Change happens slowly, then all at once. We’re witnessing an “all at once” change right now in remote work. The quarantine was just the catalyst; conditions have been ripening for a remote work renaissance for quite a while. Fast Internet connections became ubiquitous enough for remote work to be as productive as working from the office. Software as a service blurred the line between the “corporate network” and the outside world. Zoom made videoconferencing easy, and Slack made instant messaging fun. But all the technology in the world can’t solve the hardest problem with remote work: fairly compensating employees. Facebook is currently facing this problem. Though they are going to allow remote work on a scale unprecedented in the company’s history, they are coming under fire for requiring employees who move to lower cost-of-living areas to take pay cuts. Having built a 15-person, remote-first company, I’ve sat on both sides of this table. It’s healthy - for both the employee and employer - to pay people based on their location. There’s a trick to setting this up, however, and retroactively changing employment agreements isn’t it.
There are two schools of thought when it comes to remote compensation: the “Basecamp school,” and the “Buffer school.” Basecamp pays everyone in the company the 90th percentile of San Francisco market rates. Buffer, on the other hand, uses a formula to calculate salaries that takes into account an employee’s role, seniority, experience, and location. The location factor has a lot of impact: Buffer employees in expensive locales can receive up to $22,000 in additional salary to offset their higher cost of living. I modelled our remote compensation model after Basecamp, and it remains one of my biggest professional mistakes. The model worked well when we were small and all of our employees lived in expensive coastal cities, but things changed as we grew. The structure bred resentment as employees in expensive areas felt perpetually underpaid. The company’s growth was hamstrung, buried under a mountain of payroll pressure. Everyone lost - except for the employees living in cheap locales.
To put things in perspective, let’s break down the numbers. An employee lives in San Francisco, and makes $160,000. After income taxes and housing costs, she can expect to take home $78,000.1 If she moved to Las Vegas, she would take home $107,0002 - an effective raise of 37% just for moving. To equal the take home pay in Las Vegas, the San Francisco employee would need to make a base salary of $210,000. What’s more, a dollar goes a lot further in Las Vegas than it does in San Francisco. Compensation is much more than a raw exchange of dollars for time. It’s a reflection of how much a company values its employees, and a significant driver of employee happiness. It doesn’t matter if all employees are paid the same base salary if their take-home pay is dramatically different. Calculating salaries based on location gives the employer the financial leeway to ensure that the part of an employee’s salary that actually matters - their take-home pay - is sufficient and fair.
This is why I’m such a fan of the Buffer compensation model. It works because of its transparency. Compensation is a social contract. It states: “We value your contribution to the company at $X per annum.” By making explicit how much of an employee’s salary is derived from their contribution versus their location, Buffer can fairly compensate employees based on their location. This has to be described upfront when the employee joins. What employers cannot do is retroactively apply a location-based compensation model to existing employees if they start to work remotely. This is Facebook’s mistake. A Facebook employee in Menlo Park is still the same employee once they move to Tulsa. Cutting pay breaks the contract that Facebook made with their employees when they were first hired, and is a transparent cash grab that will lead to higher turnover and, in the long run, cost the company money. It costs tens - sometimes hundreds - of thousands of dollars to hire and train a new employee. This is very likely to be less than the delta between a Silicon Valley salary and a cost-of-living adjusted salary. The right thing to do would be to grandfather existing employees starting to work remotely, and transition all new employees to a salary calculated based on location.
It is true that this model privileges the employees who get grandfathered in. They could move and receive an instant 37% raise. From my experience, however, this rarely happens. People build ties to the places they live. They build families there, they make friends, they have favorite coffee shops and hiking spots. If already well-compensated - which Facebook employees are - I believe that it’s unlikely very many grandfathered Facebook employees will move. And even if they did - who cares? The company made a commitment to pay a certain amount, and they should honor it. The goodwill created by grandfathering employees will outweigh the cost savings created by cutting pay. 
1 Assuming that the employee is a single filer, and paying $2,500 per month in rent.
2 Assuming that the employee is a single filer, and paying $1,000 per month in rent.