

THE SIMPLIFICATION RESPONSE
None of this was presented as retreat. The Council described it as simplification to boost competitiveness. The process began with the Draghi and Letta reports and with a declaration in November 2024 calling for a simplification revolution, and it was preceded by a directive in April 2025 that postponed the obligations by two years and became known as Stop the Clock.

THE SIMPLIFICATION RESPONSE
What an institution does when it cannot match the variety of what it faces
Article 3 of 8 · Series I of III · Published 26 August 2026 · Analysis → Forecast → Recommendations
How this series measures things. Every article applies the same three questions to its subject. Concentration: how many genuinely independent alternatives exist, once shared upstream origins are traced rather than counted. Criticality: what stops if this fails, and how quickly. Substitution time: how long until an alternative actually functions. Article 2 established that institutional capacity is the most concentrated dependency in the system. This article examines what that dependency does when it runs short.
1. The Signal
On 18 March 2026 a directive came into force that removed most European companies from the scope of sustainability reporting.
Directive (EU) 2026/470, published in the Official Journal three weeks earlier, narrowed the Corporate Sustainability Reporting Directive to undertakings with more than 1,000 employees and net turnover above 450 million euros. Both conditions must be met at once.
The previous test was different in kind, not only in level. It applied to large public-interest undertakings above 500 employees, where large meant exceeding two of three thresholds: a balance sheet total of 25 million euros, net turnover of 50 million, or 250 employees on average.
A three-part test with a two-of-three rule became a two-part test with an and. The number moved, and so did the shape of the question.
The same instrument narrowed the Corporate Sustainability Due Diligence Directive to undertakings above 5,000 employees and 1.5 billion euros of turnover. Listed small and medium enterprises were exempted entirely. Sector-specific standards were removed. Companies below 1,000 employees that sit in the value chain of a reporting undertaking were designated protected entities and given the right to refuse information requests that go beyond a voluntary standard.
Firms already reporting under the first wave received transitional relief for two financial years. Member states have until 19 March 2027 to transpose the reporting changes, and until 26 July 2028 for the due diligence ones.
None of this was presented as retreat. The Council described it as simplification to boost competitiveness. The process began with the Draghi and Letta reports and with a declaration in November 2024 calling for a simplification revolution, and it was preceded by a directive in April 2025 that postponed the obligations by two years and became known as Stop the Clock.
Article 2 argued that institutions facing more than they can process do not seize up — they economise on variety. This is what that looks like when it is finished: a rule that once asked a nuanced question of many firms now asks a simple question of few, and the change arrives under a positive name with a competitiveness rationale attached.
This article is about that response. Not whether this particular instance was good policy, which is a separate argument and one on which reasonable people differ. About the mechanism, its three recurring forms, who pays for it, and why it does not run backwards.
2. The Mechanism
There is a result in cybernetics that governs this and is almost never cited outside it.
A control system must possess at least as much variety as the system it regulates. Variety here means the number of distinct states it can recognise and respond to. If the regulated system can present a hundred meaningfully different situations and the regulator can distinguish twenty, the remaining eighty are not handled badly. They are not handled at all, because they are not seen as distinct.
An institution facing more variety than it can match has exactly two options. Acquire more variety, which means more people, better training, more time, better data infrastructure. Or reduce the variety it is required to handle.
Article 2 established why the first option is slow: assessors are trained rather than hired, and the substitution time runs to years. When variety must be matched now and capacity arrives in five years, the second option is not a preference. It is the only move available.
The three forms
Variety subtraction is not a single manoeuvre. It recurs in three distinguishable forms, and telling them apart matters because each one hurts a different party.
Form one: the category replaces the case
An assessment that once considered particulars is replaced by a classification that considers membership. The question stops being what is true of this situation and becomes which box this situation falls into.
The Omnibus does this twice. Protected entity is a new category defined purely by headcount, and everything inside it receives the same treatment regardless of what those firms actually do. Sector-specific standards were removed, which is the same move seen from the other side: the sectoral distinctions that made the assessment finer were the part that could not be scaled.
The category is cheaper because membership can be checked and particulars cannot. That is the whole of its appeal, and it is a sufficient reason on its own.
