Bad Incentives: Pest Control as a Market Failure, and What the Disclosure Modelling Actually Shows

1. Introduction: an economic problem in entomological clothing

Most of the recurring failures in urban pest management have been explained, in this journal and elsewhere, in biological and structural terms. Dispersal between units defeats unit-scoped treatment. Resistance defeats chemistry. Registration economics narrows the toolkit.

Underneath all of those sits a simpler problem that receives far less attention. In multi-tenant housing, the person who pays for treatment is frequently not the person who captures the benefit, and the person who bears the cost of an inadequate treatment frequently had no say in choosing it. That is not a technical problem. It is a textbook description of an externality, and it produces exactly the outcomes economics would predict.

1.1 Why this framing is useful

Because if the failures are economic, then technical improvements will not fix them. Better products, better training and better protocols all operate inside an incentive structure that rewards under-treatment. Changing the structure is a different kind of intervention, and the modelling literature reviewed here suggests it may be the higher-leverage one.

2. Prevalence and why it is poorly described

Any economic analysis needs a baseline, and here the data problem resembles the one described for urban rats elsewhere in this journal.

The prevalence of infestations in major United States cities is high, although only poorly described.¹ The best available figure comes from a population survey rather than industry reporting: in 2014 the New York City Community Health Survey estimated the annual prevalence of bed bug infestations at 5.1 per cent city-wide, and as high as 12 per cent in some neighbourhoods.¹

Two observations. First, one household in twenty at city scale is a substantial market, and one in eight in the worst-affected areas is a public health condition rather than a nuisance. Second, the more than two-fold variation between the city average and the worst neighbourhoods is the first appearance of a pattern that recurs throughout this paper: the burden is not distributed evenly, and the averages conceal that.

3. The externality

An externality exists when an economic decision imposes costs on parties who are not part of the transaction. Pest control in attached housing generates one continuously.

3.1 The mechanism

The chain is documented rather than hypothetical. Bed bug infestations have an infectious nature that creates spillover effects, and those spillovers produce highly nonlinear cost-benefit relationships.¹

The legal literature describes the resulting dynamic at the individual level with unusual clarity. In one documented case, a tenant unable to afford a professional pest control agent, whose landlord was unwilling to pay for treatment, experienced a worsening infestation, and her eventual move may have spread the bugs to her new building. The move did nothing to address the original infestation, which could have expanded to other units through the walls or through her discarded furniture.²

Every decision in that sequence was individually rational. The tenant could not fund treatment. The landlord declined an expense he was not clearly obliged to bear. The tenant moved to escape an intolerable situation. The aggregate result was an infestation in two buildings instead of one.

3.2 Why relocation is the costly externality

Tenant relocation is the transmission mechanism that converts a building-level problem into a market-level one. It is also the response most strongly incentivised when the alternative is living in an untreated unit.

This is worth stating plainly because it inverts a common industry assumption. Tenants who move are not being irresponsible. They are responding sensibly to a situation where the party with the authority and the resources to treat has chosen not to. The externality originates upstream of the tenant's decision.

4. Modelling infestation as contagion in a market

The most rigorous attempt to quantify any of this comes from a modelling study published in the Proceedings of the National Academy of Sciences.

The authors introduced a mathematical model to study the spread of bed bugs and to predict the costs and benefits of policies aimed at controlling them.¹ The methodological move is the interesting part: the model adapts the traditional susceptible-infectious-susceptible framework, originally developed for infectious diseases, to capture housing market dynamics.¹

4.1 Why SIS is the right borrowing

The analogy holds because the underlying structure matches. Units are susceptible or infested. Infested units transmit to susceptible ones through dispersal and tenant movement. Treatment returns an infested unit to susceptible rather than to immune, because nothing about having been treated protects a unit from reinfestation.

That last point deserves emphasis. In epidemiological terms there is no acquired immunity here. A treated unit is exactly as vulnerable as an untreated one, which is why prevalence in a market can remain stable even under substantial continuous treatment expenditure.

5. What the disclosure analysis found

The policy the model evaluates is disclosure: a requirement that landlords notify prospective tenants of recent infestations in a unit.¹

5.1 The result

While disclosure aims to protect individual tenants, the results suggest these policies also reduce infestation prevalence market-wide. Disclosure results in some initial cost to landlords but leads to significant savings in the long term.¹

5.2 The mechanism behind the result

The paper's own explanation is that if bed bug prevalence decreases following more proactive disclosure legislation, fewer treatments and less tenant turnover could increase the value of rental properties.¹

Disclosure works, on this account, by changing the payoff to under-treatment. A landlord who must disclose a recent infestation to the next prospective tenant has acquired a direct financial interest in genuine elimination rather than apparent quiet. The information requirement converts an externality into a cost the decision-maker actually faces.

