Until global and remote work is the norm and as productive as on-site work, there is a local market. And the local market is one of the factors you're competing within/against (employee/employer perspective).

Setting one market wage worldwide as an employer ensures that you'll not get much (any?) talent from the high-cost locations.

I don't see why; it just means you have to set the global wage equal to the ceiling, not the average of all local market wages.

Then it is the individuals choice of paying a portion of the cost for living in a high-cost location, or choosing to live in a lower-cost location and pocketing extra money. This reduction is associated with increased employee flexibility pertaining to their choice of lifestyle. It also might encourage a redistribution of individuals out of the high-cost locations. That would have some benefit (if this was industry standard) to the public in those locations as it could reduce market-demand for housing through redistribution of demand among the broader market.

I could accomplish a similar effect by merely offering 50% more than the prevailing Argentinian developer wage while still realizing substantial savings over SF rates.

>So, we punish Ann for moving by reducing her salary?

I don't see a problem with reducing Ann's pay. If rent is only $300 instead of $3000 the pay should reflect that.

If that sounds like a problem, imagine that the company pays rent or for hotel stay for visiting employees. If hotels cost $400 a night in SV and $40 a night in Argentina, should the company pay $400 a night in Argentina in order to 'equalize the pay'?

It does seem odd that an employer can use a person's lifestyle as a basis for adjusting their pay. Almost a violation of privacy.

If you want 'presence' in a certain location you'll be paying market rates and those already factor in what it costs to live there.

Would the same thing apply for moving to a different neighborhood of the same city? A different apartment/house in the same neighborhood? Gaining or losing a roommate?

In which case, realistically Ann should get paid the same amount as Sofia. No problem hiring all Sofias if Ann's situation in Sillycon Valley does not confer any benefit to the company at all.

The reason Ann can ask for $170k a year is because there are other companies in Silicon Valley that are not distributed companies, and that are competing for her, because they can't hire Sofia's.

If this company hires her anyway, despite not getting any benefit from her location, the company is making a stupid decision. They shouldn't hire her, they should hire two Sofia's elsewhere instead.

Salaries are both compensation for work performed, but also the cost of accessing certain labour markets. If you want to hire a bus driver in Manhattan, that person is going to cost you a lot more than hiring a busdriver in Hicksville. Not because there is a difference in skill, but because you need to pay a premium to get access to people who can work in Manhattan. Silicon Valley works exactly the same way, you need to pay a premium to access its labour force. And if you don't need access to it, don't pay the premium!!

Many software companies hire remotely and are rediscovering this. We don't need, as a function of the business, to have someone in e.g. central Tokyo and central Kansas, but it turns out that we often end up having that, or the strong possibility of that given the ex ante distribution of individual candidates.

Almost every distributed shop I worked with has a formal or informal discretionary bump in "the formula" to accommodate high-cost-of-living candidates. It is generally closer to $10 ~ $20k rather than "double."

American companies used to (and some foreign companies still) adjust salaries based on family size, too, which was also primarily a way to maintain access to desirable candidates for non-managerial work who had gotten an early start on family formation. (This was subsequently illegalized in the US, AFAIK, but "city in which you live" is not in general a protected class in US labor law.)

It is somewhat weird though. I think it would make more sense to pay employees with dependents less.

- Those who for medical reasons cannot have children.

- Those who for (lack of) religious reasons do not have lots of children (families with 6+ children have a different distribution of religious beliefs than the overall population).

- Homosexuals.

On the other hand, paying less to people with dependents (why?) effectively discriminates against:

- Heterosexuals.

- People who for religious reasons have lots of children.

- Disabled dependents (not a protected class in the US, but maybe it should be!).

0. http://www.touchngo.com/lglcntr/akstats/statutes/title18/cha...

1. http://www.jmls.edu/clinics/fairhousing/pdf/cook-county-huma...

The idea is that because men on average earn more money, or at least did until recently, women generally move when their husbands get a new job. This means that men are able to compete for the best wages in a global market, but women are only able to compete for the best wages in their local market. Which, depending on their industry, often means only a handful of options, and thus severely limited bargaining power.