Form two: the threshold rises above capability
A rule that applies to everyone above a line survives by moving the line, and the line moves to wherever the institution can still cope rather than to wherever the policy rationale sits.
This is the clearest form in the Omnibus. Five hundred employees became one thousand. Two hundred and fifty within a two-of-three test became one thousand within an and. For due diligence, the line moved to five thousand employees and 1.5 billion euros.
The stated rationale is that the largest undertakings have the greatest influence on their value chains and are best equipped to absorb the cost. That argument is coherent and may well be right. It is also exactly the argument one would construct after the fact for a line drawn where capacity ran out. The two are indistinguishable from the outside, which is the subject of section 4.
Form three: the obligation stops binding
The rule remains on the books and ceases to compel. Enforcement is not withdrawn in name; it is withdrawn in effect, through exemptions, transitional relief, rights to withhold, or the substitution of a voluntary standard for a mandatory one.
The Omnibus does this in several places at once. First-wave firms received relief for two financial years. Protected entities gained a right to refuse requests exceeding the voluntary standard. Undertakings below the new thresholds may report against a voluntary standard instead, which converts an obligation into an option while leaving the subject matter formally addressed.
This form is the hardest to observe, because nothing is repealed. A reader of the statute book would find the obligation intact.
What the three have in common
Each reduces the number of distinct situations the institution must recognise. None of them reduces the number of distinct situations that exist.
The variety does not go anywhere. It stops being the institution's problem and becomes somebody else's.
Applying the instrument from Article 1 makes the position plain. Concentration: for most of these determinations there is exactly one body. Criticality: high, and rising as more of economic life requires a permission. Substitution time: years. This is the dependency profile of a single-source supplier with an eighteen-month qualification cycle, and it sits underneath every plan that requires an institution to say yes.
3. Subtheme One — Who Pays
Simplification is routinely discussed as though its costs and benefits fell evenly. They do not, and the distribution is predictable enough to plan around.
The typical case is genuinely better off
This must be said first, because an argument that treats simplification as pure loss will be wrong about why it keeps happening.
A firm that clearly sits below the new thresholds has been relieved of a real and substantial cost. A firm that clearly sits above them now faces a shorter, more predictable set of requirements. Both get faster answers and spend less on determining what applies to them. The relief is not rhetorical.
Roughly four fifths of previously covered undertakings fall outside the new scope, according to widely reported estimates of the change. Most of them experience this as the removal of an obligation they found burdensome and whose marginal contribution was unclear. That is a real benefit to a large number of parties, and it is why simplification is popular and why resisting it is politically expensive.
The atypical case pays
The cost falls on situations that do not fit the surviving categories, and it falls in three distinguishable ways.
— The case that was better than its category. A firm with 900 employees and unusually high exposure to a supply chain risk is now indistinguishable from a firm with 900 employees and none. It has lost the ability to be assessed on its particulars, which for the firm doing something well is a loss of the only mechanism that recognised it.
— The case that fell just outside. Thresholds create cliffs, and a cliff has a side. At 1,000 employees the obligation is total; at 999 it is absent. Neither number describes anything real about the underlying situation, and both are now decisive.
— The party the rule was protecting. Where an obligation existed to protect somebody other than the regulated firm, narrowing the scope narrows the protection. That party is usually absent from the simplification debate, because the debate is conducted between the regulator and the regulated.
The asymmetry of attention
The beneficiaries of simplification are numerous, identifiable and organised. The costs fall on parties who are diffuse, frequently unaware, and in some cases not yet affected.
That asymmetry does not indicate bad faith. It is a structural property of who shows up. A consultation on reporting thresholds will hear extensively from firms near the threshold and hardly at all from the people the reporting was meant to inform. The resulting rule reflects the balance of participation rather than the balance of interest, and it does so through an entirely legitimate process.
The general form, beyond this case
The same distribution appears wherever variety is subtracted, in any domain.
A benefits system that replaces individual assessment with a categorical test serves the standard claimant faster and leaves the unusual one with no route at all. A permitting regime that adopts categorical exclusions clears the routine application in weeks and offers the atypical project no proportionate path. A platform that replaces human review with automated classification handles the ordinary case instantly and gives the edge case an appeal to the same classifier.
