Hysteresis Effects in Market Competition Restoration

1. Introduction

Hysteresis effects in market competition restoration describe a situation in which competitive conditions do not automatically return to their former state even after the conduct or event that weakened competition has ended.

In ordinary economic theory, one might expect that removing an anticompetitive practice—such as exclusionary conduct, discriminatory access, a merger, predatory pricing, or a restrictive agreement—would allow competitors to re-enter and restore competition. Hysteresis challenges that assumption.

The essential idea is:

The market's present competitive structure depends not only on current conditions, but also on the path by which the market reached those conditions.

Thus, a dominant undertaking may acquire advantages during a period of exclusion and retain them after the exclusionary conduct stops.

Examples include:

• customers becoming locked into an ecosystem;
• rivals exiting and losing scale;
• data advantages accumulating over time;
• network effects becoming stronger;
• distribution channels becoming unavailable;
• employees and technical expertise moving to the incumbent;
• interoperability standards becoming dependent on the incumbent;
• brand loyalty becoming entrenched;
• switching costs increasing; and
• suppliers becoming dependent upon the dominant firm.

Consequently, termination of the original anticompetitive conduct may be insufficient to restore effective competition.

2. Meaning of Hysteresis in Competition Law

Hysteresis is borrowed from economics and physics.

In a competitive-market context, it can be represented conceptually as:

Initial competition → exclusionary event → competitor exit/lock-in → structural change → conduct ends → competition remains weak

The important feature is that:

the market does not retrace its previous path when the original conduct is reversed.

For example:

1. Firm A and Firm B initially compete.
2. A adopts exclusionary contractual arrangements.
3. B loses customers and exits.
4. A gains scale and data.
5. A's ecosystem becomes the industry standard.
6. The contracts are subsequently terminated.
7. B does not automatically return.
8. Customers remain locked into A.
9. Suppliers continue depending on A.
10. Entry remains commercially unattractive.

Therefore, a remedy that merely says "stop the exclusionary conduct" may fail to recreate the competitive conditions that existed before the infringement.

3. Why Hysteresis Matters to Competition Law

Traditional competition analysis often focuses on:

• the conduct;
• market power;
• effects;
• foreclosure;
• consumer harm; and
• prospective remedies.

Hysteresis introduces another question:

"What irreversible or difficult-to-reverse structural effects did the conduct create?"

This is particularly important in digital markets because competitive advantages can accumulate rapidly.

A platform can simultaneously control:

• users;
• data;
• applications;
• advertising;
• payments;
• cloud infrastructure;
• APIs;
• identity systems;
• search/discovery;
• technical standards.

Once competitors disappear, restoring the prohibited conduct to zero does not necessarily restore those competitors.

4. Main Mechanisms Creating Hysteresis

A. Exit Hysteresis

A competitor excluded from a market may permanently leave.

Suppose:

A dominant platform forecloses rival B for five years.

B may have:

• dismissed employees;
• terminated infrastructure contracts;
• lost investors;
• abandoned R&D;
• lost customers; and
• discontinued the product.

Even if foreclosure ends, B may not possess the resources necessary to return.

Thus:

temporary exclusion → permanent competitive loss.

B. Network-Effect Hysteresis

Network effects create especially strong hysteresis.

The value of a platform may increase as its user base grows.

If Firm A reaches critical mass:

more users → more developers → more applications → more users.

A rival cannot necessarily reverse this process merely by receiving equal access.

The incumbent may have crossed a competitive tipping point.

C. Data Hysteresis

Historical accumulation of data can generate persistent advantages.

For example:

exclusive access → more data → better algorithm → better service → more users → still more data.

Removing exclusivity does not erase the accumulated dataset.

Therefore, a remedy may need to address the stock of accumulated competitive advantage, not merely future conduct.

D. Switching-Cost Hysteresis

Customers may become dependent on:

• proprietary formats;
• APIs;
• software;
• cloud architecture;
• subscriptions;
• stored data;
• trained employees;
• complementary products.

Even after an abusive restriction disappears, customers may rationally remain with the incumbent because switching has become expensive.

