What buildings cost, and who pays

Situation

In 2024, the global wellness real estate market reached a volume of 584 billion US dollars. The Global Wellness Institute reported the figure in its 2025 report “Build Well To Live Well” (GWI, 2025). The institute expects 15.2 percent annual growth over the next five years and a doubling to 1.1 trillion US dollars by 2029. The historical growth rate over the preceding period ran considerably higher, around 19.5 percent. The GWI report cites 2024 once as 548 billion and once as 584 billion in its own text without reconciling the difference; the headline figure is 584.

Independent of the wellness sector, the real estate services firm JLL published a stock analysis in November 2024 (JLL, 2024, “Opportunity through Obsolescence”). It examined 776 million square meters of office space across 66 markets. Between 322 and 425 million square meters of that stock require substantial capital expenditure to meet current standards. 78 percent of this globally obsolescence-affected office product sits in the United States and Europe, with 44 percent in the US and 34 percent in Europe. The share of required Capex that concentrates in the two regions runs at 83 percent.

Two independent sampling logics converge on the same picture. One provider measures the volume of the wellness sector, the other the Capex gaps in conventional office stock. Both read the same pressure.

Regulation is moving in parallel. In April 2024 the EU adopted the revised Energy Performance of Buildings Directive (Directive (EU) 2024/1275, EPBD Recast). It entered into force on 28 May 2024, with the deadline for national transposition on 29 May 2026. For the first time, an EU directive uses Indoor Environmental Quality as a defined legal term (Article 2). The directive does not impose a binding minimum across the European stock. A conditional IAQ monitoring requirement appears in Article 13, restricted to non-residential buildings in the Zero-Emission-Buildings or deep renovation case, and only where technically and economically feasible. The directive sets no Europe-wide indoor air mandate.

Finding

The economic structure behind the market movement is an externality. The developer decides the ventilation rate and the materials that shape indoor air. The cost of those decisions lands elsewhere. It lands with the employer whose staff perform less well, and with the health insurer that pays for treatment. In residential buildings the tenant carries the direct health consequences for the family.

Joseph Allen at the Harvard T.H. Chan School of Public Health calls this the split incentive. He ran cognitive tests on office workers randomly assigned to different indoor air conditions in the 2016 COGfx study (Allen et al., 2016, Environmental Health Perspectives). Cognitive scores in the “Green+” condition with elevated ventilation rates ran 101 percent above the conventional reference, and 61 percent above it in the “Green” condition. Piers MacNaughton and colleagues followed up with a field study in occupied green-certified offices and found a smaller but consistent effect of 26.4 percent (MacNaughton et al., 2017, Building & Environment).

MacNaughton worked the economic side of these findings with Allen in 2015 (MacNaughton et al., 2015, International Journal of Environmental Research and Public Health 12(11):14709-14722). Doubling the ventilation rate in a typical office building costs less than 40 US dollars per person per year in additional energy. The expected productivity gain on the employer side runs at 6,500 US dollars per person per year. At roughly 1:160, the return far exceeds the cost even under far more conservative assumptions. It is still made against fresh air in the majority of cases, because the cost sits with the builder and the gain does not.

The market level has begun to show the economic consequences of this externality. Natasha Sadikin, Irmak Turan and Andrea Chegut reviewed rental data from the CompStak database in 2021 at MIT, filtering for health-certified offices (Sadikin, Turan & Chegut, 2021, MIT, “The Financial Impact of Healthy Buildings”, CompStak data). Tenants in certified buildings pay effective rents that run 4.4 to 7.7 percent higher per square foot compared with comparable conventional buildings. Katharina Minkow and Franz Fuerst at the University of Cambridge revisited the same question in 2025 on a larger dataset (Minkow & Fuerst, 2025, Journal of Environmental Management). They studied a sample of health-certified US offices against a large comparison sample of non-certified offices in the same markets, using a hedonic model with propensity-score matching. Across all model specifications they found a rent premium of 4 to 6 percent. In the same paper they conclude that location and neighborhood factors contribute more to rent than certification itself.

