Refracting Green

I characterize the current push towards marijuana legalization in America as “refracting green”.  Refraction–or the change in direction of a wave due to a change in its transmission medium–is the phenomenon that leads us to misjudge an object’s position due to distortions caused within our environment.  Many policymakers, having realized that marijuana is a valuable commodity and not nearly as dangerous as our other legal drugs (nicotine, alcohol, caffeine, etc.), are hoping to cash in on this market by “formalizing” it–in other words, legalizing the drug, regulating its production and distribution, and taxing the resulting product.  Conversely, this formalization is also supposed to ensure control where the “war on drugs” could not.  Participants in the marijuana industry see it similarly; legalization ought to bring more profits by allowing access to customers who were out of reach before (due to the robust network characteristics of illegal marijuana distribution networks) and minimizing prosecutorial risk.  The pursuit of money–particularly in the realm of policy making–has distorted (and made impossible) the goal of control.  It has also made it impossible for the burgeoning, legal marijuana industry to operate legally.  How could this be?  It’s all about the taxes.

How should marijuana be taxed?  The answer depends on the goals of policy makers and on the structure of the existing local (informal market) for marijuana.  One of the least-considered issues is how the current, informal market will interact with newly created, formalized markets; researchers often view illegal and legal markets as separate entities, but they are really all part of a unified commodity production market.  I’ll provide an overview of the Washington approach, then offer my thoughts on how the Oregon market requires a unique approach (subsequent post).

Washington

Washington’s marijuana law is a sterling example of how to ensure actor non-compliance. Here’s why.  Washington adopted a three-tier distribution system similar to the much-maligned alcohol market (though producers can also hold processor licenses under newly adopted regulations) and has elected to impose a 25% tax at each stage of production/processing/retailing, as well as local sales tax to end consumers.  These rates were codified by I-502 and are, therefore, not alterable by the agencies tasked with implementing marijuana legalization.  Washington’s state budget agency (OFM) has projected that growers will receive $3 / gram, processors will receive $6 / gram, and the pre-tax retail price will be $12 / gram.  Retail sales taxes average 8.87% in the state.  Since the OFM doesn’t tell us if the “pre-tax retail price” includes the 25% excise tax, we end up with a wide range for the predicted average consumer price:  an ounce of legal marijuana should cost consumers between $370 and $465.  The company hired to provide analysis and policy recommendations–BOTEC–has produced a number of white papers analyzing the economics and consumer decision making processes at play in this newly formalized market.  As they note,

“Anecdotal evidence suggests that consumers will pay a higher price for legal, tested, and regulated marijuana than for black market product. What is not clear is how much more consumers would be willing to pay. If the price premium to purchase from a legal operator is set too high, consumers at some point will prefer to purchase from an illicit operator if doing so will save them significant money.”

Time will only tell if these retail prices are low enough to lure consumers in; more information on Washington’s informal market is needed to assess the viability of state-sanctioned marijuana sales.  My research on the Oregon marijuana market–which is only separated from Washington by the Columbia River–demonstrates that the average price per ounce in Oregon’s informal market is $177.  Will consumers willingly pay 200% more to legally purchase their marijuana from a store?

To make matters even more complicated, BOTEC has recently suggested that the original OFM analysis of retail prices is not accurate for compliant retailers due to federal income tax laws and because compliant producers will need to charge more per pound ($3125 vs. $2500) to earn modest profits from their trade.  After providing various income scenarios for growers, processors, and retailers, BOTEC argues that:

“Federal income taxes will have a much larger impact on pricing behavior, especially when we start to consider implications for IRC Section 280E. Retailers, unlike producers and processors, have many costs that might be disallowed for tax deduction purposes under this law provision. In theory, if Section 280E were to be strictly applied, the retail marijuana industry simply could not exist because the economics could not support income taxes calculated on gross margins. Under this theory, the excise tax the retailers pay to Washington would not be deductible. Using the average sized retailer as an example, the federal income tax would be $417,400, but the retailer would only net $392,200 before payment of any excise tax or federal income tax. Thus, even excluding the retailer excise tax, the retailer would not have enough cash to pay the federal income tax (assuming strict application of Section 280E).”

