Critism of analysis of local marijuana taxes in California

“an independent California cannabis tax expert . . . said that . . . the city (Lemon Grove) could bring in $560,000 at 4 percent and $1.12 million at 8 percent.”

$560,000 X 2 = $1.12 million.

So if the city taxed it at 16%, they would bring in $2.24 million, and if they taxed it at 32%, they would bring in $4.48 million, and if they taxed it at 64%, they would bring in $9.96 million?

Well, there wouldn’t be a huge drop-off going from 4% to 8%, but there would be at least SOME bit of migration to the black market and to lower-tax jurisdictions.  So revenues would not exactly double.  They’d be less.

Close enough for government work?

How tax competition can threaten marijuana revenue

How tax competition can threaten marijuana revenue

Pat Oglesby, opinion contributor

Here’s what to know about cannabis tax competition.

280E is constitutional

280E allows marijuana sellers to deduct only cost of goods sold, so they can’t deduct advertising and marketing expenses.  Taxpayers say the application of 280E to marijuana businesses is unconstitutional.  Off the cuff, I disagree.  (Here’s the Tax Court decision (finding 280E constitutional) and dissent in this case.) Continue reading “280E is constitutional”

Tax marijuana by potency? $50 reward for finding a jurisdiction that does for cigarettes.

“I’ve challenged people over and over, including at conferences and on twitter, to name a jurisdiction that taxes tobacco by nicotine content, and no one has.” On 4 September 2018, I wrote that, and it’s still true.

Make that “unconcentrated tobacco” or “cigarettes.” Now that vaping has caught on, I wouldn’t rule out taxing vaping material by nicotine content.  It’s concentrated, so it’s homogeneous.  And it’s workable to tax marijuana concentrates by THC, like Canada.

But taxing unconcentrated plant material is SUCH a bad idea that no jurisdiction does it for analagous cigarettes.  Sampling allows cheating and manipulation.  These products are not homogeneous.  If that’s a deal killer for tobacco, why isn’t it for cannabis?

So maybe this is a publicity stunt.  The first message RECEIVED naming a jurisdiction that taxes cigarettes by nicotine content gets a $50 reward.  Write

How notches or discontinuities in Illinois’s weed tax are an unnecessary vestige of the primitive Scales of Justice “range” method of liquor taxation

The notion that marijuana flower is best taxed by measuring its THC content refuses to die.  Illinois backed into a THC tax on flower with a rule that smokeable marijuana with THC over a certain percentage (35% or so) is taxed at a higher rate ad valorem – 25% versus 10% for less potent stuff.

Why didn’t Illinois tax by THC directly?  To answer a different question, no jurisdiction in the world taxes tobacco by tar nicotine content.  ($50 prize for first person to identify such a current tax.)  THC in flower is impractical to measure, so a wide “range” of up to 35% should cover it all.  Flower rarely exceeds 35% THC; most smokeable concentrates do.

Leaving aside the weakness of ad valorem taxes, why didn’t they just tax raw plant material at one rate and smokeable concentrates at another?  The rush of getting legislation to move, according to one report, without getting into the fine points.

So there’s a notch.  34.999% hits you with one tax bill; 35.001% hits you with a tax bill 150% higher.  (Gaming possibilities, are listed in the BOTEC report for the Washington State Liquor and Cannabis Board, like spraying flower with concentrates to land comfortably below 35%.)

Notches, or discontinuities in tax rates producing marginal rates over 100%, are  usually suboptimal.  Some may be unavoidable, like the UK VAT threshold.  With 499,999 pounds sterling in turnover, there’s no VAT.  With 500,001, you owe VAT on each transaction.  Now that’s awkward, and calls for prediction (if it’s not based on the previous year’s numbers,) but how else could you work it?  To have identical transactions taxed differently would be weird.  One customer sneaks tax free in before the 500k threshold, but the next customer in line that day pays the tax.

