[RealityVoid]

Well, in one sense, the Luddites were right, it did hurt THEIR job. Society, as a whole, was better off, but not them.

And growing automation means a growing number of people that are affected by it. Where is the tipping point where a great enough number of people can't produce competitive economic output and a relatively small percentage produce the most value? What happens then? Growing inequality can mean trouble for all, regardless who invests more effectively.

[Fantastic_T]

The Luddites were wrong in the sense that they and their children for the most part had far better jobs than they did before the power loom. Automation has always caused improvements in the occupations that exist.

[ska]

I'll try and dig up a cite later, but I've read analysis that this is just wrong. The "croppers" themselves mostly had very much worse jobs than before, as did their children and in many cases their grandchildren.

The multi generational aspect is perhaps surprising, but the first generation shouldn't be. Automation may create "better" jobs in some sense, but as a rule they aren't usually accessible to the people who are displaced by it.

[Fantastic_T]

Please do dig it up, because I find it highly unlikely that in an environment where the average wage increases by 20% over a generation, that the children of those who had been in one of the many occupations that had been made obsolete by automation were not better off than their fathers were at the best of times.

If it was a general rule that the children of those who lose their job to automation are worse off than their parents, the majority of the population would have seen its quality of life degrade over the last two hundred years given most occupations that existed at the beginning of the period have been made obsolete or mostly obsolete since that time.

[Someone]

Average wage can go up while national income goes down.

Even if national income goes up, too, changes in the distribution of income still may make things worse for large groups. If you pay your robot maintenance team of 10 each $100k a year to be able to fire 30 workers who used to make $30k a year, total wages go up, average wage goes up, but the group of workers feels worse. Even if they could do the new robot maintenance job, not all of them can get one. Moreover, they now know that others make $100k a year. That makes the $30k they can make if they are lucky enough to get hold of one of the remaining 'worker' jobs look less.

This more or less already has happened for blue collar workers. They make more money as individuals, but there are far fewer of them.

Yes, there are replacement jobs and, historically, many of them payed better, but that isn't a guarantee that that will keep happening.

For me, the big question is: will there be replacement jobs this time, and will they pay well enough? If not, taxing capital (aka a 'robot tax') for me seems the best solution.

So, what are those replacement jobs? I wouldn't know. On the other hand, I doubt people knew at the start of the industrial revolution, either.

[Fantastic_T]

>Average wage can go up while national income goes down.

Only sustained increase in national income can support sustained wage growth. Wage growth that comes out of increasing labour's share of national income is capped, at whatever capital's share of national income is. Once labour's gone from 60% to 100% of national income, no more growth can come from this source. Wage growth that comes out of growth in national income has no natural limit, since national income has no inherent cap.

What this means in practice is that wages will grow far more from growth in national income than from labour growing its share of national income.

In other words, growing the pie is a far more effective strategy for fostering wage growth than trying to increase labour's share of the pie.

A real world demonstration of this principle is the last 200 years of wage growth history: wages have grown something on the order of 20X since the early 1800s. That is something like 10X more than the maximum growth that would have been possible through only increasing labour's share of the pie.

If growing labour's of the pie comes at the expense of even a slight bit of long term average economic growth, then on a long enough time scale, workers will be worse off, as they lose out on a recurring exponential boost for a one-time boost. The former is guaranteed to exceed the latter over the long run.

>If you pay your robot maintenance team of 10 each $100k a year to be able to fire 30 workers who used to make $30k a year, total wages go up, average wage goes up, but the group of workers feels worse.

It's important to understand that automation does not only help highly skilled workers. Mechanisation/automation make complex work more simple to do, and makes expensive equipment cheaper to buy. This means things like starting an advanced business become far easier and more accessible to the masses.

Historically, we've seen increases in productivity, which is mostly a result of increasing automation, be associated with increases in not just average wages, but median wages as well. The typical person, meaning the masses, are better off from automation, and any policy that slows the rate of automation, like taxing those who invest in automation, harms the masses.

[ska]

See my other message about the problems of looking at population averages for something like this.