Working from home: Too much of a good thing?
BibTeX
@article{behrens2021working,
title={Working from home: Too much of a good thing?},
author={Behrens, Kristian and Kichko, Sergey and Thisse, Jacques-Fran{\c{c}}ois},
journal={Available at SSRN 3768910},
year={2021}
}
Abstract
We develop a general equilibrium model with three primary production factors—land, skilled, and unskilled labor—and three sectors—construction, intermediate inputs, and final consumption—to study how different intensities of telecommuting affect the efficiency of firms that embrace home working, as well as its impact on the whole economy. In doing so, we pay particular attention to the effects of increasing working from home (WFH) that go through changes in the production and consumption of buildings: more WFH reduces firms’ demands for office space, but increases workers’ demand for living space since additional room is required to work from home. We find that more WFH is a mixed blessing: the relationship between telecommuting and productivity or GDP is ∩-shaped, whereas telecommuting raises income inequality. Hence, WFH is not a panacea since an excessive downscaling of workspaces may be damaging to all and exacerbate economic inequality.
Notes and Excerpts
Our key results can be summarized as follows. First, it is profit-maximizing for firms to implement a partial WFH strategy, that is, the working time is split between home and office, which is beneficial to both the skilled and unskilled. Second, WFH is a mixed blessing: when the WFH share steadily increases, more telecommuting increases skilled and unskilled workers’ productivity and GDP when there is no excessive downscaling of office space, whereas its starts to decrease productivity and GDP beyond some threshold.
Commuting is modelled as a tax.
Following the urban economics literature, we model commuting costs using an iceberg τk ≥ 1, which amounts to lowering the worker’s income.3 Since the opportunity cost of time increases with income (Small, 2012; Koster and Koster, 2015), it is reasonable to assume that τs > τℓ . Without loss of generality, we normalize τℓ to 1 and set τs ≡ τ > 1
According to BLS data (https://www.bls.gov/news.release/cesan.nr0.htm), Americans spent about 16-17% of their income on transportation in 2019. This includes all expenses, not just commuting. Mas and Pallais (2017) find that American workers are willing to give up 8 percent of their wage for the option of WFH, thus suggesting that commuting likely represent a large share of the total transportation budget.
When a skilled works home, she provides one efficiency unit of home-work; when she commutes to the firm, she provides 1/τ < 1 efficiency units of office-work.