Dutch Pilot Inspections Rose to 145. Who Pays for That?

Inspections of Dutch pilot coffeeshops reportedly went from 8 in 2023 to 145, and fines turned up within the first year. Both numbers have been read as signs of a system doing its job. Neither has come with a price tag.

My argument is simple. A rising inspection count is a cost line as well as a success signal, and it belongs in the cannabis policy debate next to tax revenue and crime. If keeping a closed supply chain honest takes effort that grows quickly, then that effort is part of what a regulated market costs.

A caveat first. The 8 and the 145 come from our own earlier summary, not from a document I have been able to check. The year, the scope and the municipal split are still being confirmed against the primary evaluation report. Where I am inferring rather than citing, I will say so.

What I think is missing from the coverage, ours included, is the question of how many people and how much money sit behind those figures.

The Number Everyone Read as Good News

When we wrote up the Dutch Cannabis Pilot Report: Regulated Supply Chain Diversifies Market, the tone was cautiously positive. The April 7, 2026 one-year assessment landed in the same place: an apparent success, with some fines issued. More inspections, in that reading, means the authorities are paying attention.

I do not think that reading is wrong. I think it is incomplete. The jump from 8 inspections in 2023 to 145 is the figure doing the most work, and it comes from our own earlier summary. It is pending confirmation against the primary report.

It may also not be like-for-like. If the 2023 figure covers a few weeks after launch and the later one covers a full year, the rise says little about scaling. The same goes if more growers and shops entered the system over that time. Scales fast could mean the effort per outlet climbed, or only that there were more outlets.

Then there is a definitional problem. Is 145 a count of inspections, of inspected sites, or of inspection visits? Those are different units with different costs.

The argument has to survive whichever answer turns out to be true. It does, because the question it asks, what the effort costs per licensed outlet, can be asked of any of those units. It just gets a different answer for each.

What the Reports Count and What They Leave Out

What is documented is this: the evaluation work and the April 7, 2026 assessment report inspections and some fines. The primary source is the RAND and Trimbos evaluation, and the September 9, 2026 report followed that earlier assessment. Those are the documents the experiment's public story rests on.

What I cannot find, in the coverage I have seen, is the denominator. I do not see the number of inspectors, the hours they spent, or the budget behind the inspection figures. That is my read of the coverage, not a claim about every document that exists. A municipal report or a ministry answer to a parliamentary question could easily hold what the headline evaluation does not.

Who inspects matters a great deal. Municipal enforcement, the NVWA, the police and a ministry each carry different costs, and each publishes in a different place. A count that mixes them is hard to price.

The fines reportedly issued in the first year are a real data point, but I cannot yet say how many there were, how large, or for what kind of violation. A handful of paperwork penalties and a few serious breaches would tell very different stories.

If the cost data exists, I would look first at municipal records and at the Ministries of Health (VWS) and Justice and Security. It is unlikely to be in the evaluation's headline chapters.

Why a Closed Chain Is Expensive to Keep Closed

The experiment is built as a closed chain. Licensed growers supply participating coffeeshops in selected municipalities, and nothing is meant to enter or leave outside that loop. Every handoff is a place where the chain can leak.

Here is my inference, and it is an inference: a closed chain has to be policed at each link. That means the grower, the transport and the shop. If so, enforcement effort should rise with the number of licensed participants, and probably faster than the number of shops alone, because each added grower adds its own links.

Compare the older model. The coffeeshop front door was tolerated, while the supply behind it was not licensed or inspected as a legal activity. The old system tolerated a gap in the middle. The pilot closes that gap and takes on the job of watching it. Different design, different enforcement bill.

There is evidence that looking finds things. Reports of violations among growers in the ongoing experiment suggest inspection is turning up real issues. I will not quote figures I cannot confirm.

None of this says inspection is wasteful. Catching violations is what a closed chain pays for. My point is narrower: this is a recurring operating cost, not a one-off launch expense. It will show up every year the pilot runs, and a policy debate that counts only tax revenue and crime is missing a column.

A Yardstick: Enforcement Cost Per Licensed Outlet

Here is the yardstick I would propose: total annual enforcement spend divided by licensed outlets. Alongside it, two companion figures: inspections per outlet and cost per inspection. All three need the same small set of inputs.

• Inspector full-time equivalents
• Hours per inspection
• Lab and testing costs
• Fines collected
• Number of licensed growers and shops

Per outlet beats a raw count because a raw count has no scale. 145 inspections means one thing across a few dozen outlets and something quite different across hundreds. A count without a denominator cannot be compared with anything.

The comparison I care about is across models. A closed-chain pilot, a state license regime with seed-to-sale tracking and a tolerated-sales model all enforce differently. They can only be set side by side if each reports the same inputs.

I cannot fill in the Dutch numbers yet, and I would rather leave the table blank than invent them. The point is that the denominator should be published. Without it, one regulator's rigor and another's laxity look identical on paper.

The Strongest Objection: It Is Start-Up Cost, and Illicit Markets Are Not Free

The best objection is that this is start-up cost. Early inspection is front-loaded. Rules are new, licensees are untested, and effort may fall as compliance settles. If later figures drop, my claim that effort scales fast weakens. I accept that, and it is why one year of data cannot settle anything.

The second objection is also fair. The illicit market costs enforcement and policing money too, and that is rarely tallied either. A regulated market could carry a visible inspection bill and still come out cheaper than the hidden one it replaces. I would not be surprised.

Where does the idea survive? Neither counter removes the need to report the cost. They only make the comparison impossible to settle without the numbers. An offset you cannot measure is not an offset, it is an assumption.

So what would settle it? Three things:

• Multi-year inspection counts, to see whether effort levels off.
• Staffing and budget lines behind those counts.
• A like-for-like baseline for policing the unlicensed supply in the same areas.

If the third is unobtainable, as it may be, the first two still allow a clean comparison between regulated programs.

Questions to Put to Your Own State's Program

You can run this test on your own state's program. Ask the regulator for the inspector headcount, the annual number of inspections and the enforcement budget. Then divide by the number of licensed outlets.

Ask who pays. Is enforcement funded by license fees, by cannabis tax or by the general budget? That decides who really carries the cost of a regulated supply chain, and whether a growing inspection effort is paid for by the industry or by everyone else.

Ask whether the numbers are published by year and by municipality or county. Without that split you cannot tell a real trend from a change in how many outlets were open.

Treat the Dutch pilot as a test case. Its evaluators are in a position to publish cost data that most programs never release, and readers can reasonably ask for it.

Rules and legality vary by place, so check local law and read the regulator's own published reports before drawing conclusions.

The practical takeaway is modest. Any claim that a regulated market works should come with a per-outlet enforcement cost, and readers are entitled to ask for it. Tax revenue and crime figures tell half the story. What it costs to keep the system honest tells the rest.

The open question is whether the Dutch inspection effort levels off or keeps growing as the pilot matures. Later reports and municipal records will show which. Until then, 145 is a number waiting for a denominator.

Browse our seed collection.