United States v. RealPage — and the Architecture That Makes Tenants Carry Everyone Else’s Decisions

The RealPage case is not only about whether an algorithm influenced rent.

It is about what happens when a system is built from the perspective of those who profit from housing while excluding the people whose lives are most directly shaped by it.

Case Agenda

• The One-Sided Market: landlord data without tenant reality
• Data Is Not Neutral: missing people shape the outcome
• Housing as Shelter, Not Merely an Asset
• Extraction Capacity Disguised as Market Value
• A Recommendation Is Directed Energy
• When Advice Becomes Market Infrastructure
• RealPage’s Responsibility for Foreseeable Influence
• Landlords Cannot Own the Income and Disown the Decision
• Formal Discretion Does Not Erase Authorship
• Property Managers and the Offloading of Discernment
• When “Market Intelligence” Replaces Human Intelligence
• Investors Cannot Treat Expectation as Due Diligence
• Returns Must Be Traced Back to Their Human Cost
• Regulators Are Not Spectators to the Market
• Foreseeability Should Have Produced Earlier Regulation
• The Algorithm Has No Conscience, but Humans Gave It Power
• Automated Distance Does Not Remove Human Responsibility
• Everyone Participates, but the Tenant Carries the Weight
• Professional Language as a Shield from Consequence
• Following the Harm Backward Through the Decision Chain
• Government Must Not Breadcrumb Responsibility
• Authority Without Full Duty of Care Becomes Performance
• The Missing Function: Legal and Governmental Triage
• Housing Must Be Prioritised as a Basic Human Need
• Why Basic Needs Require the Highest Level of Scrutiny
• The Missing Tenant Data: affordability, wages and displacement
• Consent Under Dependency Is Not Free-Market Choice
• Pricing Versus Stewardship
• Recommendations Become Investments When They Alter Reality
• Software That Organises Behaviour Is Not a Passive Messenger
• The Central Failure: Nobody Wants to Own Their Part
• Responsibility According to Knowledge, Power, Benefit and Control
• Distributed Responsibility Is Not Diluted Accountability
• Tenants Should Not Carry Responsibility for Paying Under Necessity
• The Remedy Must Include the People Excluded from the System
• Financial Settlement Without Structural Correction Is Incomplete
• Tenant Representation, Transparency and Challenge Rights
• The Central Principle: No system should price a basic human need primarily through data supplied by those who profit from making it more expensive

If RealPage collected information primarily from landlords, property managers and property owners, then the system was already built from one side of the relationship. It may have had extensive data, sophisticated modelling and large-scale market visibility, but volume does not automatically create balance. A system can contain millions of data points and still remain structurally narrow if every point reflects the same set of interests.

When you collect information from only one side, you do not receive a complete market.

You receive a coordinated perspective.

You receive what landlords want to earn, what investors expect to return, what property managers are told to optimise, what occupancy rates are considered acceptable and what level of increase the market may be able to absorb.

But where is the tenant?

Where is the person deciding between rent and food?

Where is the family whose wages have not risen alongside housing costs?

Where is the worker whose job is unstable?

Where is the parent living close to a child’s school?

Where is the disabled tenant whose home is not easily replaceable?

Where is the person who cannot simply move because every comparable property has been priced through the same logic?

If a system is designed to influence the cost of housing, tenant information should not be treated as optional context. It is foundational information.

Housing does not exist merely as an asset.

It exists as shelter.

Any system attempting to determine or recommend its price should therefore understand both sides of the relationship: the economics of maintaining property and the human limits of the people required to pay for it.

Without that balance, the system does not calculate value.

It calculates extraction capacity.

It asks how much more the market can carry without properly asking who is carrying it.

Data Is Not Neutral When the Missing People Are the Ones Paying

RealPage may argue that it gathered market information and produced pricing recommendations. But information is not neutral merely because it is numerical.

The choice of whose information is collected already determines the direction of the system.

If landlords provide the data, landlords define the market.

If investors establish the expected return, investor appetite becomes the standard.

If property managers are rewarded for increasing revenue, increased rent becomes evidence of performance.

If tenants are absent from the data architecture, affordability becomes an afterthought rather than a design principle.

That absence is not empty.

It actively shapes the output.

The tenant’s income, vulnerability, stability, family structure, dependency on location and realistic alternatives may never enter the calculation. Yet those are the realities through which the recommendation becomes lived consequence.

This is one of the great failures of modern automated systems: they often claim comprehensiveness because they possess enormous quantities of information, while remaining incapable of seeing the people excluded from the dataset.