To make a profit under the new legal regime, producers, processors, and, especially, retailers will be forced to either sell marijuana for exorbitant prices (compared to the black market) or become creative in their marijuana acquisition programs.  Washington’s tax structure is essentially nudging would-be legitimate marijuana operators into the black market to remain profitable.

It gets worse.  Washington has set a limit on the total amount of licensable growing space for the state, which will be divided amongst licensed producers; it amounts to 2 million square feet or, roughly, 46 acres.  There is no distinction made between outdoor and indoor production–if it’s used to grow marijuana, it counts toward the 46 acres.  Is this enough to meet Washington’s marijuana demand?  To find out, we need to consider the use patterns among specific demographic groups.  First, no one under 21 years of age will be allowed to use.  Second, the use rates differ between age groups (21-25 are most likely to use).  Third, use amounts differ dramatically between user types (light and heavy).  Since no study has assessed the amount consumed by individual Washington users, we’ll also have to substitute data I’ve collected on Oregon users (grams / month).  This can be formalized as:

Yearly Demand = Light User Demand Model + Heavy User Demand Model

where:

Light User Demand = ((21-25 POP * % Users) * Light Use Amount) + ((26+ POP * % Users) * Light Use Amount)

Heavy User Demand = ((21-25 POP * % Users) * Heavy Use Amount) + ((26+ POP * % Users) * Heavy Use Amount)

or

218,437.905 lbs. = 17,727.33 lbs. + 200,710.575 lbs.

The Washington marijuana market demands approximately 218,000 pounds of product each year.  With the capped growing space policy (46 acres), each acre licensed for marijuana production will have to yield a minimum of 4,748 pounds of finished product to meet the state’s yearly demand.  And those are perfect conditions (which do not exist in Washington!).  Throw in a conservative 15% loss rate for pests, molds, and bacteria and each acre will need to produce close to 5,500 pounds.  Many outdoor marijuana farmers in Southern Oregon have suggested that each plant should have a footprint of 20 feet x 20 feet (400 square feet).  This reduces the spread of disease, maximizes air flow, and allows farm workers enough space to operate without damaging plants.  Using this calculation, each outdoor acre can safely support 109 marijuana plants; the entire state would have 5,014 legal marijuana plants.  Each of these plants would have to yield over 50 pounds of finished marijuana to meet consumer demand.  Have you ever seen a 50 pound marijuana plant?  No one has.  How about a 25  pound plant (218 plants per acre)?  No one has.  A more likely scenario–involving large-scale green houses and costly supplementary lighting–is a 10 pound average per plant @ 109 plants per acre:  this would result in about 50,000 lbs. of finished marijuana and a 168,000 pound shortfall.  Where will it come from?  Well, who’s located nearby and grows a lot of pot?

It’s no wonder that BOTEC’s president–Mark Kleiman–responded to questions about marijuana “leakage” out of Washington as “a non-problem”.  The Washington system was designed to underproduce expensive marijuana; it will not be economically feasible for individuals to ship legally produced marijuana out of the state (a win for advocates of marijuana control in the state).  On the other hand, it forces the legitimate marijuana industry to rely on illegally produced marijuana, both from within and outside of the state–which limits the likelihood of collecting excise taxes and a full reporting of marijuana-related incomes (BOTEC calculates two compliance rates for their tax estimates:  100% and 97%–neither seem likely).

This approach strikes me as a 20th century solution for a 21st century problem.  We have much more effective tools available to us to enable compliance and protect new participants in the formal market.  Things are taking off at the Oregon legislature and I’ll post my thoughts on appropriate ways forward soon.