E. Reputation and Brand Hysteresis

Long-term exclusion can alter consumer perceptions.

A rival that disappeared for several years may be perceived as:

• unreliable;
• technologically obsolete;
• less secure; or
• less innovative.

The incumbent's historical market position can therefore continue influencing demand.

F. Supply-Side Hysteresis

A dominant undertaking can capture:

• skilled workers;
• suppliers;
• distributors;
• developers;
• manufacturers;
• infrastructure providers.

Even when exclusion ends, these parties may remain economically dependent on the incumbent.

5. Hysteresis and Barriers to Entry

Hysteresis is particularly important for understanding barriers to entry.

A temporary exclusionary strategy can transform the market itself.

Before exclusion

After prolonged exclusion

Thus, the original barrier may disappear while the consequences remain.

6. Hysteresis and Market Definition

Hysteresis may also affect market definition.

A market that appears contestable under a static analysis may actually be difficult to enter because historical events have altered competitive conditions.

For example:

A market may technically contain several firms, but only one possesses the accumulated data, infrastructure, users and interoperability relationships necessary for effective competition.

Competition authorities therefore need to examine dynamic competitive constraints, rather than simply counting current competitors.

7. Hysteresis and Article 102 TFEU

Under Article 102 TFEU, hysteresis can be relevant to abuse of dominance involving:

• exclusionary contracts;
• refusal to supply;
• tying;
• rebates;
• predatory conduct;
• discriminatory access;
• interoperability restrictions;
• self-preferencing; and
• leveraging into adjacent markets.

The central issue is whether conduct has produced durable foreclosure effects.

The authority may need to consider not simply:

"Is the conduct continuing?"

but:

"Has the conduct already altered market structure sufficiently that competition cannot recover without intervention?"

8. Hysteresis and UK Competition Law

Under the Competition Act 1998, particularly Chapter II prohibition analysis, hysteresis can inform assessment of:

• foreclosure;
• exclusionary abuse;
• barriers to entry;
• actual or potential competition;
• effects of contractual restrictions;
• refusal of access; and
• effectiveness of remedies.

The UK's modern digital-market framework makes this particularly significant because digital market power can be reinforced by:

• network effects;
• economies of scale;
• data advantages;
• interoperability;
• ecosystem lock-in.

9. Hysteresis and Merger Control

Hysteresis is also relevant to merger remedies.

Suppose a merger eliminates a significant competitor.

Even if the merger is later unwound, competition may not return to its pre-merger state.

The acquired business may have:

• integrated its technology;
• lost independent employees;
• migrated customers;
• abandoned competing products;
• transferred intellectual property; and
• integrated databases.

Therefore:

divestiture after integration ≠ automatic restoration of pre-merger competition.

This is one reason competition authorities may prefer structural remedies before irreversible integration occurs.

10. Hysteresis and Remedies

This is perhaps the most important practical implication.

A conventional behavioural remedy might order:

"The undertaking must cease discriminatory conduct."

But if hysteresis has already occurred, this may be inadequate.

Effective restoration may require:

Structural remedies

• divestiture;
• separation of business units;
• removal of exclusive contracts;
• restoration of independent distribution;
• transfer of assets.

Access remedies

• interoperability;
• API access;
• data portability;
• non-discriminatory access;
• technical standards.

Data remedies

• data portability;
• data-sharing arrangements;
• restrictions on data combination;
• data silos.

Ecosystem remedies

• multi-homing;
• default-choice mechanisms;
• switching tools;
• interoperability;
• identity portability.

The objective is not merely to stop the violation, but to restore contestability.

11. Important Case Laws

The following cases do not necessarily use the word "hysteresis" as a formal legal doctrine. Rather, they establish principles concerning persistent foreclosure, market structure, entry barriers, network effects, switching costs, accumulated advantages and effective restoration of competition.

1. United States v. Microsoft Corp. (2001)

The Microsoft case is one of the strongest illustrations of hysteresis-type reasoning.

Microsoft used its operating-system position to restrict competing browsers, particularly Netscape.

