Corrections appreciates acuminous commentary. A comment by “Student1776” was left in response to our post
“Don’t Legalize Marijuana”. It was thoughtful enough to warrant its own post in response. Corrections is willing to repeat this activity for insightful commentary.
The post by Student1776 follows:
The writer of the blog makes an excellent point that the use of taxation as a means to control the extent of use. There are at least two additional pertinent elements. First, illegal status tends to constrain use differentially to areas where law is less respected or at least less effective. In the US the illegal status of marijuana or crack cocaine or other drugs tends to confine sales, distribution and to some degree consumption to ghetto areas. Making marijuana or crack cocaine legal would allow the expansion of sales and distribution to all areas - for example the suburbs - likely resulting in expanded use and an expansion of adverse social consequences both in terms of numbers of users and in terms of affecting more empowered parts of the electorate whose political pressure would doubtless be made manifest. Second, as one considers the wider principle of legalization and taxation there is a different aspect to consider. Drugs affect individuals differentially for a variety of biological and other reasons. For any given drug different people have different propensities to addiction. For example, about 6-10% of people exposed to alcohol for as little as a first exposure to several exposures will over time become addicts while 90+% of people who consume alcohol can take it or leave it. These latter people with a propensity not to be addicted are said to be able to "chip" alcohol. For opioids and tobacco the same phenomenon occurs but the number of potential addicts is larger - in a broad ballpark around 20% of people being at risk for addiction with 80% of people for example being able to use an opioid for pain for an extended period without coming to exhibit drug seeking behavior (that is to "chip" opioids or cigarettes. For amphetamine type drugs the proportion of addicts to chippers is far higher - likely on the order of 30-60% depending on the amphetamine with the number of chippers being correspondingly lower (70-40%). The point is that while marijuana has a relatively low rate of severe addictive behaviors associated with it and a high rate of chipping, other drugs are less benign. People who are addicted do continue to show some level of "rational" behavior in the sense of price responsiveness but addicts are by the nature of the disease incompletely rational. Addicts will engage in extreme and irrational activity including violence and law breaking at felony levels. The issue is that while drugs are illegal and distribution is highly restricted there are many people who are potential addicts who are simply not exposed to the addictive drug and as a consequence are able to live productive lives and not be addicted. Pricing issues might to some degree permit a similar constriction of the potential exposure of addicts but for many potential addicts not exposed at this time because of limits of distribution, their financial characteristics may allow them to surmount limitations on exposure due to pricing. It is also noted that if taxation is raised to too high a level then the taxation itself will be surmounted and we will have just created a new kind of illegality. (Not unlike the current cigarette smuggling industry). The bottom line is that the Corrections blogger is correct that taxation could be used to help control exposure, one can anticipate far wider distribution and for the highly addictive substances far higher levels of addiction and associated social damage in localities and social circles currently out of the loop of distribution.
Four points have been made that it is incumbent upon Corrections to dispute, in one form or another.
1) Illegality has differential effects on area that taxation does not have. Specifically, Student1776’s conjecture is that currently use is kept in relatively indigent areas—post legalization, it could spread to wealthier areas. This, in turn, may lead to more users, and empower “undesirable” political classes.
Concerning the first point, Corrections agrees and considered this point. However, it is important to note that there is a reason that use was constricted to indigent areas to begin with. Specifically, that these areas had low opportunity cost of both use and penalty. But it is worth noting that when the government makes something illegal, the black market has no means of writing enforceable formal contracts.
This means that in order to enforce contracts, firms use force. Firms that dominate the industry tend to be those best at wielding violence, rather than those with the most efficient methods. This, in turn, creates an industry that has a lower cost to producing violence on competing firms, consumers, and government.
Corrections posits that this helps propagate the very indigent communities that the author concerns himself with. Were these industries replaced with companies like Johnson & Johnson, Pfizer, and GlaxoSmithKline, we suspect that illegal activity such as drug-dealing, and the resultant violence, would be reduced dramatically.
This reduces transfers from wealthier communities, who will have to pay for less law enforcement, perhaps helping to empower and create more desirable political classes. Implicit in Correction’s understanding of the author’s reference to a desirable political class is one that does not depend on the Rothbardian “gang of thieves writ large,” viz., the government.
2) Drugs have differential addition rates. People who are easily addicted and were previously were not exposed to drugs will now have easy access to them. They will subsequently consume the drugs and become addicted.
Corrections does not dispute the idea that people have differential addiction rates. Indeed, it supplements the idea of
differential treatment effects we expounded upon in a previous post.
