Does teleworking really reduce greenhouse gas emissions? (Note)

Usefulness of the article :By presenting key concepts and analytical frameworks related to the environmental impact of buildings and transportation, this article explores the issues underlying these questions through the lens of teleworking. Highlighting the complexity of the mechanisms linked to the transportation/urban area system helps dispel certain illusions about the role of technology in the fight against climate change.

Summary:

• The use of teleworking should, in the short term, make it possible to avoid a certain amount of business travel and reduce the energy consumption typically associated with offices. However, the geographical characteristics of individuals able to telework qualify this impact.
• Teleworking involves a shift in energy consumption from offices to homes, which leads to diseconomies of scale and a loss of energy efficiency, while renovations are progressing at too slow a pace.
• In the longer term, the widespread adoption of teleworking represents a demand shock that is likely to lead to reallocations in thecommercialand residentialreal estatemarkets that are still difficult to predict.
• As household mobility patterns are also affected, a multitude of rebound and indirect effects blur the mechanical relationship between teleworking and greenhouse gas emission reductions.

The national interprofessional agreement of November 26, 2020, confirmed the changes in working patterns in France during this year, which will undoubtedly represent a turning point in the practice of teleworking in the longer term. It consolidates the legal framework for the multitude of arrangements that allow work to be relocated, particularly to the home. One of the positive consequences frequently cited is the potential for reducing greenhouse gas (GHG) emissions[1]. This hypothesis of a positive environmental impact of teleworking is part of a more general belief in the ability of technology alone to solve environmental problems. It deserves to be examined more rigorously, taking into account both direct and indirect effects and the choices made by stakeholders, to which economics can provide particularly insightful perspectives.

The issue affects two significant items in France’s GHG balance sheet[2], which therefore occupy an important place in the recovery plan presented by the government in September 2020: transport and residential and commercial buildings. Responding to this question with a robust empirical analysis would require numerous adjustments, and this work aims instead to inform the debate by attempting to complicate an issue that is less straightforward than one might think.

Let’s first establish a simplified analytical framework. In some sectors, work is carried out in an office, which may require travel from home (the « traditional » model). A whole range of technological, legal, and organizational measures make it possible to turn the home into a workplace (the « teleworking » model). After showing that, in the short term, teleworking avoids certain emissions, which are inevitably replaced by others, we will explain the difficulty of making predictions about longer-term developments.

1) Can we quantify the GHG emissions avoided through teleworking in the short term?

The answer seems intuitive: the widespread adoption of full teleworking, without commuting or physical offices, would simply cancel out all emissions related to office work. However, even in a model whose sustainability and ability to be generalized to other sectors is unknown, the question of quantification arises. We can try to look at this more closely in order to identify the mechanisms at work more clearly.

1.1 Travel less to emit less?

With nearly 39% of French GHG emissions in 2017, transport is the activity that emits the most GHGs in France (ADEME, 2018). A large part of this is due to commuting between home and work, 57% of which is done by car, according to data from the Survey on Household Environmental Practices (2016). All these figures might suggest that reducing or even eliminating business travel would lead to a significant drop in transport-related GHG emissions. However, it is necessary to break these down into several factors on which they depend:

– Transport demand (the number of trips for each type of transport);

– The average distance traveled per trip;

– Energy consumption for each type of transport; and

– The emissions factor specific to each type of energy, which establishes the correspondence between the amount of energy consumed and the amount of GHG emitted.

When considering how to reduce transportation consumption, many underlying questions come into play. Which means of transport, individual or collective, high-carbon or not, is not used? Which distances are no longer traveled? If we take into account the sectoral characteristics, level of education, or income of people who are likely to be able to telework, we can assume that the majority of them live in city centers. For example, in Paris in 2016, one in two workers worked in the same district as their place of residence (INSEE, 2016). However, in 2011, residents of urban centers emitted half as much CO2 as the average when traveling to their place of work or study (INSEE, 2011).

These geographical characteristics of teleworkers would therefore tend to reduce the environmental impact of teleworking, at least in terms of transportation.

1.2 Focus on office energy consumption

Offices have a significant environmental impact (OID, 2019). The tertiary sector accounts for 15% of total electricity consumption and 8% of GHG emissions in France. This consumption is used for lighting (12%), but above all for air conditioning, in order to maintain a certain set temperature, using heating (29%) and cooling (25%) systems. It therefore depends on the capacity to insulate, but also on the choices made for these systems. For example, the electricity consumption of an individual air conditioner is twenty times higher than that of a fan, while using refrigerants with a global warming potential 3,260 times higher than_CO2.

In this context, offices are subject to the tertiary eco-energy scheme relating to consumption reduction obligations. The text imposes quantified performance obligations on all owners of buildings larger than 1,000 m²: 40% energy savings by 2030 and 60% by 2050, compared to a reference year between 2010 and 2020. The identified levers for action are renovation, changes in usage, and automated control.

