I’ve been a vocal proponent of “remote” working for pretty much all of my career in technology – which, as it happens, started off in a fully remote gig. (Thanks, IBM!) My belief all along – as I’ve written about before – is that remote working is a fundamentally disruptive technology whose clear advantages will inevitably win out. I think the truth in this is pretty obvious and won’t belabor it further.
This week, Facebook broke open a long-running controversy over remote compensation with its announcement about letting (some) employees live wherever they want. Fundamentally, the question is how to pay remote employees “fairly” – indeed, what that word even means, or if it’s even relevant. This hits right at the core of some basic issues that make compensation such a hard topic. Lots of companies struggle with compensation, and most for non-evil reasons.
That all said, I do think there’s a right and wrong answer – colored a little bit by cold pragmatism.
Here’s the pragmatic bit: if it takes cutting wages to pry more tech jobs out of SF/NYC and into “everywhere else,” then so be it. Even after a 20% haircut, folks in Winston-Salem or Jacksonville or Little Rock or Bloomington will probably still find the starting salary of a product marketing manager at Facebook competitive. The more employers we have competing over workers’ wages, the better. Expanded opportunity is always better. So fine. Go for it.
With that caveat out of the way, let me go on: cost of living adjustments (COLAs) are deeply unfair and a bad practice. People doing the same jobs and providing the same value should be paid the same.
What has changed
Until very recently, going fully remote simply was not an option for most people. If you wanted to work at Company X, you had to be based at an office in some giant metro. That was the deal.
Yet in every software company I’ve worked for, almost everyone’s job could be – and sometimes was – done perfectly well online, remotely. Yes, all of them: product, engineering, design, ops, marketing and more. Even skeptics of remote work know this, deep down. And right now, we are seeing a real-life proof of concept for this fact that everyone has sorta known for a long time. Remote teams work just fine. The doubters have been conclusively proven wrong.
In other words, the business justification for employees needing to be in a major metro is collapsing. There are many who may nevertheless want to live in, say, NYC – but their doing so is not critical, or indeed even connected, to the company’s success. Code does not care where it’s written. Neither do user stories, design mockups or seller enablement decks. If it works in Tempe, it’ll work in Menlo Park, too.
(Because this is the internet and people love to nitpitck, let me just add that there are some employees who may actually have a business-related reason to live in a certain geographic area. In particular, I would think of direct client-facing roles, like sales or customer success. They are some rare exceptions to this rule.)
If the basic rationale for COLA-based compensation is that employees in high-cost metros need more to be made whole, what happens when their employer no longer has a preference? Said another way: the flip side of “why do employees in cheaper areas need as much money” becomes, rather, why does the company need people in San Francisco?
This question of what people “need” to afford their lifestyle comes up frequently when talking about employees in cheap/expensive locations. This is the wrong way of approaching the question because it’s fraught with unavoidably personal choices. If someone in San Francisco wants to spend a large chunk of their income on living where they do, then that’s great – and their choice. But if a remote (or office-based) employee in San Mateo does the same work as their colleague in Little Rock, why is one’s personal choice rewarded and the other penalized? What if the Little Rock employee has a child and is caring for an elderly parent, and the San Mateo employee is single and unattached? If the question is “need,” why are those circumstances not deemed relevant? In short, making COLA adjustments inevitably makes assumptions, mostly inappropriate, about what individuals “should” do with their money; and deems living in a big city to be a worthy choice while other decisions to be less so.
Gitlab, for example, comes out and says this plainly:
“If everyone is paid a standard salary, those who live in high-income areas would have less discretionary income when compared to their counterparts in lower-income communities.”
Well… yes! That’s the point.
COLAs are unavoidably a subsidy to someone and a penalty to another based on personal lifestyle choices. Wanting to live in a desirable, high-cost metro should not create a business obligation for your employer. Where you live is, and has always been, a consumption choice. That choice is only drawn in sharper contrast now.
Labor supply and demand at internet scale
The most common argument for using COLAs is that they’re required due to labor supply and demand. Namely: the market-clearing wage varies across different regions, so companies should pay based on local norms. This is the basic approach taken by such fully distributed companies like Gitlab, who helpfully makes their compensation formula fully transparent.
