Hacker News Reader

>The market for most types of human labor is soft in advanced economies, and getting softer. We don't grow the economy by assembling a huge workforce to do/build something, we grow it when (relatively) tiny teams build a new tool, create a new market or disrupt an existing industry.

As hip as this statement sounds, what quantitative evidence do we have for it? I'd blame financialization before I'd blame computerization, simply because high-tech products and services simply don't account for that large a portion of GDP.

On the other hand, the mechanization thesis does seem to suit something like what happened in manufacturing: output stayed the same or went up, but employment dropped consistently or was shifted to cheap-labor countries. There, at least, experts have examined numbers.

But again, the other big set of examined numbers is about financialization, which has showed that as the portion of the economy devoted to finance went up, and the "output" of the financial sector went up, the productivity and output of the non-asset economy stagnated.

I love the scare quotes on "output" because that's exactly it - the paper wealth that's escalating is very ephemeral.

Until you know exactly how Google is a real company, it doesn't much look like one. And my experience is that explaining this can be a challenge.

But, say, a textile worker in the 1950s had about as much marginal production as they do now. So the textile work goes to where that's good enough - for now. Eventually, it gets automated.

I think what's really decoupling is price and value-as-utility. My Dad, born in the 1930s, had three shirts, two pairs of pants and at most two pairs of shoes until he went in the the military. That wasn't unusual. Now, I can buy a decent shirt for what I make in minutes.

I think it was Megan McCardle who noticed that in the "Little House on the Prairie" books, they could not afford a third tin cup for Carrie, the baby. Yet the Ingalls were middle class for their time and place. That time was right before the big productivity boom in farming - from mechanization.

Yeah, I think of the financialisation of society in the 90s & 00s as being a massive share dilution of the middle class.

the wealth of the economy as a whole didn't grow much, but the number of shares in that economy did and were taken by one sector - massively devaluing everyone else's share.

Not just computerization but also:

1) Women entering the workplace. This increased the supply of workers without increasing demand, which drives down prices.

2) Global trade. Post WWI it was the US >> Western Europe >> Everyone else. By the 1980's Japan was resurgent and the third world started opening up factories. Unskilled factory workers had been solidly middle class, but now they have to compete with someone in the third world making dollars a day.

3) Growth of financial sector that doesn't employ nearly as many people as the GDP it creates.

> 1) Women entering the workplace. This increased the supply of workers without increasing demand, which drives down prices.

Women earning salaries led to higher consumption.

Both can be true. Consider that the majority of employees produce more value than they personally consume.

Lol at how nobody ever mentions immigration. Total labor hours haven't increased since 2000 but we've imported tens of millions of people. I'm not an economist but I think it's obvious that the average person would be having an easier time getting paid without the added population.

>> Note that the decoupling correlates closely with the computerization of society/business around the mid-1980s.

It also correlates closely with the decline of labor unions.

Agreed. In the 80's Reagan fired the striking air traffic controllers. It was the start of society's war against unions.

The US was very strange at the time in that most people viewed themselves as upper middle class. Of course they weren't but the idea let many people look down on union members as an undeserving underclass. That in turn was used by capital owners to create support for legislation outlawing activities by upstart union members against the more deserving mid and upper classes.

Some part of society has been at "war" against unions since their beginnings. To call the Reagan air traffic controller strike the start is bizarre.

To call the Reagan air traffic controller strike the start is bizarre.

Not at all. Bizarre is the need to immediately jump to such invective.

The US government deployed armed federal forces to literally kill union strikers at privately held companies nearly 80 years before Reagan used his executive powers to fire federal employees.

I stand by the phrase bizarre.

After WW II unions became a pathway for millions to enter the middle class. It was Reagan who changed that back to the pre-war situation.

Speaking with disdain doesn't make your position any stronger or more correct.

I didn't disdain your position. I thought it was odd, strange, maybe to the point of amusing as it seemed to use a particular word "war" that was a much more apt description of the state of societies interaction with unions many years before your supposed "start".

You've already revised your position, to one that I think is still overly broad and simplistic, but one that I think is common and thus not bizarre.

If I were to describe my thoughts on the particular point in history its that Reagan's action was an effect of the decline of labor unions, that occurred because in many ways they had accomplished a lot of their stated and most approachable goals. The only reason he could get away with it was because the unions had lost so much of their sway. They didn't lose their sway because he fired striking workers.

You also see that it was not confined to the US as Thatcher was doing similar things in the UK.

From an outside observer to this conversation, you (forgotAgain) are speaking with much more disdain than kasey_junk is.

The decline of labor unions are the symptom. Unionized jobs were exported now that the world was safe for global trade.

Bingo. The deal the unions negotiated in Detroit could only work when there was nowhere else to get cars. It especially couldn't work when Japanese cars were both cheaper and better.

Hi. I'm a software engineer. Show me a union whose membership improves my life, and I'll be happy to talk.

Bear in mind that improves is present-tense for a reason.

> Hi. I'm a software engineer. Show me a union whose membership improves my life, and I'll be happy to talk.

