Hacker News Reader

Comment by lupire | Hacker News Reader

The author almost noticed Simpson's paradox:

> Surprisingly, the UC Santa Cruz study suggests that employment in low wage industries is growing. The share of worker in jobs considered low-wage in 1997 grew 25 percent over the next 20 years, while the percent working in middle- and high-wage jobs declined.

The problem isn't that wages are dropping for certain work; the problem is that middle-wage jobs are being automated away, leaving only low-skill but non-automatable jobs that require organically-optimized things that all people have but machines don't, like fingers and eyes.

View on HN

the jobs that are mentioned in the article that are being hit the hardest are teachers, firefighters, caretakers and so forth. How many of those have been automated away?

It seems that tech workers in the valley have found their own analog to blaming immigrants for downward wage pressure, their particular boogeyman seems to be the robots.

No, the answer is much more trivial. As the article points out, the huge influx of capital in the area is being returned only to the tech sector, who merrily spend it and drive up the cost of living. The machines are not to blame for this one.

I think there’s a step missing in the analysis there. If tech workers are spending all this money — what are they spending it _on_? If they’re spending it locally shouldn’t that increase wages locally?

Shelter and avocado toast.

We're spending it on housing, mostly.

Rent. Are the landlords local?

All apartment complexes around here are owned by Californian corporations. So yes, they're local.

For example, Essex: https://www.essexapartmenthomes.com/about

They're even publicly listed: NYSE:ESS

Or the Irvine company, after which Irvine, California is named after because they designed the city. https://en.wikipedia.org/wiki/Irvine_Company

Irvine and the Irvine Company are both named for the Irvine Ranch and James Irvine who developed the property after donating a chunk to the University of California.

A fun fact is that the University of California, Irvine, was named that six years before the city was incorporated.

An area it's being spent on is rent at properties largely owned by property firms.

In the case of the building I live in (in "downtown" Sunnyvale), the three front office workers and two maintenance persons come nowhere near making a dent in the $4.5 million in income this one property earns yearly. Can't speak to the wages of the staff in the other >12 buildings in the portfolio, but the folks keeping this place running aren't driving Teslas.

No. They’re spending it all on housing.

To be clear the flow is: tech companies make enormous revenue but require high skill workers. To make this work they pay whatever it costs to get workers and no more - in this case that means just enough to offset the enormous cost of housing.

Since the baby boomers practically banned all housing development in the bay once their little burns were built, this result in a little see-saw battle where landlords raise the rent until tech companies are forced to raise wages, back and forth. Employees are left with an equation that looks basically like: enormous salary - obscene rent = just a bit more cash flow than you’d earn in software anywhere else.

Thus the flow of capital is from tech companies to landlords, in an arms race that will only end when the landlords have taken so much of the capital out of tech that the industry sputters and moves away.

Or until landlords are abolished like they should be.

Employees are left with an equation that looks basically like: enormous salary - obscene rent = just a bit more cash flow than you’d earn in software anywhere else.

It's probably true that the cost of minimally acceptable housing drives entry level tech salaries, but they go way up from there. A senior SWE at Google with salary+bonus+stock of $300k/year has a whole lot of disposable income even after the ridiculous housing expenses.

There are very few senior SWEs at Google, relative to tech employees as a whole.

First off, Google employees are the outliers even in SF. Second, there are many situational factors that make a huge difference as to whether 300k leaves “a whole lot of disposable income.”

The simplest is families living on one income. In most of the country this is easily achievable if you can break six figures, but in SF housing is priced for two full area incomes, so if you want to make it work with one parent at home you either have a relatively small and ratty apartment (compared to housing stock elsewhere) or you spend $6000+ per month on housing which takes a dramatic bite out of that $300k you’re taking about.

> I think there’s a step missing in the analysis there. If tech workers are spending all this money — what are they spending it _on_?

They often aren't, they are saving it (effectively, often directly, investing in the global capital markets and/or, insofar as their compensation has an equity component, directly in their own firm.)

Right there are no immigrants to speak of in silicon valley

> It seems that tech workers in the valley have found their own analog to blaming immigrants for downward wage pressure, their particular boogeyman seems to be the robots.

