Basing remote salary on geography? There may be a less contentious solution
By Ines Delclaux
Executive Summary
The increase of productivity from switching to work-from-anywhere (WFA) from work-from-home (WFH) can mean a decrease in labor needed, and thus cost, that tech companies are otherwise achieving by varying salary relative to the geography of the employee which may bring morale blowback.
Research published in the Strategic Management Journal suggests there is an increase of 4.42% in workers’ productivity when transitioning from WFH to WFA, without sacrificing quality/accuracy in that work product.
Using this study’s methodology to assess WFH versus WFA performance can give a truer picture of the alternatives open to employers for cost-cutting measures, at least for employees involved in more routine, quantifiably output driven, tasks. A report by McKinsey lays out what those jobs may be.
Tech industry under pressure to reduce overhead, but split on their options
The marked shift to work-from-anywhere (WFA) has brought a debate about geography-based salary. Although there is now a frenzy of companies cutting down remote work and calling employees back to the office, such as Meta, Google, Amazon, and paradoxically Zoom, the reality is that in 2023, when the pandemic seems like it is in the past, 20% of full-time employees in this sector work-from-home (National Bureau of Economics Research, 2023).
Most of the tech headquarters are situated in San Francisco or New York areas, which leads the list of most expensive cities to live in the United States. As a cost cutting measure, many companies (with Meta having received the most attention) have announced that they are adjusting salaries depending on their employee’s chosen geographical locations while other companies are proudly holding the line on that measure (ex Spotify or Reddit). The impact on morale of either tactic, which may well impact productivity and quality as a result, has likely not been fully explored as the alternatives have not been in play for enough time to collect solid data.
What has been explored by McKinsey, however, is the types of jobs and tasks most suited for WFA employees. By analyzing 2,000 tasks across 800 jobs, they predicted that jobs with time spent mostly in updating knowledge, interacting with computers, and analyzing information would be those with no productivity lost at all (“What’s next for remote work: An analysis of 2,000 tasks, 800 jobs, and nine countries”, McKinsey Global Institute, November 2020). In the case of WFA specifically, of course, there is likely result of productivity gains and not just no loss. Regardless, clearly tasks in line with much of tech development.
Remote work shift that increases productivity and maintains quality
With the COVID-19 pandemic, working-from-home (WFH) became a greatly accelerated reality. The benefits of remote work have been widely discussed, such as flexibility in schedule, less commute time, improvement in mental health. Less researched has been work-from-anywhere (WFA), which differs from work-from-home in that workers can choose any geographical location. Adding more geographical flexibility may bring even more benefits to workers and their employers, but there are fewer studies, however, that analyze the direct impact of increasing geographical flexibility on productivity and quality of that work.
In a recent paper published in Strategic Management Journal, researchers analyzed how WFA transition impacted productivity and quality versus WFH. The study analyzed a dataset of 831 workers in the US Patent and Trading Office (USPTO) that enrolled in WFA between 2012 and 2013 (so not only pre-pandemic but before reliable video-conferencing was widely available). The USPTO’s WFA program allowed employees that were already WFH to live and work anywhere in the contiguous US that was more than 50 miles away from the USPTO office. USPTO patent examiners review the claims of a patent and move the patent application through the examination process and issue “First Office Actions” (FOAs) that accept the claims as patentable or not. The applicants can respond by modifying the patent or appeal the decision to start a “Request for Continued Examination” (RCE).
Thus the study had their two dependent variables: (a) to measure productivity (output): the number of FOAs in a given period; (b) to measure quality (rework): the number of RCEs in a given period. The independent variable analyzed was switch from WFH to WFA. Using OLS regression or Poisson pseudo-maximum-likelihood estimation with high dimensional fixed effects, the study found an increase of 4.42% in the total number of actions when transitioning from WFH to WFA. Importantly the amount of RCEs did not increase in that same transition, meaning the quality of the work was not affected by increasing geographical flexibility (ie not sacrificing quality/accuracy).
Finding when and where work-location freedom is optimally instituted
The academic paper offers an interesting contribution to the debate: increasing productivity from WFH to WFA would allow a company to reduce number of employees and so realize comparable savings to varying salaries by geography. Consider this hypothetical: if WFA employees are 5% more productive, then 10,000 WFH employees may do the work of 9,500 WFA employees. At $150,000 average salary, 500 fewer employees result in $75mm savings. If you maintain the 500 employees but instead give a 5% average cut depending on geographical location (anywhere from 0% to 10% cut), you are also saving $75mm.
Take for example Spotify, that announced it would allow its 6,500 employees to switch to WFA in 2021. The average salary of Spotify is $110,373 a year. Their company said it would maintain their San Francisco and New York salary, but if they are 5% more productive this means that they could need just 6,175 of those employees. Doing this would save annually $36 million. Therefore, this increase in productivity with the transition to WFA from WFH could mean saving millions of dollars annually in operational costs without any loss of productivity or accuracy.
The potential increase in productivity, without sacrificing quality, when switching from WFH to WFA has big ramifications for tech employers, especially those looking to cut overhead by instituting a geography-based salary for remote employees. But it comes with certain conditions to be applicable elsewhere:
Day One: Assess if there is a type of employee, and work product, at the company that is driven by specific productivity standards. USPTO employees were performing routine, non-collaborative, output-based tasks. In jobs where productivity and quality are not so easily measured (where creativity, initiative, or success are not directly correlated to units of output), the calculus of the study may not hold as well.
After: Provided the above is the case, employers should analyze if WFA increases productivity from WFH with their own employees to date, potentially using the same methodology from the original paper and past data.
Finally: If past data provides that result, it is suggested that small pilots of a WFA program be initiated to see if the results from the study can be replicated going forward under the current work environment. Only then can a new calculus be implemented to see if geographical-based salaries, or alternatively, cuts to the workforce, are equivalent approaches to decreasing salary costs.
In the current economic outlook, where tech is particularly under pressure to “right-size” their overhead with job cuts, maintaining the best productivity possible AND overall morale can be informed by this WFA finding for tech companies of all sizes.
____________________________
Ines Delclaux is a PhD candidate at Grossman School of Medicine at New York University and a member of the GRO-Biotech Club. The research applications proposed in this article are solely the views of the author and do not necessarily reflect the views of the original academic journal article authors nor any individual member of our Editorial Board.