A one-sided system does not become fair because the calculation is complex.

It becomes a complex expression of one-sidedness.

Real value emerges when all affected parties are understood.

That does not mean every tenant should personally determine the rent. It means a responsible pricing architecture should account for affordability, wage conditions, housing scarcity, local displacement risk, tenant retention, social stability and the wider consequences of coordinated increases.

Otherwise, the system does not prevent misguided outcomes.

It manufactures them efficiently.

A Recommendation Is Directed Energy

RealPage cannot fully remove itself from responsibility by describing its outputs as recommendations.

A recommendation is not nothing.

It is directed energy.

It has an intended recipient.

It is created to influence a decision.

It is packaged through information, authority and prediction to move the recipient toward a particular action.

The stronger the reputation of the system, the more weight the recommendation carries.

The more landlords rely upon it, the more the recommendation shapes the market it claims merely to observe.

At a certain scale, a recommendation stops functioning like advice and begins functioning like infrastructure.

If thousands of landlords receive similar recommendations based upon shared or overlapping information, the output may shape the very market conditions it later presents as objective.

That means RealPage should be required to understand the likely direction of its influence.

What happens when the recommendation is followed repeatedly?

What happens when many competing landlords receive the same upward pressure?

What happens when tenants cannot meaningfully reject the price because nearby properties are being guided by comparable systems?

What happens when a recommendation becomes the standard against which property managers are evaluated?

What happens when refusing the recommendation is treated as poor revenue management?

A company cannot design an instrument to change conduct and then behave as though the resulting conduct belongs entirely to someone else.

If the recommendation affects prices, occupancy strategies and housing access, it is not merely information.

It is participation.

The law should ask not only whether landlords retained formal discretion, but whether the recommendation carried enough structural authority to shape what that discretion became.

Landlords Cannot Own the Income and Disown the Decision

Landlords may argue that they retained final discretion.

But discretion is not a moral exit.

The landlord accepted the recommendation.

The landlord chose the platform.

The landlord contributed information.

The landlord benefited from the higher rent.

The tenant did not receive an invoice from an abstract algorithm. The tenant received it from the owner or manager controlling the property.

So the landlord remains responsible.

Saying that the software suggested the increase does not remove authorship. It reveals that the landlord outsourced part of the thinking while retaining the benefit.

And the relationship itself matters.

Tenants are not external market participants with no connection to the property owner. They are the people whose payments sustain the asset, service the financing, support maintenance and produce income.

They pay the bills behind the investment.

To raise their housing costs through a system they cannot see, challenge or understand is not discretion exercised transparently. It is decision-making performed behind the backs of the people who must absorb its consequences.

A landlord may legally own the property.

That does not mean the tenant has no standing within the moral and economic architecture of the decision.

The property is an asset to one party and a home to another.

Those two realities cannot be treated as equal simply because they meet inside a contract.

One party can usually replace a tenant.

The other may not easily replace a home.

That imbalance increases the landlord’s responsibility rather than reducing it.

Property Managers Cannot Call Outsourced Judgment “Market Intelligence”

Property managers may say they followed market intelligence.

But what does that phrase actually mean?

Did they independently assess whether the recommendation was fair?

Did they examine local wages?

Did they consider tenant hardship?

Did they evaluate whether the increase reflected actual property costs or merely what the system believed the market could endure?

Did they compare the recommendation with the condition of the property and the quality of service being provided?

Or did they simply accept the number because it arrived through software carrying the appearance of expertise?

Calling something market intelligence does not automatically make following it intelligent.

Property managers are not meant to be passive conduits between software and tenants. Their role requires discernment. They sit between ownership interests, building operations, tenant relationships and the reality of local housing conditions.

If they surrender that judgment to an algorithm, they are not eliminating responsibility.

They are refusing to practise it.

They cannot own the title of manager while acting only as the messenger of an automated recommendation.

Management means interpretation.

It means recalibration.

It means understanding when a system’s output is technically possible but socially destructive.

If they cannot assess the recommendation for themselves, then the market intelligence has replaced their intelligence rather than assisted it.

And if that is the case, the law should ask why the human role remains present only when responsibility must be assigned.

Investors Cannot Treat Expectation as Due Diligence

Investors may say they expected lawful returns.

But everyone expects lawful returns.

Expectation is not investigation.

It is not enough to place capital into a company, housing portfolio or revenue-management structure and then claim innocence because unlawful coordination was not part of the stated investment thesis.

Investing is not only the transfer of money.

It is the placement of energy, influence and expectation into a system.

