Algorithmic Zoning Systems and Real Estate Market Control

1. Introduction

Algorithmic zoning systems refer to the use of automated or AI-assisted systems to determine, recommend, rank, or enforce land-use outcomes. These systems may process zoning maps, parcel characteristics, ownership data, development proposals, environmental constraints, infrastructure capacity, property values, traffic patterns, demographic information, and historical permitting decisions.

In real-estate markets, algorithms can therefore influence:

• which parcels may be developed;
• permissible building density and floor area;
• residential, commercial, industrial or mixed-use classification;
• development approvals and permit prioritisation;
• affordable-housing requirements;
• redevelopment and urban-renewal areas;
• allocation of scarce development rights;
• property valuations and investment decisions;
• location recommendations to developers;
• access to municipal or privately controlled property data; and
• rental or sales pricing after development.

The competition-law problem arises when an algorithm does more than neutrally administer zoning rules and instead controls access to land, development opportunities, information, or market participants in a way that restricts competition.

There is currently no well-established body of reported decisions specifically titled "algorithmic zoning" under competition law. Consequently, the legal framework has to be constructed from zoning/urban-development antitrust cases, real-estate competition cases, government-regulation cases, and the emerging algorithmic-pricing litigation.

2. Meaning of Algorithmic Zoning

A simplified algorithmic zoning model can be represented as:

Parcel data → Algorithmic classification → Development score → Zoning/permit recommendation → Market access

For example, an algorithm could calculate:

Parcel A → high development suitability → mixed-use → FAR 5.0 → expedited approval

while:

Parcel B → low suitability → residential-only → FAR 1.5 → ordinary approval

If the criteria are objective, transparent and legally authorised, the system may improve administrative efficiency.

The competition concerns become more serious where:

1. the algorithm is controlled by a dominant private platform;
2. competing developers must use the same proprietary system;
3. competitors supply commercially sensitive data to the algorithm;
4. the algorithm systematically favours affiliated developers;
5. zoning recommendations are manipulated to exclude rivals;
6. access to zoning or property data is restricted;
7. algorithmic decisions create artificial scarcity of developable land; or
8. competing landlords/developers use a common algorithm to coordinate market conduct.

3. Relationship Between Zoning and Competition Law

Traditional zoning is normally a public regulatory function, whereas competition law is principally concerned with restrictions on competitive conduct.

The distinction becomes important.

Traditional zoning

A municipality decides:

"This land may be used only for residential purposes."

That is normally governmental regulation.

Algorithmic private coordination

Suppose competing developers jointly provide confidential land-development information to a private algorithm and agree to follow its recommendations:

Developer A + Developer B + Developer C → common algorithm → development restrictions/pricing recommendations.

This can create a very different competition-law question.

The Supreme Court's decision in Fisher v. City of Berkeley illustrates the importance of distinguishing unilateral governmental regulation from concerted conduct by private competitors. The Court held that a municipal rent-control scheme did not itself constitute concerted action under Sherman Act §1 merely because competing landlords were required to comply with it.

4. Major Competition Concerns

A. Artificial restriction of developable land

An algorithm could designate large quantities of land as unsuitable for development.

If a private actor controls that classification, it may artificially reduce effective supply.

Potential consequences include:

• increased land prices;
• reduced developer entry;
• reduced housing construction;
• concentration of development opportunities;
• increased bargaining power of incumbent developers.

The relevant competition question is not simply whether zoning restricts construction, but who controls the restriction and whether it unlawfully excludes competitors.

5. Algorithmic Allocation of Development Rights

Development rights are often scarce.

Examples include:

• floor-area ratios;
• building heights;
• density bonuses;
• transferable development rights;
• redevelopment rights;
• commercial-use permissions;
• hotel or retail quotas.

An algorithm could allocate these rights according to a scoring system.

Competition concerns arise if the system:

• systematically allocates rights to incumbent firms;
• gives preferential treatment to affiliated developers;
• penalises new entrants;
• uses proprietary data unavailable to rivals;
• makes decisions that cannot effectively be challenged.