6. Why landlords oppose a policy that saves them money

Disclosure laws have been actively contested by landlord organisations, with opponents fearing they could stigmatise affected buildings and lower property values.¹

This is a genuine puzzle if the long-term savings are real, and we think it has at least four plausible explanations, which we offer as analysis rather than as findings.

Time horizon mismatch. The costs of disclosure are immediate and certain. The savings are distributed, delayed and probabilistic. A landlord planning to sell within five years may rationally weigh these differently from one holding for twenty.

The benefits are partly collective. Market-wide prevalence reduction benefits all landlords, including those who do not comply enthusiastically. The individual incentive to support a collective good is weaker than the collective benefit implies. This is the standard free-rider structure.

Distributional asymmetry. Landlords in low-prevalence portfolios bear compliance costs without much prevalence benefit, while those in high-prevalence portfolios gain most. Industry associations aggregate both.

Certainty of loss versus probability of gain. Stigma on a specific unit is a concrete, attributable loss. Reduced future prevalence is diffuse and cannot be pointed at.

None of these require anyone to be irrational or ill-intentioned. They describe why a policy with positive aggregate returns can face organised opposition from its own beneficiaries.

7. The concentration problem

The accompanying PNAS commentary raises a qualification that we consider the most important result in this literature for anyone thinking about equity.

After relaxing the assumption that populations mixed randomly, the costs and burdens became concentrated such that little overall benefit was realised.³

7.1 What random mixing assumes, and why it fails

A model with random mixing treats every household as equally likely to move into any unit. Real housing markets do not work that way. People move within price bands, within neighbourhoods, within networks, and constrained by credit, references and discrimination.

Under structured mobility, infestation circulates within a subset of the market rather than diffusing across all of it. Disclosure then functions differently: it protects those with the mobility to act on the information, while those constrained to the affected segment gain little, because their alternatives are drawn from the same pool.

7.2 The equity implication

The commentary notes the evidence for socioeconomic clustering of bed bug infestation and identifies structured population mobility and nonrandom bed bug incidence as a critically important line of research.³

Read alongside the New York prevalence spread of 5.1 per cent city-wide against 12 per cent in the worst neighbourhoods,¹ the picture is one where an information-based remedy delivers most of its benefit to households that already have options. That is not an argument against disclosure. It is an argument that disclosure alone is insufficient, and that policies depending on the ability to choose a different unit will underperform for the people worst affected.

8. Who is missing from the model

The commentary identifies a second limitation with direct practical consequence. The modelling focused on the costs to landlords and did not consider costs and savings to tenants or municipalities, and as the original authors themselves note, there could be a considerable increase in net savings if tenants are accounted for.³

8.1 The uncounted tenant costs

The tenant-side cost of an infestation is substantial and almost entirely absent from policy analysis: professional treatment where the tenant ends up paying, replacement of infested mattresses and upholstered furniture, laundering and heat treating all clothing and linens, temporary relocation during treatment, and in severe cases the loss of personal property that cannot be treated.⁴

Renters' insurance typically will not pay for eradication of bed bugs or bed bug related damage.⁵ This is a meaningful gap: the cost is real, foreseeable, and uninsured.

8.2 The health cost

There is also a burden that no cost model has seriously attempted to price. Residents of infested units report chronic insomnia, severe anxiety, hypervigilance, social withdrawal, and symptoms meeting clinical criteria for post-traumatic stress disorder, with children and elderly residents particularly vulnerable.⁴

If tenant costs were included, the case for early and thorough intervention would strengthen considerably, because those costs scale with duration rather than with the number of treatments purchased. The commentary's call for future work to include costs and savings to all foreseeable actors, along with price elasticities and additional externalities, is exactly right.³

9. Information asymmetry and the lemons problem

Disclosure is an information remedy, which implies an information problem. It is worth naming that problem precisely.

A prospective tenant cannot observe a unit's infestation history. A landlord can. This is the classic structure of Akerlof's market for lemons, and it produces the classic consequence: because tenants cannot distinguish clean units from recently infested ones, they cannot pay a premium for cleanliness, and landlords therefore have no market reward for achieving it.