Using Marijuana While Driving

The Oregon legalization bill tackles getting high while driving, but leaves DUII laws unchanged.  If you light up (or its equivalent) in the car, you would be subject to the following, new proposed penalty:

Section 73.  Use of marijuana while driving; penalty. (1) A person commits the offense of use of marijuana while driving if the person uses any marijuana while driving a motor vehicle upon a highway. (2) The offense described in this section, use of marijuana while driving, is a Class B traffic violation.

Because this section does not alter DUII laws in any way, it serves as an additional punishment for those who are caught using marijuana while *safely* operating a motor vehicle.  The violation is similar to speeding (more than 10 MPH over posted limit, but less than 20 MPH), following to closely, etc. and results in a $300 ticket.  As a cogent reader can surmise, a cop who pulls you over for smoking marijuana while driving has probable cause to assess your ability to drive via field sobriety tests and/or breath/blood tests.

This legalization bill does not set THC blood concentration limits (as Washington did) and makes no alterations to the DUII laws–it seems that law enforcement personnel and future recreational users could use some clarification on marijuana’s effect on driving and the current tools available on DUII enforcement.  Others simply want to know:  does marijuana legalization require alterations to current DUII laws?

Stoned Drivers

First, lets address what the research has found on individuals’ driving behavior while under the influence of marijuana (and only marijuana).  It’s really important to emphasize upfront that the trifecta of marijuana, alcohol, and driving do not mix well (Robbe 1998; Ronen et al. 2010)–even in small amounts, the combination of alcohol and marijuana dangerously impairs drivers.  There is consensus within scholarly research on several important issues:  (1) marijuana does impair driving, (2) marijuana users compensate for that impairment by driving slow (well, not that slow, but 5-10 mph slower than their uninhibited selves), and (3) alcohol intoxication is far more dangerous in commonly used amounts of each respective drug.  Research on stoned drivers finds that–when results are statistically significant (they are not usually with low doses of marijuana)–they have slower reaction times and more difficulty maintaining their lane of travel than non-inebriated, non-distracted drivers.  As could be reasonably expected, this is accentuated by higher doses.  In all real world driving and marijuana consumption studies, stoned drivers were able to safely operate their vehicles and performed better than their legally drunk (> 0.08 BAC) counterparts.  Robbe (1998) notes at the conclusion of his double-blind comparative study of marijuana, alcohol, and marijuana+alcohol intoxicated drivers in the Netherlands that the “evidence from the present and previous studies strongly suggests that alcohol encourages risky driving whereas THC encourages greater caution, at least in experiments.  Another way THC seems to differ qualitatively from many other drugs is that the former’s users seem better able to compensate for its adverse effects while driving under the influence”.  This finding is echoed in much of the literature on stoned drivers (real world and simulated driving).

Dosing Issues and Cannabinoid Interactions

Real World Dosing

Dosing in these studies, however, is an issue; the most empirically valid research protocols control for body weight and administer real marijuana (much of it grown at the National Institute on Drug Abuse pot farm at the University of Mississippi) in doses based on micrograms of active THC per kilogram of body weight.  In the highest administered THC dose studies with real world driving assessments (300 micrograms THC / kg body weight) (Robbe 1998), stoned drivers performed about as well as drivers who had blood alcohol content (BAC) levels of 0.04%–affected, but well below the legal limit.

How does this dosing regime stand up to real world conditions?  Let’s take a 170 pound person (77.11 kg) using high quality marijuana with approximately 15% THC content as an example; how much real marijuana would it take to achieve similar levels of inebriation (300 micrograms / kg body weight) studied by scholars?

