I've just written an article (waiting for publication) lamenting the same thing from almost the exact opposite angle: Lots of jobs, no money.
Here in Scotland, the government has just given £10.6m in 'support' to Amazon to encourage them to open a distribution centre (actually they already had two, but anyway...) The Scottish government is trumpeting this as a recognition of Scotland's "highly skilled workforce". Actually it's 750 jobs (gov't estimate) mostly on 1% above the legal minimum wage, with no real prospects.
And to cap it all, Amazon don't even pay tax in the UK. All their sales go through a Luxembourg company. So it seems ridiculous that the taxpayer is subsidising the operations of a highly profitable, tax-avoiding company. Effectively we're buying jobs.
Meanwhile, Borders is bankrupt (loss of 1,200 jobs), and other booksellers (Waterstones paid £2m in corporation tax last year) are struggling.
Relating it back to this post - highly skilled work requires less people. Low skilled work requires less people AND pays less money. But it's progress, and I fear it's simply an economic reality. I don't know what the solution is.
Note: I don't want to litter this post with climbdowns, moderations etc. I'm not disparaging Amazon jobs, and 750 people working instead of unemployed is good. But we can't win in a race to the lowest paid unskilled labour. And Amazon's labour can't even be offshored anyway, so who are we bidding against?
"not many jobs"? in a town of 3400 people, the new datacenter is expected to create 50 direct jobs and 150 contracting/maintenance positions. if we say the workforce comprises half the population, that means 10% of the jobs in town will be datacenter related. that sounds pretty good to me.
apple isn't going to magically turn your town from podunk, nowhere into some big metropolis.
They contracted someone to build the thing in the first place. Whoever is there for the data centre will be renting or buying a place to live, or at least staying at hotels, shopping for food and going to restaurants.
It may not solve all of the town's problems but it probably does not do any harm either.
You're right. But the problem is bigger than this. It's that the data centre eliminates a far greater number of jobs than it creates. Somewhere else in the US, there's a CD pressing factory which has just closed its doors with 150 jobs lost. And 10 retailers who sold those CDs. And the company that printed Road Atlases. It's an inevitability, but one that is overlooked by saying, "Well it's progress, we're creating skilled jobs in the businesses that are doing the displacing. Let's retool our workforce."
The Sony CD plant in Terre Haute, IN employs ~ 1,400 [1]. I know they have other plants, but I think that is it for the Eastern US. And, it's pretty much the only company in the country that produces physical media, so other CD pressing plants are pretty rare.
If they're counting jobs at Apple HQ, there are FAR more than 50 jobs associated with this data center.
They're talking about the people who report to work in the actual data center building. And that's an incredibly stupid way to measure the job creation of a cloud data center, given that its entire purpose in life is to allow remote access of its information.
True, but there are knock-on effects of creating 50 new jobs in a town of 3400. These 50 people now have money to spend locally which could result in the creation of many more low wage jobs to service the new business.
Yeah, but then, the town is also now on the hook for police and fire service (and probably a variety of other non-obvious costs) for a huge datacenter, though the company is paying discounted taxes...
Jobs in software related industries have been stagnant or in decline in the US for the last decade, and traditional narratives about the ability of workers from older industries to upgrade their skills and become "knowledge workers" seem to no longer be true.
Technology is about making things easier. Computers, in particular, are about automation. We're finally starting to figure out this automation thing in operations. Cool.
If you're not figuring out how to make you redundant the guy beside you, then the guy beside you is trying to figure out how to make you redundant.
It is incredibly short sighted to say that the sum of the jobs created by a data center is the people reporting to work in the building. This data center likely created hundreds, if not thousands of jobs. It's just that most of them aren't in hicksville, NC.
There are hardware engineers designing the facility, software engineers maintaining the software infrastructure, product managers scheming of ways to use Apple's cloud facilities, etc.
> This data center likely created hundreds, if not thousands of jobs. It's just that most of them aren't in hicksville, NC.
Well, yes. But it's worth taking note of the point that Governments subsidizing the creation of data centers in places like hicksville, NC with the idea that the jobs they create will become a huge boon to the local hicksville economy is fundamentally mistaken.
Aside from a skeleton staff to main the facility, most of the work is, as you say, create for people who will never need to set foot in North Carolina. The local jobs aren't bad for Hicksville, but they don't justify the huge subsidies, and they aren't going to ever save hicksville from economic irrelevancy.
Absolutely. But it appears that the fault falls clearly on the NC state government, which according to the article debated for less than a minute before awarding Apple $46 million in tax breaks.
The irony is that Apple basically used Republicans' blind allegiance to Job Creators to get huge tax breaks that will allow them to hire hundreds of people in the Socialist tax-happy hellhole of California, where most of the best knowledge workers live.