9.1 The attribution problem compounds it

A second information failure sits underneath. Determining who introduced an infestation, and therefore who must pay, is often very difficult in multi-unit buildings, because many tenants move in and out, some have recently travelled, and others may have brought in infested furniture.⁵

The legal literature is blunt about the consequence. Responsibilities of landlords and tenants are often unclear and vary across jurisdictions, and this uncertainty and lack of uniformity fails to incentivise behaviour that would promptly treat and prevent the spread of infestations, and does not fairly assign the costs arising from them.²

That sentence is, in effect, a diagnosis of institutional failure written by lawyers rather than economists. Unclear liability produces delay, and delay is the variable that determines both the final cost and the extent of spread.

10. The unit economics of treatment

Turning from markets to the transaction itself, the cost structure of treatment has a feature that shapes everything above.

Trade pricing indicates chemical treatment running roughly 270 to 775 US dollars per room but typically requiring two to four visits, so that three visits at 500 dollars per room reaches 1,500 dollars. Heat treatment at 2.00 to 3.50 dollars per square foot for the same room might cost 800 to 1,200 dollars in a single visit. On these figures, for moderate to severe infestations requiring three or more chemical cycles, heat often comes in at the same or lower total cost with faster resolution.⁶

We flag clearly that this source is a vendor of thermal equipment and therefore has a commercial interest in the comparison. The figures are presented as trade pricing rather than as independent data, and the reader should weigh them accordingly. We use them because we have not located independent Canadian cost data, which is itself a finding.

10.1 The structural point beneath the numbers

What matters analytically is not the specific figures but the shape: the multi-visit approach has a lower entry price and a higher total cost, while the single-session approach has a higher entry price and may have a lower total cost.

A budget-constrained decision-maker facing an uncertain outcome will choose the lower entry price. That is rational under constraint, and it is how the more expensive path gets selected in aggregate. The same logic explains why building-wide intervention, which is cheaper per unit and more likely to resolve, loses to unit-by-unit treatment, which has a smaller invoice today.

11. The principal-agent problem in service contracts

The third market failure operates between client and contractor, and the industry has been notably quiet about it.

Under a per-visit or per-unit contract, the contractor is paid for activity. Elimination ends the revenue. We are not suggesting contractors deliberately prolong infestations, and we have no evidence of that. The point is structural: the contract does not reward the outcome the client wants.

11.1 Where the misalignment binds

It binds hardest exactly where the biology demands the most. Building-wide inspection costs the contractor time that a per-unit contract does not fund. Advising a landlord that the reporting unit is not the source means recommending a larger job that the client did not ask for and may decline. Recommending that no treatment is needed, which the thermal and rodent literature both indicate is sometimes correct, means declining revenue.

11.2 The alignment that would work

A building-wide service agreement priced on outcome rather than activity aligns the parties. The contractor profits from elimination because elimination reduces their cost of servicing the agreement. The landlord buys a state rather than a series of events. This model exists in commercial contracting and is rare in residential multi-family work, which is where the biology most requires it.

12. The Manitoba position

Manitoba's framework resolves one of the three failures more cleanly than most jurisdictions, which sharpens the other two.

12.1 The liability question is settled here

As set out elsewhere in this journal, Manitoba's Residential Tenancies Branch treats pest control as a landlord responsibility and names pests including bed bugs as health risks landlords must keep under control. Provincial guidance further specifies inspection of units beside, above and below a treated unit.

That is a materially better starting position than the jurisdictional patchwork the legal literature describes.² The attribution problem that paralyses response elsewhere is, in Manitoba, largely answered as a matter of law.

12.2 What remains unaddressed

Two failures persist. There is no disclosure requirement, so the information asymmetry between landlord and prospective tenant remains intact, and with it the absence of any market reward for genuine elimination. And the principal-agent structure of the typical service contract is unchanged by any of the above.

The modelling suggests disclosure would reduce prevalence market-wide and eventually save landlords money.¹ The commentary suggests the benefits would accrue unevenly and would do least for the households most affected.³ Both findings are relevant to any Manitoba policy discussion, and neither appears to have entered one.

12.3 What an operator can do without waiting for policy

Offer building-wide agreements priced on outcome. Quote the cluster rather than the unit, and explain why. Document thoroughly enough that a landlord's due diligence position under the Residential Tenancies framework is strengthened, which converts documentation from an overhead into a product feature. None of this requires legislative change.

13. Limitations and open questions

The modelling is American and theoretical. The SIS framework and disclosure analysis were developed for United States housing markets with different tenancy law and different market structure.¹ The mechanisms transfer; the magnitudes should not be assumed to.

