Good Intentions; Bad Policy

I learn from Scott Sumner’s blog that in many California cities, residents with past marijuana convictions will jump to the head of the line for licenses to sell the drug legally — this by way of compensating them for past persecution.

Scott approves. I don’t, for two reasons:

First, if you want to compensate people for past persecution, the right way to do it is with cash, not by misallocating productive resources. If there must be licenses, they should be allocated to those who can use them most efficiently, regardless of any past history.

Second, drug dealers have never been the primary victims of anti-drug laws. They can’t be, because there is free entry and exit from that industry. Anti-drug enforcement leads to exit, which in turn leads to higher profits for those who remain — and the exit continues until the profits are high enough to compensate for the risks. One way to think about this: All those “persecuted” drug dealers were, in effect, employing the government to stifle their competition, and paying a fair price for that privilege in the form of occasionally being convicted and punished themselves.

The primary victims of anti-drug legislation are potential consumers who were deterred by artificially high prices. How do you compensate those victims? You can’t. In a population of 1000 people who have never used drugs, it’s quite impossible to identify the 200 or 300 or 400 who would have happily indulged if only the price had been lower.

This is one more reason to be diligent against bad legislation generally. Even if you believe the legislation will eventually be repealed, attempts to compensate the victims are likely to be misdirected, misguided, and socially harmful.

Share/Save

29 Responses to “Good Intentions; Bad Policy”


  1. 1 1 Bob Murphy

    Great post Steve. You are the most “out of the box” econ blogger.

    I left this comment on David R. Henderson’s analysis:

    Good points all around…

    Another thing that’s odd: I’m not sure exactly how to analyze the preference being given to the former drug convicts, but depending on the specifics, the proposal is even worse than Landsburg thinks.

    If the government is auctioning off a set number of licenses to the highest bidders, *except* that they give a margin of preference to former drug convicts, then it’s the taxpayers who pay for the compensation.

    However, if the licenses have a set price and it’s just a matter of the government selecting which people will get them, then ironically the people in society who “pay for” the compensation are current drug consumers. I.e. they have to deal with drug sellers who are less efficient than potential rivals who were excluded from the license.

    For an analogy, suppose the government took all the existing pot smokers, and made them each kick in $1. Then they randomly picked a former drug convict to win the pot of money. Would that be a sensible way to compensate for the past abuses of the Drug War?

  2. 2 2 Bob Murphy

    The more I think about it, the situation is more nuanced than I said above. But, I think I am bringing up another element that needs to be considered.

  3. 3 3 Harold

    I don’t see why this particular thing should be singled out as the a “good example” of affirmative action if it is generally not approved of.

    The statement “Given how drug pushers have been persecuted over the years, it’s about time we provided some compensation.” applies at least equally well to many other groups. If this is the rationale then affirmative action should be supported for any persecuted group.

    There seems to be two aspects to affirmative action.

    One is in the continuing presence of bias that disadvantages particular groups. Affirmative action then redresses the balance and attempts to restore what would happen in a closer to perfect market.

    The other is to compensate victims for past wrongs. As SL points out, the latter as simply a way of smuggling the costs past people so they don’t notice. It would make more sense to give a tax rebate or cash. The compensation aspect does not provide any direct benefits to anyone other than the compensated.

    The former should provide a benefit to everyone. For massive and pervasive discrimination the positive benefits may take a long time to be realised. The prejudice may have become structural in the society resulting in individuals that are actually less qualified. In this case, the immediate effect could be detrimental, although this is by no means certain. There will be conflicting effects. On the positive side the affirmative action will counteract the original bias and easily measurable parameters, such as lack of qualifications that arose from the bias. This should mean that better people are employed. On the negative side, those lack of qualifications may make the individual less productive. However, over time the positive effects will manifest as future generations are freed from the direct effects of the bias and are able to fulfil their potential.

    To my mind, affirmative action only makes sense in the continuing presence of bias or “vicious circle” effects previous bias. If compensation is deemed appropriate, then a better means of providing it should be sought.

  4. 4 4 Scott F

    “One way to think about this: All those “persecuted” drug dealers were, in effect, employing the government to stifle their competition, and paying a fair price for that privilege in the form of occasionally being convicted and punished themselves.”

