The sequence error: when an early price sets a later one

Cause of death: a price agreed in a small early market travelled, by rule, into every market the company had been saving.

Most companies choose their launch order for operational reasons. The regulator is fast. The distributor is ready. Someone knows the country. Then a price gets agreed, and it turns out that price was never a local decision at all.

Roughly 23 of 31 European countries use external reference pricing as their primary systematic pricing criterion.¹ Across economically comparable countries, 15 of 19 use it in some form, with a median basket of seven countries.² The most referenced countries are France, the UK and Germany.³

What this means in practice is that a national pricing authority does not negotiate your price from first principles. It looks up what you charged elsewhere and applies a rule.

The rules are not gentle

The rules vary, and the variation is the whole game.

Some take an average. The Netherlands sets a statutory maximum at the arithmetic mean of four countries, revised at least twice a year, and it swapped Germany for Norway in 2020.⁴ Switzerland compares ex factory prices across nine countries, applies country specific deductions, and then weights that result 50/50 against a domestic therapeutic comparison, revisiting every three years.⁵

Some take the lowest. Brazil caps the factory price at the lowest price charged for the same product across a basket of 14 countries plus the country of origin, and requires marketing in at least four of them before it will set a definitive price at all.⁶ Turkey references the lowest wholesale price among its designated source countries and then caps an originator at 60 percent of it.⁷ Greece uses the average of the two lowest distinct eurozone prices.⁸

Some take the middle. Canada's regulated basket of eleven countries sets new medicines against the median international list price. The United States and Switzerland, the two highest priced comparators, were removed from that basket in 2019.⁹

Japan's rule is the one people forget. If the calculated domestic price exceeds 1.25 times the average of the US, UK, German and French prices, it is adjusted down.¹⁰

Read those together and the sequencing consequence is obvious. In a lowest price system, a single cheap market anywhere in the basket sets the ceiling for everything. In an average system, each early market you concede drags the mean. And the clocks differ: some revise annually, some every three years, some twice a year. A price you agree today reprices markets you have not entered yet, on schedules you do not control.

The evidence that this changes behaviour

The World Health Organization's own guideline on pharmaceutical pricing lists, among the consequences of external reference pricing, "launch delays and product withdrawals in countries expected to list such products at lower prices, to avoid prices being referenced." It qualifies this as theoretical and without robust attribution.¹¹ Its recommendations on the practice are conditional, not strong, and rest on no controlled studies.

The empirical work is less equivocal.

Across 85 new chemical entities in 25 markets, only about 55 percent of possible launches happened at all. Non launch, not delay, was the norm. A 10 percent higher expected price was associated with a 2.65 percent higher launch hazard. The authors' stated mechanism: "a low price in one market may 'spill-over' to others, through parallel trade and external referencing," so manufacturers "rationally prefer longer delay or non-launch to accepting a low price."¹²

Then there is the natural experiment. When Germany introduced reference pricing as a supporting criterion in 2011, prices fell in the countries that reference Germany, by between 3.5 and 7.5 percent. But prices rose in the countries inside Germany's own reference basket, by 5.5 to 8.4 percent, because manufacturers negotiated harder in those markets in anticipation of Germany looking them up later.¹³

Nobody announced that strategy. The prices moved anyway.

The clearest evidence is a government's own words

In 2024 Germany legislated an option for confidential reimbursement amounts. The stated rationale, recorded in the consultation submissions on the draft bill, was that transparency restricted the negotiating parties because the reimbursement amounts negotiated in the German system "exert an international effect as reference prices," reducing manufacturers' revenue prospects outside Germany.¹⁴

A national legislature changed its pricing law because its own prices were propagating abroad. That is the sequence error, described by the state that causes it.

Academic work on the German withdrawals is more direct still. Of 26 products withdrawn for economic reasons between 2011 and 2016, the stated mechanism was that Germany is a reference price country for at least 18 European and non European countries, so manufacturers wanted to avoid the discounted German price becoming public in order not to accept lower prices in the countries that follow it.¹⁵ Worth noting in the same breath: the German medical profession's own account of the same withdrawals attributes them to something else entirely and does not mention international referencing at all.¹⁶ Two credible readings of one set of facts.

The workaround, and its limits

The industry answer to external reference pricing is to keep the list price high and discount confidentially. It works, up to a point, and it has quietly hollowed out the system it was designed to evade.

Only one country in the WHO European Region has transparent net price accessibility. In most, net price data are not accessible, and in some it is not even known which products are subject to a confidential agreement.¹⁷ The European Commission conceded a decade ago that publicly available prices "are often facial prices which do not take into account the managed entry agreements, as these are often confidential."¹⁸ An Italian regulator now requires applicants who cannot disclose a foreign price to disclose instead the counterparty, the execution date and the duration of the confidentiality agreement: a formal concession that the number it wants does not exist in public.¹⁹

Which produces the trap. External reference pricing depends on list prices. Confidential discounts have made list prices fictional. So companies now defend a list price they do not receive, in order to protect other list prices they also do not receive, while every payer in the chain knows the numbers are not real.

The counterweight, because the evidence is not one sided

Two official evaluations looked for the access consequences of a reference pricing change and did not find them.

The Netherlands evaluated its own 2020 basket swap and found realised price pressure lower than forecast, and on availability concluded that no difference could be established between the periods before and after the amendment, partly because of data quality and simultaneous mitigating measures.²⁰

A Canadian cohort study of 349 new patented medicines found two year launch proportions in Canada fell from 45.0 to 30.8 percent after December 2017, but that comparator countries fell similarly, from 83.9 to 70.1 percent, and found no overall negative impact attributable to the policy. The exception matters: for medicines with major therapeutic benefit, Canada fell from 45.8 to 31.3 percent while US launches rose from 91.7 to 100 percent.²¹

How to see it coming

Before the first price is agreed anywhere, build the map. Which countries reference which, on what rule, on what clock. It is public information and almost nobody assembles it before it matters.

Then ask the only question that counts: if this price becomes the reference, what does it cost me in the markets I have not entered yet? If that number exceeds the revenue from the market in front of you, the sequence is wrong, and no amount of local negotiating skill will fix it.