This week, the Canadian media dutifully reported the content of a Carleton University report “Quantifying the net impact of hybrid work on greenhouse gas emissions associated with workplace and residential energy consumption” by Farzam Sepanta, Melina Sirati, and William O’Brien. The authors’ claim that civil servant remote work lowers emissions by 25% for the National Capital Region (NCR) and 64% in Quebec. The conclusions and claims are wrong.

In the report we find that an NCR civil servant who works 5 days per week in the office will produce 6.2 tonnes of CO2-e per year while that same civil servant will only produce 4.6 tonnes under a completely remote work policy – a savings of about 1.6 tonnes. In Quebec, we see a reduction of 1.3 tonnes, taking us from nearly 2 tonnes to 0.7 tonnes (the differences between the NCR and the hydro-electric building heating used in Quebec). Using survey data from civil servants, the authors carefully account for home emissions, transportation, and electricity usage. I don’t disagree with any of their calculations, their survey methods, or their results. The problem is, they didn’t finish their analysis. When we look at policy around climate change, we have to account for all the emissions that result from that policy prescription. We cannot look at only half the policy picture, and that is precisely what the authors have done.
In addition to reducing direct emissions from work related travel, civil servants who work remotely also save money. They no longer have work-travel expenses such bus passes, parking, gas, or even dress clothes. In some cases, remote work reduces child care costs. If we take an annual Ottawa bus pass with other minor incidentals as the baseline, the lower bound on savings for civil servants who no longer have to travel to work is about $2,000 per year after tax. My guess is the average savings across the survey group in the study is higher than $2,000, but let’s start there. If civil servants work from home permanently, then that extra $2,000 shifts their income up on a permanent basis. The permanent income hypothesis tells us that most of that $2,000 will be consumed, not saved (which is especially true for civil servants since they already have a pension). Civil servants tend to be solidly middle to upper middle class, which suggests that the $2,000 will have spending patterns associated with that income group. Flying produces about 2 kg CO2-e/dollar, driving about 1.7 kg CO2-e/dollar, and electronics about 0.2 kg CO2-e/dollar. If the civil servant uses the $2,000 for an extra trip to Europe, that choice creates about 1.6 tonnes from the flight alone plus extra from local transportation. The $2,000 that made the trip to Europe possible completely erases the remote work emission gains and more. Of course not all civil servants will take a trip to Europe every year with the savings, but they will spend it on something. If it’s a road trip or using the money to help support a mortgage on a recreational property, again, most if not all of the the emission gains are lost. My guess is that the spending will produce less emissions than the remote work reductions, but that $2,000 will likely produce a significant amount of emissions. Even if it’s at the level of just consumer electronics, that still represents 400kg of emissions – undermining the authors’ reported savings by at least 25%. Civil servant spending diminishes the authors’ results further if the monetary savings are greater than $2,000, which seems entirely plausible.
That consumers spend monetary savings on carbon producing activities from climate mitigation and increased efficiency is a well-known phenomenon called the rebound effect. Norwegian researchers estimate that rebound spending eliminates about half of policy-implied emission reductions. If the Norwegian observation holds with Canadian civil servants, the authors’ conclusions are wrong by a factor of 2. The closest the authors come to acknowledging rebound is with their comment:
“… telework can be a sustainable alternative to traditional work models as long as all stakeholders consciously and voluntarily contribute to adopting sustainable behaviors and to making conscious sustainable decisions associated with different domains and aspects of their lives that can impact the emissions and the 2050 net zero goals.”
That statement does not absolve the researchers from estimating civil servant rebound effects, even if only crudely. Sure, if civil servants become ascetic monks perhaps we could make the argument that we don’t need to worry about rebound, but I live in Ottawa and I know how civil servants consume (hint, they’re just like everyone else). Statistics Canada has a wealth of data on consumer spending by income groups and Treasury Board Secretariat publishes the salaries of all civil servant classifications. Using those data sources together, the authors could have at least placed bounds on emissions from the new spending made available from teleworking.
If climate change policy is to matter, we have to be honest. We cannot tell each other half truths. In the end, the atmosphere doesn’t care where the carbon comes from. Rebound effects matter. The Carleton report and an uncritical media reminds me of Richard Feynman’s comment on the Challenger disaster:
“For a successful technology, reality must take precedence over public relations, for Nature cannot be fooled.”