> So it has always baffled me that tech companies do the exact opposite.
The primary inputs to the tech industry, especially but not exclusively consumer software industry, are culture, trends, and of course software engineering labor. All of these predominantly generated by cities, and among those, the big cities of the world.
The inputs to auto manufacture are materials, skilled manual labor, and schematics. The schematics like software, are a product of culture, and trends, and engineering research, all of which are cultural objects predominantly produced by major urban hubs. That is why auto manufacturers still design their cars in cities.
That doesn't mean that software can't be developed remotely or in cheaper COL places. But its most likely be to be designed in proximity to major cultural centers.
No, because you're missing out on that the specific culture matters. Cities produce culture in a certain sense, but they don't all produce the same culture.
There's a reason that the major tech hubs in the US are liberal cities, and more conservative ones generally don't have a strong tech sector.
Exactly, culture matters a great degree for creative work. If there were a big untapped reservoir of cultural input for tech firms being underutilized, it would be invested in. You see this a bit in secondary tech hubs like Boulder or Austin, which by dint of their history have some of whatever culture and skillsets the tech industry needs.
Big cities and metropolitan areas are generators of their flavor of culture, and that influences the kind of work objects that the local labor pool can produce.
Many of the cities on GP's list are actually major tech hubs (Tokyo, Sao Paulo, NYC, LA), and many of the other cities (Cairo, Mexico City, LA) are major creative centers for other industries especially media/entertainment industries for their respective spheres of orbit.
That's sort of a subset of the real answer: the in-demand talent that the companies need so much of are usually left of center, and relatedly appreciate some things that are more common/better in left-er areas: public transport, decent public universities, support for biking/walkability, friendliness to cultural/ethnic diversity, friendliness to LGBT people, stronger social safety net.
There are things that liberal areas are worse at, of course: housing costs are usually higher, they're not as accommodating to religion, taxes are higher, some regulations can be stifling. But most techies are less concerned with these latter things than the former.
Comparing the direct population of Columbus OH to SF is silly; it ignores the huge amount of the surrounding area, as SF proper is rather small. Some quick Googling says the Columbus MSA is ~2 M people, and the SF+San Jose MSAs are 5-7 M people.
Some quick Googling says that Phoenix metro is about 5M people.
Meanwhile, the bay area is the #1 highest COL in the continental U.S. Phoenix is ~#29.
If the argument is that tech requires a big city, I think Phoenix qualifies. If this is a hangup for you, consider Philadelphia: 5-7 M people, #21 COL.
It is obvious to me that population size is not why tech giants gravitate to SF or Seattle. (I'm confident that the high COL is also not the reason.)
Tucking 1000 employees into Houston and 1000 into Charlotte and 1000 into Toronto, then saving $200,000,000 per year seems very attractive to me. And if relocating 3000 makes sense, why not 10000?
There's obviously enough good reasons that they haven't done it. But from my seat in the balcony of the ignorant it seems awfully strange from an industry that prides itself on disrupting stale thinking.
> Tucking 1000 employees into Houston and 1000 into Charlotte and 1000 into Toronto, then saving $200,000,000 per year seems very attractive to me. And if relocating 3000 makes sense, why not 10000?
There are plenty of companies in the tech industry who already have done this - how many times have you heard about someone told by their employer that they will have to relocate to X and take a pay cut, or lose their job?
These are, however, companies whose M.O. is cutting costs to the bare minimum to increase margins. You can probably think of many that fit this model. This nearly always means they are no longer focused on growth or product innovation.
Even growth-focused companies have already moved a most of their operational and support to lower cost-of-labor areas, well before COVID19. But very few tech companies have their creative functions in those areas.
Part of it has to do with capital, but a fair portion is due to network effects.
As an example; in Boston, if you are looking to setup a Biotech shop you need to throw a rock down the street and you'll hit someone involved in the industry. In Louisville, KY? Forget about it, yes there are some folks who are in the field but as a whole?
The primary inputs to the tech industry, especially but not exclusively consumer software industry, are culture, trends, and of course software engineering labor. All of these predominantly generated by cities, and among those, the big cities of the world.
The inputs to auto manufacture are materials, skilled manual labor, and schematics. The schematics like software, are a product of culture, and trends, and engineering research, all of which are cultural objects predominantly produced by major urban hubs. That is why auto manufacturers still design their cars in cities.
That doesn't mean that software can't be developed remotely or in cheaper COL places. But its most likely be to be designed in proximity to major cultural centers.