But discontinuities for excise taxes may not be necessary.  They used to be.  They  began, I think, with the liquor tax.  Higher proof spirits weigh less than weak ones, so weighing the batch reveals the alcohol content.  But digital scales weren’t invented when the United States (for instance) started taxing liquor, so they HAD to tax by ranges with discontinuities.  The old Scales of Justice scales, which operated by adding weights “only one at a time,” were all they had – with an alcohol adaptation called “Dicas’s hydrometer.”  So they had six ranges – unavoidable then, for differentiation, but at least internally accurate, and unbeatable by sampling.


Ads and 280E

I’m giving a marijuana tax design talk to a Law School class at a major university on April 9.  One point I plan to make is that Tax Code section 280E, which makes marijuana advertising and marketing expenses nondeductible, is a good idea.  Ads enjoy protection under the commercial free speech doctrine, so states can’t ban them, but states, like the federal government, can make them nondeductible.  The professor saw some slides I put together, and didn’t think these were a useful part of an academic discussion.


CNR Vermont memo on 280E conformity

To:                   S.54 Conference Committee

From:             Pat Oglesby, Center for New Revenue

Re:                   Effectively Restricting Cannabis Advertisements via Tax Policy

Date:                March 11, 2020

Current Law

Current Vermont tax law follows Federal Internal Revenue Code §280E, allowing sellers of cannabis (whether or not licensed by the state) to deduct only their “cost of goods sold” – that is, the cost of producing or buying the cannabis they sell.

Under 280E, growers can deduct almost everything.  Nearly all their costs, such as salaries, rent, and electricity are treated as “cost of goods sold,” because they are inputs to producing cannabis.

Retailers, however, can typically only deduct only what they pay for the product itself.  They cannot deduct salaries, for instance, because that is an expense of selling cannabis, not of producing or buying it.  Similarly, 280E prohibits the deduction of advertisement costs.

As it looks at legalization, Vermont can consider carefully, ahead of the federal government, what to do about marijuana advertising.


Possible Options:

The House’s version of S.54 seeks to repeal 280E “conformity” for Vermont, and allow all selling expenses to be deducted for state income tax purposes.  This treats cannabis like every other business, which certainly seems an equitable approach – but it also would make advertising expenses tax deductible.  This would effectively create a state subsidy of cannabis advertising.

By adopting an amendment sponsored by Rep. Donahue, the House chose to completely ban cannabis advertising.  Such a blanket ban may be an unconstitutional infringement on “commercial free speech” – the case law is unsettled on the question. It may be that such a ban is permissible under Vermont’s Constitution, and that Federal courts would not take up such a case while Federal prohibition remains in place – but query how long Federal prohibition will last.  If Vermont’s blanket ban on advertisements is ultimately struck down, there will be no restriction on cannabis advertisements.

Alternative 1: Say what’s NOT deductible.

Denying deductions for advertising is a place to start.  Listing other non-deductible selling expenses like excessive salaries stifles potential constitutional challenges.

Alternative 2:  Say what IS deductible.

A more nuanced approach would be to only re-allow certain 280E-disallowed expenses, keeping all others (including advertising expenses) non-deductible.  The Legislature could be guided by broadly acceptable social goals – allowing, for example, deductions for W-2 employee pay and benefits, maybe up to a cap (say, $100,000 per person per year).  A limited list of deductions could favor labor over capital, and small businesses over large ones (“mom and pop” shops use word of mouth and the personal touch to attract customers, not capital).

Possible additional allowable deductions include 1099 “gig” workers, and service providers like lawyers.  Charitable contributions sound nice, but are often disguised promotions. Real estate interests would push hard for rent to be deductible, but parents like to keep retail spaces non-glitzy and unappealing.


A blanket ban on advertisement is likely to encounter Federal and State Constitutional challenges.  If such a ban is enacted, and then overturned, the State would not have any effective retractions on cannabis advertisements on the books, and the Legislature would have to scramble to enact them, over the objections of an industry that is likely to have more lobbying clout in the future than it has now.

By reverting to the restrictions on advertisements in the pre-Donahue House version of S.54 (i.e., audience restrictions, requiring pre-approval of all marketing materials, along with payment of a material review fee, etc.), and ensuring that cannabis advertisements are not inadvertently subsidized by allowing their costs to be deducted for state income tax purposes, S.54 would better protect Vermonters from the dangers of widespread advertisements intended to increase potentially problematic use.