300 micrograms THC * 77.11 kg = 23,133 micrograms = .023 grams THC

.023 grams / .15 grams (i.e. 15% THC per 1 gram marijuana) =

0.153 grams of 15% THC marijuana

How does this stack up to the consumption patterns of regular marijuana users?  My research suggests that regular users of marijuana consume about .46 grams per day, though the concentration of THC and the amount per “use” is unknown.  If our fictional 170 lb user were to consume the Oregon average (.46 grams) of 15% THC marijuana in one sitting, he/she will receive a dose of 900 micrograms of THC per KG body weight–or exactly 300% of the most commonly used dose used in research on stoned drivers.  How do drivers perform at that level of intoxication?  No one has systematically studied this yet and it seems very important to conduct research on the effect of real-world doses on real-world driving performance.  Is it safe to assume that all marijuana users consume their daily allotment in one sitting?  Probably not, though such an occurrence cannot be discounted and also warrants further research.

Cannabinoid Interactions

The most glaring issue with all studies of marijuana dosing is derived from our nascent understanding of the interactions amongst cannabinoids when ingested.  Unlike alcohol, every individual marijuana plant has a unique effect on its users because of the variability in  their chemical composition (particularly cannabinoids and terpenes); in this regard, there is no realistic possibility of perfect standardization across marijuana studies or perfect extrapolation of clinical studies to real world situations.  Certain cannabinoids are known to interact with each other, with the two most important being THC and CBD (which is now viewed by many researchers as more medically beneficial than THC).  CBD is not psychoactive and is not likely to affect psychomotor abilities (that aspect has only been studied in mice so far).  Additionally, CBD inhibits the psychoactive effect of THC (i.e. the anti-marijuana drug is present in marijuana!), which means that varieties of the plant with higher CBD to THC concentrations will affect end-users’ driving ability less than their low CBD to THC counterparts.  Even with less or no impairment, individuals would have higher concentrations of active THC in their blood after consuming CBD-rich marijuana products (which calls into question the validity of THC ng / ml tests).  This means that measurements of THC are problematic (in plants or human bodies) without accounting for the antagonistic effects of other cannabinoids or terpenes.  Again, this is another area that requires significant scientific advances before any standard should be adopted.

THC in the Blood vs. Standardized Field Sobriety Tests (SFSTs)

5 nanograms of active THC in 1 milliliter of blood (5 ng / ml) is the marijuana impairment limit adopted by Washington.  This level appears to be based on Ramaekers et al. (2006) study on non-driving psychomotor performance exams and active THC concentration tests (blood), the authors recommended that 2 ng-5 ng THC / 1 ml of blood be established as the “lower and upper range of a THC limit for impairment” (114).  Unfortunately, the study is problematic in many critical ways.  First, in terms of research design, the authors themselves conclude that “the present study will be criticized for the face validity of the performance tasks and their ability to reflect driver impairment or crash risk” (120).  In other words, the tests used to assess impairment do not necessarily relate to real world driving performance.  That’s problematic when constructing a legal standard, though not surprising–the study was funded by the German Federal Police Academy and the German Society against Alcohol Drugs and Driving (Ramaekers et al. 2006: 121).  Second, other researchers flatly reject the notion that THC concentrations in the blood are related to impairment; instead, they suggest that high THC concentrations in the blood simply indicates very recent use (Robbe 1998; Papafotiou 2005a, 2005b; Karschner et al. 2009).  Ramaekers et al.(2006) study has been the only one to date that has recommended a standardized threshold for active THC in the bloodstream.

This point is driven home graphically in Papafotiou et al.’s work (2005a; 2005b) on Standardized Field Sobriety Tests (SFSTs) and marijuana intoxication (though dosing differs from gold-standard studies).  Papafotiou et al. provide mean THC ng / ml results over time for their 40 research participants (Fig.1, 2005a: 109), who ingested a placebo, low dose (0.813 grams@ 1.74% THC), or high dose (1.776 grams of 2.93% THC) of marijuana.

