Since the IMF put a tariff equivalent on the barriers inside the single market, several figures have circulated for goods and for services, in speeches, reports and the press, and economists have not agreed on any of them. For the 28th regime it does not matter which one you take. None of them measures what the regime changes, and over the ten years the Commission has calculated, the regime saves paperwork, not trade: EUR 328 million to 440 million in administrative costs, spread across an estimated 308,000 companies.
The 28th regime is the proposed optional EU company form, Commission proposal COM(2026) 321 of 18 March 2026. This piece checks what the internal-tariff estimates can carry; it does not argue for or against the proposal.
In this piece
- The estimates disagree, and every cause they point to sits outside company law
- The Commission’s impact assessment files the IMF figure under company law
- Over the ten years it calculates, the Commission expects paperwork savings, not trade effects
- The barriers the tariff comparison points at are handled in other acts
- Method and data
- Sources
Read the passage yourself: impact assessment SWD(2026) 321, part 1, on the Commission’s site. Section 2.1 carries the IMF figure; section 8.4 carries the savings.
The estimates disagree, and every cause they point to sits outside company law
The best-known figure comes from a speech. On 16 December 2024 in Vilnius, Alfred Kammer, director of the IMF’s European Department, said the Fund’s estimates suggest barriers to trade between EU countries “might be as high as” a tariff of about 44 per cent for goods and 110 per cent for services, against an effective external EU tariff of around 3 per cent.
The method behind it compares how much member states trade with each other with how much they trade inside their own borders, after allowing for distance. Whatever gap is left counts as a barrier: product rules that differ, services licences, language, public buyers who prefer local suppliers, consumer habits. The method does not say which of these causes the gap, and the IMF did not attribute it to company law. When the Fund later listed what holds back Europe’s growth, internal trade barriers sat next to limited mobility of workers and capital and gaps in labour market and business regulation.
Economists do not agree on the size. Keith Head and Thierry Mayer, who have measured European border effects since 2000, argued on VoxEU in November 2025 that the estimate is likely too high and depends heavily on which dataset and which estimation choices are used. Rebuilding the calculation from the same trade data the IMF paper used, they got 32 per cent, close to a 30 per cent figure inside that paper; the 44 per cent headline comes from how sector results were added up. When they also allowed for social connections between regions, the border effect shrank sharply with every dataset and disappeared with two of them. Their reading is that much of what remains reflects business networks and differences in taste, which governments cannot change quickly.
Authors of the IMF analysis returned to the question in February 2026. They accept that a preference for locally made goods is hard to measure and ends up inside their estimates, and they propose comparing Europe with the barriers between US states to judge how much policy could remove. The title of their column says where they place the problem: fragmented product markets. An ECB Economic Bulletin article lists published estimates for goods ranging from 8 per cent to 60 per cent, depending on the countries, years and method.
The figures have travelled far from the research. Mario Draghi used one in the Financial Times in February 2025. The Commission President used one in her 2025 State of the Union address, and at the launch of the proposal on 18 March 2026 she presented the regime as an answer to fragmentation inside Europe, arguing that barriers within the Union hurt more than tariffs from outside.
What it means. The size does not need settling to answer the question here. Every cause named on either side of the debate (product rules, services licensing, public buyers who favour local firms, networks, tastes) sits outside company law. Take the lowest estimate or the highest: the 28th regime was not written to move it.
The Commission’s impact assessment files the IMF figure under company law
The Commission’s impact assessment cites the Vilnius speech and restates it (part 1, section 2.1): “The International Monetary Fund (IMF) estimates that legal fragmentation across Member States creates a non-tariff barrier equivalent to a tariff of about 44% on average for traded goods.” The IMF did not say legal fragmentation. A few lines further down, the assessment sets its own limits: “This impact assessment focuses on problems and drivers related to corporate rules, including insolvency.”
The Commission’s consultation record says the same from the other side. Summarising the discussions in its High-level Forum on Justice for Growth, the explanatory memorandum of COM(2026) 321 records that several participants “underlined that while the fragmentation of the company law was indeed a problem, many difficulties were outside the company law area.”
What it means. The impact assessment borrows the figure to set the scene and then, correctly, confines itself to corporate rules. When an internal-tariff number appears next to the 28th regime, read two separate claims: one about the single market, one about company law. Only the second is in this proposal.