    I’m very curious how you come to the conclusion that the amount of incarceration time spent in prisons was fair? Is there some equilibrium between amount of potential trafficking (and profits therefrom) and the amount of time spent in jail? It seems likely to me that many people may have been punished much more than others, perhaps based no something as simple as the shade of their mugshot. In same line of questioning, how would I know if the amount (ignoring for now the distribution) of incarceration were fair?

    Lastly, are they able to sell these licenses? If so, A) they will wind up ultimately in the most efficient producers hands, and B) it was essentially no different than a cash handout.

  5. 5 5 Steve Landsburg

    Scott F: The (ex ante) costs paid by drug dealers were “fair” in the sense that they chose to incur those costs, and hence presuambly found them acceptable. (Many others chose otherwise and dropped out of the industry.) In equilibrium, the marginal drug dealer has to be indifferent about whether or not to be a drug dealer, and in a constant cost industry (which seems to be a pretty good approximation here), all drug dealers are marginal.

  6. 6 6 Bob Murphy

    Scott F.,

    I don’t want to speak for Steve, but I think he’s “merely” doing something like the following (but I put “merely” in quotation marks because I wouldn’t have made the connection unless Steve had done it):

    Suppose a bunch of coal miners were complaining that they should get compensation from the government, because their occupation was more dangerous than the typical job. Surely the first thing we would bring up in response is to say that they are already getting paid a premium in their wages, precisely because of that.

    Now, I grant you that that’s not the end of the story. Even in a free market, mechanisms might evolve so that (e.g.) if you die in a mine collapse, the employer pays your estate an extra amount on top of your own life insurance etc.

    But, I think Steve is bringing up something that even most free-market economists (myself included) wouldn’t have remembered, when thinking about the policy Sumner brought up.

  7. 7 7 Rob Rawlings

    The logic of Steve’s argument is (I think) that whenever supply-side conditions for a good or service are made more difficult as a result of legislation anyone who chooses to continue to supply that good is voluntarily agreeing to accept these harsher condition and so does cannot be said to have lost out as a result.

    I disagree with this conclusion. Suppose I could in a world with no penalties earn $100,000 a year as a drug-dealer, and my next best earning option would be $50,000 as a liqor-dealer. Suppose the government passes laws that make drug-dealing harder so I can now only earn $75,000 a year plus I have a chance of jail time if caught. If I evaluate the risk of jail-time at $15,000 a year I will still choose to deal drugs even with severely reduced earnings.

    In other words I think there is a supply-side equivalent of consumer-surplus and in this case it can be evaluated at $40,000 in lost earnings plus fear of jail-time.

  8. 8 8 Josh

    But wait… why isn’t this symmetrical? Aren’t there in theory also fewer consumers competing for the illegal drug since it’s illegal to consume and thus you also have less competition from fellow consumers?

    I guess dealers get harsher sentences so maybe you’re right.

  9. 9 9 Steve Landsburg

    Rob Rawlings: Suppose I could in a world with no penalties earn $100,000 as a drug-dealer and my next best earning option would be $50,000 as a liquor-dealer”. This can be true only if you have some unusual talent that is valuable only to drug dealers and not liquor dealers — where the word unusual is key here. If there are many with the same talent as you, then they all leave the liquor industry and enter the drug industry, which increases profits for the remaining liquor dealers, reduces profits for the remaining drug dealers, and tends to continue until profits are equalized across the industries. The assumption here (which I alluded to earlier with the jargon “constant cost industry”) is that people with such unusual talents are rare, so that the typical drug dealer cannot earn substantially more as a drug dealer than as a liquor dealer. This is a pretty standard assumption in scenarios like this one, and though I don’t know exactly how accurate it is here, I’d be willing to bet it’s pretty close to accurate. Therefore the costs of anti-drug legislation would tend to primarily on buyers, not on sellers.

    In the extreme case: drug dealers earn $75 K; so do liquor dealers (this is enforced by the fact that any difference in profits would be eliminated by movement back and forth between industries.) Now new anti-drug legislation reduces profits for drug dealers to $65K; then drug prices have to rise to restore the drug dealers’ profit back to the initial $75K (this assumes that the $75K profit in the liquor industry is unchanged). Thus the anti-drug legislation does no harm to sellers (at least ex ante).

  10. 10 10 Enrique

    This is a great post. I would only add that politics is often more about virtue-signaling (or “ideological rent-seeking” — a term I coined here: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=915929) than anything else.

