Wednesday 16 July 2014

The Price of Weed

Tim Worstall's likely right about the price of marijuana in a fully free market. He notes that the price of marijuana per kilogramme, in Malawi, where much international tobacco production takes place, runs as low as $3, so a pack of 20 joints would be maybe about $0.50. He makes a few predictions as consequence:
So, yes, I feel confident in predicting that a pack of fully legal cannabis joints, one that is able to benefit from the full effects of global capitalism, would cost something like 50 cents instead of the predicted $50.
I would also go on to make two further predictions. Currently the proposed and enacted taxes on cannabis are ad valorem. That is, they’re based upon the value of the cannabis itself. As the price crashes as a result of trade I am certain that those ad valorem taxes will be replaced by unit taxation. So many dollars per pack of joints, rather than x or y percentage of the value of them. The second is that full liberalisation of the cannabis market will be followed milliseconds later by calls for trade barriers to be imposed. For, obviously, if Malawi were to be allowed to compete then all and every US based producer will go bust. Which, given the interplay between the localisation movement and the legalise cannabis movement isn’t something that will be welcomed. Legalisation would therefore be followed by the “Smoke Local” campaign.
Ad valorem taxation doesn't make sense for products whose purported negative effects are far more related to the quantity consumed than to the price of the goods that are consumed. It's pretty unlikely that a $600 bottle of Grange does a hundred times more harm to anybody than a $6 bottle of a more economical brand. You could perhaps see its being applied if the only thing that really mattered in price setting were the strength of the product, and if consumers were better judges of product strength than were the revenuers. But in the more plausible case where there can be strong gradations in product quality for the same level of active ingredient, you want to go with unit excise lest you shift the market towards cheaper products that are, perhaps, harsher on the lungs. So unit excise based on some measure of standardised THC equivalent could work: stronger joints would draw a heavier tax, just as spirits draw more excise the greater the alcohol concentration.

I'd done some very rough ballparking a few years ago (see here and here) on potential revenues from excise on marijuana were it legalised, and were it taxed on a per-unit basis with the tax set to maintain a constant price to consumers before and after legalisation. I am not asserting that there are social harms equivalent to that tax rate that the government does well by internalising through tax but rather that, for some reason, voters see demand abatement as a good and that maintaining constant prices, or at least avoiding any big drop in prices and the consequent Campbell Live specials, would be one way of avoiding the screaming-voter problem.

I assumed an autarkic domestic market as I expected that international obligations would be seen as constraining against either import or export. In that case, the domestic production cost can be a decent starting point, albeit with the caveats I'd then made and that Worstall also warns about: benchmarking prices on small-scale domestic production may be misleading if a legal domestic market provides economies of scale and better production technologies, like ones that don't require hiding the plants from police helicopters or in hydroponic-equipped closets.

If international political considerations didn't prevent trade, the numbers I'd been working with would be really rather out. I'd assumed that politicians would set unit excise to maintain costs to consumers at about the same level as it is now, with the treasury then getting much of the return from legalisation rather than growers. But it's pretty hard to imagine Kiwi growers getting prices down to anything like $0.50 for a pack of 20.

But now we have a problem with excise. The excise rate that would maintain the cost to consumers while allowing imports would destroy any domestic industry, or at least the lower end of it. If current prices are on the order of $300/ounce, and if prices from Malawi are more on the order of $3-$4/kg, and if there are about 35 ounces in a kilogram, and if domestic production costs by the time it hits retail are somewhere maybe around $100-$150/ounce, well, the per-ounce excise has just jumped from the $150-$200 range to basically $299/ounce.

Options? Setting a higher import tariff than the domestic excise might sound appealing, but it would be pretty distortionary: the price that keeps Malawian product at at least $300 makes product imported from elsewhere pretty expensive as Malawi defines the bottom of a very wide price range. Another option would be to set excise at the very bottom of the current selling price for the very worst quality product. Imported product would swamp the bottom end of the market, but there'd potentially be room for higher-priced local speciality products.

You could alternatively think of combinations of excise and retail minimum pricing if supporting local industry were important enough for you to accept either rents and rent capitalisation to cannabis retailers (if retailing licences are constrained) or to put up with (or try to constrain against) all the ways that retailers would chisel on minimum pricing otherwise.

I'm happy with whatever meets the voter participation constraint.

[Update: thanks to Peter for pointing out a typo in comments, now corrected.]


  1. Worstall, not Warstall

  2. I speak not from experience, but I understand this stuff is really easy to grow in this country - so I assume the price would need to be low in order to get many buyers