My proudest accomplishment — threatened by the corona virus

The 2017 Tax Act stopped carrybacks of losses generally, primarily as a revenue raising ploy.  Now in a mad dash to Do Something about the corona virus, Congress is going to allow carrybacks so corporations can get cash.

That brings to mind filling out a form last year, in connection with my 50thcollege reunion, that describes my proudest accomplishment.  That was dreaming up Sec. 172(b)(1)(D), the so-called corporate equity reduction transaction rule, preventing the “carryback of excess interest losses attributable to corporate equity reduction transactions.”

Here’s the story.  At the height of the Leveraged Buy Out (LBO) craze in the late 1980s, Senator Bentsen asked staff (me) to Do Something about LBOs, which were making headlines and out of favor.  But I had to be sure not to Spook the Markets.  I was reading Forbes or Fortune in my office one day when a transaction caught my eye.  An LBO artist bought a profitable company in a merger transaction that incurred huge debt.  The debt created huge interest expenses.  Those deductible interest expenses were sot big that the new operation had losses.  Those losses were carried back to the profitable years of the profitable company, creating a TAX REFUND.  That looked like the Code was subsidizing LBOs.  That didn’t sound right, so we drafted up something that Senator Bentsen introduced; Mr. Rostenkowski put it in the House bill and it sailed into the law without any opposition rearing its head. The rule was estimated as picking up $2.2 billion in revenue over five years.

A side note:  To make sure we didn’t Spook the Markets, Senator Bentsen had me hold a Staff Briefing to float the notion.  Famous journalist Jeff Birnbaum, with the WSJ then, left the briefing in the middle once he understood the plan.  The Markets were Not Spooked, and the provision stayed in the Code until it got subsumed into a broader rule in the 2017 Tax Act.  I hope it comes back.





Experience in Colorado; Naivete in California

Localities in Colorado, which has been allowing local marijuana taxation for years, are figuring it out.  In California, which legalized six years later, they are still stumbling:

Here is a nuanced and thoughtful proposal from Aurora, about to increase retail sales taxes.

COFFMAN: A hike in marijuana tax can save critical Aurora services – Sentinel Colorado;

“Under Lawson’s proposal, there would be a 25% tax increase on the local sales tax rate on recreational marijuana raising the tax rate from 4% to 5%.  This increase would still allow our 24 recreational marijuana stores to keep a competitive tax rate.  For example, even with the increase, Denver would still be half a percent higher than Aurora at 5.5%.

“The proceeds from the marijuana tax increase is estimated to be around $1 million in the first year replacing the $1 million lost revenue from the photo red light program to fully restore the funding shortfall for these critical Nexus programs.

“As Aurora’s new Mayor, I strongly support Council Member Lawson’s proposal and I will encourage our City Council to pass it.”

Mike Coffman is mayor of Aurora.


In California, meanwhile, a blunt approach in Costa Mesa doesn’t distinguish among retail, growing, and testing taxes:

Look, there is some room for local retail taxation in any locality.  Consumers aren’t totally mobile in the long run.  Prohibitionists say there are negative externalities with consumption, and they can’t be persuaded.

Processing and distribution are different.  They create jobs, with no negative externalities (right?), and they will drift to low- or zero-tax jurisdictions.

And here’s the thing:  A tax on testing (also mobile in the long run) is a sure sign that lawmakers don’t have the long-term best interests of the community in mind.  Testing is the kind of enterprise most localities want – clean, high-paying jobs.  Those are positive externalities.  Taxing testing is both primitive and unprincipled.

Tax growing?  That’s pretty mobile, too.  Is there a negative externality?  Smell?  Some people like the smell.


More at How tax competition can threaten marijuana revenue;


Continue reading “Experience in Colorado; Naivete in California”

$100 prize for guessing Colorado weed flower tax AMR for Q1 2021

What will Colrado’s Average Market Rate for flower be as of January 1, 2021?  It’s $1,316 today.  Best guess wins $100.  Enter at

In case of tie, earliest sent message wins.  (Hint:  All AMRs have been whole dollar figures so far – no pennies.) Entries close when Colorado releases the price, at which time winner will be announced and check for $100 issued.  One guess per entrant, for now.  Employees of the State of Colorado are ineligible.

