As soon as some of Silicon Valley's tech giants said they would allow indefinite remote work—and in some cases adjust pay for those who moved out of the San Francisco Bay Area—both employees and employers began to consider the impact of these decisions.
National data covering all industries would suggest that someone moving from the Bay Area to a lower cost locale—quite literally anywhere else in the US—might expect a pay cut. For example, a move to Portland, Oregon, might results in a reduction of 20% or more, based on national pay surveys. Such theoretical cuts have many potential implications. Tech workers would need to assess the impact on their lifestyles when moving to lower cost of living locations. Companies "headquartered" outside of the Bay Area could better compete for talent against Silicon Valley-"based" companies that just moved in. Human resources functions would be tasked with the tough administration and messaging of pay changes.
But is that really what the data tells us is likely to happen?
Not so fast
Don't panic, just yet. Our research shows that even before the impact of COVID-19, and the (sometimes reluctant) adoption of distributed work as a viable concept, tech companies were already paying employees outside of Silicon Valley much more than some might think. Less than Silicon Valley, yes—but not as low as some data sources might show.
The chart below shows that after controlling for job characteristics (e.g., job level, job function) and individual characteristics (e.g., years of experience, tenure), pay differentials relative to Silicon Valley are relatively narrow, and far less than what conventional national data may suggest. Take the example of Seattle. National data would suggest that a job paid $100,000 in San Francisco would be paid about 13% less in Puget Sound. However, our research indicates that the current pay differential is smaller - closer to 6% less. So instead of expecting a $13,000 pay cut, the hypothetical reduction would be closer to $6,000.
By analyzing the predictors of pay for US-based employees of high tech companies using a multivariate regression, we were able to isolate the "all else equal" coefficients for various geographies. We used Mercer | Comptryx, a full census survey, which has a standardized data architecture. This standardization means that an entry-level professional in industrial design means the same thing across companies. This provides an ability to control for characteristics like the job level of an employee, the type of work, their tenure and career movements (e.g., promotions) within a company, with high degrees of reliability and validity. Our base pay regression r-squared (a measure of how well our model explains variation in pay) is 84%, and all of the geographic differentials shown in the chart above are statistically significant.
What this means
Pay cuts won't be as big as some might think
Our data confirms that tech companies have largely been paying to their own industry standards, separate from what generalized national data tell us.
As an employee, this may mean that any pay cut you were worried about may not be as big as you think. Our experience also tells us that companies rarely cut pay, but that's a separate topic.
As a tech industry employer, if you actually do plan to reduce pay, make sure you understand the implications (cost and otherwise) before diving in. Planned for cost savings from the arbitrage of hiring in lower cost labor markets may not pay off as much as you were envisioning. Employers also need to prepare for increasing expectations for pay transparency from both current and prospective employees.
Companies "located" outside of Silicon Valley will face more competition
Companies that paid "local" market salaries in cities like Atlanta or Chicago may have a harder time competing with tech companies—even for functional, non-tech jobs. Not only did such tech companies already pay higher than what local data would suggest, but because those same tech companies are now open to hiring employees across the country, they may end up driving up "local" wages for tech industry jobs.
The upside for employees who lived in cities and couldn't or didn't want to move to tech hubs like Silicon Valley is they now have greater access to employment opportunity beyond "local" or "regional" employers. The tech job market will soon be national, and "local market rates" will be replaced by some variant of "Silicon Valley tech rates less 10%".
Human resources will actually be disrupted
The orderly sourcing of campus hires and the yearly assessment of geographic pay differentials will give way to a whole new set of challenges.
Talent acquisition and geographic flexibility will shift from "where does the company have offices" to "where does the talent exist". On the one hand, candidate pools and pay will be less about city address and more about availability and capability. On the other hand, there are address verification and tax withholding implications of where people physically work.
That first employee moving from Illinois to Iowa shifts from being the burden of obtaining an employer tax identification number in a new state, to the opportunity of being able to hire from a diverse slate of candidates across all 50 states.
The implications for employee experience in this distributed model will bring its own set of opportunities and challenges.
Value propositions that literally centered around shared space, amenities, and perks like caffe lattes and laundromats have an opportunity to be reset. Building a trusting community culture, fairly managing performance for flexible schedules, and sustaining innovation across distance and time zones will become opportunities for employers to differentiate themselves. At the same time, achieving such lofty goals will take more effort when everyone is not at the same few same campuses.
Employers have a narrow opportunity to re-imagine their talent strategy to gain a competitive advantage. I think it is time to begin to abandon long held conceptions on the role of location in talent strategy and pay strategy. What do you think?
Thanks to Jin Woo Chang and Jonathan Cottrell for their collaboration on this research, and to Roger Sturtevant for his guidance. In our next post, we plan to cover sector differences in geographic differentials, e.g., do SaaS firms pay differently from hardware firms. Ultimately, we will publish a research paper on the weighting of factors that predict pay across high tech.