That quote seems more like political-speak than actual hard numbers. The article also said that the state government gave Apple $46 million in tax breaks, while local government cut their property taxes by 50 percent and personal taxes by 85 percent. I'm hard-pressed to see how the taxpayers aren't subsidizing this datacenter.
Money is fungible; tax breaks are no different in function than charging the full tax rate but handing Apple a pile of cash equivalent to what they'd have saved with the tax break.
It's a subsidy. The argument that they made more with the subsidy than they would have without it is separate from the idea that tax breaks don't amount to a subsidy (which was yummyfajitas assertion).
I keep seeing new articles in this type of tone. Like they know what's wrong but are afraid to say what they want for fear it won't be published. What this article is describing is capitalism working effectively, maybe too effictively to be sustainable. Capitalsim is all about using technology to produce profits more efficiently. By building a data center in a small town Apple is efficiently using its techology to produce profits. This isn't something to be complaining to Apple for doing.
Technology's effictiveness is outstripping captialisms mechanism for job creation. The unemployment problem probably cannot be fixed without drastic socio-economic reform.
In such an economy, unemployment isn't the problem - resource allocation is. With decreasing numbers of jobs, the old "get a job to get money to get food" solution breaks down. An alternative solution could be for the government to provide a base level of necessities to the entire population. For example, it is within our current technology level (the US overproduces food, and various farms lay unused) to provide a 2000 calorie per day ration to our entire population. By replacing minimum wage laws and welfare checks, which simply drive inflation, with explicit grants for necessities (food, shelter, clothing) we can maintain higher levels of unemployment without perverting the capitalism-based incentive structures in other areas of our economy. That is, we want people to be motivated to better themselves by the prospect of those traditional capitalist rewards, and we don't want a solution which simply gives raw money away because a percentage, perhaps a significant percentage, will be misspent on luxuries rather than necessities and would likely reduce motivation to seek gainful employment.
TLDR: A two-tier economy based on enabling luxuries for those who work and providing a ration of necessities for all members of the society, scaled by our technology level, will enable us to sustain the higher levels of unemployment driven by technological advances.
> An alternative solution could be for the government to provide a base level of necessities to the entire population.
I like this idea and the idea of the post-scarcity society in general, but one thing I've never seen explained or examined is how this is actually accomplished. How is the government producing these necessities and distributing them? Could you elaborate further?
Today, we have a government that distributes free money to people by spending money taken from other people -- social security moves income from the working class to the elderly, medicare moves income to the impoverished. These systems move money from haves to have nots, but money isn't the best thing to move because it's too indirect, too flexible, too fluid. The owner of money can buy food to feed their children or toys to feed their appetite for entertainment. Money can be used to acquire many of Manfred Max-Neef's fundamental human needs.
That's a problem when you have a society built on capitalism where the reward is supposed to be attained by hard work and ingenuity. Welfare, social security checks, tax credits, bailouts, and minimum wage laws all share one thing in common: the creation of a moral hazard by taking money from one group and giving it to another. The moral hazard, of course, is reduced motivation to perform work, be creative, etc., and to instead become dependent on the money from the government.
We see this parasitic behavior in established corporations and in multiple generation welfare recipients. We see it when the government allows itself to take advances from pension accounts or other "lock boxes".
Ok, but that's not your question, you asked, "How is the government producing these necessities and distributing them?"
Well, let's split your question up into three parts:
1. What is the government doing today?
2. How can the government, which produces nothing except laws enforced by violence, acquire these necessities?
3. How can these items be distributed equitably?
Today our governments are wasting mind-blowing quantities of money on social programs and bailouts which transfer raw money to new owners and in the process create moral hazards which are assuaged by continuing transfers of money. The government, today, attempts to satisfy the needs of the impoverished and unemployed through programs like Section 8 rent-controlled housing, unemployment insurance, food stamps, and I'm sure something exists for clothing.
Now, I suppose that if the Economist sat down and figured out how much money we've seized and reallocated it would be more than enough to buy each citizen a ration of N calories per day per year, for their lifetime. Now, that statement is true for some N (0 is a solution). Similar statements exist for clothing: N issues of outfits of some standard quality per person per year, as well as for shelter: N months of rent at some standardized housing quality.
If we can define national minimum standard values for Nfood (e.g. 2000 calories per day), Nclothing (e.g. 3 outfits per year), and Nshelter (e.g. 12 months of 400sqft per person) and allow these values to scale by Real GDP/person so that improvements in the economy lift everyone's living standard and contraction reduce standards, then the government can simply pass laws enabling citizens to claim these perks from authorized vendors.