Here’s some background: Continue reading “$100 prize for guessing Colorado weed flower tax AMR for Q1 2021”

California Marijuana Weight Taxes as of January 2020 — and Alaska, Maine, Colorado and Nevada

For the record, here are weight tax rates for cannabis in U.S. states as far as I know.  I’m preserving the documents here, since the rates will change.


California adjusts its marijuana tax rates for inflation. Here is a copy of the rates as of early 2020.

Flower, $9.65 per ounce; leaves, $2.87 per ounce; and fresh cannabis plant, $1.35 per ounce.[1] New categories can be created by regulation as needed.

[1]; copied at

2020 January Tax Guide for Cannabis Businesses


Here are Alaska’s rates, from, Average Market Rate for Unprocessed Retail Marijuana | Department of Revenue – Taxation:

“Effective January 1, 2019, sales and transfers of marijuana are subject to new tax rates. Mature bud/flower are taxed at $50 per ounces and trim is taxed at $15 per ounce. A new rate has been added at $25 per ounces for bud/flower that is considered immature or abnormal. This is bud/flower that did not fully mature or develop, contains seeds, or failed testing. Clones are taxed at a flat rate of $1 per clone and not on the estimated weight.”


Here are Maine’s rates, from

(Title 36, §4923: Excise tax imposed has a copy.)

1.  Excise tax on marijuana flower and mature marijuana plants.   A cultivation facility licensee shall pay an excise tax of $335 per pound or fraction thereof of marijuana flower or mature marijuana plants sold to other licensees in the State.

[PL 2019, c. 231, Pt. B, §7 (NEW).]

2.  Excise tax on marijuana trim.   A cultivation facility licensee shall pay an excise tax of $94 per pound or fraction thereof of marijuana trim sold to other licensees in the State.

[PL 2019, c. 231, Pt. B, §7 (NEW).]

3.  Excise tax on immature marijuana plants and seedlings.   A cultivation facility licensee shall pay an excise tax of $1.50 per immature marijuana plant or seedling sold to other licensees in the State.

[PL 2019, c. 231, Pt. B, §7 (NEW).]

4.  Excise tax on marijuana seeds.   A cultivation facility licensee shall pay an excise tax of 30¢ per marijuana seed sold to other licensees in the State.



Here are Nevada’s, with fair market value to be multiplied by 15 percent:

As of 1-1-2019:

Click to access Marijuana-Fair-Market-Value-Jan-1-2019.pdf


Here are newer rates:

Fair Market Value at Wholesale Summary Jan 20 (Final Draft).


Here are Colorado’s — 15 percent of the numbers below:

Average Market Rate for Unprocessed Retail Marijuana | Department of Revenue – Taxation

Period Rate: January 1 to March 31, 2020

 Retail Bud Rate  $1,316 per pound
 Retail Trim Rate  $350 per pound
 Retail Immature Plant Rate  $9 per plant
 Wet Whole Plant Rate  $191 per pound
 Seed Rate  $5 per seed
 Trim Allocated for Extraction Rate  $247 per pound
 Bud Allocated for Extraction Rate  $299 per pound



Fear of a federal marijuana crackdown in Utah?

What stopped state-run marijuana sales in Utah?  Fear of a federal crackdown?  Or something else?

“Officials in Utah last year abandoned a plan to operate state-run medical marijuana dispensaries, in part because of fears the federal government might intervene or withhold unrelated funding as punishment,” says a Boston Globe story by Dan Adams and Felicia Gans.

I don’t think those fears were warranted.  I suspect they were manufactured. Continue reading “Fear of a federal marijuana crackdown in Utah?”

Exaggerating mj taxes in California?

California story:  “The CDTFA says its analysts determined a 60% markup on wholesale collected by distributors must be raised to 80%. It will ensure the tax paid on all products is equal to 15% as required by law.  Cultivation fees are also going up. Basically, Blurton said that means a product that’s $60 today will be $70 tomorrow.”

