Job Density and Why It Beats Job Count

The two cities, side by side

New York employs about 121,000 software developers at a median of $166,830. San Jose employs 87,350 at $213,110. New York has 38 percent more of them and pays $46,280 less in the middle of the range.

If more jobs simply meant a better market, that could not happen. It happens because job count mostly measures how big a city is, and New York is big at everything. The count is real and it is answering a question about the city rather than about your occupation.

That gap is the whole subject of this article. One number tells you how many doors exist and a different number tells you how hard anybody behind them will compete for you. Most people planning a move only ever look at the first.

The number that explains it

Take your occupation’s share of local employment and divide it by its share nationally. A result of 1.0 means the city holds exactly its fair share given its size. Anything above that means the place is denser in your work than the country as a whole.

New York’s software developers come out at 1.18, which is barely above average. A metro of 9.4 million jobs holds a lot of everything, so almost nothing stands out. San Jose comes out at 7.23 on a labor market of 1.1 million.

That is the difference between a city that happens to contain your work and a city built around it. The calculation takes four published numbers and a minute, and it explains pay gaps that job counts cannot. Nothing else on a jobs page carries that much information.

Why concentration pays and size does not

Employers in a dense market have to compete for the same people. Where dozens of firms need one skill in one place, leaving is easy and keeping you costs real money. Where one or two employers need it, they set the price and there is nothing to test it against.

The market also survives any single employer. In a dense market losing your job is an inconvenience and you interview somewhere down the road. In a thin one it can mean selling a house, which is a completely different category of event.

And pay information circulates in a dense market because people know each other. That circulation is the precondition for pay moving at all. It does not exist where your occupation is scattered thinly across a big city’s economy.

The small places this reveals

Boulder, Colorado has a labor market of 178,320, which is smaller than most people’s mental list of technology cities. Its concentration of software developers is 3.68, higher than San Francisco’s 2.72, with a median of $164,560. Fewer than eight thousand developers produce that density.

The pattern turns up everywhere once you look for it. Huntsville, Alabama concentrates aerospace engineers at 44 times the national rate. Morgantown, West Virginia concentrates registered nurses at 3.60 times its share.

None of those places appears near the top of any ranking, because rankings sort by count and these are small metros. They are invisible by construction rather than by oversight. That is precisely why the pools of candidates there are shallower.

Why every best-cities list looks the same

Those lists sort by absolute numbers, which returns the largest metros in the country whatever the occupation happens to be. The method guarantees the answer before the question is asked. Population and opportunity are not the same quantity.

A list recommending New York, Los Angeles and Chicago for software, for nursing, for accounting and for logistics is not telling you anything about those professions. It is telling you that those three cities are large, which you already knew. That is a fact about geography rather than about careers.

Treat any ranking that returns the same cities for every occupation as a population ranking wearing a job-market label. The useful version sorts by concentration and returns different answers for different work. If every occupation gets one answer, something is wrong.

Doing the calculation yourself

You need three published numbers plus one you already have. Your occupation’s employment in the metro, that metro’s total employment, and the same two figures nationally. Divide the first pair, divide the second pair, then divide one result by the other.

Above 1.0 and the place holds more of your work than its size implies. Above 2.0 and you are looking at a genuine specialist market rather than a large city that contains some of your colleagues. Above 3.0 you are almost certainly looking at a cluster.

It takes a minute per metro and it will point you at places no ranking will. Run it on five or six candidates rather than one, because the interesting results are the ones you would not have thought to check. Familiar cities were going to score how you expected.

What count is still good for

Count buys volume and variety, and both are worth having. A large market has more openings in raw terms and more kinds of employer, which matters if you want optionality or you are changing sector. Both of those are legitimate reasons to prefer size.

New York is not a bad place to be a developer by any reasonable standard. It is a place where being one is unremarkable, and the median reflects exactly that. Nine million jobs means almost every skill is already present in quantity.

The honest summary is that count tells you how many doors there are and concentration tells you how hard anyone behind them will compete for you. Which of those matters more depends on whether you are optimizing for options or for price. Most people never make that choice explicitly at all.

Density is why remote work did not flatten pay

The obvious prediction was that remote hiring would erase geographic pay differences. If location stops mattering, the San Jose premium should collapse toward the national median fairly quickly. Plenty of confident commentary said exactly that would happen.

It largely has not, and concentration explains why. A dense market is not simply a pile of jobs in one place. It is employers who compete with each other, colleagues who talk, and a salary level everyone in the market knows, and all three survive the office closing.

A company hiring remotely into that market is still hiring against those employers and still has to pay near what they pay. The exception is a company that prices by the worker’s location rather than its own, which is the arrangement worth asking about explicitly before accepting any remote role. The two policies produce very different numbers for one job.

What to ask about a place before you commit

Ask how many separate employers within commuting distance need what you do. Not how many jobs exist, but how many distinct organizations. Three substantial employers and one dominant one are completely different markets at identical concentration.

Ask whether the concentration comes from one industry or several. Software in San Jose spans hardware, platforms, finance and defense contractors, so a downturn in any one of them is survivable. Aerospace in a single-installation town is not.

And ask whether the second career in your household has anywhere to go. A metro that is superb for you and thin for a partner is a worse move than one that is merely decent for both. Dual-career households are where the concentration calculation most often gets skipped entirely.