Jeremy Gabe and Michael Rehm linked Sydney office leases to the measured energy performance of the same buildings in 2014 and analyzed them in a hedonic regression (Gabe & Rehm, 2014, Journal of Property Investment & Finance 32(4):333-351). They found no significant rent differential by performance. Six location and quality factors explained more than 85 percent of gross rent, and energy performance added nothing of its own. Their core sentence: their study provides “evidence against the common assumption that rent premiums at the asset scale reflect tenant willingness to pay for energy efficiency”.

Gabe and Rehm measure energy efficiency under the Australian NABERS scheme, not health certification under WELL or Fitwel. They do not deny the existence of an observable asset premium. They deny that an observed premium can be read as tenant willingness to pay for the property in question without further identification. That caution carries directly into the young healthy-building premium literature. The MIT and Cambridge results constitute a strong signal without amounting to proof of causality. Notably, Minkow and Fuerst point to the same problem from within the pro-evidence camp, by attributing more rent variation to location and neighborhood than to certification. Gabe has not relativized this since. His follow-up work shows that labeled green buildings in Sydney use net-lease structures four to five times more often than unlabeled ones, which further confounds gross-rent comparisons. MIT measures the premium in effective rent, not gross rent. That eases only the concession and net-lease part of the caveat. The location and quality selection that drives the Gabe and Rehm result remains untouched.

Whether this market movement holds up depends less on any single premium study than on whether the national cost of poor indoor air runs large enough to generate the political pressure that resolves the externality.

France remains the only large EU country with its own macroeconomic estimate. Guillaume Boulanger calculated the social cost of indoor air pollution in 2017 for ANSES, the French agency for food, environmental and occupational safety, based on six pollutants (Boulanger et al., 2017, Environment International 104:14-24). His result lands at roughly 19 billion euros per year, around one percent of French GDP. Pierre Kopp served as the economist on the team and no separate Kopp study exists; what grey literature occasionally cites as “Kopp 2016” traces back to the same ANSES work. ANSES itself frames the figure as an illustrative order of magnitude rather than a definitive accounting. It cannot be added to outdoor-air cost estimates, because the cohorts and damage logics overlap.

The British Building Research Establishment calculated the cost of poor housing for the National Health Service in 2021 (BRE, 2021, “The cost of poor housing in England”). First-treatment costs run at 1.4 billion pounds per year, with 857 million of that attributable to cold-related conditions. The broader social measure reaches 18.5 billion pounds when productivity losses and follow-on costs are included.

A comparable global figure cannot be built methodologically. The WHO Europe office assessed the total cost of air pollution in the European region at 1.6 trillion US dollars per year (WHO Europe, 2015). The figure blends outdoor and indoor air monetarily and cannot be split into two ledgers. The mortality side is separately attributed, with roughly 482,000 deaths from outdoor air and 117,000 from indoor air for the reference year 2012, but the cost side is not. The 1.6 trillion serves as macro background. It does not serve as indoor-specific cost evidence, and it does not enter the cost stack in this analysis.

The structural situation matches two historical precedents in which externalities between cause and harm were resolved only through regulation.

The tobacco complex is one. The US states reached the Master Settlement Agreement with the four largest cigarette manufacturers in 1998. Steven Schroeder summarized the history in 2004 in the New England Journal of Medicine (Schroeder, 2004, NEJM 350:293-301). The manufacturers pay the states 206 billion US dollars across 25 years, essentially to refinance Medicaid costs for tobacco-related disease. The externality had built up across four decades during which the epidemiological evidence was uncontested.

The closer parallel to buildings is lead paint, because it sits inside the wall. Elise Gould worked out the economic balance of removing lead paint from US housing in 2009 in Environmental Health Perspectives (Gould, 2009, EHP 117(7):1162-1167). Per dollar invested, the societal return runs between 17 and 221 dollars depending on the damage category considered. The US ban on lead paint in housing came in 1978. The epidemiological evidence at that point was roughly seventy years old.

For buildings, the epidemiological evidence has been settled for forty years. It carries two structural advantages over tobacco and lead paint. First, the correction is cheap relative to the benefit, and the 1:160 ratio MacNaughton calculated runs far wider than in either precedent. Second, the resolution is arriving not solely through regulation but in parallel through the market, with premiums for certified buildings and a measurable repricing risk for the rest.