While you'll no doubt continue telling the deplorables that they're racist and xenophobic for failing to welcome their currency and cost-of-living arbitraging foreign-replacements with open arms, could you at least refrain from pretending that the basic laws of economics (i.e supply and demand) no longer exist?

Immigrants increase the supply of labor, and thus lower wages. That's not a boogieman. It's a fact.

Presumably those immigrants also increase demand for food, housing, services, transport, etc. Whether the increased demand outdoes the increased supply or vice-versa is going to vary situation by situation.

> That's not a boogieman. It's a fact.

Nope....

Immigrants are more likely to start businesses than the local population, thus creating jobs. Mostly that's small service and retail businesses. But also Google and Tesla.

Plus, obviously, a population increase also drives demand. In fact, for every job that somebody has, exactly one job was somehow created, on average.

In your theory, the US could radically lower unemployment by simply splitting up.

Yes and SV is famously flooded with immigrants.

> As the article points out, the huge influx of capital in the area is being returned only to the tech sector, who merrily spend it and drive up the cost of living.

Most of the cost of living increase is due to bone-headed housing polices happily voted in by existing residents, the vast majority of which do not work in tech. I personally find it very sad that we've prioritized rewarding a tiny, vocal minority at the expense of literally everyone else, but it seems a large number of people would rather point fingers at the villian-of-the-day than change anything.

> I personally find it very sad that we've prioritized rewarding a tiny, vocal minority at the expense of literally everyone else

you mean the tech industry? I mean, sure NIMBYism is a big problem but the tech industry gentrifies everything it touches.

Are the policies bone headed?

I'm not American so may not have this perfectly right but the Detroit of the past seems to be a good analogue to SV. Lots of well paid jobs in a single industry. When the eventual downturn occurs you don't want to be left with a ghost town.

CA voted in an effective freeze to property taxes back in the 1970s, with property tax increases being capped at 1% per year and the rate being transferable to your children. This has created a landed aristocracy in CA cities of those paying well less than the normal rate in taxes, badly distorting the market and local economies (no property tax, no services), resulting in lots of silly policy choices.

Detroit enjoyed ~3 generations of prosperity before declining. It would be silly to deny the opportunities provided by the tech industry and the wealth it generates to millions of people just because the area might eventually decline.

The SV policies aren't about diversification -- they are about preventing housing from being built.

To be more clear;

The lack of diversification in employment makes the residents as a whole more vulnerable to the next sector based bust (see dotcom crash).

If there is a larger supply of housing then more properties will be put on the market in such an event. This could exacerbate any fall in prices or result in a large number of abandoned properties which can be pretty damaging to a neighborhood.

We get a sort of similar effect here in Australia. A mine opens in the outback somewhere and prices rise rapidly in the nearest town(s). Then a whole heap of new housing gets built to meet the demand and make some profit. One day resource prices fall, the mines cut back on operations, and suddenly there is a housing oversupply and the whole local market falls pretty harshly.

If you were a local resident who lived there the entire time you would probably be feeling pretty shortchanged by the end.

The correct answer is probably to build some new housing supply but definitely not as much as the peak demands.

The correct answer is probably to build some new housing supply but definitely not as much as the peak demands.

Why is that? It doesn't change the fact that once the mines are gone, they're gone. The value of existing homeowner's investment has been preserved but nothing's been done to buffer the economy or help people pushed out of rentals. It's just transferring wealth to rentiers.

Diversifying the economy -- having an environment more encouraging to industry, not less -- is the right answer.

In the mining example I gave it's probably more clear cut than the SV case.

A whole heap of people are working high wage jobs in essentially the middle of nowhere. The local towns don't have the capacity to replace many lost jobs and aren't the ideal place for all of the now unemployed workers to end up. You don't want a whole heap of empty houses in an area with no jobs.

Diversifying the economy is definitely a step in the right direction and is likely more feasible in an area like SV. On the flipside you've got to be careful that the other industries are valuable in their own right and don't exist solely due to the rivers of money in the area.

Detroit is a ghost town, but the suburbs are fine. The story of Detroit is one of urban decay. Detroit's population has declined, but the metro area's population has been steadily growing. Detroit's downtown is not seeing a resurgence because there basically isn't one. It's a ton of 1950's low density housing stock.