Want to join the Industrial Workers of the World?

You seem to have missed the the "show me" part.

In what way in joining the Wobblies going to improve my life?

We don't need a union per se, but a professional association (aka guild) would be fantastic. As professionals we need to stop letting employers be the sole arbiter of competency, and we also need to institute a code of engineering ethics that is independent and based on sound theory and practice (rather than whatever the business wants at any given time).

Imagine being SEOA (Software Engineer Association of America) Level 1 certified and never having to do Fizz Buzz at an interview again. Or a Glassdoor-like employer database that gives you a clue what you might be walking into. Independent health insurance coverage, etc. There could be lots of great ways it would improve the average engineer's life.

Also imagine not being able to get a job because you tried to implement some new and cutting edge process or program and were black balled for life.

That's also what guilds do.

I shudder in horror. While I personally am your average classically trained CSSE grad, many of my friends and family are autodidacts who created for themselves very successful careers in IT and software engineering. I cannot see how a guild would help them, and frankly the deplorable state of modern IT certifications plus the similarly large pool of friends and family who do _and_ don't test well would make me continue to Fizz Buzz (and more) potential recruits no matter who says they've got programming chops.

The ACM already has a reasonably sound code of ethics.

Why are tech writers and programmers treated differently in the law? Recruiter collective lobbying: http://andreas.com/faq-overtime.html

"The recruiters accepted our exemptions because they were going after bigger fish: the engineers. The computer engineers who work as W-2 contractors are generally earning $100-300/hr and often work 60-80 hours per week. The recruiters were afraid that if such workers were entitled to overtime pay, the companies may cut overtime work, and therefore the recruiters would lose their 30-50% share of that money (this can be much as $6,000 per week to a recruiter for a single worker.)

Engineers have always refused to organize or even to be aware of their interests: they think recruiters are their friends. One engineer said to me: "Engineers think they're so smart that no one could do such a thing to them. Wow. They got really screwed." The law is an annual loss of as much as $50-75,000 dollars per each engineer. Yep, it's legal. The recuiters wrote a law to take away their money. Silicon Valley engineers were plundered by their "friends"."

Since the US has essentially no social safety net to speak of, the primary source of welfare ends up being workplace injury-related benefits. This produces a powerful secondary-gain situation that incentivizes injury for those who work physical jobs.

I don't think people intentionally get hurt of course, but rather when they do sustain an injury it becomes economically advantageous to prolong it as much as possible since the job is hard and their pay for working is so low. Usually it's the only way for menial laborers to get a paid vacation, and (sadly) getting permanent disability benefits is like early retirement.

Well put. This was a topic explored in detail by This American Life (with Planet Money) in the episode Trends With Benefits:

http://www.thisamericanlife.org/radio-archives/episode/490/t...

You've articulated nicely the conclusion I drew listening there (and, well, I'm sure they articulated pretty well themselves.)

> Note that the decoupling correlates closely with the computerization of society/business around the mid-1980s.

Something I have only recently been realizing is just how much of a business transaction is not automated yet. There are still many, many jobs that involve updating Excel spreadsheets.

I think people tend to equate "manual labor" with physical labor. There are a ton of business/cubeville jobs that will get automated in the next 20 years.

It will take that long because businesses can still make money with manual business processes, they just can't scale. And it takes a special kind of business owner to care, to want to grow beyond a successful business to a billion dollar enterprise.

It also correlates closely with the advent of smart mechanized labor (mostly dumb robots), which of course requires computers, but I think more of a problem. You need a fraction of the number of people to put together a car as you needed in the 70's.

I would argue it's not unions, it's not just computers, it's not globalization, it's not physical jobs being expensive to hire in the US, it's that you just need fewer people to do the same stuff period.

Across the economy you have examples where low to medium skilled labor is being replaced. First it was manufacturing, next it will be low skilled white collar and so on. Where do those people with low to medium skill jobs go?

The fundamental question is whether the information tech revolution is going to have the same effect as the industrial revolution, where eventually everyone ends up better off, and we create broad prosperity. And even if we can build that society, change is happening pretty fast so we may not have the time to create it.

It's pretty easy to see a dystopic future where you've only got high education/skill engineering/AI/whatever people at the top, and a large mass of low to medium skilled labor that have been entirely displaced by the inventions of the former. It's not clear how we avoid that, and I'm not convinced the solutions the authors advance are effective.

"I don't think people intentionally get hurt of course, but rather when they do sustain an injury it becomes economically advantageous to prolong it as much as possible since the job is hard and their pay for working is so low."

I have a friend who is a private investigator. His typical day is following a person like this around, filming from a distance, eventually gathering up enough evidence to prove the fraud. Business is booming apparently.

We're both kind of guessing :) but I'd suspect financialization before I suspected computerization. This is total observer bias, but what I have seen is some productive asset get sold and resold, and the accumulated debt strangles it. Computerization took the form of sheet metal shaping, CNC, order management and mainly tended to reduce error more than it replaced skilled labor.

We also had the big China buildout which idled a lot of factory labor. I'd be willing to expect that to reverse course over the next decade.