My reading is therefore that the turning point has arrived. I see a constellation that I cannot find in the history of comparable externalities. On one side the market is paying a measurable rent premium for certified buildings for the first time. On the other side, with the EPBD Recast in 2024, a legal term for indoor environmental quality enters the European framework for the first time, while the scientific evidence available since Ulrich 1984 has not relativized itself in the interim. For the position to hold, two hard conditions need to be met: the EPBD transposition has to land cleanly in the large member states, even if that requires going beyond the minimum of Article 13; and the rent premium has to be reproducible in two consecutive quarters with clean causal identification. Both are still open, and the next few quarters will show whether they are met.

Research context

Roger Ulrich showed in 1984 in Science that postoperative patients in hospital rooms with a view of trees had shorter stays and required less potent analgesics than patients facing a brick wall (Ulrich, 1984, Science 224(4647):420-421). The finding has held up across many replications. Out of that line, a research base on the influence of built environments on health and performance has grown over four decades.

The gap is not in epidemiological evidence. It lies in translating that evidence into two model classes that would carry the market effect.

The first model class is asset pricing. As far as I can read the methodological documentation, the major property indices still do not treat IEQ as a separate variable. The MSCI Global Real Estate Index is the leading example; a corresponding variable does not appear in its methodology papers. Cashflow models for institutional investors work with Net Operating Income, which partially absorbs the energy cost and therefore the IEQ contribution, but does not isolate it. As long as the data does not sit standardized in the indices, premium estimation remains tied to studies like MIT and Cambridge, and the identification weakness that Gabe and Rehm describe stays in place. A standardized IEQ variable in the major property indices would move premium estimation out of single studies and into index data.

The second model class is actuarial. No standard policy today links a tenant’s health risk to the indoor air quality of the landlord’s building. The evidence base already supports such a model, but none uses it yet. Only such a model, treating IEQ data as a predictor, would push the externality into the insurance price. Whoever offers it first carries a competitive advantage that the others can close only through their own data work. That is the gap in which a forward-looking market participant can build a lead.

The block is not technical but cultural: the data sits in safety and health categories rather than asset categories.

Consequences

For asset owners and REIT managers: Class-A buildings with health certification outperform the comparable market by 4.4 to 7.7 percent in effective rent according to MIT, and by 4 to 6 percent according to Minkow and Fuerst. These premiums are plausible under the identification caution from Gabe and Rehm, but they are not conclusively causal. How much of this premium remains durable inside net-lease structures and cashflow-effective rents is the question every serious REIT analyst should answer concretely for the 2026 reporting period at the latest. Holding stock without a certification path means carrying a growing repricing risk that continues to climb even independent of EPBD transposition, because the market is already processing it.

For family offices and institutional investors without day-to-day asset management: the spread between certified and non-certified stock will widen across the next five years. An allocation that does not reflect that spread lives on the assumption that rent premiums will close again. That is a bet against every market data point currently available, and it needs its own well-argued thesis.

For developers and builders: Doubling the ventilation rate costs roughly 40 US dollars per person per year in additional energy. The expected productivity gain on the tenant side runs 160 times higher. Whether this design choice enters the brief is not an ESG question. It is a question of whether the building drops out of the competitive set on cashflow grounds across its standing period. Planning to minimum code today means planning against the building’s own asset value in ten years.

For insurers and reinsurers: indoor air quality is becoming a modelable risk factor. Whoever begins in the next eighteen months to draw IEQ data into their own claims books as a covariate can offer, by 2027 or 2028, a product the competition does not have. The data is sufficient; what is missing is the institutional decision to treat indoor air as an actuarial variable, and the firm that makes that decision first holds the lead until others close it through their own data work.

For national EPBD implementers: the 29 May 2026 deadline is running. The conditional obligation in Article 13 does not suffice to resolve the externality described in this analysis. Member states that pull IEQ standards into building codes beyond the minimum create a location advantage for their own asset classes, because they reduce repricing risk and underwrite the healthy-building premium through regulation. France and the United Kingdom have already supplied, through the ANSES and BRE calculations, the figures on which a building-code reform can be economically grounded. Germany does not have those figures. As long as they remain absent, the national reform will stay below the ceiling the EU framework permits.