Not following the math.  I don’t know what figures to use, but I’m having trouble seeing why a product that’s “$60 today will be $70 tomorrow” on account of taxes.  Looks to me like it should go up a lot less.

  1. Cultivation per-ounce tax

Let’s say that $60 is for an eighth.  The cultivation tax on an eighth went from $1.16 (1/8 x the old rate of $9.25 per eighth) to $1.21 (1/8 x the new rate of $9.65).  That’s a nickel more tax, but maybe it gets marked up in reselling to . . . a dime?  A quarter?

  1. Retail excise tax of 15% of imputed price.

What is the new wholesale price that gets multiplied by 15%?

Say wholesale pre-tax and pre-markup price is and was $25.  The old markup was $15 (60% of $25); the new markup is $20 (80% of $25).  So the pre-tax price on which tax is calculated goes from $40 to $45.  The tax goes from $6 (15% of $40) to $6.75 (15% of $45).

More realisitically, maybe, say wholesale pre-tax and pre-markup price is and was $36.  The old markup was $21.60 (60% of $36); the new markup is $28.80 (80% of $36).  So the pre-tax price on which tax is calculated goes from $57.60 to $64.80.  The tax goes from $8.64 (15% of $57.60) to $9.72 (15% of $64.80).


Hard to see why the after-tax retail price should go up by $10 when one tax goes up by a nickel and the other goes up by around a dollar.  There are markups on the taxes, but not 1,000% markups.  Maybe I’m missing something.  I’ll ask some friends in California.



Center for New Revenue Person of the Decade

Steve Shay, when he was at the Obama Treasury, worked on and pushed through regulations to slow down “corporate inversions”– companies shifting headquarters overseas – over vehement corporate objection.  That should have made him Center for New Revenue Person of the Year then, but the idea of having a Person of the Year needed to percolate.

He has kept fighting for sound tax policy by writing and speaking, over and over, and is just this month out with a capillary-level article explaining how the Trump Administration gave away more to multinationals, by simple regulatory action, than tax professionals though possible.  The Democrats can fix that in 2021.

So Steve Shay is the inaugural Center for New Revenue Person of the Decade.

Draft defense of weight tax on marijuana vs. CA LAO

Working on a response to the LAO report (subject of December 18 post on this blog) for a paper that won’t go final before April 10 – a defense of the weight tax in California (people there call it a cultivator tax, which reflects the confusion – it’s actually collected downstream).  Would welcome comments and pushback.

  1. Weight tax isn’t the worst tax; it’s the best – and clearly better than ad valorem (percentage of price) excise taxes on cannabis. Look at federal alcohol and tobacco taxes.  And all 50 states have non-ad valorem excises on tobacco; all but government monopoly states  have non-ad valorem excises on all three forms of beverage alcohol.

California’s weight tax is not working, but the few problems I’ve heard about (there may be more) can be readily fixed.

a.  Flash frozen stuff needs a separate category – fine.Maybe Colorado’s “wet whole plants” category plugs right in.

b.  CA’s collection mechanism for the weight tax is Byzantine.  Simplification – to tax payment by first distributor, for instance, seems doable.

  1. Federal marijuana legalization won’t happen until there’s a proven method of collecting taxes that doesn’t put the federal government taxing every retail sale in every state.That’s too far-flung a network to supervise, so the tax will need to be collected upstream of retail.

An upstream ad valorem tax is too tricky to work.  Look at Colorado and Nevada, whose nominal ad valorem upstream taxes (15% of producer price) are de facto converted to weight-based taxes.

State taxes will be the model for federal taxes, and federal legalization won’t happen without new taxes (at least if 280E is repealed).  Ad valorem (percentage of price) taxes aren’t that model.  Switching from weight to ad valorem by California would sow confusion in an already glacial federal legalization process.

  1. The LAO’s problems with the weight tax are not stated, except:

“[T]he state has not established any mechanism for consistent, direct third‑party verification of the weight of harvested plants.” How hard could that be?  As @DaveSilberman tweeted, “Any experienced cannabis consumer can tell by sight whether the seller is overstating weight.”  That’s the beauty of a weight tax – it’s verifiable.