See also:

http://econpapers.repec.org/paper/izaizadps/dp7343.htm

In academia we usually refer to this as the "two-body problem". It sometimes results in a spouse earning less because they can't negotiate as effectively; on the other hand, it also often results in a spouse being hired who wouldn't otherwise have been. Statements like "I'd love to work here, but my husband will need to find a job too... can you hire him as a lecturer?" are far from uncommon.

For example, an engineer may get hired at a time when there is a high supply and relatively low demand, and is only able to negotiate a salary of $100k. 5 years later, because of the relative shortage of engineers, a junior engineer is hired at $130k, while the the more senior engineer is now only making $120k.

Of course, trying to come up with a master system to adjust for something like that will be impossible, as this article points out (using a hypothetical move from a higher cost location to a lower cost location). The company will not adjust the senior engineers salary until he puts up a fight and comes in with competing offers.

What this suggests is perhaps that the most effective and fair system is no system at all, one that remains dependent individual actors (or businesses) trying to get the best deal that they can for themselves. This is, in practice, what is happening now, but we wouldn't have any expectation of every company trying to create some sort of formula to determine what is "fair."

This means sometimes that you will be hurt by environmental factors (say you quit or get fired in a down market in your field), but other times you may cash in (maybe the senior guy at your company quits and you have the opportunity to leverage that into a much higher salary for doing his job).

Trying to make some master system that is "fair" and will work in all circumstances just seems silly. The market is perhaps a good arbiter of value on a broad scale over long time periods, but in more localized environments and shorter time spans, it is far more random. Thus, the best personal system in my mind is one in which we can take advantage of the randomness when possible, and learn to handle the inevitable downsides as well.

However, employers who wait for their engineers to squawk and bring competing offers are playing a dangerous and inefficient game, IMO. I've absolutely given raises when the local market put my long-time employees under market. (And sometimes the way you find out is a competing offer, but if you're hiring, you generally know the market.)

On your other point about companies being inefficient, I think we should stop worrying about whether employers or companies in aggregate are being efficient. In fact, I want them to be inefficient. It gives people more opportunities to take advantage of inefficiencies.

I suspect we have some sort of bias in which we apply fairness rules that are good for individuals in a community and apply it to "the system" which is not just a scaled up version of individuals interacting which either. We like to think big companies operate as a unified entity, but really, there's a lot of shit going on under the hood that may lead to very weird "irrational" decisions, many of which an individual could take advantage of.

Generally, I just prefer people to be more opportunistic, rather than trying to make sense out of non-sensical systems.

My point about inefficiency was a note to employers -> try not to be inefficient and risk-seeking in this particular area.

The remote contracting market, especially for fixed rate contracts, is not driven by locations. It's driven by who in the globe can do a particular job at a particular price, and there's a wide variety in how much that translates to in "goats" for different developers.

Traditional remote jobs are still driven mostly by paying for a fixed amount of time working on the job per week. As such, they're not hiring for a "job to be done"—they're hiring you as an individual. To do that, they just need to pay more than the next best option, which is frequently determined by the local labor market.

Given that, option 1 makes a lot of sense, with the catch that you have to have a clear policy so it's known ahead of time. Also realize that Ann might try to get a better replacement offer and then you'll have to match to the "new" market for that location.

In that case it would likely not seem absurd because presumably there would be some extra inherent value in luxury car drivers (e.g., uber drivers).

That's the point actually. Driving a luxury car is equally as irrelevant as location. There really isn't a good reason for a company to reward employees for how those employees spend money outside of work.

If you're buying a car and it comes with a high power engine that you don't care about, you still have to pay the premium for the engine. You can't pay less because you tell the seller that you won't utilize the extra power. The car has a higher market price because there are people that will pay extra for the high power engine.

This statement is still correct and less loaded if you remove the words "get to".