The court recognised that exclusionary conduct could reinforce Microsoft's position by:

• limiting rival distribution;
• strengthening application barriers to entry;
• protecting the operating-system monopoly.

Hysteresis significance

Once Netscape's distribution opportunities were substantially reduced, merely stopping particular exclusionary practices would not necessarily recreate Netscape's former competitive position.

The case therefore illustrates:

exclusion → loss of distribution → weakened rival → reinforced incumbent → persistent market power.

2. United States v. AT&T (1982)

The AT&T litigation concerned the structure of telecommunications markets and the relationship between monopoly control and competitive entry.

The eventual structural remedy separated major telecommunications activities.

Hysteresis significance

The case demonstrates why structural intervention may be necessary where monopoly control has become embedded in:

• infrastructure;
• distribution;
• network access; and
• related markets.

A purely behavioural prohibition may be insufficient when the market architecture itself perpetuates dominance.

3. United States v. Terminal Railroad Association of St. Louis (1912)

The Supreme Court dealt with control over critical railroad terminal facilities.

The terminal arrangement created a substantial obstacle for competitors requiring access to essential infrastructure.

Hysteresis significance

Control over a bottleneck facility can change the competitive structure of an industry.

If rivals are denied access for a sufficiently long period, their competitive capabilities may deteriorate.

The case therefore provides an early illustration of:

infrastructure control → foreclosure → weakened competitors → durable market advantage.

4. United States v. Griffith (1948)

Griffith involved the use of power in one market to obtain or reinforce advantages in other markets.

The Supreme Court's reasoning demonstrates the importance of examining how an existing market position can be used to restrict competition elsewhere.

Hysteresis significance

Once rivals have been excluded from an adjacent market, restoring formal access does not necessarily restore their lost customer relationships and competitive capabilities.

This is particularly relevant to modern digital ecosystems.

5. Lorain Journal Co. v. United States (1951)

A dominant newspaper attempted to prevent advertisers from dealing with a competing radio station.

The Supreme Court treated the conduct as an unlawful attempt to preserve monopoly power.

Hysteresis significance

The case demonstrates how exclusionary conduct can prevent an emerging competitor from obtaining the scale necessary to establish itself.

The important lesson is:

Preventing competitive development can have consequences extending beyond the duration of the exclusionary conduct itself.

That is essentially a hysteresis problem.

6. Aspen Skiing Co. v. Aspen Highlands Skiing Corp. (1985)

Aspen Skiing is particularly important for persistent competitive effects.

A dominant ski operator terminated cooperation with a smaller rival even though the joint ticket arrangement had previously benefited consumers.

The Supreme Court found the conduct relevant to monopolisation.

Hysteresis significance

A dominant firm can use control over a complementary relationship to deprive a rival of access to customers and distribution.

Once the rival loses:

• customers;
• reputation;
• commercial relationships; and
• scale,

restoring the original arrangement may not automatically recreate the previous competitive conditions.

12. European Competition Law Cases

7. Hoffmann-La Roche v Commission (1979)

The Court of Justice recognised the importance of loyalty-inducing exclusivity arrangements employed by a dominant undertaking.

The case established foundational principles concerning abuse of dominance and exclusionary effects.

Hysteresis significance

Long-term exclusivity can alter purchasing behaviour and make customers dependent on the dominant supplier.

Once that dependency develops, simply terminating exclusivity may not immediately restore competitive purchasing patterns.

8. Michelin I v Commission (1983)

Michelin concerned loyalty rebates and the use of economic incentives by a dominant firm.

The Court was concerned with mechanisms capable of restricting customers' ability or incentive to switch.

Hysteresis significance

Rebate systems can produce cumulative effects.

A customer repeatedly incentivised to concentrate purchases with the dominant undertaking may become commercially dependent upon it.

Thus:

incremental loyalty → cumulative dependence → reduced rival opportunities.

9. British Airways v Commission (2007)

British Airways involved loyalty-inducing rebate arrangements.

The Court upheld the concern that such arrangements could foreclose competing airlines.

Hysteresis significance

The importance of the case extends beyond the rebate itself.