This may transform regulatory scarcity into competitive advantage.

6. Information Advantage and Data Concentration

Algorithmic zoning can create a valuable information infrastructure.

A platform controlling:

• parcel ownership data;
• zoning histories;
• permit applications;
• development proposals;
• land valuations;
• infrastructure information;
• environmental assessments; and
• planning decisions

may obtain a significant informational advantage.

If competing developers cannot obtain equivalent data, competition may be affected.

This resembles the broader antitrust problem of information asymmetry combined with control over an important digital infrastructure.

7. Algorithmic Exclusion of Competitors

Suppose an algorithm ranks development applications.

If the algorithm consistently gives higher scores to companies that:

• already possess large land portfolios;
• purchase the platform's other services;
• share more proprietary information;
• have existing relationships with the platform; or
• are affiliated with the algorithm operator,

smaller developers could be disadvantaged.

The legal analysis would examine:

1. relevant market;
2. market power;
3. exclusionary conduct;
4. foreclosure;
5. competitive effects;
6. legitimate regulatory justification; and
7. causal connection between algorithmic conduct and competitive harm.

8. Algorithmic Zoning and Collusion

This is particularly important.

Imagine five major developers independently entering data into the same proprietary zoning/development platform.

The platform receives:

• future construction plans;
• target rents;
• vacancy projections;
• land-acquisition strategies;
• planned developments;
• development timing.

If the system subsequently recommends that competitors reduce construction or maintain similar development strategies, the algorithm could facilitate coordination.

The underlying principle is increasingly relevant in algorithmic-pricing cases: technology does not automatically immunise otherwise unlawful coordination.

The DOJ's RealPage litigation alleges that competing landlords supplied competitively sensitive information to a common pricing system and received algorithmic recommendations affecting rents.

Although RealPage concerns rental pricing rather than zoning, its analytical significance for algorithmic zoning is considerable.

9. RealPage Litigation as an Emerging Algorithmic Competition Model

United States and Plaintiff States v. RealPage, Inc.

The DOJ sued RealPage in 2024 under Sherman Act §§1 and 2, alleging that its software facilitated coordination among competing landlords and that RealPage maintained monopoly power in commercial revenue-management software.

The allegations concerned the use of competitors' non-public information, algorithmic recommendations and mechanisms encouraging landlords to follow those recommendations.

The DOJ subsequently amended the case to include major landlords.

Relevance to algorithmic zoning

The analogy is:

RealPage

Competitor data → common algorithm → pricing recommendations → reduced independent decision-making

Algorithmic zoning

Developer/land data → common algorithm → development recommendations → potentially reduced independent development decisions

The two situations are not legally identical. But RealPage demonstrates how conventional antitrust concepts can be applied to algorithm-mediated coordination.

As of November 2025, the DOJ announced a proposed settlement requiring RealPage to end certain practices involving competitively sensitive information and pricing alignment.

10. Case Law

Case 1: Community Communications Co. v. City of Boulder, 455 U.S. 40 (1982)

This is one of the most important cases for algorithmic zoning.

Boulder adopted a municipal ordinance restricting the expansion of a cable television company. The Supreme Court rejected the argument that municipal regulatory authority automatically provided Parker state-action immunity.

The Court emphasised that municipal action requires a clearly articulated and affirmatively expressed state policy to displace competition before state-action immunity can apply.

Principle

A municipality cannot necessarily avoid antitrust scrutiny merely by describing an anticompetitive measure as regulation.

Algorithmic-zoning relevance

If a municipality delegates substantial zoning functions to an algorithmic system, questions may arise regarding:

• statutory authority;
• scope of delegation;
• state policy;
• transparency;
• whether the algorithm merely implements regulation or creates independent restraints.

The case is especially relevant where algorithmic zoning restricts market entry.

11. Case 2: City of Columbia v. Omni Outdoor Advertising, 499 U.S. 365 (1991)

In Omni Outdoor Advertising, a billboard company alleged that a competing billboard operator had influenced municipal officials to adopt zoning restrictions that disadvantaged competitors.

