A+| A| A-
An Alternate to Survey Methods to Measure Work from Home
In the times of COVID-19, it is desirable to know jobs that can be performed from home. A most common way to estimate jobs that can be performed from home is by carrying out surveys. However, unavailability of surveys in the home country compels researchers to estimate a work from home index using other countries’ surveys. This can potentially lead to large measurement errors. We believe that to overcome these challenges rating-based methodology provides a reasonable alternative which can easily be replicated in any country. Using rating-based methodology and statistical tools like inter-rater reliability we attempt to provide a robust index of WFH for India.
The authors thank Alankar Gupta for his excellent research assistance. This article is circulated as a working paper under the same title at SSRN website.
In unprecedented times like the COVID-19 pandemic, when “social distancing” is a new normal, knowing jobs that can be performed from home is useful for policymakers and researchers. The available literature extensively uses United States (US) worker-based O*NET surveys to estimate jobs that can be performed from home (Baker 2020; Dingel and Neiman 2020; Koren and Peto 2020; Mongey and Weinberg 2020). Dingel and Neiman (2020) propose a methodology, based on O*NET surveys, to estimate jobs that can be performed from home for countries other than the US. However, estimates based on O*NET surveys cannot be generalised as the task content of the occupation may vary significantly in other countries (Dicarlo et al 2016; Lo Bello et al 2019; Saltiel 2019, 2020). Using the STEP (Skill Towards Employability and Productivity) surveys, Saltiel (2020) estimates workers who can work from home for 10 developing countries. In the case of unavailability of a reliable work from home index (WFHI), researchers have relied on O*NET surveys for their respective countries. For instance, Chatterjee et al (2020) have used Dingel and Neiman (2020) estimates (based on O*NET surveys) for India and estimated that 16% of the jobs can be performed from home using the National Sample Survey, 2011–12 data. This article will show that WFHI obtained using O*NET surveys has an upward bias in the case of a developing country like India and can lead to misleading conclusions.
Given this background, we propose a rating-based robust approach to measure WFHI which is a cost-effective and timesaving alternative to countries where indigenous O*NET-like surveys do not exist. To check the reliability and robustness of ratings, we borrow statistical tools like inter-rater reliability (IRR) from psychology literature. We believe ratings-based WFHI, with significantly higher IRR, provides a reliable estimate in the absence of surveys and that this methodology can easily be extrapolated to other countries given the availability of basic dictionary of occupational titles and data on employment share.
Dear Reader,
Or
To gain instant access to this article (download).
(Readers in India)
(Readers outside India)
