Friday 1 July 2011

And in prison news...

New Zealand's prisons go smoke-free today. Some inmates are smoking their nicotine patches. I'm not sure that's an improvement.
Inmates have not been able to buy tobacco, matches or lighters since June 1, and cigarettes are now in hot demand.

"Some prisoners have made their normal allowance last more than a week, by smoking half at a time," assistant general manager of prison services Brendan Anstiss said.

"Tobacco has become a more valuable commodity. Those who do have cigarettes are more likely to keep them to themselves."


Corrections Minister Judith Collins announced the health and safety-based smoking ban last June, giving a year for prisoners to get used to the idea and try to quit.

The impending ban has already forced a price hike for tobacco on the prison black market, with a packet of cigarettes fetching up to $300, according to sources.
Ok. So no new cigarettes into prison starting 1 June, with no smoking allowed after 1 July. And the price rises to $300 within a month? I wonder if there's any time series data on those prices. You could likely use them to back out a measure of prisoner time preferences. A non-smoking prisoner could get his allocation in on 31 May then has a choice to sell or hold the cigarette on each day up to 30 June, knowing the value goes wonky 1 July after consumption is banned. Every day the price is higher than the prior day. And it reaches $300 by 30 June. That suggests pretty high discount rates by non-smoking cigarette sellers in the black market, or systematic underestimation of how much of the prison's stock was being depleted in each period. It would be cool to run Hotelling's pricing model to back out implied discount rates.

Meanwhile, in Canada, female prison guards in male prisons are being given training in how to avoid having sexual trysts with prisoners. Roissy would likely claim it's mostly due to the alpha nature of many prisoners; I'd worry about the extent to which guards' power was being exercised over prisoners in unwelcome fashion.

Previously in prurient Canadian prison news...


  1. Gary Becker's still working, right? This clearly calls for an integration of the economic theory of crime with the theory of rational addiction. For one thing, it is quite obvious that we should expect a reduction in offending by smokers relative to the offending by nonsmokers. I'm sure Becker can come up with a bunch of more counterintuitive predictions.

  2. Except for criminals who want to quit smoking and have no other way to constrain themselves, right? They'll be marginally more likely to offend to get the free quit smoking package.

    Way more interesting would bed work on US panel data given some states banned prison smoking and other didn't.

  3. According to my memory, the concept "inability to constrain oneself" makes no appearance in the theory of rational addiction. Am I misremembering?

  4. Use Hotelling to infer discount rate or interest rate? (Though I suppose the real trouble with using Hotelling is that the empirical prediction never quite seems to show up.)

  5. @Lemmus: Yes, and that would be a test too.

    @Ryan: Presumably prisoners can access loans on normal markets via relatives outside of prison, even if at pawn shop / payday loan rates. So it's a test of discount rates.

  6. Hotelling's just no arbitrage, right? If you have a high discount rate but you can borrow & save at the normal interest rate, then you still hold onto those cigarettes as long as the RoR exceeds the interest rate -- you just borrow a ton.

  7. Yeah, but you can only get so many cigarettes into the place - there are quotas on how many you can import into the prison, or at least that's how I read the story. If that's right, there's a fixed and binding number of cigarettes in the system on the start date that start being smoked down over time. By the last day, the price of a pack has risen to $300.

    Hotelling is arbitrage over time, yes. A cigarette in the pack for sale tomorrow must be worth at least the selling price of a cigarette today plus the interest rate you'd earn on that sold cigarette. Now, if prisoners have access to anywhere near normal interest rates (money in prison over time worth about as much as money out of prison over time), there's no way we should see a ramp up to $300 per packet within a month. Or, I want to start giving money to prisoners to invest in within-prison ventures. I'm pegging this one as high discount rates by those selling cigarettes or perhaps systematic underestimation of the depletion of reserves over time.

  8. I'm going with the latter (systematic underestimation). Right, I'm not assuming they can get more cigarettes (it's an exhaustible resource), but when I write down the problem, I get either (a) the change in price over time is equal to the interest rate, or (b) we hit some sort of non-negativity constraint and so they can't borrow more money or they sell all the cigarettes on one day or something along those lines, and that this answer is independent of the prisoners' discount rate(s).

    I don't think we should be shocked that we're getting a really weird result. Prices have historically fallen over time, but Hotelling says resource rents must grow exponentially over time, and there's only so long we can wave our hands talking about declining extraction costs before I call shenanigans on the whole business.