The investor expects the system to produce more than what was originally placed into it. That expectation creates pressure.

More growth.

Higher yield.

Better occupancy.

Increasing rents.

Lower costs.

Faster returns.

Those expectations travel through ownership groups, property managers, software providers and tenants.

The tenant experiences the final pressure, but the pressure may have originated far above the lease.

Investors therefore carry a duty to investigate what produces the return.

What business model supports it?

What behaviours are being incentivised?

What risks are being transferred downward?

What legal boundaries are being tested?

What human need is being monetised?

Who is unable to opt out?

An investor cannot benefit from the architecture while refusing to examine its consequences.

If they performed no meaningful due diligence into how the increased returns were being generated, then their expectation of legality is not evidence of responsibility.

It is evidence of distance.

A lawful return is not created by calling it lawful in advance.

It is created through structures capable of producing value without unlawfully coordinating markets or extracting beyond what the underlying service justifies.

Regulators Are Not Spectators to the Market

Regulators may argue that housing prices are ordinarily established by the market.

But regulators are not external observers of the market.

Their name already describes their function.

They regulate.

They stand between commercial behaviour and the people affected by it.

They establish what information can be shared.

They define what coordination becomes unlawful.

They examine whether competition is genuine.

They intervene when market power threatens public welfare.

They are responsible for translating complex commercial activity into boundaries that ordinary people can trust.

If an algorithmic pricing system expanded across the rental market while regulators failed to examine the information being pooled, the recommendations being generated and the dependence those recommendations were creating, then regulatory failure belongs inside the case.

The housing market did not suddenly become algorithmic in secret.

Revenue-management software was marketed, purchased, implemented and scaled over time.

Why did regulators wait until the architecture had already influenced large numbers of rents?

Why was there no earlier requirement for transparency?

Why were tenants not informed when algorithmic systems materially affected their price?

Why were landlords not required to disclose which data was being used?

Why were independent audits not standard?

Why were housing systems allowed to coordinate informationally before the law determined whether that coordination undermined competition?

The state cannot repeatedly arrive after scale and present itself as though it had no opportunity to act earlier.

Foreseeability should create public duties as well as private duties.

If the risk of algorithmic coordination was visible enough to prosecute later, it was visible enough to regulate sooner.

The Algorithm Carries No Conscience—So Why Is It Being Given Power Over Shelter?

The algorithm does not carry legal consciousness.

It does not carry moral consciousness.

It does not carry ethical consciousness.

It does not experience the consequence of its recommendation.

It cannot be evicted.

It cannot miss a meal.

It cannot move a child out of school.

It cannot lose proximity to family.

It cannot become homeless.

It cannot be ashamed of the harm it causes.

It cannot be persuaded by compassion unless compassion has been deliberately translated into its design.

So why is it being trusted to influence the price of shelter?

The answer is not that algorithms should never assist decision-making. They can process information, identify trends and improve administration.

But a tool without conscience should never become the final authority over a basic human need.

The more consequential the decision, the more human responsibility must remain active around it.

Instead, many institutions use automated systems precisely because they create distance.

The landlord sees a recommendation.

The property manager sees a dashboard.

The investor sees a return.

The regulator sees a market movement.

The tenant sees the bill.

The algorithm becomes the place where everyone’s responsibility disappears.

But the tool did not choose to deploy itself.

Humans built it.

Humans purchased it.

Humans supplied the data.

Humans followed the recommendation.

Humans collected the money.

Therefore, the responsibility remains human.

Everyone Participates, but the Tenant Carries the Weight

The full chain is visible:

RealPage produces the system.

Landlords provide data and adopt the recommendations.

Property managers implement pricing decisions.

Investors reward higher returns.

Regulators permit the architecture to expand.

The algorithm processes the inputs.

And the tenant pays.

Every other participant can describe their role through professional language.

Recommendation.

Discretion.

Intelligence.

Expected returns.

Market forces.

Regulatory limitations.

Model output.

The tenant does not receive the protection of terminology.

The tenant receives the consequence.

That is how responsibility is currently distributed in many systems: the people with titles receive explanations, while the people without power receive outcomes.

It is the same structure frequently found between governments and civilians.

Public institutions make decisions through procedures, policies, committees, models and departments. Each person inside the chain can explain why their role was limited. Each actor can point to another stage of the process.

But the civilian still carries the accumulated result.

The rent rises.

The service disappears.

The benefit is delayed.

The tax increases.

The appeal is rejected.

The home is lost.

Systems often provide language to those making decisions and silence to those living beneath them.