I have no idea why a fully-remote company would want their developers to live in Silicon Valley so much that they would pay a premium for it. Why even bother being fully-remote then?

It's like wanting developers to drive a luxury car so much they would pay a premium for luxury car owners.

You try to make it sound ridiculous by changing location with the car they drive, but that is intellectually dishonest because it's not clear remote work is as efficient. There is a reason Google doesn't allow it.

But to a remote company, it's just as irrelevant as a luxury car. It's an artificial segment of the market to them. They don't need to enter the frenzy. There is a equilibrium of global supply and demand that is lower than SV equlibrium. Replace your SV hires with the others at that price, and save money. Capitalize on your remote work, instead of tossing it aside.

If the SV developer's value to the company is the same but they get paid more, then you're saying to all the other developers who work there, "You produce enough value that we could pay you this much, but we don't want to. We prefer to pay you less, simply because we can."

This is, of course, their right, but that sort of thing tends to make people unhappy, and unhappy employees tend to be unproductive, and then gone.

Several people here are taking it as a given that a technology company has to hire some quota of people from the Valley.

Or, more accurately, not to pay for the high power engine. Choose and pay for the car independent of the irrelevant factor of engine power. If the best value car (from your perspective) happens to be one with a high power engine, then you'll get that one. But you don't pay more for one, especially when your prospective market is broad (the entire world).

Just like with the car analogy, if you need to put together a fleet of vehicles immediately, you will have to buy anything that is available, including the one with the bigger engine.

That's self-induced due to your treatment of the market. Suppose you'll pay SV devs 170k.

Do you have any idea how easy it would be to replace those spots by offering 160k (regardless of location)? And you'll save money.

Most devs, FYI, do not live in SV area. The market for remote work is huge.

For a company with all remote employees, location is indeed irrelevant. Unless you're claiming there are no developers in the whole rest of the world that would do the job equally well for cheaper, which seems unlikely. Or unless you can back up the claim that developers in certain locations are simply better than ones in other locations, which seems to be what you're implying.

As an anecdotal counter example, I have moved several times in my career, and don't recall it ever affecting my productivity.

> If you're buying a car and it comes with a high power engine that you don't care about, you still have to pay the premium for the engine. You can't pay less because you tell the seller that you won't utilize the extra power. The car has a higher market price because there are people that will pay extra for the high power engine.

I don't see what that example has to do with anything.

Ann is paid $170,000 in the Bay Area because that what all the other developers of her caliber are willing to exchange their time and expertise for.

A company would pay Ann $50,000 if it could. They only hire people who they think they will eventually make more from their labor than they will pay the person.

If you are Google, you will make roughly $1M per employee currently. For Software Developers that number is even higher, for Sr. Software Developers that number is even higher. The fact that how you get paid and how much value you generate don't have much of an effect on each other is basic capitalism.

What's smart is to be Ann, get a $170k salary and move to Costa Rica and keep the $170k because at the end of the day, she wouldn't have been hired if she didn't make more than $170k of value for the company.

Well...sort of. Ann is paid $170K because Bay-area companies have bid up the cost of Bay-area developers to $170K.

Ann doesn't really have a say in this, and it has nothing to do with how much value she adds to the company (although obviously a company wouldn't fork over 170K if she didn't add at least that much value to the company).

Now if Ann moves to Costa Rica, she is now a Costa Rica-based remote developer. And I suspect that Costa Rica-based remote developers have not been bid up to the same level as Bay-area based developers. So her employer would certainly be justified in seeking a downward salary adjustment.

This would mean hiring companies would need to accommodate remote workers; are they willing to for the cost benefit? My experience says no (except for <100 companies that are remote first).

The Bay area seems to have a positive-feedback loop going on. Highly-paid programmers have bid up the cost of housing to astonishing heights.

So when companies want to attract more developers, they have to offer even higher salaries, which lead to housing costs being bid up even more.

This... is not the primary driver of the insane housing costs in the area. A survey from a year or two back found that tech workers make up ~8% of SF's population.