Repeated incentive structures can alter:

• travel-agent behaviour;
• customer allocation;
• rival sales opportunities.

Once those relationships are redirected, a rival may not immediately recover merely because the rebate programme ends.

10. Intel v Commission (2017)

The Intel litigation is particularly important for modern analysis of exclusionary rebates.

The Court of Justice required consideration of the circumstances relevant to whether rebates are capable of producing foreclosure.

Hysteresis significance

The case reinforces the importance of assessing actual competitive effects and the ability of rivals to compete, rather than treating formal termination of a practice as sufficient evidence of restored competition.

This is particularly relevant where exclusion has produced long-lasting scale and customer effects.

11. Google Shopping (Google and Alphabet v Commission, 2024)

The Google Shopping litigation concerns preferential treatment of Google's comparison-shopping service within general search results.

The case is highly relevant to digital-market hysteresis.

Search visibility can influence:

• traffic;
• consumer awareness;
• scale;
• merchant relationships;
• data accumulation.

Hysteresis significance

If a rival comparison-shopping service loses traffic over an extended period, ending preferential treatment does not automatically restore:

• its user base;
• merchant relationships;
• accumulated data;
• brand recognition; or
• algorithmic relevance.

The case therefore illustrates how digital foreclosure can generate path-dependent competitive effects.

13. Competition-Restoration Problem

A useful analytical model is:

Stage 1 — Competitive equilibrium

Several firms compete.

↓

Stage 2 — Exclusionary conduct

Dominant firm restricts rivals.

↓

Stage 3 — Competitor deterioration

Rivals lose:

• customers;
• data;
• scale;
• investment;
• distribution.

↓

Stage 4 — Market tipping

Network effects and switching costs strengthen the incumbent.

↓

Stage 5 — Conduct ends

The authority prohibits the original conduct.

↓

Stage 6 — Hysteresis

Competitive conditions remain degraded.

↓

Stage 7 — Remedy must address accumulated effects

Intervention focuses on restoration of contestability, not merely cessation.

14. Hysteresis and Digital Markets

Digital markets are especially vulnerable because competitive advantages accumulate rapidly.

Consider a hypothetical AI platform:

Exclusive data access

↓

more training data

↓

better model

↓

more users

↓

more developer integrations

↓

more API usage

↓

more revenue

↓

more computing capacity

↓

better model

The initial exclusion may subsequently become unnecessary because the accumulated advantage has become self-reinforcing.

This is a classic positive-feedback hysteresis mechanism.

15. Hysteresis and Data Portability

Data portability can be understood partly as an anti-hysteresis remedy.

Suppose users remain with a platform because their historical data is trapped there.

If the platform is prohibited from future exclusion but users still cannot transfer their data, the historical lock-in survives.

Data portability can therefore reduce the persistence of historical market power.

However, portability alone may not be enough where:

• network effects remain strong;
• data cannot easily be transferred;
• interoperability is absent;
• users face switching costs;
• the incumbent controls essential complements.

16. Hysteresis and Interoperability

Interoperability can be even more important where network effects exist.

If users can communicate across platforms, the incumbent cannot rely as heavily on its installed user base.

Thus:

interoperability → lower switching costs → reduced network-effect advantage → greater contestability.

This is why interoperability remedies can function as competition-restoration mechanisms, rather than merely behavioural restrictions.

17. Hysteresis and Algorithmic Competition

Algorithmic systems can intensify hysteresis.

An incumbent's algorithm may learn from:

• historical transactions;
• user behaviour;
• search queries;
• pricing;
• interactions;
• supplier data.

A rival entering later lacks equivalent historical data.

Therefore:

historical market success becomes future algorithmic superiority.

The competitive advantage is no longer simply the result of present conduct; it becomes embedded in the technological system.

18. Hysteresis and Killer Acquisitions

Hysteresis also helps explain concerns over acquisitions of nascent competitors.

A dominant platform may acquire a potential rival before it becomes a major competitor.

Once integrated:

• the rival's independent trajectory disappears;
• its technology may be absorbed;
• its employees may be reassigned;
• its customers may migrate;
• its independent innovation path may terminate.