Over the ten years it calculates, the Commission expects paperwork savings, not trade effects
Section 8.4 of the same assessment gives the headline. Administrative savings, estimated with the Standard Cost Model (the Commission’s method for putting a price on forms and procedures), come to EUR 328 million to 440 million over ten years for the companies that choose the regime. The executive summary puts the expected number of those companies at 308,000.
| Saving | Commission estimate | Where it comes from |
|---|---|---|
| Setting up the company | EUR 550 to 1,300 per founder, once | Fast-track registration capped at 48 hours and EUR 100 (section 6) |
| Share transfer | EUR 1,780 to 2,850 per transaction | Digital share transfers (section 8.4) |
| Financing round or other capital operation | About EUR 1,100 per transaction | Online meetings and capital increases without in-person notary steps (sections 6 and 8.4) |
| No paid-in capital at incorporation | About EUR 400 per company | Only where a member state requires paid-in capital today; EUR 55.2 million over ten years (section 6) |
| All administrative savings | EUR 328 to 440 million over ten years | An estimated 308,000 companies (section 8.4, executive summary) |
Swipe sideways to see the full table.
Dividing the Commission’s total by its own company count gives roughly EUR 1,065 to 1,430 per company over ten years. That division is ours; the Commission publishes no per-company figure.
None of these is a trade effect, and the assessment does not claim one. The savings come from registration, capital, share transfers, the timing of tax on employee stock options, and closure procedures. That is the limit of what has been calculated. The European Parliament’s research service noted that the assessment leaves some expected benefits unquantified, additional investment flows among them. If the regime pays off more in the longer run, that is where it would show: a company that is easier to invest in across borders, not goods that are cheaper to sell across them.
What it means. On the Commission’s own calculated horizon, the regime is a paperwork saving. For a startup that raises money several times and moves shares between investors, EUR 1,100 a round and up to EUR 2,850 a transfer is real money, and the case for the regime can rest on it. It does not need a tariff comparison, and it cannot carry one. Our reads of how formation would work and of the stock option scheme go through where these savings come from.
The barriers the tariff comparison points at are handled in other acts
Some single market barriers are being taken apart this year, in acts that have nothing to do with company form. On 23 June 2026 the Council and the Parliament provisionally agreed a regulation creating one EU interface for posting-of-workers declarations, which member states can choose to use in place of their own portals. The Council’s announcement names burdensome posting procedures as one of the “terrible ten” barriers in the Commission’s single market strategy, and calls the deal one of the first deliverables of the “One Europe, One Market” roadmap. On 15 September 2026 the Commission proposed a Fair Labour Mobility package, including a European Social Security Pass for requesting social security documents, such as the certificate a posted worker carries, fully online.
The 28th regime is one item in the same “One Europe, One Market” agenda that EU leaders endorsed on 19 March 2026. It answers one kind of friction on that list, the cost of running a company under 27 sets of corporate rules.
Judge the 28th regime by what it touches: the cost of being a company, not the cost of crossing a border.
Method and data
We downloaded the impact assessment (SWD(2026) 321, parts 1 and 2) and its executive summary (SWD(2026) 322) from the Commission’s site and searched the text for “tariff”, “IMF” and “44”. The IMF figure appears once, in section 2.1 of part 1. The savings figures are in sections 6 and 8.4 of part 1 and in the executive summary. The per-company range under the table is our own division of the Commission’s total by the Commission’s own company count. The IMF wording is from the text of the Vilnius speech of 16 December 2024. The consultation quote is from the explanatory memorandum of COM(2026) 321 as sent to the Council (ST 7498/26).
The newsletter
Follow the 28th regime from the documents
We track procedure 2026/0074(COD) through the Council working party, the committee amendments and the Commission’s own papers, and say what each change means in practice. One email every two weeks or so.
Subscribe on the28thregime.euSources
- European Commission, SWD(2026) 321 final, impact assessment report accompanying COM(2026) 321, 18 March 2026: part 1, part 2
- European Commission, SWD(2026) 322 final, executive summary of the impact assessment: PDF on the Commission’s site
- European Commission, proposal COM(2026) 321 of 18 March 2026, as sent to the Council, ST 7498/26 on the Council register; procedure 2026/0074(COD)
- IMF, Alfred Kammer, “Europe’s Choice: Policies for Growth and Resilience”, speech, Vilnius, 16 December 2024: imf.org
- IMF, Alfred Kammer, opening remarks at the press conference on the economic outlook for Europe, 17 October 2025: imf.org
- Keith Head and Thierry Mayer, “No, the EU does not impose a 45% tariff on itself”, VoxEU, 13 November 2025: cepr.org
- “EU barriers to scaling up: The case of fragmented product markets”, VoxEU, February 2026: cepr.org
- European Central Bank, “What is the untapped potential of the EU Single Market?”, Economic Bulletin article: ecb.europa.eu
- European Parliamentary Research Service, briefing on the impact assessment, EPRS_BRI(2026)788143
- Council of the EU, “Council and Parliament agree on digital declaration system for posted workers”, press release, 23 June 2026: consilium.europa.eu
- European Commission, “Commission proposes ambitious measures to strengthen fair labour mobility”, 15 September 2026: employment-social-affairs.ec.europa.eu