Switzerland's CO₂ tax on thermal fuels: Does it give property owners any financial pressure to go green?

Our new study examines whether Switzerland's carbon tax creates real financial incentives for property owners to replace fossil heating systems in the rental market—and finds the answer is, so far, mostly no.

Introducing our new working paper, "Carbon pricing and owner incentives for low-carbon investment in the Swiss rental housing market." (Available at SSRN: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=6849802.)

This paper is a chapter of Nina Lustenberger's thesis, developed with Angelika Brändle and myself. Nina drove the project, including the data work, the empirical design, and the initial writing. Huge congratulations to Nina!!

Here is my TL;DR.

Switzerland has had a CO₂ levy on thermal fossil fuels since 2008, with the rate raised several times to reach 120 CHF per tonne of CO₂ by 2022. The idea is to make fossil-heated buildings more expensive to run, so owners replace their heating systems with cleaner alternatives.

But here is the catch. In owner-occupied housing, the person paying the heating bill is the same person who decides whether to invest in a heat pump. In rental housing in Switzerland (and similar systems like Germany), it is not. Tenants typically pay heating costs through ancillary charges (charges on 'Nebenkosten'), while landlords control investment decisions. This institutional mismatch—the classic split-incentive problem—means higher energy costs don't automatically reach the decision-maker.

So does the levy actually change the financial calculus for rental property owners? And if not, what would it take?

A quick note on why we look at financial outcomes rather than heating system replacements directly. Tracking actual replacements is also important. Instead, here we ask an arguably more fundamental question: does the levy achieve its policy goal by affecting landlords' returns or risk exposure? These are the financial signals that would rationally come before any investment decision. If the levy isn't showing up in net rents or vacancy costs they face, there is little reason to expect it to show up in heating system choices either.

Findings 1: The levy doesn't reduce landlord returns.

We compare net rents in newly signed rental contracts between fossil-heated and non-fossil-heated apartments, before and after the 2018 and 2022 levy increases, across more than 118,000 contracts. If the levy were biting into landlord returns, net rents in fossil-heated apartments should grow more slowly than in comparable non-fossil ones.

We find no robust evidence of this. Net rent growth in fossil-heated dwellings did not fall significantly relative to non-fossil alternatives after either levy increase. The levy, in short, does not appear to be hurting landlords' rental income.

• Note 1: What follows from this? The levy-induced cost increases are likely borne elsewhere—most plausibly by tenants via higher ancillary charges—but our data don't directly observe ancillary costs, so this remains an inference rather than a finding.

• Note 2: There is heterogeneity across rent segments. Some negative net rent trends (= costs are felt by the owners) appear in lower-rent apartments, while higher-rent fossil properties held their ground or even outperformed non-fossil alternatives. Maybe in tight housing markets, fossil-heated landlords could even raise net rents relative to non-fossil competitors? (Switzerland's housing markets are broken with the vacancy rate as low as 0.1% in Zurich, but this is a topic for another evening.) None of the segmented estimates are statistically significant, but they point to the importance of market conditions in mediating the levy's reach.

Implications? A carbon levy that leaves rental income intact gives landlords little financial reason to act.

Findings 2: Vacancy risk for fossil-heated properties hasn't risen meaningfully either.

A second channel through which the levy could motivate landlords is vacancy risk. If tenants increasingly prefer non-fossil-heated apartments, fossil-heated properties might sit empty more often. We analyze annual vacancy cost shares for more than 2,600 multifamily investment properties and again find no robust evidence of increasing risk for fossil-heated buildings after the levy increases.

• Note 1: Some small positive effects appear in certain specifications, but they are not statistically significant and sensitive to sample restrictions.

• Note 2: In city centers and suburban areas, there is actually a slightly negative effect on vacancy for fossil-heated properties—reflecting the persistently tight Swiss rental market where demand holds up regardless of heating type, or so we interpreted.

Implications? As long as Swiss rental markets remain tight, landlords might face little pressure from tenants voting with their feet either.

Takeaway

Price signals alone, it seems, are not enough.

This is a case where a well-intentioned policy instrument runs into a structural wall. The split-incentive problem in Swiss rental housing is real, and it blunts the transmission of carbon pricing to the decision-makers who actually control heating system investments.

What would help? Better policy design (clearer, more transparent levy communication at the unit level); higher stringency (price signals need to be large and credible); and complementary instruments such as subsidies, renovation loans, and regulatory pressure. Switzerland could also look to Germany, where the 2023 CO₂ Cost Sharing Act directly tackles the split-incentive problem by requiring landlords to bear a rising share of carbon costs depending on the building's emissions intensity — making the levy impossible to simply pass through.

For tenants in fossil-heated apartments: the evidence suggests you are bearing most of these costs already.

P.S. On Data and methods

A snapshot of the average net rent trend for the fossil-heated vs. non fossil-heated properties (new contracts only) based on the SSRN working paper.

• The study draws on two datasets provided by Wüest Partner AG, covering multifamily buildings held by institutional investors across Switzerland. The dataset covers roughly 15% of all such rental dwellings nationally. The first is a contract-level dataset of 118,715 newly signed rental agreements (2016–2024), with detailed dwelling and building characteristics. The second is an annual property-level panel of 2,658 buildings tracking vacancy costs over the same period.

• We use a difference-in-differences framework comparing fossil-heated (treated) and non-fossil-heated (control) properties around the 2018 and 2022 levy increases, with propensity score matching to improve group comparability.