The Supreme Court addressed the interaction between municipal zoning decisions, private lobbying and antitrust immunity. The case involved a market in which the incumbent had substantial market power and allegedly worked with municipal authorities to restrict competing billboard construction.

Principle

The case demonstrates the difficulty of challenging regulatory restrictions where private actors participate in government decision-making.

Algorithmic-zoning relevance

Imagine:

dominant developer → supplies algorithmic planning data → algorithm recommends restrictive zoning → municipality adopts recommendation.

The critical questions would include:

• Who designed the algorithm?
• Who supplied its data?
• Who controlled its parameters?
• Did the incumbent influence the criteria?
• Were competing developers given equal access?
• Was the algorithm merely advisory or effectively determinative?

Thus, algorithmic lobbying could create a technologically sophisticated version of the concerns examined in Omni.

12. Case 3: Racetrac Petroleum, Inc. v. Prince George's County, 786 F.2d 202 (4th Cir. 1986)

Racetrac challenged the denial of a zoning exception necessary to operate a gasoline station and alleged antitrust violations.

The Fourth Circuit upheld the judgment against Racetrac, applying principles concerning state action and zoning regulation.

Principle

The mere fact that zoning decisions affect competitive opportunities does not automatically convert them into antitrust violations.

Algorithmic-zoning relevance

An algorithm that denies a developer's application cannot automatically be characterised as anticompetitive merely because the decision harms that developer.

There must be a legally relevant theory of competitive harm.

This requires separating:

legitimate land-use regulation

from

exclusionary manipulation of land-use regulation.

13. Case 4: Oberndorf v. City and County of Denver, 963 F.2d 174 (10th Cir. 1992)

The Oberndorf litigation arose from an urban-renewal project in Denver.

Landowners alleged that the urban-renewal plan eliminated competition among buyers and developers in a particular real-estate market and violated Sherman Act §§1 and 2.

The Tenth Circuit upheld summary judgment for the defendants. The record included claims that the urban-renewal plan eliminated competition among buyers and developers, while the court also considered municipal state-action immunity.

Principle

Urban redevelopment and zoning measures can have substantial effects on real-estate competition without necessarily constituting an antitrust violation.

Algorithmic-zoning relevance

This case is highly useful where an AI system is used for:

• urban renewal;
• redevelopment-zone designation;
• selection of redevelopment developers;
• identification of "blighted" areas;
• land assembly.

An algorithm classifying a neighbourhood as "high redevelopment priority" could substantially affect property values and competitive opportunities.

The legal issue would remain whether the governmental decision falls within valid regulatory authority or involves actionable exclusionary conduct.

14. Case 5: McLain v. Real Estate Board of New Orleans, Inc., 444 U.S. 232 (1980)

McLain involved an alleged agreement among real-estate brokers to maintain fixed brokerage commissions.

The Supreme Court held that the alleged real-estate brokerage activity could satisfy the interstate-commerce requirement for Sherman Act jurisdiction.

Principle

Real-estate transactions can fall within federal antitrust jurisdiction where the requisite interstate-commerce connection exists.

Algorithmic-zoning relevance

This is important because algorithmic land-use systems may initially appear "local."

But real-estate markets are economically interconnected with:

• interstate investment;
• financing;
• construction materials;
• mortgage markets;
• institutional investors;
• property-management services;
• technology providers.

Consequently, a digitally coordinated real-estate market cannot necessarily escape antitrust scrutiny merely because individual properties are locally situated.

15. Case 6: Fisher v. City of Berkeley, 475 U.S. 260 (1986)

Berkeley adopted rent controls for residential property.

The Supreme Court held that the municipal rent-control system did not constitute concerted action under Sherman Act §1 because the restrictions were imposed unilaterally by the government rather than through an agreement among landlords.

Principle

Government-imposed regulation and private concerted action must be distinguished.

Algorithmic-zoning relevance

This distinction is essential.

If:

Government → algorithm → mandatory zoning decision

the conduct may be analysed primarily as governmental regulation.

But if:

Private developers → common algorithm → agreed development restrictions

the conduct can raise conventional antitrust concerns.

