Inequality

[The Minsky Moment]

PART I

Standard economic theory tells us the the price of labor in a free market economy should be equal to its marginal product.  The intuition behind it is as follows: let's say Firm A decides to keep wages lower then the value of the marginal product that results from that labor -- then Firm B can make money luring away workers from Firm A by paying a little more and reaping the benefit of the marginal product.  That competitive process should continue until equality is reached.

That's the theory, what is the reality?

Let's say you look back historically and plot labor productivity (output per worker) vs. worker compensation.  Economic theory suggests that over time, these should move in tandem.  And in fact, in the US, from 1947-1982, that is exactly what one sees.  If you graph it out (as can be found in the third chart here: http://economistsview.typepad.com/timduy/2012/04/distributional-impacts-of-monetary-policy.html) the two lines move together.  But after 1982, the two lines diverge: labor productivity rockets up but worker compensation stagnates.  Worker wages cease to reflect the productive value the workers provide and the gap has steadily increased over time.  Put another way, while worker productivity in the US has almost doubled since 1973, real median wages over the same period increased only about 4%. 

That's the problem - it has received extensive coverage and attention - the question is why?

[Josephus]

I think this thread needs naked women.

[DGuller]

Should labor productivity really drive the wages?  Let's say I own a farm, and have 100 migrant workers (all of them legal, as far as other interested parties are concerned) collect the harvest.  Then, one day, I take out a loan to buy a John Deere tractor, and hire one guy to operate it and do the job of the 100 legal migrant workers. 

That one guy is 100 times as productive compared to migrant workers, but why should he be paid 100 times as much?  If I pay him 100 times as much, then I'm worse off after buying the tractor, because now I have to find some additional money to pay for the interest on the loan that I took out to buy the tractor.

[Jacob]

I think this thread needs naked women.

Maybe later.

But first, let's hear what Joan has to say. Some of us asked him for this.

[PDH]

I think this thread needs naked women.

potm

[Syt]

DG is onto something here, I think. I seem to recall from my childhood that in the early to mid 80s automatization of production was a very hot topic in Germany (i.e. substituting labor with capital). This would have increased productivity enormously. With wages not increasing as much, production costs should go down, i.e. the goods can be brought to the market for cheaper (at least as far as labor is involved - not talking raw materials) - and they retain affordable for people whose income didn't go up as quickly as productivity.

Which could lead into a consumerist society which we of course don't have.

Or something. I'm not an economist.

[crazy canuck]

Should labor productivity really drive the wages?  Let's say I own a farm, and have 100 migrant workers (all of them legal, as far as other interested parties are concerned) collect the harvest.  Then, one day, I take out a loan to buy a John Deere tractor, and hire one guy to operate it and do the job of the 100 legal migrant workers. 

That one guy is 100 times as productive compared to migrant workers, but why should he be paid 100 times as much?  If I pay him 100 times as much, then I'm worse off after buying the tractor, because now I have to find some additional money to pay for the interest on the loan that I took out to buy the tractor.

JR is talking about median wages of people that are employed.  I dont think the figure takes into account the people that are not employed.

Second, the standard theory, as explained by JR, suggests that if you kept the sole remaining worker at the same salary a competitor would lure him away with slightly higher wages (there is a lot of room for an increase given the greater productivity) and you will be stuck with the tractor and no one to drive it and so you will also need to raise your wages accordingly until some kind of equilibrium is reached.


JR, awaiting party 2. 

[alfred russel]

Should labor productivity really drive the wages?  Let's say I own a farm, and have 100 migrant workers (all of them legal, as far as other interested parties are concerned) collect the harvest.  Then, one day, I take out a loan to buy a John Deere tractor, and hire one guy to operate it and do the job of the 100 legal migrant workers. 

That one guy is 100 times as productive compared to migrant workers, but why should he be paid 100 times as much?  If I pay him 100 times as much, then I'm worse off after buying the tractor, because now I have to find some additional money to pay for the interest on the loan that I took out to buy the tractor.

If I buy the tractor, and the farmers around me buy the tractor, I'm going to think about slashing the wages of the guy I hire. After all, there are now 100 farm workers for every job.

[DGuller]

JR is talking about median wages of people that are employed.  I dont think the figure takes into account the people that are not employed.

Second, the standard theory, as explained by JR, suggests that if you kept the sole remaining worker at the same salary a competitor would lure him away with slightly higher wages (there is a lot of room for an increase given the greater productivity) and you will be stuck with the tractor and no one to drive it and so you will also need to raise your wages accordingly until some kind of equilibrium is reached.


JR, awaiting party 2. 

Two things, as far as I can see.  Let's assume that driving the tractor is not a rare skill, at least not to the extent that being able to bend forward and backwards for 12 hours a day without keeling over is.  There are 99 of the former employees waiting to take the seat behind the tractor if the current guy is not happy.

But, let's assume that those 99 other guys eventually found other jobs, or that immigration police found them, so they're not desperately competing for the tractor job anymore.  Let's also assume that competitors have tractors on their own.  They still can't bid up the tractor driver's wage 100 times, because all of the farm owners have to spend capital to get the tractor, whereas they didn't have that capital outlay previously.  Surely capital costs have to eat into some of that 100x productivity gain.

[crazy canuck]

Should labor productivity really drive the wages?  Let's say I own a farm, and have 100 migrant workers (all of them legal, as far as other interested parties are concerned) collect the harvest.  Then, one day, I take out a loan to buy a John Deere tractor, and hire one guy to operate it and do the job of the 100 legal migrant workers. 