And that housing stock overseen by a corrupt government that embezzled from the public, and a school board presided by a guy who seemingly can't write. It's a city whose politics are scarred by _multiple_ race riots.

In contrast, SV is a network of small suburban towns, all of which are happily constructing new office spaces but virtually no new housing. Rent prices are 3k USD a month, highest in the nation. An economic downturn shifts the local market from 'absurdly unaffordable' to 'merely very expensive'. And due to California's own political schemes, a permanent drop in housing prices would probably help local governments balance the books if it meant they could pay folks less while still taking in the same amount of tax revenues due to Prop 13.

  bone-headed housing polices happily voted in by existing residents

"Residents" have no such authority.

Cities and counties don't even have an initiative or referendum power open to residents, generally.

But most communities do have an elected city council that's supposed to represent the will of the voters. In my community any attempt to build new housing results in a lot of current reidents showing up at planning meetings and complaining about how the new development will ruin their quality of life, take away street parking, and cause gridlock.

Oddly, I only know of two people in my condo associate that work in tech, and both of us support new housing, the ones that are most vocally opposed have been here for 20+ years. Ok, maybe that's not so odd, the people that have been here the longest don't want more change.

  a lot of current residents showing up at planning meetings

Do you go to such meetings? Do you watch them on cable access or view the tapes at City Hall? (Heck, into the 1990s, the only available video of San Jose city council or SCC Supervisors meetings were if we taped it ourselves.)

In reality, almost all such meetings are lightly attended at best unless a "progressive" interest group transports then in.

Almost all Bay Area city council try to portray themselves as "progressives". Even in stodgy San Jose, not one councilmember is a Republican.

Not all of them, but I go when I can. The last one I went to was packed full of neighborhood residents overwhelmingly against the development who were worried about traffic and parking. Though the development was right next to a Caltrain station, which is exactly where mid to high density development should be happening.

  The last one I went to was packed full of neighborhood residents

Please let me know the date of that meeting; I'd like to check out the tapes.

They do in San Francisco. There's the power of the referendum (for instance, Prop C 2016), but more importantly, there's the power of the "community review" [1], with which a dedicated neighbor can stall construction for years by forcing reviews and appeals. The rules in other Bay Area cities are slightly different in construction, but not in result.

[1] https://hoodline.com/2015/11/sf-planning-101-how-community-i...

  They do in San Francisco.

Hence the "generally". Those particular 47 square miles excluded.

  There's the power of the refererendum (for instance, Prop C)

Prop C was a charter amendment, not a referendum.

"Referendum" has a specific legal meaning in CA and is almost never used because the signature threshold is the same as for Initiative statues that you can bend however you want. Referendum takes just as much work with far less utility.

  a dedicated neighbor can stall construction for years by forcing reviews

Again a random "dedicated neighbor" has no power to "force" anything... even in SF. In SF, especially, it's money that talks.

There was a study which I cannot locate right now which discussed that one of the substantial impacts of a Walmart moving into a town was that all the "community leadership" type jobs left. Things like accountants and office admins and other business professionals. This was because the small businesses Walmart replaced were the ones that used these services locally. Walmart did these things at it's corporate office far away from the community.

So those who did mid-level things like accounting or office administrative work could no longer find work locally and either had to move or take a lower paying service worker position at Walmart.

Essentially what that reveals is that most of these mid-level jobs are susceptible to the Balassa-Samuelson (BS) effect [1]

Which means that most jobs that do not require physical labor onsite are either being rapidly outsourced to cheaper markets or automated if possible. This ends up being that group in the middle of the pay range and likely accounts for the reduction in overall wages for that group.

[1] http://www.bankpedia.org/index.php/en/87-english/b/23137-bal...

The Walmart study sounds quite interesting, and makes sense. Source?

It's not wrong, a LOT of tech companies have opened up support offices in Texas. The inexplicable thing is why engineering stays in SV.

That's like asking why banking stays on Wallstreet. It's because that's where the biggest and most influential firms are headquartered.