Even if authorities later require divestiture, restoring the original entrepreneurial and technological trajectory may be difficult.

19. Hysteresis and Competition Authorities

Competition authorities should therefore distinguish between:

Conduct cessation

"Stop doing X."

and

Competition restoration

"Make it possible for effective competition to emerge again."

The second is considerably more demanding.

A regulator should examine:

1. How long did exclusion continue?
2. Which competitors exited?
3. What assets did they lose?
4. Did network effects strengthen?
5. Did customer switching costs increase?
6. Did data accumulate?
7. Did distribution relationships change?
8. Did suppliers become dependent?
9. Has the market tipped?
10. What remedy can reverse these effects?

20. Measuring Hysteresis

Potential indicators include:

Structural indicators

• competitor exits;
• market-share persistence;
• concentration;
• entry rates;
• exit rates.

Consumer indicators

• switching rates;
• multi-homing;
• churn;
• contract duration;
• customer acquisition costs.

Digital indicators

• API dependence;
• interoperability;
• data portability;
• developer concentration;
• default settings.

Innovation indicators

• R&D entry;
• patenting;
• venture investment;
• product launches;
• technological diversity.

A competition authority could compare:

pre-conduct competitive conditions → post-conduct conditions → post-remedy conditions.

If the final market remains materially different from the initial competitive state, hysteresis may be present.

21. Hysteresis Versus Ordinary Persistence

The concepts should not be confused.

Ordinary persistence

A firm remains dominant because it continues to offer a superior product.

Hysteresis

A firm remains dominant partly because past competitive events have altered the market in ways that continue affecting present competition.

The distinction is important because competition law should not punish legitimate success.

The relevant question is whether persistent market power results from:

innovation and efficiency

or from

historically induced foreclosure and structural lock-in.

22. Legal Test for Hysteresis Effects

A useful analytical framework is:

Step 1 — Identify the original competitive structure

Who competed before the relevant conduct?

Step 2 — Identify the intervention or exclusion

What conduct altered competitive conditions?

Step 3 — Identify accumulated effects

Did the conduct cause:

• exit;
• lock-in;
• data accumulation;
• network effects;
• loss of scale?

Step 4 — Test reversibility

If the conduct stops today, can competitors realistically return?

Step 5 — Identify path dependence

Does the current market depend materially upon historical exclusion?

Step 6 — Select a restorative remedy

What intervention can recreate contestability?

Step 7 — Monitor recovery

The authority should measure whether:

• rivals enter;
• switching increases;
• market shares become contestable;
• innovation improves;
• dependency decreases.

23. Key Legal Principle

The central principle can be expressed as:

Competition law remedies should sometimes be concerned not merely with terminating anticompetitive conduct, but with reversing or neutralising the durable competitive distortions that the conduct has already created.

This is particularly important in:

• digital platforms;
• cloud computing;
• AI;
• telecommunications;
• payment systems;
• operating systems;
• online marketplaces;
• data-driven services;
• infrastructure markets.

24. Conclusion

Hysteresis effects in market competition restoration describe the persistence of competitive distortions after the original anticompetitive conduct has ended.

The central problem is path dependence.

A period of exclusion can produce:

• competitor exit;
• customer lock-in;
• network effects;
• data accumulation;
• distribution foreclosure;
• supplier dependence;
• loss of innovation;
• increased switching costs.

Once these effects become embedded, simply prohibiting the original conduct may not restore competition.

The lessons from cases such as Microsoft, AT&T, Terminal Railroad, Lorain Journal, Aspen Skiing, Hoffmann-La Roche, Michelin, British Airways, Intel and Google Shopping demonstrate the broader legal importance of foreclosure, durable market power, access, network effects and effective competitive constraints.

Accordingly, modern competition enforcement should distinguish between:

ending an infringement and restoring competitive conditions.

Where hysteresis is substantial, effective enforcement may require interoperability, data portability, access obligations, removal of accumulated exclusionary advantages, structural separation, divestiture, or other remedies specifically designed to restore contestability.