highlight a common problem in the social sciences -- lots of people don't understand statistics well, and you end up with descriptive papers or papers with statistics misued (its easy to teach someone how to press the SAS or SPSS buttons, but hard to teach them to understand the numbers). The same is probably true with other groups on genetics and neuroscience, who read the to-sell story too much on what is almost always overexaggerated finds (as I've noticed in some of the blogs -- including this one, no offence meant -- there is a big difference between blog entries and papers!).

The only real solution is that people either know a lot more, or people from different areas co-operate better.

observa said...

Harry, I caught a news report yesterday stating Australians gambled $15 bill last year, $8.7 bill (58% of that total) on pokies. So about 3 out of every 5 discretionary gambling dollars is punted through pokies and that definitely isn't from the middle to top end of town. Most middle class folks hardly ever gamble, period. That's because they're smart enough to know they're not really gambling but donating.