Being ordinary has become cheaper. Being unusual has become more expensive, and it became more expensive without anyone deciding that it should.
4. Subtheme Two — Why It Does Not Reverse
Simplification behaves like a ratchet. It advances readily and returns rarely, for two reasons that compound each other.
It is invisible while it happens
Capacity-driven simplification is indistinguishable from three other things that look identical from outside: genuine deregulation on the merits, modernisation, and efficiency reform.
All four produce the same artefacts. A higher threshold. A shorter form. A removed requirement. A category where an assessment used to be. And all four are announced in the same language, because the language of improvement is available to each of them and the language of capacity failure is available to none.
No institution announces that it has narrowed a rule because it ran out of assessors. It announces that it has reduced burden, improved competitiveness or streamlined a process, and each of those claims is usually true as far as it goes.
One test distinguishes them, and it is worth applying every time. Ask whether the obligation survived with a new date and a smaller scope, or was removed on its merits after an argument about whether it should exist. Deregulation contests the purpose. Capacity-driven simplification preserves the purpose and shrinks the reach, which is why the stated objectives usually remain in the preamble of the instrument that guts them.
The Omnibus passes this test in the direction of capacity. The Green Deal objectives were not withdrawn. The reporting requirement was not argued to be pointless. Scope was narrowed, dates were moved, and the stated purpose survived intact in the recitals.
Reversal requires the capacity that was missing
The second reason is arithmetic and harder to argue with.
Restoring individual assessment requires assessors. Restoring a lower threshold requires the capacity to process everyone above it. Restoring discretion requires people qualified to exercise it and a process capable of defending it on appeal.
Every one of those is the capacity whose absence caused the simplification. The condition for reversal is the condition whose failure produced the original move, and there is no mechanism by which subtracting variety generates the capacity to add it back.
Meanwhile the capability atrophies. A body that has not performed individual assessments for six years has lost the staff who knew how, the internal precedent that made them consistent, and the appellate record that made them defensible. Rebuilding is not a matter of restoring a budget line. It is a matter of rebuilding a professional practice, which takes longer than building one from scratch because it must first overcome the institutional memory of not doing it.
What reversal actually requires
The historical cases where variety was restored share a pattern, and it is not gradual improvement.
A failure visible enough to be undeniable, attributable to the specific simplification rather than to circumstances, with an identifiable victim, occurring inside the political attention span of the body responsible.
Those four conditions rarely coincide. The cost of variety subtraction is diffuse and delayed by construction, which is precisely what makes it available as a response. A simplification whose harm was immediate and concentrated would not have been adopted.
5. What Most Analysis Gets Wrong
That simplification is deregulation
Deregulation contests whether an obligation should exist. Capacity-driven simplification preserves the obligation and narrows who it reaches. The two produce similar artefacts and imply opposite forecasts: a deregulatory cycle can reverse when politics change, while a capacity-driven one cannot reverse until capacity exists.
That it is a failure
It is the correct move for an institution that cannot acquire variety in the time available. The alternative is not fine-grained assessment for everyone; the alternative is a queue that grows without limit, which serves nobody and eventually produces the same simplification under worse conditions. Criticising an institution for economising misidentifies the constraint.
That the burden was removed
The variety did not vanish. A firm no longer required to report still has the exposures the report would have described, and its counterparties still need to know about them. The information demand reappears as investor questionnaires, customer requirements, lender covenants and insurance conditions, in a form that is less standardised and frequently more expensive than the obligation it replaced.
That this is a European story
The evidence here is European because the European Union documents its own process unusually well: dated proposals, published trilogue outcomes, recitals stating rationale, and an official journal recording exactly what changed. That legibility is a property of the archive rather than of the phenomenon. The same mechanism operating in a jurisdiction that publishes less would be harder to observe and no less present, and we would rather state that limitation than generalise past our evidence.
6. Base, Stress and Extreme
Four paths, with our probability assessment and the condition that would falsify each. Probabilities sum to one hundred.
