Working Paper: NBER ID: w29431
Authors: Janice C. Eberly; Jonathan Haskel; Paul Mizen
Abstract: The impact of an economic shock depends both on its severity and the resilience of the economic response. Resilience can include the ability to relocate factors, for example, even when new technologies or skills are not yet at the ready. This resilience per se buffers production and has an economic value, which we estimate. The COVID-19 pandemic caused a widespread decline in recorded GDP. Yet, as catastrophic as the collapse was, it was buffered by an unprecedented and spontaneous deployment of what we call “Potential Capital,” the dwelling/residential capital and connective technologies used alongside working from home. Together potential capital and labor working from home provided additional output margins and capacity. We estimate the contribution of this capital, and the remote work that it facilitated, to have roughly halved the decline in GDP in the US reducing the fall in GDP to 9.4 log points in 2020Q2 at the trough of the recession. Similar effects are seen in the 6 OECD countries for which data are available, output fell by 14 log points, but would have fallen by 26 log points had only workplace inputs been available. Accounting for the contribution of “Potential Capital” also revises downwards estimated total productivity gains in the business sector during the pandemic from 8 log points to 5 log points in 2020Q2. We also find an output elasticity of domestic non-dwellings capital to be similar to that of workplace ICT capital, reflecting its role as productive capital. Turning to the future, changes in working from home depend upon relative costs, relative technologies and, crucially, the elasticity of substitution between home and work tasks. We estimate that that elasticity to be more than unity, meaning that the growth of ICT will raise the share of work done remotely.
Keywords: COVID-19; economic resilience; working from home; potential capital
JEL Codes: E01; E22; O47
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Potential capital at home + labor working from home (J24) | Economic resilience during COVID-19 (F69) |
| Home capital utilization (E22) | Resilience of GDP (F69) |
| Domestic non-dwelling capital elasticity (E20) | Workplace ICT capital elasticity (J29) |
| Use of potential capital (G31) | Mitigation of expected decline in output (E23) |