Finally, it should be borne in mind that energy consumption accounts for only 20 to 35% of the GHG emissions of a commercial building. The majority of emissions come from the construction of the building and the extraction of the materials needed for it.Life cycle analysis, which attempts to quantify all incoming and outgoing material and fluid flows at each stage of a product’s life cycle, therefore shows that the emissions avoided by teleworking are most likely to appear in the longer term, through a reduction in office construction.

2) Is consumption at home or at the office equivalent?

2.1 What are the emissions at home?

With teleworking, consumption is shifted to homes. It is therefore now necessary to examine in more detail the emissions caused by the latter, which represent 16% of the total, particularly due to energy consumption. However, this consumption is far from optimal in relation to the technical frontier, as shown by the Energy Performance Diagnostics (DPE) study, which assesses the energy performance and GHG emissions of a home. This tool is at the heart of public policy on energy management in French housing stock, the characteristics of which can be studied thanks to regular surveys conducted by the Ministry for Ecological Transition.

Of the 29 million primary residences as of January 1, 2018, only 1.9 million (6.6% of the housing stock) are considered energy efficient (DPE labels A and B). In contrast, around 4.8 million homes (17% of the housing stock) have EPC labels F and G, commonly referred to as « energy-inefficient » (CGDD, 2020). The latter are mainly houses built before 1949, located in rural and/or mountainous departments, a profile that could potentially interest workers wishing to take advantage of teleworking to leave the cities: indeed, the desire of urban French people for detached houses has been a persistent phenomenon since 1945. Conversely, in departments where heating requirements, which play a major role in the EPC, are lower, energy-intensive housing is rarer. Most of these are occupied by the most modest households. While there is certainly a correlation between those who live in these homes and those who would not be able to telework there anyway, the example of Paris, with 31% of its homes classified as « energy inefficient, » should give pause for thought about the supposed environmental virtues of « all teleworking. »

The importance of housing consumption and the gap between current consumption and the optimum level justify the prominent place given to the thermal renovation objective in the recovery plan (2 billion out of the 30 billion earmarked for the « Ecology » component). In fact, means-tested and results-based assistance (« MaPrimeRénov' ») is now available to all, comprehensive renovations are encouraged, and households are better supported. However, the obstacle seems to be less financial than informational: 5 to 9 million homes could be renovated without public assistance but are not, even though these are financially profitable energy-saving operations (Trésor Eco, June 2020).

2.2 Economies of scale for offices

Offices also allow for economies of scale and investment in more efficient technologies for the three levers of action mentioned above: (i) renovation (insulation materials), (ii) usage control, and (iii) automated management (e.g., large amounts of data can be collected, and privacy issues are less sensitive). Beyond consumption volume, they also make it possible to choose the energy source used, in particular through Price Purchase Agreements ( private electricity purchase contracts), which secure investments in green energy production facilities. According to the Sustainable Real Estate Observatory, final energy consumption in residential buildings represents 226 KWHEF/m²/year and 48_{kgCO2/m²/year} of GHG, compared to 179 KWHEF/m²/year and 16 for offices.

Thus, the shift of work from the office to the home could lead to diseconomies of scale, resulting in suboptimal aggregate consumption. More generally, the phenomenon of home-based activities is likely to lead to an increase in GHG emissions that are not directly related to work (meal delivery). Furthermore, although digital consumption still accounts for a minor share of GHG emissions (2%)[4], its very strong growth following the dematerialization of working relationships should contribute to greater awareness and visibility of the issues involved.

3) Between rebound effects and rebalancing: the limits of forecasting

In the short term, the negative effect of teleworking on GHG emissions is therefore not so obvious. This is even more the case when we look further ahead, trying to anticipate the effects of subsequent rebalancing and taking into account the heterogeneity of the economic agents concerned. The number of workers and hours that can be sustainably subject to teleworking will obviously be a decisive factor.

3.1 Real estate market adjustments

The most anticipated impact of teleworking relates to the commercial real estate market. The work of Bergeaud and Ray (2020) has shown that the widespread adoption of teleworking could result in less complementarity of space in the production function of companies: hiring an additional worker would no longer necessarily require the acquisition of additional space. This would therefore allow for a better allocation of production factors for businesses and a better allocation of workers in terms of space by reducing the constraints on access to high productivity and the agglomeration effects associated with cities (Hsieh, 2019).

The shock in rental demand for office space could thus lead to more vacant buildings being converted into housing, while the territorial rebalancing of residential real estate demand could lower housing prices in central areas with excess demand. These factors would contribute to a reduction in construction (which, as we have seen, accounts for the largest share of a building’s GHG emissions) and limit urban sprawl, with a view to achieving Zero Net Artificialization[5] (ZAN). ZAN also has objectives for reducing GHG emissions and, above all, capturing them. One of the knock-on effects could be a reduction in the amount of space allocated to parking (in Paris, half of the 140,000 surface parking spaces are expected to disappear by 2026[6]), freeing up land in city centers.