Path | P | What it looks like | What would falsify it |
Continued subtraction | 50% | Simplification spreads across regulatory domains. Thresholds rise, categories replace assessment, voluntary standards displace mandatory ones | Two or more major regimes lowering thresholds or restoring individual assessment while inbound volume rises |
Two-tier settlement | 30% | Machine assessment for the standard case funds restored human judgement for exceptions. Variety returns, but only where data infrastructure allows | Automated assessment deployed without any corresponding increase in exception-handling capacity |
Private substitution | 15% | The subtracted requirement reappears through lenders, insurers, customers and investors, unstandardised and unevenly enforced | Contractual and financing requirements narrowing in step with the statutory ones |
Capacity restoration | 5% | Institutions acquire enough qualified capacity to restore fine-grained assessment at scale | This is the falsifier for the series as a whole rather than a scenario needing one |
The third path is the one most likely to be underestimated. A requirement removed from statute does not remove the underlying information need, and the private version arrives without the standardisation, the proportionality or the appeal route that the public version carried.
7. Forecast — One Year, to mid-2027
Transposition runs late, and lateness functions as further deferral
Probability 0.70 · Confidence: Medium-High
Member states have until 19 March 2027 for the reporting provisions. We expect a material number to miss it, for the same reason the original obligations were deferred: national administrations are the layer where the capacity shortage actually sits, and a directive that simplifies at European level still requires national implementation capacity to deliver it.
Late transposition is not a neutral delay. It extends the period in which firms cannot determine what applies to them, which is the condition that made the original regime expensive.
Second-order effect. Firms that invested in readiness before the scope narrowed hold stranded capability. Some will monetise it by selling assurance services to those still in scope; more will quietly stand it down, which removes private capacity from the system at the same time as public capacity is short.
What would weaken it. Transposition proceeding broadly on time, which would indicate that the constraint was European-level negotiation rather than national administrative capacity.
8. Forecast — Three Years, to 2029
The simplification method spreads to further regulatory domains
Probability 0.70 · Confidence: Medium-High
The omnibus is now an established instrument: a single act amending several regimes at once, justified by competitiveness, delivering higher thresholds and narrower scope. Sustainability and digital have been through it. We expect at least one further domain to follow by 2029.
The mechanism that spreads it is precedent rather than ideology. Once a method for reducing scope has survived trilogue and produced a usable instrument, it becomes the cheapest available response to any regime that is behind schedule.
Second-order effect. Compliance planning shifts from interpreting rules to forecasting scope. The operative question for a mid-size firm stops being what the rule requires and becomes whether the rule will still apply to it in three years, which is a question no compliance function is equipped to answer.
What would weaken it. New regimes entering application on their original terms and timetables, particularly where they require new assessment infrastructure to exist first.
9. Forecast — Five Years, to 2031
The subtracted requirement reappears privately
Probability 0.60 · Confidence: Medium
By the early 2030s we expect information requirements removed from statute to be substantially reconstituted through lending covenants, insurance conditions, customer qualification and investor screening.
The Omnibus anticipates this and tries to prevent it. Protected entities were given the right to refuse requests exceeding the voluntary standard, which is an explicit attempt to stop the obligation trickling down the value chain. That provision constrains reporting undertakings; it does not constrain a bank, an insurer or a customer, and those are the parties with the leverage to ask.
Second-order effect. The private version is worse on the dimensions that mattered. It is unstandardised, so a firm answers the same question in five incompatible formats. It carries no proportionality, so a small supplier faces a large customer's template. And it has no appeal route, because a commercial counterparty owes no reasons.
What would weaken it. Financing and commercial requirements narrowing in step with the statutory ones, which would indicate that the demand was regulatory in origin rather than genuine.
10. Forecast — Ten Years, to 2036
The ratchet holds
Probability 0.65 · Confidence: Medium-Low
Over a decade we expect the thresholds set in 2026 to persist or rise, and we do not expect a return to the pre-Omnibus position. Not because the policy argument was settled, but because reversal requires the assessment capacity whose absence produced the simplification, and nothing in the intervening decade is likely to generate it.
The forecast is falsifiable in a specific way: it fails if any EU act restores mandatory sustainability reporting to a threshold at or below 500 employees before the end of 2036.
We hold this at lower confidence than its probability suggests. The mechanism is strong and the horizon is long, and a decade is enough time for a visible failure with an identifiable victim to appear. That is the one thing that reliably restores variety, and it cannot be forecast.