For a variety of reasons, [0] it's nearly impossible to build new housing in San Francisco. I get the impression that the situation is similar in much of the Bay Area.

In San Francisco, for the past decade or so for every new unit of housing created, roughly five people have entered the city. [1]

Add supply to take care of the backlog and meet expected medium-term demand, and you'll see prices stabilize (and maybe return to less-insane levels).

[0] The least of which is Rent Stabilization. :)

[1] You could make the argument that "If noone was able to pay the insane prices, the prices wouldn't be insane." This is true, but -frankly- there are lots of very highly-paid people out there. As far as "highly-paid" people go, tech workers really aren't all that highly-paid. :) (I know of decent-to-good engineering sales [2] folks working at bigcos that make between 2 and 10x what I understand mid-level Google engineers to make.)

[2] That is, salesmen that also have a technical background, can handle crunchy sales and configuration questions, and can even do real, deep troubleshooting of the product they're selling.

I'm in the wrong business.

nod nod

If you've both the people and technical skills required to do engineering sales, you're probably seriously limiting your earnings by remaining a programmer.

Sadly, I forget what these positions are actually called. Best of luck in your hunt! I hope you find and fill such a position and are both happy and richly rewarded in it. :D

Edit: terminology

Yep. Obviously. I hope that you didn't think that I thought otherwise.

From what I've read about the topic, it seems like a few things are true:

* Housing costs are rising radically throughout the Bay Area

* No major Bay Area city is building to meet demand in the area

* Some Bay Area cities (notably, SF, MV, SJ, and others) are actively impeding new residential construction with a variety of pleasant-sounding excuses

It's true that SF's fucked-up housing policy doesn't necessarily mean much for the rest of the Bay Area. However, it's a sad fact that the landowners in much of The Area have -correctly- surmised that they stand to make a shitload of money if they fail to build to meet demand.

Fuck housing that's reasonably priced when there are pockets to be lined and fortunes to be made, amirite? :(

The latter was driven by a strong belief in many Bay Area startups and companies that remote work doesn't work well, and that you get more output per head and more productivity if you sit next to your developers. That's what the premium is for in that case, usually with a tech stack that is relatively easy to find people for (e.g. RoR, PHP, Python).

The former was driven by rare knowledge with few customers, all of whom have good reasons to be relatively price insensitive to get the job done.

Pick your strategy accordingly...

That $170k is what it's worth for her, as a person, to switch jobs for.

I've always wondered if this is even possible. Imagine you are a Trust Fund Kid who happens to be a skilled developer as well as a giant troll. You could under-bid on salary offers. But would the company still hire you? Would you be able to prove your chops in spite of the negative signal of valuing yourself below market? (Of course, you are probably rare and won't have an effect on the overall market for developers, but I still think it's a fun thought experiment.)

Companies decide to make an before they pick a salary offer

The point of the article is that, assuming equal employees, you are in fact saying that one has a higher economic value to the company based only on where they choose to live.

Salary may naturally vary by location, but wealth accumulation (in dollars, not goats) should be the same for equivalent contributors across the distributed organization.

He's basically advocating for global market value. The side effect here is that individuals are free to maximize their wealth by choosing to live where they can best maximize their wealth. This allows the individual to chose what is valuable to them - dollars/goats or the intangibles of a location, e.g. sunshine, culture, freedom, family, etc.

This makes some sense for the individual where you have global scarcity. This enhances the competition for the individual's skill where geography in many cases presents a real barrier. For example in non-distributed workplaces you are essentially up against a sort of implicit collusion to minimize rates, the local market demand for your skill.

That seems like a really, really bad idea.

I'm no expert on labor law, but why wouldn't it be?

Why would labor law ban the increase of one's salary based on size of one's family?

I actually don't know what would happen if I moved to Argentina. I presume my salary would get dropped. Maybe not.. depends on if I can convince someone else to give me an offer at my previous salary for remote work... but then what of my Argentinian friend?