Therefore, the identity of the algorithm operator and the legal source of the decision are critical.

16. Case 7: Supermarket of Homes, Inc. v. San Fernando Valley Board of Realtors, 786 F.2d 1400 (9th Cir. 1986)

The case concerned allegations of monopolisation in the San Fernando Valley real-estate market.

The Ninth Circuit discussed the requirements for monopolisation under Sherman Act §2, including:

1. monopoly power;
2. willful acquisition or maintenance of that power; and
3. antitrust injury.

The court also considered attempted monopolisation principles.

Algorithmic-zoning relevance

A private algorithmic zoning platform could potentially create §2 concerns if it becomes sufficiently powerful in a relevant market and uses that position to exclude competing platforms or developers.

Possible exclusionary mechanisms include:

• refusing access to critical zoning datasets;
• discriminatory API access;
• preferential ranking;
• interoperability restrictions;
• tying zoning analytics to unrelated services;
• discriminatory licensing;
• algorithmic downgrading of competitors.

17. Case 8: Eastman Kodak Co. v. Image Technical Services, Inc., 504 U.S. 451 (1992)

Although not a zoning case, Kodak is highly relevant to algorithmic real-estate ecosystems.

The Supreme Court recognised that substantial market power can arise in an aftermarket even where the primary equipment market appears competitive.

Algorithmic-zoning application

Suppose developers can theoretically choose among many property-management systems, but after adopting a dominant zoning platform they become dependent upon:

• proprietary zoning data;
• API access;
• historical permit data;
• compliance software;
• development scores;
• proprietary analytical models.

Switching may become difficult.

This creates potential algorithmic lock-in and raises questions about:

• switching costs;
• aftermarket power;
• interoperability;
• data portability;
• exclusionary conduct.

18. Essential-Facility-Type Concerns

An algorithmic zoning system may become economically indispensable where it controls access to:

• municipal zoning databases;
• parcel-level planning data;
• development permits;
• development-right inventories;
• infrastructure-capacity information;
• building compliance systems.

The important competition-law question becomes:

Can the operator lawfully refuse competitors access to an indispensable digital infrastructure?

The answer depends heavily on market power, substitutability, regulatory authority and the particular jurisdiction.

A simple claim of "essential facility" is not automatically sufficient. Modern antitrust law generally requires considerably more than mere usefulness.

19. Discriminatory Algorithmic Zoning

Algorithmic zoning may produce discriminatory outcomes even without explicit discriminatory instructions.

For example, a system might use historical variables such as:

• property values;
• historical development rates;
• prior permit approvals;
• neighbourhood investment;
• infrastructure quality.

If historical inequality is embedded in those variables, the algorithm may reproduce the same distribution of development opportunities.

Competition-law implications can include:

• discriminatory access to development opportunities;
• exclusion of new entrants;
• geographic foreclosure;
• differential treatment of competing developers.

Other bodies of law, including constitutional, administrative, housing and civil-rights law, may be more directly applicable depending on the facts.

20. Algorithmic Zoning and Market Foreclosure

Consider a city where a private company operates the dominant zoning-analysis platform.

The platform controls:

Data → Analysis → Development ranking → Permitting recommendation

If competing developers cannot obtain equivalent information, the system may create a foreclosure mechanism.

Possible foreclosure chain

Control of data

↓

Superior algorithm

↓

Better development predictions

↓

More successful land acquisition

↓

Larger development portfolio

↓

More data

↓

Improved algorithm

↓

Greater market power

This creates a data-development feedback loop.

21. Network Effects

Algorithmic zoning systems can display network effects.

More users generate:

• more parcel data;
• more development outcomes;
• more permit information;
• more market information;
• more predictive accuracy.

That can lead to:

more users → more data → better algorithm → more users.

A dominant system could therefore become increasingly difficult for rival systems to challenge.

Competition authorities may examine whether this represents legitimate innovation or exclusionary conduct.

22. Algorithmic Zoning and Vertical Integration

Suppose a company operates:

1. zoning software;
2. property-data services;
3. brokerage;
4. construction financing;
5. property management; and
6. a development company.