What would weaken it. Material expansion of qualified assessment capacity, measured in trained staff rather than in budget, in a way that makes restoration administratively feasible rather than merely desirable.
11. Signals to Watch
— Threshold movements in any regime, and the direction. Upward is subtraction; downward would be the first genuine counter-evidence in this series
— Categorical rules replacing individual assessment: categorical exclusions in permitting, categorical eligibility in benefits, automated classification in enforcement
— Voluntary standards displacing mandatory ones, which is form three and the hardest to see in a statute book
— Transposition and implementation delay at national level, which is where capacity shortage becomes observable
— The reason given for each change. Competitiveness, burden reduction and modernisation are the available vocabulary; capacity failure is not, and its absence carries no information
— Private reconstitution: lender covenants, insurance conditions and customer questionnaires asking for what the statute stopped requiring
— Assessor, examiner and inspector headcount and training pipelines, which determine whether any of this can ever run backwards
12. Recommendations — Individuals
The practical consequence of variety subtraction is that your particulars matter less and your classification matters more. That is uncomfortable and it is actionable.
Immediate — 30 days
For each institutional relationship that matters to you, find out which category you are in. Not whether you qualify in principle, but which box the form puts you in, and where the boundaries of that box sit. Most people know their eligibility and not their classification, and under a categorical regime the classification is the whole of the outcome.
Build — 12 months
Where you sit close to a threshold, find out which side and by how much. Cliffs are invisible until you are on the wrong side of one, and the distance to the edge is usually knowable in advance and almost never checked.
Where your circumstances are genuinely unusual, stop relying on being assessed on your merits. The mechanism that recognised particulars is the one being removed. This is not a counsel of despair; it means documenting your position in whatever standard form exists, because a standard form that is examined beats an accurate account that is not.
Position — 3 years
When a life decision could be arranged either to fit a standard category or to require individual assessment, and the two are otherwise close in value, take the category. The assessed route is not being refused. It is being made slower and less certain, and slowness compounds in ways that are hard to see at the point of decision.
Avoid. Assuming that a removed requirement means the underlying question has stopped being asked. It usually reappears from a private party who owes you no reasons and offers no appeal, and the private version is frequently harder to answer than the public one it replaced.
Why this works. You cannot restore an institution's capacity to look closely. You can find out how you are classified, how close you are to an edge, and whether your situation can be expressed in the form that survives. All three are within reach and all three are usually unexamined.
13. Recommendations — Business
Most compliance functions track what the rules require. Very few track whether the rules will still apply to them, and under an active simplification agenda the second question drives more value than the first.
Immediate — 60 days
Map your position against every threshold that determines a regulatory obligation: headcount, turnover, balance sheet, emissions, volume, jurisdiction. Record the distance to each boundary and the direction you are moving. A firm growing through a threshold and a firm that has just been exempted by one face opposite problems and neither usually knows which it is.
Then check the second-order exposure. If your customers are in scope and you are not, you may receive their obligations by contract. If you are in scope and your suppliers are not, you may be unable to obtain what you need to comply.
Build — 12 months
Treat scope as a forecastable variable rather than a fixed input. Build the compliance plan against the obligation as it stands and record separately which parts are threshold-dependent, so that a scope change can be absorbed as a decision rather than as a shock.
Do not stand down capability the moment scope narrows. Firms that dismantled readiness after the Stop the Clock directive and again after the Omnibus have now paid twice for nothing. Capability that is retained can be sold to those still in scope, offered to customers as assurance, or redeployed when the private version of the requirement arrives.
Position — 3 years
Assume the requirement returns through commercial channels. Build the ability to answer the substantive question once, in a form that can be reshaped for a lender, an insurer, a customer or a regulator, rather than building to whichever template is currently mandatory. The underlying question is stable; the party asking it is not.
Where you are a supplier, expect to be classified rather than assessed. Fitting a recognised standard is worth more than being genuinely good in a way no template captures, and that is a statement about the assessment mechanism rather than about merit.
Avoid. Building a competitive position that requires an institution to appreciate a distinction. Any advantage that depends on being assessed on particulars is exposed to the next categorical rule, and the categorical rule will not be aimed at you and will not consider your case.