I think all developers should get offered the same amount independent of location. As remote work takes off (and I think it will...), this will become the norm. Why? Because of how market works. If Google will pay me 170k to work from my house, then smaller companies need to start competing with that.

Ironically, the big companies are still way behind on this. I work at Canonical... our devs are all fully remote. I have people from nearly every continent working on my project. There's very very little friction to remote teams, except sometimes you need to talk to someone who's asleep. But that honestly is not usually that bad. You keep close teams in similar time zones and everything just works out. We use hangouts and irc and email... hell, the disincentive to have useless meetings probably saves us more time than we spend working around the relatively small problems remote work introduces.

(also, I'm kinda glad to get away from whiteboarding crap... a whiteboard is not a good tool for pretty much anything in development... a shared text window is almost always superior)

I think the article puts things a little unfairly by saying only "Option 1: Ann's pay is reduced". Yes that is what's happening, but if Ann knows what her compensation will look like in Buenos Tiempos, she should be able to figure out roughly if she will take home more money by moving there.

Then again, what is the difference between "remote workers" and "offshore development"? The latter doesn't seem to have worked out too well so far?

All developers will get the same if not for regulatory barriers (working permit and alike).

Assuming you pay people in their own currency (which is useful to make sure they don't pay unnecessary fees, and also that they get a consistent pay each month). How do you keep salaries consistent globally?

Do you adjust them each month to compensate fluctuations? Sucks for those that just got a pay cut.

I don't know the answer, I think the way Buffer does it is the best of the options I've come across so far. I'd like a better solution.

Other parts of the Buffer payment formula simply illegal. The practice of paying people more for each dependent [0] violates many localities' (e.g. Cook County, IL [1]) anti-discrimination ordinances.

0. https://open.buffer.com/transparent-salaries/ 1. http://www.jmls.edu/clinics/fairhousing/pdf/cook-county-huma...

The way I look at is that employment is a mutually beneficial agreement. Both parties should have the option to end the agreement when some part of the situation changes. When Ann moves and for some reason thinks that it's unfair that she won't be making an SF salary in Ohio, she can leave the company. Same goes for the employer when an employee moves from Ohio to SF. Saying "you're working remotely and we will pay you Ohio market rate but if you move to SF, we cannot afford that, so you will either have to accept Ohio market rate in SF or leave to find another company that will pay you SF market rate" sounds like the only fair option to me.

Where are you buying your goats?

Differential pricing cannot last in a fungible market. The companies that offer $100k and get $300k value from each employee will eventually out-compete those that offer $100-200k and get $300k value.

It's the same mechanism behind free markets defeating other forms of discrimination (I use this in the general, non-legal sense). If most companies pay women less because that's the market, eventually the non-discriminating outliers will do better and the market adjusts to the more efficient reality.

The timing may be unknown, but the movement towards a non-discriminating equilibrium in fungible markets is not.

Just because you don't value location doesn't make the location premium any less real. The simple fact is that there is a lot more demand for developers in the bay than there is in South America.

A company could certainly overpay the devs outside of SV so the numbers look the same, but it's fiscally irresponsible.

> Perhaps you are (inadvertently) either using location as a lever to underpay employees, or using it as an excuse to overpay people you like but perhaps do not justify the cost for the business?

I think this hits the mark. Companies that push heavily for remote work generally aren't doing it because they genuinely care for the benefits it brings employees. They're pushing for it because it saves them money over operating with an entirely local team (when local means a high market-rate area like SF), precisely because they can discriminate based on location when negotiating salaries.

In a non-free market, such as a command economy, prices could be set by someone(s). But he doesn't like the idea of an arbitrary entity (the company, in his case) setting the number of "goats" (living standard) you get paid. He wants developers to get to negotiate that. But they already are, and globally developers and companies have settled on the prices we have.

I suspect that some analysis would show that developers want a certain standard of living globally, and beyond that, they are pretty happy. I'd love to be paid the living standard of a millionaire, but I'm also okay to be paid what other developers are getting paid. Until all of us decide refuse anything less than living like millionaires, we'll get paid a middle class salary.