Hacker News Reader

Comment by CriticalCathed | Hacker News Reader

This is going to be a shock to all of those people who were willing to give tax incentives for Amazon to build distribution warehouses in their city in exchange for "jobs."

Though, this is only the beginning. A few thousand jobs now, the majority of them later.

View on HN

How do you know overall jobs will decrease?

It's also totally possible that each warehouse will be made twice as efficient per employee... but also deliver twice the merchandise... maintaining the same total number of jobs.

Kind of the same way that automation of the past 100 years across the economy has still resulted in record-setting employment right now.

Curiously, what does the homelessness & (true, ie not unemployment) jobless rate look like in that graph? Of course though, they're tough metrics.

More importantly though, we're on the cusp of some pretty hairy water. Imagine we drop all of the truck drivers, taxi drivers, Uber and Lift drivers tomorrow. How many millions of people will that be - where can they go?

I'm not against automation in the slightest, but to think it won't have an impact on employment seems farfetched. Imagine a robot tomorrow, humanoid capable of basic tasks. How many millions of people would that displace almost instantly?

I don't think the past 100 years even remotely indicates how we'll handle this as a society. We aren't talking about taking away one job. We're talking about taking away a skill category. No more drivers. No more basic manual labor.

And that's not even talking about how growth would need to expand to handle the increase in workforce. Imagine our workforce doubles tomorrow - who is going to consume it all? Do we even want more consumption?

I don't think it's as cut and dry as looking at the street cleaners and how they handled it when horses stopped being used transportation. We're talking about fundamentally categories of work becoming obsolete, and having no jobs that some people qualify for.

With that said, I'm curious to live in a world where the population is forced to be educated. No low skilled jobs might have a benefit of (basically) no low educated people. However I am worried about how it will affect the current populace.

Looking at the state of my apartment I could really use a fulltime maid/cook. I can't afford that right now, but if the difference between the average income for the low income group and the high income group goes up enough, that would be a possibility (assuming computer programmers continue to get paid a high income, whereas maids do not).

> I don't think the past 100 years even remotely indicates how we'll handle this as a society. We aren't talking about taking away one job. We're talking about taking away a skill category.

We did it with farmers. 80% of jobs in early 1800's, close to 40% in 1900, 20% by 1930, 5% by 1965, and 1.7% now. A whole "skill category" like driving doesn't even come close to farming.

Automation, and even displacing drivers, is going to continue happening gradually, not overnight. E.g. only the easiest driving jobs in the easiest areas are automated first, then gradually more difficult ones... There's no "overnight". It'll take many years, and it will be fine just like it has each decade over the past century.

[1] https://azizonomics.files.wordpress.com/2013/01/farmjobs.jpg

So I take it your view is we should not plan / discuss how to handle it? Take it on good faith that it'll be fine?

My view is that the loss of driving jobs can be handled normally, via existing free market and education institutions -- that it's no different from the loss of other job categories over the past 50 years.

That this isn't special, that we're not "on the cusp of some pretty hairy water", but rather that this is just par for the course.

I just don't see any statistics that suggest the scale or rate will be anything close to larger/faster than our economy can normally handle. If someone can show those comparative statistics exist, then I'll change my mind.

That's fair. It would be interesting to have data that compares all of the job classes out there and the number of workers out there, against the worker education.

Ie, how many millions of food service workers, manual laborers, and drivers (for trucks, taxis, etc) could move upward in career depth? Where would we expect them to be absorbed?

I definitely think this is hairy waters, in the sense that I want to see studies done. Planning for how it can be handled. Etc. Calling it fine in my view would suggest we don't even need to study it - which seems unfair. This degree of automation, assuming we can get over some AI hurdles of course, seems unparalleled - at least enough to warrant real data.

Just a quick point: they don't necessarily need to move upward. They can move laterally. They can work in package delivery, in restaurants, at Starbucks, in health care, in tourism, in solar panel and home battery installation, in environment cleanup sites, in converting homes to greater energy efficiency, and so on. Honestly a lot of the new jobs 20 years from now we have a hard time guessing what they'll even be (if you magically know, you'll make billions investing), which makes "planning" for it a bit handwavy anyways. (Remember, nobody "planned" for web developers to be an exploding new career.)

Well that goes back to the original point though. Which is, that no education jobs will theoretically dwindle. If we have machines capable of handling all, repeat all no education manual labor jobs; things like picking fruit, delivering boxes, etc. Where will those people go?

They will be forced to acquire more education or skill oriented trades.

The concern is if humanity will be able to adapt to something it has never seen. Where creativity and thinking oriented jobs become more and more required.

Although having a very low unemployment rate is better than a high unemployment rate, it doesn't help paint the whole picture of the American workforce. The other metric that ought to be mentioned along with unemployment is the labor participation rate, which has been trending worse for a while now [1]. Considering the two together, the economy isn't as rosy as it looks. [2]

For example, if I were to be chronically unemployed and quit looking for jobs altogether, I would actually help the unemployment rate look better since I'm no longer part of that metric.

Part of the problem can be attributed to automation in jobs and job duties, thus requiring a smaller labor force. For those who were laid off by automation in manufacturing or other sectors, many would likely have a hard time finding a job in their established career. Expecting most of such workers to get a job an an on demand skill like maintaining machinery or software development is a herculean task. Most likely they are the ones who fall prey to the poor labor participation metric.

[1] https://www.investopedia.com/ask/answers/061515/what-key-dif... [2] http://www.startribune.com/even-with-unemployment-at-50-year...

They seem to be trying to convert these lost employees into "delivery companies"[0]. As much as I dislike the Uberfication of the world, seems like a clever way of shifting human capital from warehouse to delivery.

[0] https://arstechnica.com/information-technology/2019/05/amazo...

Most of those tax breaks have specific empployment figures associated with them.

Bu there's no penalty if this target are not met.

Or the penalty is "cheap" aka worth it.

Aren’t the tax breaks often tied directly to the number of jobs? I believe that’s how the incentive structure for Amazon’s new NoVA headquarters works for instance.

If the company doesn’t create the job, the tax break is reduced by a certain percentage per job.

> Aren’t the tax breaks often tied directly to the number of jobs?

Usually, no, though the public justification for the tax breaks is usually in terms of expected jobs.

This is false, there are almost always specific job number requirements that must be met to be eligible for tax credits or the credits are tightly tied to hiring. The Joliet, IL Amazon warehouse incentive deal is a typical example. Amazon can credit employee withholding taxes against corporate income tax. If there are no employees hired, then there is no withholding tax, and no tax credit.

All of these situations are unique, but most of the time the benefits are awarded on a per job basis. So if the plan is to hire 2000 people, the reward will end up being divyied out depending on how many of the 2000 people are hired.

It shouldn't be a shock. There's overwhelming data supporting the idea that tax incentives for low paying jobs never end up paying for themselves.

Reply on news.ycombinator.com