Why this works. Institutional throughput has concentration of one and a substitution time of years. The rule from Article 1 applies unchanged: you cannot remove the dependency, and you can convert it into a form that preserves alternatives and buys time.
14. Recommendations — Capital
Scope risk is a distinct exposure from regulatory risk and is rarely modelled separately. Regulatory risk asks whether the rules change. Scope risk asks whether they still reach the asset, and it moves faster.
Immediate — this quarter
Identify holdings whose thesis depends on a regulatory requirement existing. Compliance service providers, assurance firms, reporting software, data vendors and certification bodies all have revenue that is a function of scope. When four fifths of the addressable population leaves scope in a single instrument, that is a demand shock arriving through a legal channel rather than an economic one.
Build — 12 months
Separate two exposures. The risk that an asset fails to meet a requirement is a compliance risk and is usually priced. The risk that the requirement stops applying, taking a revenue stream with it, is a scope risk and usually is not.
Then look at the other direction. Where a public requirement is withdrawn and the underlying information need persists, the demand migrates to private channels. The firms positioned to serve lenders, insurers and corporate buyers may capture what the compliance vendors lose, and that transfer happens on a different timetable from the legislative one.
Position — 3 years
Treat threshold structure as a jurisdictional characteristic. Two jurisdictions with identical stated policy and different thresholds produce different populations of regulated firms, and the threshold is a better predictor of what a business will actually experience than the policy is.
Watch for scope risk becoming a disclosed and named risk in offering documents. When it is priced explicitly it stops being an edge, and the interval before that happens is where the return sits.
Avoid. Reading a simplification package as a durable reduction in regulatory burden. The obligation was narrowed rather than resolved, the underlying demand persists, and it tends to return through channels that are less standardised and harder to hedge than the statute it replaced.
Why this works. Capital's advantage is the ability to reallocate before a repricing rather than after it. Scope changes are unusually forecastable, because they are proposed, negotiated and published months in advance in documents almost nobody in the market reads.
15. What Would Change Our Mind
Each forecast above carries its own weakening condition. Three developments would undermine this article's argument as a whole.
— Thresholds move downward while inbound volume rises. A regime that broadens its reach while under load is doing the opposite of what the mechanism predicts, and two such cases would be decisive rather than anecdotal.
— Individual assessment is restored at scale and survives appeal. Not pilots and not discretionary carve-outs, but fine-grained assessment reinstated in a high-volume regime and sustained for more than one cycle.
— Assessment capacity expands materially, measured in qualified staff. This is the precondition for everything above, and its absence is the load-bearing assumption of the whole article.
A limitation we record rather than manage. The evidence in this article is European and so are three of the four forecasts. The mechanism is claimed as general, and the case material is one jurisdiction chosen because it publishes its own reasoning in unusual detail. That is a real weakness in the argument and not merely a caveat: a mechanism demonstrated in one archive has been demonstrated in one archive. We would rather a reader hold this at the confidence the evidence supports than at the confidence the writing suggests.
Founder's Lens
[ EDITORIAL GATE — WRITTEN BY HAND BEFORE PUBLICATION. Never generated. Replace this marker with the founder's text, or record a suspension. ]
16. Bottom Line
An institution that cannot match the variety of what it faces has two options, and only one of them is available on the timescale it needs. It cannot acquire capacity in the time available. So it subtracts variety instead, and it does so in three recurring forms: the category replaces the case, the threshold rises above capability, and the obligation stops binding while remaining on the books.
The variety does not disappear. It stops being the institution's problem and becomes somebody else's.
The typical case is genuinely better off, which is why this happens and why resisting it is expensive. The atypical case pays, and the parties who pay are diffuse, often unaware, and absent from the process that decides. The change arrives under a positive name with a competitiveness rationale, because that is the vocabulary available, and it is usually accurate as far as it goes.
And it does not run backwards. Restoring variety requires exactly the capacity whose absence removed it, and the practice atrophies while the rule is dormant. The historical exception is a visible failure with an identifiable victim inside the political attention span — four conditions that rarely coincide, because a harm that concentrated would have prevented the simplification in the first place.