These vendors must be compensated for providing these goods, so let's let them participate in a dutch auction for each geographic statistical area. Competition among vendors drives down prices for the government who then pays winners with money reallocated from existing social and corporate welfare programs. A standardized dutch auction process and participation requirements would need to be implemented and audited to protect against quality shortcuts, graft, and corruption.
To recap: (1) the government is already spending all of the money necessary to do this, poorly; (2) the government can pass laws allowing citizens to claim their rations of food, shelter, and clothing from authorized vendors who themselves participate in an audited dutch auction (instead of no-bid contracts and similar corruption).
The last part, equitable distribution, that is the hard part. Encouraging the states to pass uniform laws isn't hard -- just attach money as a carrot and use the 14th amendment as a stick. Obtaining the resources isn't hard -- we're already spending them.
Equitable distribution is hard because you need to decide "who gets" and "have they been given enough".
In my opinion, the only defensible answer to the first decision is everyone gets access to the current rations of food, shelter, and clothing. There are two groups of people in the first decision: those paying into the system and those not paying into the system due to calamity. Clearly, the system should payout to the needy. But deciding who is needy is hard and fraught with potential corruption and complexity. So just simplify and allow everyone a claim. The truly non-needy will simply forgo their claim.
Up to this point, the changes necessary are straightforward and workable within our current legislative environment. However, our answer to the second decision will require a significant change in the national mood. Specially, we will need to track how much each person has received to reduce the impact of the inevitable criminal element that shows up whenever price controls are implemented -- and let's be clear, goods sold to the government for free distribution to the people are a form of price control (hence the strict limits on what is given out: only food, clothing, and shelter, only at known grades, only at fixed quantities, and only to authorized people).
There are 3 possibilities:
a) We trust everyone not to double-spend their claims. Let's all laugh for a moment.
b) We implement a national centralized bio-metric database to track claims. It will likely be derided as Mark of the Beast stuff.
c) We implement a national distributed crypto-currency for claim spends using wallets distributed by the social security administration on a per-SSN basis. Foodcoin. HN revolts.
If our society can solve the last part, the rest sort of falls into place. By providing for the bare necessities freely, we create a system with far fewer moral hazards and without undermining capitalism in the other areas of our economy. Our population can choose to live with the bare necessities, or choose to attain higher rewards via capitalism. We already have the resources, we're just not allocating them appropriately for a society with high levels of unemployment.
This is getting pretty long but one last point -- cultures with food safety and more education have lower birth rates. Lower birth rates are desirable as unbounded growth would quickly undermine my proposal, so clearly we would also want to provide minimum claims to education. Given that the future is a computational knowledge economy where continuous learning is as critical as regular access to food, it makes sense to offer continuing education in grants of Neducation, but I'm not sure how to standardize quality.
I'm happy to hear your thoughts on these ideas, or suggestions to improve them.
I absolutely agree. And it's all well and good talking about startups creating jobs, but really, what YC startups have created any meaningful number of jobs (aside from highly paid developers (Dropbox, Heroku) or sales staff (Airbnb)? I'm not taking away from their success in any way, I just think we need to recognise that there's a fundamental problem on the horizon. I suspect the solution lies in fewer people working and contributing a greater proportion of their income. Obviously that's unpalatable to everybody.
Do you remember that article from Mark Andreesen that "Software is eating everything"? It's a reality in the West that is often neglected and may lead to major changes in the (industrial-age) model of employment. There are other sectors with similar problems: Green energy, where people assume more jobs will be created although its obvious that it actually requires less human effort.
I absolutely agree. And it's all well and good talking about startups creating jobs, but really, what YC startups have created any meaningful number of jobs
What's a meaningful number of jobs? Creating jobs is only worth doing if it creates 10 jobs? 100 jobs? 1000 jobs? 10,000 jobs? None of these jobs can be "created" overnight, it requires years of building a company to get to the point where you can employ 10,000 people, much less employ an additional 10,000 people.
The US civilian labor force is 154.2 million[1]. At 9% unemployment (13,878,000 of the total labor force), 10,000 jobs reduces the number of unemployed people by 7/100ths of a percent, so even that seems like it's not good enough.
The point being that every little bit helps, and we're going to have an easier time getting people employed in handfuls than trying to get massive companies (of which there are fewer, over all) employing an additional 10,000 or 100,000 people -- nevermind that an economy that has a large number of smaller employers is going to be stronger and more resilient than a small number of massive employers. A business that employs 4 people is better than that business not existing at all.
It's true that new companies create less jobs than they used to. Previously they used to create 7.5 jobs/company, today only 4.5[1]. that means 2.3 million jobs instead of 3 million.