That is why the law must not be impressed by procedural distance.

It must follow the consequence backward.

Who benefited?

Who knew?

Who had authority?

Who could have intervened?

Who chose not to?

Who relied upon someone else’s judgment?

Who supplied the data?

Who was missing from the process?

Who paid?

The Government Should Not Breadcrumb Responsibility

Government should be the one body that does not breadcrumb.

It should not distribute small fragments of protection while leaving citizens to assemble the whole system themselves.

It should not create partial rules, obscure procedures and disconnected remedies that only specialists can understand.

The government should be the most transparent, comprehensible and trustworthy architecture in the system.

People should know where responsibility sits.

They should understand what rules apply.

They should be able to challenge decisions without requiring extraordinary resources.

They should know which institution will protect them before the harm becomes catastrophic.

But this requires individuals inside government to extend themselves beyond the minimum requirements of their roles.

A title does not produce duty of care automatically.

A regulator who refuses to investigate beyond a narrow mandate is not fully regulating.

A legislator who understands the gap but waits for political convenience is not fully legislating.

An official who benefits from authority while refusing the labour required to make the system coherent is preserving a position rather than serving its purpose.

That is where government becomes performance.

The offices remain.

The procedures continue.

The language sounds official.

But the system does not reach far enough to protect the people it claims to serve.

A government that provides only fragments of protection while expecting civilians to carry the consequences is not governing completely.

It is administrating incompleteness.

Housing Is a Basic Need, Not an Ordinary Pricing Experiment

The most serious question is why an algorithm was permitted to influence the cost of a basic human need without the highest level of scrutiny.

Housing is not a luxury market in the ordinary sense.

People can choose whether to purchase certain goods.

They cannot choose indefinitely whether to have shelter.

Maslow’s hierarchy places shelter within the foundational conditions required for human security. Whether one accepts every feature of that model or not, the underlying principle is obvious: a person cannot meaningfully pursue education, work, health, family stability or personal development while lacking a safe place to live.

Housing should therefore be triaged differently from optional consumption.

The closer a system comes to controlling food, water, shelter, healthcare or physical safety, the greater the public duty attached to it.

Governments should prioritise the systems that influence basic needs before they prioritise the systems that merely improve convenience, profitability or speed.

Yet housing regulation is often reactive.

Prices rise.

People are displaced.

Homelessness increases.

Families move repeatedly.

Communities fragment.

Only after the harm becomes widespread does the system begin investigating the machinery behind it.

That is a failure of prioritisation.

A government unable to identify shelter as a higher priority than the convenience of algorithmic pricing does not understand the order in which public responsibility should operate.

The issue is not only whether RealPage violated competition law.

It is why a housing system was permitted to experiment with one-sided data and automated recommendations before proving that the basic need would remain protected.

The Missing Tenant Data

A truly balanced system would not rely solely upon what landlords want, investors expect and market conditions appear capable of supporting.

It would also incorporate:

• local wage levels;
• average household income;
• tenant turnover caused by rent increases;
• displacement risk;
• homelessness pressures;
• transportation costs;
• school and childcare dependency;
• disability-related housing needs;
• the availability of comparable properties;
• the condition of the property;
• maintenance performance;
• energy costs;
• tenant satisfaction;
• and the proportion of income already being consumed by rent.

This would not eliminate the commercial interests of landlords.

It would place them within reality.

A recommendation should not ask only:

How much can this property earn?

It should also ask:

What does this increase do to the people sustaining it?

That is the difference between pricing and stewardship.

One extracts the maximum permitted by conditions.

The other recognises that the conditions themselves include human lives.

Without tenant data, the platform cannot credibly claim to understand the market in full. It understands ownership.

It understands revenue.

It understands supply.

It understands competitive behaviour.

But the market is not complete without demand as lived by people rather than represented only through what they ultimately agree to pay.

A tenant paying an increased rent is not necessarily evidence that the price was fair.

It may be evidence that the tenant had no viable alternative.

Consent under dependency cannot be interpreted as free market enthusiasm.

Recommendations Become Investments When They Alter Reality

RealPage’s outputs should be understood as investments of influence.

Every recommendation places energy into a direction.

It says:

Raise.

Hold.

Discount.

Wait.

Accept vacancy.

Prioritise revenue.

Each recommendation participates in constructing the future market.

The company should therefore be judged not only by what it intended the recommendation to be, but by what it could reasonably foresee the recommendation becoming at scale.

Did the recommendation encourage landlords to behave similarly?

Did it reduce independent judgment?