The problem with this approach comes from your definition of the “labor supply.” For an HQ-bound company, that labor pool is something like everyone in a 20-mile radius. For a remote company, however, it becomes “the internet.” (Or, for practical purposes, all internet denizens in legal jurisdictions you can hire in. More on that in a moment.) From both sides’ perspectives, the candidate in Fort Wayne is not “competing” just with those in her local area; she’s competing with the one in Memphis and Spokane and Shreveport, too. When the company adjusts downward to meet a market-clearing level for the candidate’s location, they’re really talking about how much the company can penalize employees and get away with, not what a person in that role deserves. In subsidizing one employee’s geographic preference over another, they’re stating an organizational preference that makes no business sense.
Some people will see in this model a “race to the bottom.” I do not – I see it as expanding opportunity. If there are people elsewhere who can do a given job perfectly well for less compensation expense, it obviously makes business sense to let them. I can clearly understand why this would freak out a Facebook engineer making half a million a year in Menlo Park. Given the rarified talent required at most remote companies, whatever they wind up paying the person in Bangor or Reno who takes that job will still be a good, high middle-class income.
So then, the question then becomes – just how far does this go?
There’s a slippery slope counter-argument here that, taken to its logical conclusion, going fully remote means that good tech jobs will disappear from the U.S. entirely and go overseas. After all, folks in Hyderabad and Kiev and Lagos are just as smart too and could do most of these jobs. Literally since the 1990s, we’ve heard dark warnings that the software development industry was on the verge of offshoring all engineering roles. (Seriously, this was considered a big problem in the late 90s.) But not only has it not happened, but demand for U.S. engineers and other knowledge workers has grown considerably. I think this will continue to be the case. I think there’s something about building products aimed largely at American audiences and companies that makes American employees valuable. For the purposes of remote hiring, national boundaries are meaningfully less porous than state ones, mostly for regulatory and tax issues.
(From a policymaking perspective, I think it makes sense for U.S. leaders to set up regulatory obstacles to large scale offshoring of high-wage jobs. There are interesting arguments about how globalization has hurt working-class employment in America without offering any alternative model. But that is for a different blog post, and different from the level of analysis of the firm.)
When it comes to compensation, a lot of Americans would say that of course it doesn’t make sense to pay Puja in Bangalore or Mustafa in Izmir the same as a product designer in Palo Alto. But in this, I sense a lot of self-interested thinking – Puja and Mustafa probably feel differently. Indeed, if they are providing the same value to the company, I think the right thing to do is to pay them the same wage. If this means that Puja and Mustafa live like kings (/queens), then so be it. Good for them. Paying radically lower wages for offshored teams amounts to penalizing people based on an accident of their birth, which is even less defensible than doing so because of their lifestyle choices.
Wouldn’t this explode compensation costs? Well, I don’t know. It’ll raise them, sure. But Gitlab, for example, has raised something like half a billion dollars in VC money. They can afford to pay their people. What about Microsoft or IBM or any of the other major companies with hundreds of thousands of employees in low-wage countries making less than their American counterparts? I truly don’t know the answer there. What’s right is right – people should be paid equitably, not on the basis of nationality. There is a purely pragmatic argument that it’s better for, say, Bangalore for lots of people to be paid less than Americans (but still very well by local standards) than for many fewer to be paid U.S. wages. That’s not a question for me to adjudicate. But is it fair? No, it is not.
Where there is a defensible business justification for requiring an employee’s location, the employer should pay them a local wage. But if not – and for most software roles, this is the case – then the labor pool has just gotten much wider and deeper than ever before.
There will be winners and losers as this disruption shakes out over the next several months, and probably years. No one quite knows where we’ll land. But to the extent that a handful of giant coastal metro cities’ stranglehold on lucrative tech company jobs is broken, our society – and industry – will be better for it. Remote working is better for companies and teams. It’s better for employees, families and communities. Not everyone loves remote working, but not everyone likes working in an office, either. (I don’t recall anyone’s personal preference ever mattering before.)
Remote working isn’t going to crash down on us like a tidal wave. For the foreseeable future, most companies will still be hesitant about it. I am optimistic, but only guardedly so, about how the COVID crisis will accelerate remote working. I could be wrong, but I think this will advance in small steps, not giant leaps. But here’s what I do know: I wouldn’t want to be holding a long-term commercial real estate lease on a Silicon Valley office park right now.
- FAQs on Remote Product Management
- Don’t Work “Remotely”
- Software engineering as a sign of the future