Every time I see articles like this -- pointing out how large full-time admin and operations teams are now replaced with smaller teams and easy-to-manage infrastructure -- I wonder if something similar will happen to software engineering.
In fact, I am sure that something like this must eventually happen to engineering. The cost involved in designing and developing new services will be reduced to only that which is essential. I just dont know what form it will take.
Will newer, easy-to-use libraries reduce the work of five engineers to one? Will basic application development knowledge become so ubiquitous that engineers are only needed for the harder, more obscure challenges? Will all of engineering be reduced to only the most creative work, with all the logic and plumbing provided automagically by intelligent programming environments?
Devs are used to learning new technologies regularly. Libraries that reduce the work of 5 people to the work of one are common place. Before RDBMSs, database queries had to be manually programmed on a case-by-case basis, including disk access etc. When one bottleneck has been solved, devs simply move on to other, more awesome tasks.
Unfortunately most of the population does not share this constant learning philosophy.
Yes, we all yearn for the days when cloud computing data centers were run by thousands and thousands of operations engineers. Sadly, those times are no longer here - mostly because they're a figment of your imagination.
But don't let that stop you from assuming for no good reason that the software industry is about to die. People have been making that assumption for 10 years or so now, so you have plenty of good company.
I didn't say that a single data center employed thousands of engineers, or that the software industry is dying. But don't let that stop you from having an imaginary argument.
One thing is to deeply understand the business domain and how to work with business domain people. That always helps.and those skills seems much less prone to outsourcing and automation.
There's two different ways of looking at this sort of issue though.
One is from the perspective of a participant in the job market, from that point of view it isn't such a huge problem. Predicting where efficiency improvements will shift factor intensity from labour to capital is something that most people here are able to do for their own field. I'm sure, for instance, that every system administrator reading this has already taken into account the effect of cloud services on their career. Smart people will build their expertise in managing cloud systems, move into a tighter integration with developers (DevOps), or specialise in an area that is more resistant to being moved into the cloud for performance and security reasons.
That's great, but it's a little like developing an optimal musical chairs strategy, if there are 10 such jobs today and 9 tomorrow then one person isn't going to have a job. Even if he has also tried to adopt an optimal strategy.
That brings me to the second view, which is viewing the entire system from the outside. That can be in the role of an economist, a citizen, whatever. From that macro point of view, what is important is only the following: Will the net job losses in this particular job eventually be outweighed by countervailing job gains in other parts of the economy?
Historically, that has always been the case. Ever since this became a topic of discussion in the early industrial revolution it has happened time and time again that new jobs pop up that no-one could have imagined a generation ago. The question is though, is that inevitable? Will there come a point where job losses from efficiency will outpace the gain in new jobs? I have no idea what the answer to this is.
If it were to come to pass that huge swathes of the population simply don't have any valuable labour to exchange, then the basis of an economy and society based on the exchange of labour for resources would unravel.
If all but 20% of programming jobs will become redundant, what advice would you give to the remaining 80% of programmers? "Starve because there's too many of you to fit in the 20%"?
Around here in Oregon, I've seen a fair amount of similar stuff over a few decades. A big company moves into the area. Everyone is thrilled to be building the new facilities. Then, a few locals are hired, a few skilled employees are brought in from their other areas, and everyone else is left scratching their heads.
After a few years, the municipal tax break comes to an end. The big company needs new technology anyway, so it's cheaper just to build a new facility in a new town with new tax breaks, than it is to upgrade the existing one.
Here in Scotland, the government has just given £10.6m in 'support' to Amazon to encourage them to open a distribution centre (actually they already had two, but anyway...) The Scottish government is trumpeting this as a recognition of Scotland's "highly skilled workforce". Actually it's 750 jobs (gov't estimate) mostly on 1% above the legal minimum wage, with no real prospects.
And to cap it all, Amazon don't even pay tax in the UK. All their sales go through a Luxembourg company. So it seems ridiculous that the taxpayer is subsidising the operations of a highly profitable, tax-avoiding company. Effectively we're buying jobs.
Meanwhile, Borders is bankrupt (loss of 1,200 jobs), and other booksellers (Waterstones paid £2m in corporation tax last year) are struggling.
Relating it back to this post - highly skilled work requires less people. Low skilled work requires less people AND pays less money. But it's progress, and I fear it's simply an economic reality. I don't know what the solution is.
Note: I don't want to litter this post with climbdowns, moderations etc. I'm not disparaging Amazon jobs, and 750 people working instead of unemployed is good. But we can't win in a race to the lowest paid unskilled labour. And Amazon's labour can't even be offshored anyway, so who are we bidding against?