No Canadian prevalence data. The New York figure is the best available and it is a 2014 survey of a different city in a different country.¹ We have found no equivalent population-level prevalence estimate for any Canadian city.

Cost figures are vendor-sourced. The treatment economics in §10 come from a company selling thermal equipment.⁶ We have flagged this in the text and in the chart note. Independent Canadian cost data would materially improve this analysis and does not appear to exist publicly.

Our explanations in §6 are inference. The observation that landlord organisations oppose disclosure is reported.¹ The four candidate explanations are our reasoning, not tested propositions.

The principal-agent argument is structural. Section 11 describes an incentive shape. We have no data on how contractor behaviour actually responds to it, and we have been careful not to allege that it does.

14. Conclusion

Pest control in rental housing is not a market that happens to have problems. It is a market whose structure generates them.

The externality is real and documented: spillover effects give infestation highly nonlinear cost-benefit properties,¹ and the individually rational decision to economise on treatment reliably produces a worse aggregate outcome.² The information asymmetry is real, and the legal literature identifies unclear and non-uniform responsibility as actively failing to incentivise prompt treatment.² The principal-agent problem is inherent in paying for visits rather than for outcomes.

The best available modelling indicates that a disclosure requirement reduces prevalence market-wide and produces long-term savings for the landlords who oppose it.¹ The accompanying commentary indicates that once realistic, structured mobility replaces the random-mixing assumption, the aggregate benefit largely evaporates,³ which means the remedy works best for households that needed it least.

For Manitoba, liability is already allocated more clearly than in most jurisdictions, and that is a genuine advantage. What is missing is any mechanism by which a landlord who actually eliminates an infestation is rewarded for having done so rather than for having kept it quiet. Until something supplies that, the industry will keep selling visits, buildings will keep buying the cheapest adequate-looking response, and the entomology will keep being blamed for an arithmetic problem.

References

1. Xie, S., Hill, A.L., Rehmann, C.R. & Levy, M.Z. (2019). Dynamics of bed bug infestations and control under disclosure policies. Proceedings of the National Academy of Sciences, 116(14). doi:10.1073/pnas.1814647116. Source for the SIS model adapted to housing market dynamics, the market-wide prevalence reduction and long-term landlord savings under disclosure, landlord organisation opposition and stigma concerns, the nonlinear spillover characterisation, and the 2014 New York City Community Health Survey prevalence of 5.1 per cent city-wide and up to 12 per cent in some neighbourhoods. https://www.pnas.org/doi/10.1073/pnas.1814647116
2. Note: Don't Let Them Bite: Defining the Responsibilities of Landlords and Tenants in Bedbug Infestations. George Washington Law Review, 80, 243. Source for the documented case of a tenant unable to fund treatment whose move may have spread the infestation, and for the finding that unclear and non-uniform allocation of responsibility fails to incentivise prompt treatment and does not fairly assign costs. https://www.gwlr.org/wp-content/uploads/2018/04/80-Geo.-Wash.-L.-Rev.-243.pdf
3. Uncovering the hidden cost of bed bugs. Proceedings of the National Academy of Sciences commentary. doi:10.1073/pnas.1902404116. Source for the concentration of costs and loss of aggregate benefit once random mixing is relaxed, the socioeconomic clustering of infestation, the omission of tenant and municipal costs from the original model, and the call for inclusion of price elasticities and additional externalities. https://www.pnas.org/doi/full/10.1073/pnas.1902404116
4. Bed Bug Laws and Tenant Rights guide. Source for the enumerated tenant-side cost categories including furniture replacement, laundering, temporary relocation and property loss, and for the reported psychological burden including insomnia, anxiety, hypervigilance, social withdrawal and symptoms meeting PTSD criteria, with children and elderly residents particularly vulnerable. https://www.readyourlease.ai/guides/bed-bug-laws
5. Bed Bug Infestations: A Guide for Tenants. Nolo. Source for the difficulty of attributing introduction in multi-unit buildings, the tendency for landlords or their insurers to bear costs in multi-unit settings, and the observation that renters' insurance typically does not cover bed bug eradication or related damage. https://www.nolo.com/legal-encyclopedia/landlord-responsibility-bed-bugs.html
6. Bed bug exterminator pricing, PestPro Thermal Systems. Vendor-published trade pricing for chemical treatment at 270 to 775 USD per room over two to four visits, heat treatment at 2.00 to 3.50 USD per square foot, and per-visit service pricing. Note that this source manufactures thermal treatment equipment and therefore has a commercial interest in the comparison. https://pestprothermal.com/bed-bug-exterminators-prices/