The algorithm could theoretically favour land that benefits its affiliated businesses.

This creates a vertical foreclosure problem.

For example:

zoning platform → favours affiliated developer → rival developer receives lower development score → rival loses access to attractive projects.

Relevant competition questions include:

• discriminatory treatment;
• self-preferencing;
• tying;
• foreclosure;
• refusal to deal;
• data advantages.

23. Algorithmic Zoning and Merger Control

Algorithms can also affect merger analysis.

Suppose two major developers merge and the combined company controls:

• large land banks;
• proprietary zoning datasets;
• development algorithms;
• permitting analytics.

The merger could increase control over both:

physical assets

and

digital infrastructure.

Competition authorities could examine whether the transaction creates:

• increased barriers to entry;
• data concentration;
• reduced development competition;
• foreclosure of rival developers;
• control over critical planning information.

24. Algorithmic Land Banking

A sophisticated algorithm can identify parcels likely to receive:

• rezoning;
• density increases;
• infrastructure investment;
• transit access;
• redevelopment designation.

A company with superior predictive technology could acquire those parcels before competitors.

This can produce algorithmic land banking.

The conduct is not automatically unlawful.

The competition concern becomes stronger if the firm uses market power or exclusionary agreements to prevent competitors from accessing equivalent information or opportunities.

25. Algorithmic Zoning and Real-Estate Pricing

Zoning algorithms can indirectly affect prices.

For example:

Reduced permitted density

→ fewer housing units

→ lower effective supply

→ increased scarcity

→ higher land values/rents.

Thus, even where the algorithm does not directly set prices, it can influence competitive conditions through supply-side control.

This differs from algorithmic rent-setting, but the RealPage litigation demonstrates why regulators are examining algorithms that influence independent market decisions. The DOJ alleged that RealPage's system used competitors' sensitive data and algorithmic recommendations in ways that reduced independent pricing decisions.

26. Algorithmic Coordination Among Developers

A particularly serious scenario would involve:

• Developer A;
• Developer B;
• Developer C;

all supplying confidential information to one algorithm.

The algorithm knows:

• where each developer intends to build;
• when construction will begin;
• expected capacity;
• expected sales;
• anticipated prices;
• land acquisition plans.

If recommendations systematically discourage independent expansion, the system may function as a coordination mechanism.

This resembles the legal concern identified in algorithmic rental-pricing litigation, although the legal analysis must be conducted separately for zoning and development decisions.

27. State-Action Problem

Government involvement is particularly important.

Under the state-action doctrine, conduct attributable to the state can receive protection from federal antitrust liability under certain circumstances.

Community Communications v. City of Boulder establishes that municipal action does not automatically receive state-action immunity merely because it is governmental in character.

Oberndorf similarly demonstrates the importance of examining whether municipal conduct was undertaken pursuant to clearly articulated state policy.

Thus:

"The city uses an algorithm"

does not itself resolve the antitrust question.

The legal inquiry must examine the statutory framework and nature of the delegation.

28. Private Algorithm Provider vs Government Algorithm

A useful distinction is:

29. Evidence Required in an Algorithmic-Zoning Competition Case

A competition authority or court would likely need to examine:

A. Algorithmic evidence

• source code;
• model architecture;
• training datasets;
• input variables;
• weighting systems;
• ranking criteria;
• automated decision rules;
• human overrides.

B. Commercial evidence

• market shares;
• developer participation;
• land ownership;
• entry barriers;
• switching costs;
• pricing;
• development volumes.

C. Communications

• emails;
• contracts;
• developer communications;
• software-provider communications;
• internal strategy documents.

D. Data flows

The central question may be:

Whose data entered the system, who could access it, and how did it influence the resulting decision?

30. Possible Defences

An operator or municipality may argue that the system:

• improves planning efficiency;
• reduces corruption;
• standardises permitting;
• improves infrastructure planning;
• promotes environmental sustainability;
• reduces administrative costs;
• prevents incompatible land uses;
• implements democratically enacted zoning policy.