That one guy is 100 times as productive compared to migrant workers, but why should he be paid 100 times as much?  If I pay him 100 times as much, then I'm worse off after buying the tractor, because now I have to find some additional money to pay for the interest on the loan that I took out to buy the tractor.

If I buy the tractor, and the farmers around me buy the tractor, I'm going to think about slashing the wages of the guy I hire. After all, there are now 100 farm workers for every job.

You are assuming that all 100 can drive the tractor.

[The Minsky Moment]

We have been automating production since the early 19th century.  Among the classical economists that looked into the question, some theorized exactly as DG and Syt have done in this thread - i.e that greater automation and use of capital in the production process would render workers redundant and hence lower wages and employment.  The most rigorous defender of this position was Karl Marx.  He used a modified Ricardian model to predict that wages would tend to fall to subsistence levels. 

He was wrong.

To get an idea why he might have been wrong, think about what happens if Farmer A implements some efficiency creating technology - an E Machine - , so that now a single worker can produce 20 bushels a day instead of 10.  Let's say A used to pay the worker a $1 per day so that before his labor cost was 10 cents a bushel and now it is 5 cents a bushel.  If we assume there is free competition, then Capitalist A and his neighbor Capitalist B are price takers (i.e. they are too small in themselves to affect the price of wheat).  Capitalist B now sees an opportunity - he can hire away A's worker and pay up to $2 a day and produce more bushels and make more money.  Competition forces up the wage.

The objection to the story is that if everyone does this, production will increase so much that it will bring down the price of wheat and thus the marginal revenue product of labor from the farmer's perspective.  In addition, depending on the price elasticities this means that production will not increase as much is technologically feasible, which means that fewer workers will be demanded than before.  This objection holds but there is a big counterveiling force that also must be taken into account.  As a result of the E Machine, fewer farm workers are now needed to farm wheat, but more workers are now required to design, build, market and transport E Machines.  Also since the Farmer-Capitalists are making more money, they will do things like re-paint their barns and buy something nice for the wife and maybe a sports car for thenselves.  The E-Machine workers and inventors will do similar things and so on.  This all creates demand for more commodities and the labor to produce them and that demand then produces more spin-off demand and so on.

But how do we know whether these counterveiling considerations will fully compensate for the impact of higher wheat productivity.  The short answer is that Say's Law says that it will.  But does Say's Law actually hold?  JS Mill answered that question over 150 years ago.  It does hold, tautologically in fact, but only on the assumption that there are no financial assets.

That last bit gives a hint of where problems may arise . . .

[The Brain]

Put your dick on the table already. Enough stalling.

[crazy canuck]

JR is talking about median wages of people that are employed.  I dont think the figure takes into account the people that are not employed.

Second, the standard theory, as explained by JR, suggests that if you kept the sole remaining worker at the same salary a competitor would lure him away with slightly higher wages (there is a lot of room for an increase given the greater productivity) and you will be stuck with the tractor and no one to drive it and so you will also need to raise your wages accordingly until some kind of equilibrium is reached.


JR, awaiting party 2. 

Two things, as far as I can see.  Let's assume that driving the tractor is not a rare skill, at least not to the extent that being able to bend forward and backwards for 12 hours a day without keeling over is.  There are 99 of the former employees waiting to take the seat behind the tractor if the current guy is not happy.

Bad assumption.  I cant think of a period of technological advancement where all employees could do the new job.  One industry I know about that was impacted directly by technological advancement was the lumber industry here.  When the mills invested heavily in robotics and computerization the unions went crazy over the fact that many of their older members (who remained employed because they has seniority) could no longer do their jobs because they didnt understand the new tech.  Some went for retraining and picked it up but others simply took retirement packages and the newer "tech knowledgable" employees took their place.

But, let's assume that those 99 other guys eventually found other jobs, or that immigration police found them, so they're not desperately competing for the tractor job anymore.  Let's also assume that competitors have tractors on their own.  They still can't bid up the tractor driver's wage 100 times, because all of the farm owners have to spend capital to get the tractor, whereas they didn't have that capital outlay previously.  Surely capital costs have to eat into some of that 100x productivity gain.

Sure, its not a perfect correlation.  I dont think JR is arguing for that.  The problem is that two trend lines have in a sense de-coupled.

[DGuller]

The objection to the story is that if everyone does this, production will increase so much that it will bring down the price of wheat and thus the marginal revenue product of labor from the farmer's perspective.

My objection is that higher capital costs are ignored in the calculation, or at least are not shown to be irrelevant.  Now that you have an E Machine, you've got to take care of both capital and wage costs, where as previously you only had wage costs.  The labor cost of the bushel may have gone down from 10 cents to 5 cents, but the capital cost of the bushel had to have gone up by x cents, where x > 0.

[grumbler]

I was always under the impression that the divergence between productivity and compensation was explained by exactly the factor that DG brought up:  that productivity is increasingly a product of investment in workers and in capital improvements, rather than labor investments per se.  Someone can't hire away DG's tractor operator without buying a tractor, so DG only has to pay his tractor driver enough wages to keep competitors from hiring him away, not enough to make his wage increase equal to his productivity increase.

I'm not sure the "traditional theory" I "think I remember" necessarily accounts for the timing of the divergence in compensation versus output, though.  It's been a long time since i have seen anything on the topic, and I could be misremembering.