  11. 11 11 Josh

    Can this be thought of from a transaction cost perspective? We are taught that it doesn’t matter who actually remits transaction costs. But here since dealers and consumers are treated differently as far as punishments go, effectively the downward effect on price of consumers not consuming due to their own perceived legal risk is not as great as the increasing effect on price seen by dealers. Maybe I’m not thinking about this quite right.

  12. 12 12 Rob Rawlings

    #Steve (#9).

    Thanks for the detailed response. I now agree with your conclusions.

  13. 13 13 Ken B

    I like the post, and agree with the conclusion, but want to stick in a contrarian suggestion. Ponder a bit on point 9. Were the costs which were imposed on drug dealers as a whole similarly continuous and divisible into small parts? No. Unlucky Bob served 4 years while lucky Steve, Ken, and Scott served none, leading to an average cost of 1. Making the usual assumptions about diminishing returns we see that 4 +0+0+0 is not the same as 1+1+1+1, so wasn’t Bob overcharged?

  14. 14 14 Steve Landsburg

    ken b: Well, bob bought into a probability distribution, and received a draw from the distribution he boughth into. If the distribution was (4,0,0,0) and that’s worse than (1,1,1,1), there would have been accordingly more exit and higher drug prices.

  15. 15 15 Josh

    Thinking about this more, I think I have a clearer explanation for why I might disagree.

    Your assertion basically is that since drug dealers get paid a premium to offset the risk of imprisonment, they really shouldn’t have any complaints, since they freely choose to enter the market as a dealer. This premium happens because the supply curve shifts to the left. But obviously there is also a shift in the demand curve to the left because consumers are also at risk, but we think it doesn’t shift as much as the supply curve because dealers get harsher punishments. But what if the risk preferences of dealers and consumers are inherently different? What if, even _if_ the punishments are greater for dealers, that the demand curve shifts enough to the left to wipe out the premium? That is, dealers by their mental makeup may respond differently to different risks than consumers. And if this happens, then both consumers and dealers are basically out of more quantity at the original price. I’m not saying that is what’s happened, but is there a way to show it hasn’t?

  16. 16 16 Ken B

    Josh 15
    Let’s assume that’s true. Then who were the losers due to the prohibition? They include the buyers and the tax-payers. This undercuts the notion that the jailed dealers are the particular, involuntary, losers.

  17. 17 17 Josh

    I don’t think it does. If your assumption is that the dealers aren’t losers because they’re paid a premium for the increased risk they take on, my contention is that price premium doesn’t necessarily exist, though it probably does. You can take away that price premium due to a demand curve that shifts just enough left to offset the also left-shifting supply curve, such that the new equilibrium price is back to what it was without the prohibition, albeit at a lower quantity demanded. So dealers in this scenario get NO risk premium paid to them while also still selling less and assumingly still subject to the same legal risks. And consumers also get less quantity though at a “normal” price.

  18. 18 18 Nathan

    You wrote:

    The (ex ante) costs paid by drug dealers were “fair” in the sense that they chose to incur those costs, and hence presuambly found them acceptable.

    But California isn’t giving compensation on an ex ante basis, it’s giving it on an ex post basis. Only those who actually were convicted get the benefit. Those who won the drug dealer lottery by never getting caught get no such benefit. As such, it does indeed seem to compensate some of the losers of the drug war.

  19. 19 19 Josh

    My argument hinges on the idea that we assume the reason dealers aren’t victims is solely because they’re paid a premium for the risk they take on. I’m simply saying that risk premium isn’t necessarily being paid if the demand curve also shifts to the left in such a manner as to offset it.

  20. 20 20 Neil

    Were the people who did consume also the victims of the anti-drug laws because of their lost consumer surplus? Or were they simply compensating the dealers for risking jail in order to provide them with the product?

  21. 21 21 Scott F

    @Steve, Bob, Rob, and Josh: Thanks for the responses. Makes sense to me now.

    However, if one group of people was disproportionately incarcerated (or incarcerated more severely) and therefore saw that possibility as more costly, isn’t it probable that people from that group were leaving the industry (say to liquor selling), even though they would have profited more than others who stayed in the industry?

    Now that I’ve laid it out, and even though this seems likely to me, I see now that the policy still doesn’t help these people. However, it seems to me that these marginalized suppliers (assuming they exist) are also victims of the policy, and should be included alongside the consumers.