I think it's a fair question though. Obviously talent is tough enough to recruit that you don't tell engineers they're relocating and expect to keep them. But you can choose to build offices outside one very influential city:

- Google bought the Port Authority building, which is like 2 empire state buildings in terms of office space. They have plenty of engineers working there: https://www.wired.com/2010/12/google-nyc/ This year they bought the building next door: https://www.blog.google/inside-google/company-announcements/... - Apple has a large campus in Austin: http://fortune.com/2016/09/01/apple-austin-campus/ - Amazon's obviously diversifying "HQ" locations

The question is why companies hiring engineers by the thousands every year keep pushing to open more offices in the valley. Why is it ideal to shift HR and accounting away from the west coast but keep engineering local?

Because that's where the money is.

VC are in SV and they want founders/CEOs close by. And founders/CEO want their engineers close by.

But honestly even if you really need your headquarters in SV (for VC money, network, saying you're a "Silicon Valley company", etc) your best bet is to open offices in other locations to hire engineers. Not only they can have a better standard of living even if you pay them less, but you will have less competition to hire and retain engineers.

VCs are lazy and dont like traveling more than an hour

Not really. The money mostly goes to landlords.

VC -> Startup -> Employee -> Landlord

Time to relearn the basic classes of economic goods, and their corresponding pricing behaviours: commodities, wages, rents, interest, public goods.

Land, monopolies, skills and talent endowments, capital, and power all return rents. Likely risk as we'll.

Most of books 1, 2, and 5 of Smith.

https://en.wikisource.org/wiki/The_Wealth_of_Nations

One point Smith makes is that land rents tend to eat all other forms of revenue over time, since (a) the supply does not increase, (b) the use of land is necessary for every enterprise and (c) over time all the land that can be put to some use is and then it's just about competition over it.

Rents are defined as goods or services in which the seller captures not merely costs but consumer surplus.

Smith talks about land specifically. Not sure where you're going here.

Consumer surplus is a modern definition. It follows from Smith, though somewhat indirectly.

https://en.wikipedia.org/wiki/Economic_rent

If you look up Transparent California, some public employees make pretty good income. In San Jose, there are quite a few of them making >$500k. Similar with other cities. And that doesn't count the lifetime pension, health, early retirement some of them get.

https://transparentcalifornia.com/salaries/2017/san-jose/

>And that doesn't count the lifetime pension, health, early retirement some of them get.

That website explicitly includes benefits, what are you talking about?

Firefighters - fire retardant buildings, sprinklers? Anecdotally I've heard that we need a lot less fire fighting these days because fires don't spread the same way anymore. I would be interested in seeing real data though.

Caretakers - all those robotic vacuums that are starting to become ubiquitous. Certainly not a total replacement but it lowers demand for humans.

But you don't need to automate a job to depress it's wages, just increase the supply of potential workers by automating many of the other jobs.

> Anecdotally I've heard that we need a lot less fire fighting these days because fires don't spread the same way anymore.

Hmm. In California? No.

[Ok, the bit about "spread[ing] the same way" may be relevant, but not in the implied direction. And yes, the need for large numbers of firefighters in an emergency undoubtably makes planning/costs tricky.]

I assume you are referencing the forest fires? As I understand it city fire fighters are a rather different profession.

(But no, no one involved in that discussion was from California)

Re firefighters: my understanding is that local fire departments from all over the state send assistance to large fires, see, e.g., https://abc30.com/local-fire-departments-assisting-in-wildfi...

I'm confused by "(But no, no one involved in that discussion was from California)" in the context of a discussion about salaries in Silicon Valley...

Edit: and from out of state too: https://www.stgeorgeutah.com/news/archive/2018/11/11/cgb-13-...

That's me admitting that my already dubious source was not from California... unfortunately we don't get to choose how high quality and relevant our sources of anecdotes are.

"Cost of living" increases hit hard, but it's not a wage drop. It's an extra point included in the article, not the thesis, which is about wage drops.

The article does NOT say that teachers and firefighters are seeing wage cuts.

> Workers in low and middle income jobs, such as teachers and firefighters or nannies and cooks, are key to the local economy.

teachers and firefighters are middle income, and are the core of that bracket who have demand inelasticity. Store clerks do not enjoy that, so they just lose their middle-income jobs.

> low-skill but non-automatable jobs that require organically-optimized things that all people have but machines don't, like fingers and eyes.

Machines already have those. They're just more costly and clunky than humans, often by a relatively small margin.

Anyone thinking low-skilled / low-paid jobs are not automatable is deluding themselves.

Reply on news.ycombinator.com