As the figure illustrates, marijuana smokers (an important, additional caveat) experience huge spikes in their THC/blood concentrations immediately after ingesting the drug (55-70 ng / ml, dose dependent).  Within 20 minutes, that concentration is reduced to 12.8-13.8 ng /ml, and after 100 minutes, all participants were below the legal limit established in Washington (3.18 & 3.72, respectively). How could this be?  The answer is remarkably simple.  Active THC in the bloodstream has not made it to receptor sites in the body yet and, therefore, have no effect on a person.  As active THC in the bloodstream is absorbed by the body (and declining in concentration), intoxication increases.  For this reason, Robbe (1998) notes this important finding from his study:

“Inter-subject correlations between plasma concentrations of marijuana and driving performance after every dose were essentially nil, partly due to the peculiar kinetics of THC.  It enters the brain relatively rapidly, although with a perceptible delay relative to plasma concentrations.  Once there, it remains even at a time when plasma concentrations approach or reach zero.  As a result, performance may still be impaired at the time that plasma concentrations of the drug are near the detection limit.  Therefore, an important practical implication of the study is that it is not possible to conclude anything about a driver’s impairment on the basis of his/her plasma concentrations of THC and THC-COOH determined in a single sample (S77)”.

In addition to measuring THC to blood concentrations, Papafotiou et al. (2005a; 2005b) also subjected their study participants to separate driving performance tests (at three time points) and three SFSTs at two time points (horizontal and vertical gaze nystagmus, walk and turn, and one leg stand tests).  At time point one, (30 minutes after smoking; all participants had more than twice Washington’s legal limit of THC in their system), all participants were “not significantly impaired in comparison to the placebo condition,” though no SFSTs were conducted (2005b: 175).  At time point two (80 minutes after smoking; participants’ average THC concentrations were below Washington’s legal limit), “driving was significantly impaired, as indicated by the ‘straddled the solid line’ variable..and the ‘straddled barrier line variable'”.  In other words, most stoned participants would pass Washington’s blood test standard, had trouble maintaining their lane of travel, and failed their SFSTs–however, that was the extent of their impairment.

Here’s the really important piece that may explain why Washington adopted a scientifically invalid standard.  Comparing the results of the SFST battery and driving impairment standards at 80 minutes after smoking resulted in the following findings:  “of the participants who were [scored as] impaired on the driving task, 88.5% were correctly identified as impaired but only 38.5% of participants who were not impaired on the driving task were correctly identified as not impaired” (176).  This means that stoned drivers who could not maintain their lane of travel were correctly identified as stoned 88.5% of the time, BUT 61.5% of non-impaired drivers failed the SFSTs and were accused of stoned driving when they really were not–even though all drivers were likely to pass the Washington THC / blood standard.  Even worse, a remarkable 0% of non-impaired drivers were identified 105 minutes after smoking.  Others have demonstrated the lack of face validity and scientifically tenuous nature of SFSTs (Rubenzer 2008); this finding corroborates that criticism in a demonstrative manner.  Field sobriety tests appear valid as a predictor of one form of marijuana impairment (inability to maintain lane), but are dangerously ineffective–0%!–at protecting citizens from false/erroneous accusations of impairment.  Restated, the THC / blood standard is invalid as a measure of impairment, but has the potential to protect citizens from DUII arrest if they wait an hour after smoking to drive.

Summary

People who consume marijuana products containing the full range of active cannabinoids experience diminished psychomotor abilities, but, like alcohol, moderate doses do not impair their driving ability enough to consider them unsafe.  That said, we need further research on “real world” dosing and its effects on driving ability.  There is no reliable lab measurement or standard (i.e. blood tests) to quantify marijuana impairment, in part due to the multitude of active constituents in the drug and because common reference tools (i.e. THC ng/ml) are not correlated with impairment in any meaningful way.  SFSTs remain the most effective tool for law enforcement to assess driving impairment; paradoxically, the problematically high “false positive” findings of SFSTs in research settings suggests that drivers suspected of marijuana use may require a more efficacious mechanism to demonstrate their lack of impairment.  In short, current DUII laws are more than sufficient to deal with marijuana intoxication–we do not need new standards for this particular drug.