    I still, however stand by my assertion that this legislation is only problematic if the licenses can’t be sold. Furthermore, if we take Steve’s analysis that the consumers were the most hurt, then would the appropriate policy (if one wanted to make reparations) be a subsidy? This would help both consumers and marginalized suppliers, or am I missing something?

  22. 22 22 iceman

    Harold #3 – I think one can view affirmative preferences as OK in theory but unworkable in practice, e.g. figuring out who owes how much cash compensation to whom. Sumner may view this drug issue as a unique case providing for a practicable if imperfect solution because of an identifiable group of (alleged) victims who have paid an identifiable price.

  23. 23 23 Harold

    #22. He may have that view, but he does not say it.

  24. 24 24 iceman

    Well that’s my answer to how one example could be singled out.

    This is a tangent but another news story made me think of this blog – when asked whether converting a midtown NYC hotel worth $170mm into a shelter was the best way to serve the homeless with that kind of $, the city responded that they view this not as a matter of cost but of equity (every neighborhood should participate).

  25. 25 25 GregS

    I take the point about drug dealers being compensated with higher profits, but I have this nagging objection. What changes if we’re talking about a prohibition on, say, critical speech about the government rather than drugs? Couldn’t we make the same argument that jailed journalists under oppressive regimes aren’t really victims? It’s just as attractive to be a mechanic or a farmer or an accountant, so we know anyone who chooses to become an intrepid journalist is amply rewarded for their career choice. One could argue that the drug dealers earn higher profits (measured in dollars) while the intrepid journalist doesn’t. But the journalist earns “abnormal profits” in terms of higher esteem (internal and external, possibly even seguing into a future political career), since nobody else is brave enough to do what he’s doing. I think someone could make a pretty strong case that the jailed journalists aren’t “really victims” because they were compensated for the risks they took, but this just seems morally obtuse to me.

    Supposing there is a difference between the drug dealer who flouts drug prohibition and the intrepid journalist who flouts the prohibition on criticizing the regime. Producers of psychedelics didn’t really earn much money on what they produced; see the story of William Leonard Pickard or the Brotherhood of Eternal Love. It seems like these people distributed drugs despite the legal sanctions because they thought there was some kind of social benefit to taking LSD. Are these people more like intrepid journalists, or are they more like jailed cannabis, heroin and cocaine dealers?

  26. 26 26 SJA

    Moving those with marijuana convictions to the front of the line also presumes that they were indeed prosecuted primarily for marijuana-related reasons. They may have been charged with violent offenses and then entered plea agreements for lesser charges involving marijuana.

  27. 27 27 Advo

    If anything, the drug dealers are victims of legalization.
    After all, the legal drug business is very different from the illegal drug business. It has much lower margins and the required skill sets differ substantially.
    For one thing, tight accounting practices are a must in the legal pot shop, less so for a lower-level drug dealer.

    In fact, the reason why many/most of those people became drug dealers in the first place was that they were unable/not motivated to perform/adhere to the very tasks and standard that form the core of running a successful business.

    Illegal businesses with their sky-high profit margins are much more forgiving of sloppy business practices than legal ones with efficient competition.

  28. 28 28 Harold

    #24. Sure, this is not a terrible thing to do. Even if those compensated would be happier with money, and even if business would be slightly better run if left open to normal competition, this does at least ensure that some compensation gets to the victims (if we were to ignore SL’s main point that these are not victims.) This may be the best we can do, and maybe the price of slightly reduced efficiency is worth it. If we are wrong and they are not victims at all, no great harm is done.

    #27. You comment reminds of Soprano’s style euphemisms like “retirement plans”.

    Greg #25. “Couldn’t we make the same argument that jailed journalists under oppressive regimes aren’t really victims?”
    I thought of a flippant reply to the effect that we can wait a couple of years and we can have some home examples, but maybe that would not be helpful.

    On a more serious note, I think this is a good point. Perhaps we can add externalities into the equation – the journalists produces positive externalities whilst the drug dealer produces negative externalities. However, that does not address the point about the personal rewards.

  29. 29 29 ZT

    Aren’t the primary victims actual consumers, rather than potential consumers? Actual consumers bore the cost of enforcement. Potential consumers bore the opportunity cost of using drugs, which must necessarily be less than the cost of enforcement, otherwise they would have chosen to use the drugs anyway.

    (There are, of course, negative externalities to the drug war, but I don’t see that as a flaw in your argument since those externalities weren’t born primarily by drug dealers.)

Leave a Reply