The practical consequence is narrow. Find out how you are classified rather than whether you qualify. Find out how close you are to a threshold and on which side. Where your situation is unusual, stop relying on being assessed on your merits, because the mechanism that assessed merits is the one being removed. And when a requirement is withdrawn, expect the underlying question to return from a private party who owes you no reasons.
The world is not becoming ungovernable. It is being governed at lower resolution, one reasonable simplification at a time.
Forecast record
Four forecasts, one per horizon, each with a threshold, a named verifier and a resolution date, recorded before the outcome is known.
Horizon | Forecast, resolving yes or no | P | Resolves |
1 year | The European Commission has opened infringement proceedings against at least five member states for failure to transpose Directive (EU) 2026/470 within the deadline of 19 March 2027 | 0.60 | 31 December 2027 · ec.europa.eu infringement decisions |
3 years | A further EU omnibus-style act narrowing scope or raising thresholds in a regulatory domain other than sustainability or digital enters into force | 0.70 | 31 December 2029 · eur-lex.europa.eu |
5 years | The share of large EU undertakings reporting sustainability information voluntarily, or under lender or customer requirement, exceeds the share reporting under statutory obligation, as recorded in a Commission or ESMA review | 0.55 | 31 December 2031 · Commission or ESMA published review |
10 years | No EU legislative act has restored mandatory sustainability reporting to a headcount threshold at or below 500 employees | 0.65 | 31 December 2036 · eur-lex.europa.eu |
Correlation and concentration, recorded rather than assumed away. All four resolve against European institutions, and the second, third and fourth share a parent cause in the simplification agenda. These are not four independent observations. The five-year forecast additionally depends on a review being published in a form that permits the comparison, which is a resolution risk rather than a forecasting one, and we mark it as the weakest of the four on that ground.
Directional statements elsewhere in this article carry no threshold and are deliberately excluded from the record.
Sources
Figure | Class | Source |
Directive (EU) 2026/470 published in the Official Journal 26 February 2026, in force 18 March 2026 | Measured | Official Journal of the European Union |
CSRD scope narrowed to undertakings above 1,000 employees and €450m net turnover, both conditions required | Measured | Directive (EU) 2026/470; Council press release, 24 February 2026 |
Previous test: large public-interest undertakings above 500 employees, large defined as exceeding two of three thresholds — €25m balance sheet, €50m turnover, 250 employees | Measured | Pre-Omnibus CSRD and NFRD transposition |
CS3D scope narrowed to above 5,000 employees and €1.5bn net turnover | Measured | Directive (EU) 2026/470 |
Listed SMEs exempted; sector-specific standards removed; protected entities below 1,000 employees granted the right to refuse requests beyond the voluntary standard | Measured | Directive (EU) 2026/470 |
Transposition deadlines: 19 March 2027 for reporting provisions, 26 July 2028 for due diligence | Measured | Directive (EU) 2026/470 |
Stop the Clock: Directive (EU) 2025/794 of 14 April 2025 postponing obligations by two years | Measured | Official Journal |
Approximately four fifths of previously covered undertakings fall outside the new scope | Reported estimate | Widely reported analyses of the Omnibus; not an official figure |
Origin of the simplification agenda: Draghi and Letta reports; Budapest declaration of 8 November 2024 calling for a simplification revolution | Measured | European Council record |
One figure in this table is marked differently from the others and should be read differently. The four-fifths scope reduction is a reported estimate rather than an official statistic. It is widely cited, it is directionally consistent with the threshold change, and we have not located a primary source that states it. We use it once, describe it as an estimate, and build nothing on it.
In this series
— Previous: Article 2, The Complexity Generation Gap — what happens when producing complexity becomes almost free.
— Next: Article 4, When Capacity Concentrates — what changes when the historical limit on central authority stops binding.
— The method behind the Chaos Index and this series: /methodology
THRIVE IN CHAOS
Decision Intelligence for an Uncertain World
Analysis → Forecast → Recommendations · Signal → Meaning → Action → Stability
Signal Over Noise · thriveinchaos.ai
AI intelligence system with human editorial oversight.
Forecasts are probability-based analytical assessments, not certainties. This material supports independent judgment and does not constitute financial, legal or investment advice.
Join the newsletter
Be the first to read our articles.