3.2 Changes in mobility behavior

A more pessimistic scenario would be one with little impact on mobility, particularly due to the greater distance between home and work as a result of the separation of these two locations. In simple terms, the increase in distance would offset the reduction in travel frequency. On the other hand, the time saved on work trips could then be used for more leisure travel, similar to the « compensatory mobility » highlighted by J.-P. Orfeuil and D. Soleyret (2002). This growth in consumption following the reduction of limits on the use of a technology is a typical illustration ofthe rebound effect or Jevons paradox. For example, Falch (2012) shows that in Denmark, although teleworking has reduced motorized commuting by 105 km per week, these gains have been partially offset by a 77 km increase in personal travel, resulting in a rebound effect of 73% (77/105).

Thus, residents’ mobility behaviors will have a significant impact on GHG emissions. The government must therefore support this change by promoting a decoupling strategy aimed at disconnecting travel demand from environmental pressures, particularly through modal shift. This is all the more necessary given that lower public transport use may, in the longer term and in the absence of fiscal rebalancing, lead to lower investment by transport companies and therefore to greater use of private transport for other journeys.

3.3 Climate change and retroactive effects

Another key variable to consider is climate change. On the one hand, the increase in average temperature could limit the comparative advantage of offices over housing in winter. On the other hand, the increase in heat waves could lead to the transformation of housing and the acquisition of air conditioning equipment that is potentially less efficient than that used in offices, for which companies have the expertise and financial resources to make optimal investments.

The consequences of teleworking on productivity, inequality, and wages are extremely difficult to predict, and although the pandemic sheds light on certain aspects that appear to be ceteris paribus, the longer-term retroactive effects will lead to unexpected general imbalances (Trésor-Eco, November 2020). The emergence of third places of activity or the appearance of « telemigrants«  working in France from abroad, for example, are complex puzzles with as yet unknown consequences and modalities. A systematic review of 39 empirical assessments of the environmental impact of teleworking concludes that the many ambiguities inherent in the issue make it impossible to provide a definitive answer (Hook et al., 2020).

Conclusion

In the short term, teleworking is likely to have intuitively negative but difficult-to-quantify effects on GHG emissions. However, it is important to bear in mind the impact of residential energy consumption, which is far from negligible, especially in winter. Finally, in the longer term, a whole series of mechanisms and readjustments are likely to make the transformations brought about by the widespread adoption of teleworking less clear and predictable, in ways that have yet to be determined. That is why it is important to remain vigilant, without venturing at this stage to make overly assertive speculations.

This article raises more questions than it answers. Teleworking represents a shock to demand in several markets, including individual and collective transport, as well as offices and housing. Identifying and quantifying these shocks would help public authorities to develop tools to manage supply adjustments. Reference can be made to the proposals put forward by the Citizens’ Climate Convention concerning changes in mobility practices, or by France Stratégie to promote the renovation of private housing (Aussilloux, 2020). The financing of the latter could also be included in negotiations between social partners on the reimbursement of teleworking costs: the reduction in emissions at workers’ homes would thus be offset by a monetary transfer.

Finally, it should be noted that technology alone cannot solve environmental problems unless there is a profound transformation of institutions and the behavior of stakeholders.

References:

Aussilloux, V., Baïz, A. (2020).Increasing investment in the energy renovation of private housing stock, France Stratégie.

ADEME (2018), Mobility and transport, key figures.

ADEME (2019), The hidden side of digital technology.

Bergeaud, A., & Ray, S. (2020). Macroeconomics of teleworking. Bulletin de la Banque de France, (231).

General Commission for Sustainable Development. (2020). « Housing stock by energy consumption class. »

Citizens’ Climate Convention (2020). Proposals of the Citizens’ Climate Convention.

Falch, M. (2012). Environmental Impact of ICT on the Transport Sector. Telecommunication Economics, 7216, 126-137.

Girard, A., & Stoetzel, J. (1947). French people’s desires in terms of urban housing: a survey. Paris, PUF-Ined.

Hsieh, C. T., & Moretti, E. (2019). Housing constraints and spatial misallocation. American Economic Journal: Macroeconomics, 11(2), 1-39.

Insee Flash No. 11. (2016).

Levy, D., & Le Jeannic, T. (2011). Urban dwellers emit half the average amount of CO2 when traveling to work or school.

Merly-Alpa, T., Riedinger, N., & Mathieu, B. (2020). Housing stock by energy consumption class.

Observatoire de l’Immobilier Durable (2019). Barometer of the energy and environmental performance of buildings.

Orfeuil, J. P., & Soleyret, D. (2002). What interactions exist between short- and long-distance mobility markets? Recherche-Transports-Sécurité, 76, 208-221.

Trésor-Éco No. 261 (June 2020), « The construction and renovation of private housing in France. »

Trésor-Éco No. 270 (November 2020), « What do we know today about the economic effects of teleworking? »