We do, however, dispute the characterization that people who were previously not exposed to drugs will now have easy access to them. In the reading of Corrections, a very large proportion of the United States currently has access to crack cocaine, for instance. They may buy crack cocaine at some price. The question is in what form they pay the price. Corrections conjectures that most of the author’s potential users would pay in the form of search costs, rather than dollars. This can be translated into dollar form, of course.
Why would we assume that people who are easily addicted and have easy access to a product would become addicted? If these individuals know their propensity to become addicted, then they have the most to lose from trying before addiction. We should therefore see these very people shy away from drug use. We should expect to observe tests for whether or not someone has the propensity to become addicted. In the absence of such tests, we might expect the proper incentives to exist to develop them. In the absence of the possibility of such tests, one would expect risk-averse consumers to shy away from the product in general, as large swaths of the non-chipping population likely do for readily available drugs such as crack cocaine.
Whether or not one accepts this previous point, the first point remains: translate search costs into tax costs and it would appear the same costs from suburban individual’s point of view.
We might add that addictive substances such as crack cocaine were a technological advancement on cocaine. It allowed for small cheap, unobtrusive, mass distribution in a way cocaine did not, a point lifted from Roland Fryer, Paul Heaton, Steven Levitt, and Kevin Murphy’s 2005 NBER working paper “Measuring the Impact of Crack Cocaine.” It is not immediately clear whether or not this phenomenally destructive drug would have been as prevalent as it was were cocaine itself not illegal. On the one hand, cocaine is an input into crack cocaine production. On the other hand, cocaine is a substitute to crack cocaine itself, one whose competitive availability is likely increased much more than crack cocaine’s, in light of crack’s popularity being in part due to cocaine’s inefficient illegal distribution system.
3) Student1776 conjectures “People who are addicted do continue to show some level of "rational" behavior in the sense of price responsiveness but addicts are by the nature of the disease incompletely rational. Addicts will engage in extreme and irrational activity including violence and law breaking at felony levels.”
Corrections does not see violence and law breaking at felony levels as an irrational behavior. Take, for example, Lawrence Katz, Steven Levitt, and Ellen Shustorovich’s 2003 American Law and Economics Review article “
Prison Conditions, Capital Punishment, and Deterrence”. The homicide rate within a state is negatively impacted by higher within-prison death rates (from murders, illness and AIDs, poor conditions). Indeed, the impact of one additional prison death ranged from -0.1 to -0.8 fewer homicides across the author’s specifications.
Murder, rape, robbery, aggravated assault with a firearm, and burglary of residence are not excluded from the list of crimes for which sentence enhancements significantly change behavior. Daniel Kessler and Steven Levitt found, in their 1999 Journal of Law and Economics article “
Using Sentence Enhancements to Distinguish between Deterrence and Incapacitation” that those crimes declined by four percent more than California’s non-enhanced crimes, compared to the difference in decline of the same crimes in other states (without sentence enhancements) and their comparable crimes (a difference-in-difference-in-difference model).
To use the same author a third time, in the Journal of Political Economy (1998) Steven Levitt’s paper “
Juvenile Crime and Punishment” shows that not only are juveniles likely deterred by differential sentences, but that an increase in the relative punitiveness of adult crimes compared to juvenile crimes causes decreases in crime when juveniles reach the age of majority. This comparison, comparable to a regression discontinuity comparison, indicates that juveniles around the age of majority appear to be making their criminal decisions based on sentences—-another rational act.
Corrections sees criminals as behaving rationally—-if not during their crimes (say, during a drug-induced state), then in the decisions leading up to the crime (the decision to become intoxicated).
4) It is also noted that if taxation is raised to too high a level then the taxation itself will be surmounted and we will have just created a new kind of illegality. (Not unlike the current cigarette smuggling industry).
Corrections concurs. We see the result as only a partial privatization. If this is the case, then at the limit legality will change nothing because the tax will be too high and the market will remain a black market. In reality, Corrections expects that a portion of the market will be legal and taxed. To take the author’s example, we posit that New York spends vastly less on enforcement and gains vastly more revenue from cigarettes by making them legal and taxing them than it would by making them illegal and paying to enforce the law (rather than be paid).
The relationship is indeed positive, which is itself phenomenal. If high schoolers were myopic and unresponsive to future police presence, we would have expected a drop in graduation rates to result in an increase in police next period--an increase in bad high school students should make more police next period a necessity. The fact that we see this indicates that the difference between our two effects is rather large. At a first glance, while the relationship is only near significant, it would appear tantalizing. For those concerned about the outliers, the relationship remains positive dropping them from a fixed effects panel data regression.