Referenced by section in the bill are in bold followed by my commentary:

_____________________________________________________

Section 29:  Grounds for refusing to issue license…

(2Ba): that the applicant is in the habit of using alcoholic beverages, habit-forming drugs, marijuana, or controlled substances to excess.

This is completely indefensible under Oregon law and is simply bad public policy.  How would the OLCC  be able to track an individual’s use of any type of drug in excess?  Who gets to define what “in excess” means?  Will a self-admitted 5-cup-day-coffee drinker be denied a marijuana production/processing/retail license?  What about chain smokers (the cigarette variety)?  And here’s another beauty a few lines down:

(2Be and 2Bf):  has maintained an insanitary establishment and/or is not of good repute and moral character

Nowhere in the bill does it require that producers, processors, or retailers abide by state or county health codes, which calls into question the usefulness and/or applicability of 2Be. Again, 2Bf poses intractable issues as well:  what does this mean and how could it ever be legally enforceable?  This is a verbatim transcription of one of the worst sections of the failed 2012 voter initiative and it makes zero sense to include it.

Take a look at Section 29 2Bd:  doesn’t this cover everything that a licensing agency would need for legally defensible decisions?  I’m not an attorney, but it seems to cover the situation.

The same issues are present in Section 30 (grounds for cancellation or suspension of license) as well and should be removed.

_____________________________________________________

Section 57:  Homemade marijuana extracts prohibited.

You can’t do this if you define marijuana as “all parts of the plant Cannabis family Moraceae, whether growing or not; the resin extracted from any part of the plant; and every compound, manufacture, salt, derivative, mixture, or preparation of the plant or its resin” (ORS 475.005).  ANYTHING made from marijuana is considered marijuana.  Plus, it just seems like bad form to allow individuals over the age of 21 to grow marijuana but not use it how they deem appropriate.

_____________________________________________________

State Senator Floyd Prozanski has a legalization bill ready for introduction at the February 2014 legislative session.  The bill–you can get your copy here–is entitled “Control, Regulation, and Taxation of Marijuana and Industrial Hemp Act”.  The legislature has been interested in passing a thought-out legalization bill before a less articulate version is passed by Oregon voters (and public opinion polls suggest that will occur in November 2014):  this just might be “the one”.

I’ll say from the outset that I think this is a tremendous piece of public policy.  It’s well-thought out, responsible, and, in most cases, informed by the best science available.  This bill makes Washington and Colorado’s legalization bills look rudimentary (which it should–thanks to them, we have the advantage of hindsight).  That said, there are a number of items that need to be addressed for this bill to work:  I’ll go into the details in subsequent posts.  I really don’t like having marijuana placed under the regulatory scope of the Oregon Liquor Control Commission.

Legalization is a complex topic when you get into implementation;”back of the envelope calculations” often fail to consider the structure of the existing marijuana market and, in so doing, can mire the formalization process with unnecessary impediments.  Marijuana at the store can’t be significantly more expensive than its black market counterpart.  The greatest mistake of the early proposals are high tax rates applied at various stages of the marijuana production process.  This is inappropriate for a number of reasons, but, most importantly, simply because the brunt of tax revenue will be generated through income tax collection, not marijuana-specific taxes.  In fact, at the early stages of legalization more revenue is likely to be generated by having no tax at all on marijuana itself.  I’ll get into more details on this particular point in a future post.

Here’s an overview of the analysis I’ll provide on this bill:  part 2 will detail the small issues that need to be remedied in the bill; in part 3, I’ll tackle the “drugged driving” section of the bill and apply the most recent scientific evidence to offer some recommendations on how to handle this particular issue; part 4 will get into the nuts-and-bolts of appropriate tax structures by demonstrating the unity between formal and informal markets.  I will have these sections up in the next couple weeks (way before the legislative session begins on January 27th).