Did it make higher rents appear inevitable?

Did it standardise revenue expectations?

Did it transform private decisions into coordinated outcomes?

Did it allow landlords to participate in collective market pressure without directly speaking to one another?

If so, the software did more than describe reality.

It organised it.

A company that organises market behaviour cannot retreat into the language of passive recommendation once the organised behaviour produces harm.

The Central Failure Is That Nobody Wants to Own Their Part

That is the foundation beneath the case.

No actor wants to own the whole consequence.

RealPage says it recommended.

Landlords say they decided independently.

Property managers say they followed intelligence.

Investors say they expected legality.

Regulators say the market established the price.

The algorithm cannot speak.

The tenant pays.

Every participant accepts the benefit of their role while narrowing the responsibility attached to it.

That is what must change.

Responsibility should be allocated according to:

• who created the tool;
• who supplied the information;
• who excluded necessary perspectives;
• who purchased the system;
• who followed the recommendations;
• who received the increased revenue;
• who financed the architecture;
• who possessed regulatory authority;
• who failed to intervene;
• and who absorbed the harm.

Everyone is responsible according to the part they actually played.

That does not mean everyone is equally responsible.

It means nobody is permitted to disappear.

Distributed Responsibility Is Not Diluted Accountability

RealPage may carry substantial responsibility if its system pooled sensitive information, shaped pricing and materially reduced genuine competition.

Landlords may carry responsibility for adopting the system and implementing the outputs.

Property managers may carry responsibility for failing to exercise independent judgment.

Investors may carry responsibility where their expectations and governance pressures rewarded the conduct.

Regulators may carry responsibility for delayed oversight and incomplete safeguards.

Government may carry responsibility for maintaining housing scarcity and allowing a basic need to become vulnerable to opaque pricing systems.

Tenants should not be blamed for paying, because payment under the threat of losing shelter is not equivalent to meaningful participation in constructing the price.

This is not an argument that everyone is responsible and therefore no one should be punished.

It is an argument that every actor should carry the weight corresponding to their knowledge, power, benefit and opportunity to prevent the outcome.

The law should not allow corporate chains to turn responsibility into dust simply because the decision passed through many hands.

The Remedy Must Include the People Excluded from the System

A financial settlement alone will not repair the architecture.

Any remedy should include:

• restrictions on sharing competitively sensitive landlord information;
• independent oversight of algorithmic pricing tools;
• clear disclosure to tenants when such systems influence rent;
• meaningful limits on automatic increases;
• auditable records showing how recommendations were produced;
• tenant representation in the design and regulation of housing technology;
• affordability and displacement assessments;
• penalties connected directly to proven harms;
• accessible routes for tenants to challenge algorithmically influenced increases;
• and public reporting on the relationship between pricing recommendations and local wages.

The system should not be redesigned only by the same actors who built and benefited from it.

Tenants must be present.

Not as symbolic consultees.

As recognised participants whose information, experience and basic needs are structurally necessary to any platform claiming to understand housing.

The correction for one-sided data is not a more sophisticated version of the same perspective.

It is the inclusion of the perspectives that were missing.

Conclusion

The RealPage case exposes a wider pattern in modern governance and commerce.

Those with power increasingly offload judgment onto systems.

Those systems produce recommendations.

The recommendations become decisions.

The decisions create consequences.

The consequences fall upon people who were never consulted.

Then every actor claims their role was too limited to carry full responsibility.

But housing is too fundamental for responsibility to remain hidden inside technical language.

A recommendation that changes the cost of shelter is not merely a recommendation.

It is an investment of influence into the life of another person.

A landlord who follows it owns the decision.

A property manager who implements it owns the judgment they surrendered.

An investor who benefits from it owns the pressure their expectations created.

A regulator who permitted it owns the gap left unguarded.

A government that failed to prioritise shelter owns the consequences of that failure.

And a system built without tenants cannot claim to understand the market whose weight those tenants carry.

The central principle should be:

No system should be permitted to determine the price of a basic human need using information gathered primarily from those who profit from making that need more expensive.

When every professional actor describes their role through procedure, the law must return to the person beneath it.

The tenant is not a metric.

The home is not merely a yield.

The market is not a moral authority.

The algorithm is not accountable.

And the government cannot continue allowing basic needs to be governed by systems it neither fully understands nor meaningfully controls.

Because when the recommendation rises through software, ownership and investment, but the burden lands entirely on the tenant, the problem is not only the price.

It is the architecture of responsibility that allowed everyone above the tenant to benefit without carrying the weight below.