http://www.epi.org/publication/technology-inequality-dont-blame-the-robots/ Okay, so this is really interesting. It's basically a comprehensive rebuttal to about 30 years of economic papers which have posited that increasing automation directly correlates with a rise in wage inequality. More robots, fewer monotonous middle-class jobs.
It's long and heavily academic (though still perfectly readable) with many references, so I figured I'd offer a tl;dr version.
First off, here's their section titles, which actually serve as a good summary of the points:
1. Technological and skill deficiency explanations of wage inequality have failed to explain key wage
patterns over the last three decades, including the 2000s.
2. History shows that middle-wage occupations have shrunk and higher-wage occupations have expanded
since the 1950s. This has not driven any changed pattern of wage trends.
3. Evidence for job polarization is weak.
4. There was no occupational job polarization in the 2000s.
5. Occupational employment trends do not drive wage patterns or wage inequality.
6. Occupations have become less, not more, important determinants of wage patterns.
7. An expanded demand for low-wage service occupations is not a key driver of wage trends.
8. Occupational employment trends provide only limited insights into the main dynamics of the labor
market, particularly wage trends.
The key point they're making is that apparently there's little link between JOB polarization (demands for higher-skilled and lower skilled work at the expense of middling skill but easily automated work) and WAGE polarization (income inequality).
Job polarization is apparently a longstanding issue, going back as far as they have data and is something that makes sense in terms of technology gradually changing the nature work over time, but it's a slow phenomenon and has actually been dropping - there was little job polarization in the 2000's. Interestingly, they also posit that yes, educational requirements for the workforce as a whole have risen, but that increased education has largely kept pace with technological demands for higher skilled workers.
For instance, they quoted figures that the service industry is only 13% of the current workforce and this is growth of only 2% since the mid-late 70's. The growth in skilled high-tech demand is also less than expected. An example is given of retail publishing moving to e-books and internet sales. You might add a few techs, but mostly you just go from low-skill bookstore workers to low-skill warehouse workers and truck drivers (I think that IRL, you would still lose some overall employement in that scenario but it was something of a throwaway example).
They also talk about "occupational" job polarization. This is job polarization that specifically relates to people being in the "wrong field" (due to industries becoming obsolete or highly automated) and claim that this has much less correlation and impact than is supposed.
Wage polarization on the other hand, has only been seriously happening in the past three decades. Even when both types of polarization have been happening, the measureable trend curves of the two phenomenon do not synch well at all, which seriously undermines any claims of direct correlation.
Essentially, correlation is not causation.
The one thing I found partially lacking was an overall picture of automation on the labour market as a whole. Going back to the idea of the bookstore being automated, in that scenario you may lose a small number of jobs. A small erosion across ALL sectors may result in an overall surplus of labour, which exerts downward pressure on most, if not all wages.
They even reference "High unemployment" as one cause of wage polarization without going into alternative explanations for why there's an ever-growing surfeit of labour in all fields. All you need is a few super-rich guys to get big bonuses for layoffs and then it looks like you have U-shaped polarization, when in fact the link is indirect rather than direct and deals with a very small sample size (and high-end wage growth is certainly heavily affected by other factors, such as globalization).
They do sort of address that when they show that the kind of occupation you're in has had a declining relevance to wages over time, albeit that's in a roundabout way. So maybe it's a non-issue after all. But I would have liked a bit more clarity there, I think.
Some other numbers on inequality would have also been good. It has after all been commented on that even even most of the supposed 1% aren't actually SUPER CRAZY rich. It's that .01% or even that .001% that are really off the chart. Could it be that wages for just about everyone are falling and that the super-rich are accruing such disproportionate wealth as to make the income distribution curve look U-shaped when really it's more like a backwards L?
Other than that, their criticisms of previous studies DO seem valid and are quite well-researched, but what I can't say is whether their opponents are right but just did sloppy, superficial research to prove their argument, or are significantly wrong. Or we may be dealing with a messay array of hybrid reasons. I.E. That increasing automation IS still responsible for income inequality, but only in an indirect way, and only in combination with several other large factors.
They don't really suggest an alternative explanation for the rise in inequality, but I think that's normal for this sort of pure academic study - they're focusing exclusively on debunking a long-held correlation and no more than that.
They don't try to advance a competing claim and are only "fact-checking" actually makes me trust their findings more (if perhaps leaving me a bit unsatisfied). They do state that in future they're going to look at trade gaps and globalization and their effect on wages as a possible alternate explanation. I'd really like to see that when it's published.