So the California tax folks must have higher priorities than this “mechanism.”  (The actual problem, if any, is not stated – just lack of a “mechanism” for figuring out how many grams something weighs.)  California’s enforcement needs a lot of work done.  And third party verification is a strange hook to hand your hat on – its is inferior to state audit verification, which the Dominion of Canada does with its per gram taxes on flower and leaves.

The LAO doesn’t get to any distinction for THC taxes between raw plant material and processed products – and whether that distinction, rather than the number of grams on the scale, isn’t the real tricky part of a weight tax.

  1. The British Indian Hemp Drugs Commission revealed at least six categories of weight-based taxes in use in 1893., page 8. This is the tried and true method.  All these little ad valorem taxes were enacted because they were so easy to draft.  Price collapse, bundling, related party deals (and phony prices) – ad valorem taxes are primitive.



First glance at LAO mj tax report

Updated 18 December 2019:

On Twitter, I saw a California LAO report on marijuana taxes, just released:; H/T @Johnschroyer

I started to skim it, but this brought me up short:

“a potency‑based tax has two key advantages with regard to administration and compliance. First, testing labs verify the potency of cannabis products. (As we understand it, these tests rarely find substantial differences between labeled and actual cannabinoid content. More commonly, products fail lab tests for other reasons, such as high levels of pesticides.) In contrast, the state has not established any mechanism for consistent, direct third‑party verification of the weight of harvested plants. Second, THC content appears on the labels of all cannabis products, allowing for further verification opportunities upon retail purchase. In contrast, once a cannabis product has entered the manufacturing process, there is no way to verify the weight of the raw plant material used to make it.”

That strikes me as wrong, in seeming to say says measuring potency is easy; measuring weight is hard.  It overlooks lab shopping in other states,  “[T]esting labs. . . rarely find substantial differences between labeled and actual cannabinoid content”?  No.  That’s certainly not true in Washington State, where lab shopping is a serious problem, as noted in the Washington State document the LAO cites.

So the California tax folks must have higher priorities than working on this “mechanism” for vertifying weight.  California’s enforcement needs a lot of work done, and finding weight-tax cheaters doesn’t seem to need a lot of “mechanism” — just auditing.. Anyway, “third party verification” seems inferior to state audit verification (wouldn’t the state be the “second party”?), which Canada does with its per gram taxes on flower and leaves.

And once liquor has been manufactured and taxed alcohol content, does anyone take advantage of  “further verification opportunities upon retail purchase”?  It’s a lot easier for consumers to measure alcohol content than THC content of retail purchases.

Weight, it seems to me, is much simpler to verify than THC, especially and prohibitively in raw plant material.  No one taxes THC in plant material, because it’s not homogeneous — just as no one taxes tobacco by tar and nicotine content.  The report doesn’t directly criticize (in what I read) the possible Achilles’s Heel of weight taxes – the line drawing between categories.  For sales of processed material, weight of the inputs (flower, trim, etc.) before processing can be a pretty good proxy for what “we are mad at” — intoxication.  To be sure, weight is a blunt instrument, but it’s state of the art.

The report doesn’t propose (in what I read) any distinction for THC taxes between raw plant material and processed products.  I’m all for the Canadian system — tax raw plant material by the gram of flower or bud, with an ad valorem minimum tax, but tax processed products by THC.

That’s enough for now.  At some point, I’m going to take a close look at the LAO report, because I’m guest lecturing at the Duke Law Tax Policy Class on marijuana next spring and I need a paper to present.  For the Center for New Revenue, border tax adjustments for carbon taxes need work a lot more than marijuana taxes do.  Maybe I’ve missed stuff in the LAO report.

A final nit: The LAO report states, “[T]he state has not established any mechanism for consistent, direct third‑party verification of the weight of harvested plants.” How hard could that be?  As stated in this @DaveSilberman tweet, “Any experienced cannabis consumer can tell by sight whether the seller is overstating weight.”  That’s the beauty of a weight tax – it’s verifiable.