The "Why" in Wage Segregation

[muon2]

Interesting indeed. It gives some rationale for this analysis I posted last year. I was looking at it much as the sources in the article looked at income growth compared to educational attainment. What is new in the article for me is the idea of sorting between firms, not just the rise of new jobs connected to high-skill professions where productivity and wages grew most.

It also reinforces what I've been telling colleagues since I looked at this data. The problem of income inequality is deeper than a problem of the 1% and tax code. It's a problem of an educated workforce in the information age. What I can now add to my explanation is that the information age itself as aided the employer in creating the income spread.

I'm cross posting this from US General Politics since it seems relevant here, too. This chart should help guide one's attempts at a solution. The drift of the upper three quintiles away from the lower two quintiles has been generally slow and steady over the last 35 years. The decade before Reagan isn't appreciably different than the decade after, despite considerable differences in national policy.

It seems clear that wealth has migrated towards the skilled professions, even in the middle class, as the global information age has progressed. The greater the skills required, the more rapid the increase in wages. To me that suggests the most effective changes would direct more resources towards education for the skills needed in the current economy, not large scale wealth redistribution or an investment in jobs in less skilled sectors from economies of the past.

Here's a better chart in response to Ernest. This is also from the historical household income data at the US Census. I found that the top of the second quintile (40%) was the most stable in real dollars, only increasing 5% from 1969 to 2014, so I used that to compare the other quintiles. The bottom quintile remained almost unchanged compared to the second quintile during that span of years and is very close to half the second quintile.

The growth is in the upper three quintiles. The middle quintile grew about 17% compared to the bottom two quintiles. Since the bottom two quintiles had little growth in real dollars, that 17% is close to the growth in real dollars since 1967.  The fourth quintile grew at 35% compared to the bottom two quintiles, or about double the rate of the middle. The limit for the upper 5% grew at 54% compared to the bottom two quintiles, or about triple the rate of the middle. My apologies for the year sequence which looked fine until the software rendered it to a bitmap.

[muon2]

I run into too many parents who say they want their kids to get a good career, so I tell them they have to push at home and at school for the basics of English and math. Not just as they were taught, but as those subjects are now applied in the information age. I then hear about how they weren't any good in those subjects, but they got a job anyway. The economic connection between the foundational skills in school and the 21st century workplace is often lost at the time when it is most needed.

[muon2]

This article is a nice little bit of nonsense. What they don't address is that in the mid-20th century, employers didn't price labor naively; labor really was relatively equal in terms of individual productivity. On an assembly line, one person might be able to install 8 widgets an hour while another person might only install 7, but there's not that much difference. In a world where everyone has a subscription to the Washington Post, the marginal productivity of the foreign service bureau really wasn't that much different than the domestic bureaus because everyone bought the same paper, and would have had it delivered no matter what. It's not like today where each individual article drives or fails to drive traffic based on hits. One writer may go viral and get 50k hits, while another struggles for a measly 500. That's a big difference between 8 versus 7 widgets per hour.

In other words, it's not the measurement of productivity that changed, it's the productivity itself that became more unequal.

Which is why things like improving education won't fix the problem (although education in underprivileged areas certainly needs to be improved). The "skilled labor" marketplace is inherently unequal and fails to provide the kind of mass stability and cohesion that the former unskilled labor marketplace did. It is, by its very nature, geared towards winners and losers, often on a large scale. It is not geared for social or individual stability.

My view is that it would be better to invest in higher wages and protections for unskilled service professions. Unionize the Wal-Marts and the McDonald's and the janitors of the nation, and have these jobs pay living wages. The brilliance of the mass manufacturing (or farm) labor was that even a D student could make a good living.

If this is so, shouldn't we be able to model the current labor situation on that from the 18th or 19th century? In the pre-assembly line era there were also larger differences in productivity, since more of manufacturing depended on personal ability. I would also suggest that there were fairly large differences in productivity in pre-mechanized agriculture. The rates of production between two crop pickers could be quite different, even a factor of two or more. That's much more than the 7 vs 8 widgets of the assembly line.

[muon2]

This article is a nice little bit of nonsense. What they don't address is that in the mid-20th century, employers didn't price labor naively; labor really was relatively equal in terms of individual productivity. On an assembly line, one person might be able to install 8 widgets an hour while another person might only install 7, but there's not that much difference. In a world where everyone has a subscription to the Washington Post, the marginal productivity of the foreign service bureau really wasn't that much different than the domestic bureaus because everyone bought the same paper, and would have had it delivered no matter what. It's not like today where each individual article drives or fails to drive traffic based on hits. One writer may go viral and get 50k hits, while another struggles for a measly 500. That's a big difference between 8 versus 7 widgets per hour.

In other words, it's not the measurement of productivity that changed, it's the productivity itself that became more unequal.

Which is why things like improving education won't fix the problem (although education in underprivileged areas certainly needs to be improved). The "skilled labor" marketplace is inherently unequal and fails to provide the kind of mass stability and cohesion that the former unskilled labor marketplace did. It is, by its very nature, geared towards winners and losers, often on a large scale. It is not geared for social or individual stability.

My view is that it would be better to invest in higher wages and protections for unskilled service professions. Unionize the Wal-Marts and the McDonald's and the janitors of the nation, and have these jobs pay living wages. The brilliance of the mass manufacturing (or farm) labor was that even a D student could make a good living.

If this is so, shouldn't we be able to model the current labor situation on that from the 18th or 19th century? In the pre-assembly line era there were also larger differences in productivity, since more of manufacturing depended on personal ability. I would also suggest that there were fairly large differences in productivity in pre-mechanized agriculture. The rates of production between two crop pickers could be quite different, even a factor of two or more. That's much more than the 7 vs 8 widgets of the assembly line.

How could two crop pickers working on the same farm have a very large differential? Two to one would seem to be an upper limit if comparing a young healthy man with an older woman or inexperienced child.

Either way, both farm and assembly line unskilled work had a much lower variation in productivity potential than skilled labor. Of course, capital could and did multiply unskilled productivity, but any machine taken advantage of by one unskilled worker can easily be given to all the others, so in the end there's no firm/farm/level reason to have huge differences in pay scales. With skilled work, the source of the productivity multiplier is with the worker him or herself...

That's why I was curious about how labor costs went in the pre-assembly line era when production used a lot of skilled labor.