These can be legitimate objectives.

However, their existence does not necessarily resolve whether a particular exclusionary mechanism violates competition law.

31. Compliance Framework for Algorithmic Zoning

A competition-compliant system should ideally incorporate:

1. Transparency

Publish the principal decision criteria.

2. Auditability

Maintain logs showing how decisions were generated.

3. Data separation

Prevent competitors' confidential information from being improperly combined.

4. Non-discrimination

Apply equivalent criteria to competing developers.

5. Human review

Provide meaningful review of high-impact automated decisions.

6. Data governance

Define who owns and accesses zoning datasets.

7. Conflict controls

Identify situations where the algorithm provider has an affiliated developer.

8. Independent auditing

Regularly test for systematic foreclosure or discriminatory outcomes.

9. Interoperability

Avoid unnecessary technical barriers preventing competing systems from accessing legitimately available information.

10. Competition assessment

Conduct antitrust review whenever the system processes competitors' sensitive commercial information.

32. Key Case-Law Principles — Consolidated

33. Emerging RealPage Principle

The most significant modern development is that competition authorities are increasingly willing to apply traditional antitrust principles to algorithm-mediated conduct.

The RealPage litigation illustrates this transition. The DOJ alleged that competitors' confidential information was fed into a common pricing system and that the resulting recommendations reduced independent decision-making.

The same conceptual framework could become relevant to real-estate development algorithms:

Competitor information + common algorithm + aligned recommendations + reduced independent decision-making = potential competition concern.

But a zoning algorithm used by a public authority remains legally different from a private algorithm used collectively by competing developers.

34. Indian Competition-Law Perspective

For India, the analysis would principally involve the Competition Act, 2002, particularly:

• Section 3 — anti-competitive agreements;
• Section 4 — abuse of dominant position;
• Section 5 — combinations;
• Section 19 — inquiry by the Competition Commission of India;
• Sections 26 onward — investigation and adjudicatory procedure.

Potential theories could include:

Section 3

Where competing developers exchange sensitive information through a common algorithm.

Section 4

Where a dominant real-estate technology platform:

• denies access to important datasets;
• discriminates between developers;
• imposes unfair conditions;
• uses data advantages to exclude competitors.

Combination control

Where acquisition of a major zoning/data platform creates substantial concentration in real-estate information infrastructure.

The exact applicability would depend upon the relevant market and the identity of the actors.

35. Algorithmic Zoning as a New Form of Market Control

The fundamental competition-law issue can be expressed as:

Who controls the algorithm that determines who can develop, where they can develop, how much they can develop, and on what conditions?

Traditional zoning controls land use.

Algorithmic zoning can potentially control:

land use + information + development opportunity + market entry + investment timing.

Consequently, the most significant future competition-law questions are likely to concern the interaction between regulatory power, proprietary data, artificial intelligence, land scarcity and developer market power.

36. Conclusion

Algorithmic zoning is not inherently anti-competitive. Automated land-use systems can improve planning, reduce administrative discretion and make development decisions more consistent.

The competition concern arises when algorithmic control becomes a mechanism for excluding rivals, concentrating development opportunities, coordinating competitors, restricting access to essential information, or leveraging a dominant digital platform into the physical real-estate market.

The existing cases provide a useful doctrinal foundation:

• Boulder addresses municipal regulation and state-action immunity;
• Omni examines the interaction between zoning and private competitive interests;
• Racetrac demonstrates the limits of treating zoning disputes as antitrust violations;
• Oberndorf addresses urban renewal and competition among developers;
• McLain confirms the interstate-commercial dimension of real estate;
• Fisher distinguishes government regulation from private concerted action;
• Supermarket of Homes provides a real-estate monopolisation framework; and
• Kodak helps analyse digital lock-in and ecosystem power.

The emerging RealPage litigation adds a modern layer: competition law can apply where algorithms facilitate the exchange of competitors' sensitive information and reduce independent market decision-making.

Accordingly, the central legal test for algorithmic zoning should not simply be "Was an AI system used?" It should be: