Thursday, February 18, 2016

NDD on the (fake) economics of the minimum wage

New Deal Demoncrat - thoughts on the minimum wage. Long quoting alert, cos I have a lot to add:

Most people who argue against the minimum wage assume labor demand is elastic, meaning the amount demanded will disproportionately change relative to price changes. Here's a simple supply and demand chart to illustrate:

In the above chart, the government increases wages from P to P2 and demand drops from Q1-Q2. In this example, the line Q1-Q2 is longer than P-P2, meaning the increase in price had a large impact on demand.

The additional paleoconservative dogwhistle we're fed is that the minimum wage also increases unemployment, by drawing extra people into job-hunting who would like to be paid the above-equilibrium minimum wage. Which kinda begs the question of how a country with a minimum wage can ever operate at "full employment", no?

And though he'll get to in a bit, I have to point out that one big problem with the above simplistic model is that there's no scale on the axes. Good luck finding a first - or second-year prof who will actually point that out, though.

However, what if demand is inelastic? In that case, we'd be looking at the following chart:

Above we see the quantity demanded decrease slightly relative to the larger increase in price. This means that even if wages increase, we still demand a certain amount of labor.

The second chart is actually far more realistic. Consider the following fact pattern: a store currently has 100 employees. Because they're a profit maximizing entity, they are already using the optimal amount of labor (give or take say 5-7 employees). Let's assume they cut their workforce by 25% because of the increased wages they must now charge. At this level of cuts they'll see shelves stocked less frequently, fewer employees helping customers finding items and longer check-out lines. This will actually decrease their sales; customers will get tired of not being helped and will become frustrated with the longer lines.

And the elasticity is important because labour is also the economy's consumers: if you can raise the minimum wage to reduce employment a bit but the net pay to all labour (area under the curve between L=0 and equilibrium) is increased, then you've just improved general welfare and boosted consumption.

Guess what? That increased consumption then feeds back into labour demand, and suddenly the equilibrium goes back above the minimum wage, and so you end up at full employment and free-market equilibrium again.

Also, when you ignore the Benjamin Bunny Preschool static model shown above and start looking at labour dynamics, you see a lot of other weird things happen. Yes, there's price stickiness in wages; part of that, though, is due to labour hoarding, which is what happens when your company faces reduced demand during an extreme economics downturn but you refuse to fire your top engineers with 20 years' experience because you've invested so much in their human capital.

Kinda hard to model that human capital in a simple supply-demand model.

It's all a lot more complex than the bullshit X diagram we get in 1st and 2nd year econ. But the paleoconservative doctrine demands teaching us kiddies that Adam Smith's 18th-century philosophical noodlings count as "economics", and that all wages just vanish into a black hole somewhere - kinda like government spending.

Another big problem with the simple "perfect competition" model lies, actually, on the supply side. At the top, the supply curve actually bends backwards, as high-paid workers trade extra hours for more leisure: this is pretty standard teaching in late-2nd year microfoundations.

But the bottom changes direction as well: each worker has to meet a minimum level of autonomous consumption to survive, and so the supply curve bends back outwards at the bottom as well. At which point you get negative externalities that reduce labour demand further. The job of a minimum wage, then, is to cut off the lower half of the supply curve to ensure the lower equilibrium is avoided.


The underlying reasoning is based on the production function, another basic micro-econ concept:

The horizontal line shows the total amount of labor while the vertical lines shows total output. Notice that as L decreases, so does output.

So, the model's argument goes, the perfect competition demand curve is entirely a function of marginal product of labour: the store supposedly hires workers just up to the point at which the wage cost is exactly paying for profit. So, no, the store won't lay off 25% of employees if the savings in wages is offset by losses in sales. Yay free markets!

(This model fails to explain why, say, black university graduates face an unemployment rate several times higher than white highschool dropouts. I mean, wouldn't black university graduates have a higher marginal product of labour than inbred toothless West Virginia trailer trash? But that's what happens when you aggregate all labour into one "aggregate worker", as in this model.)

But the important thing to note is that this simplistic supply-demand model only works for perfect competition (and I guess monopolistic competition with some modification), yet perfect competition doesn't actually exist in the real world. In the case of labour demand, labour generally faces an oligopsony: for most jobs, there are a limited number of hirers in any market, and they tend to "collude" in the wage offered.

"Collusion" in this case doesn't have to mean some sort of conspiracy, btw; it can be as simple as saying "a middle-aged mom working part-time as a cashier shouldn't make more than $X an hour", or "I don't care how good a programmer you are and how the dozens of scripts you've written have improved productivity at our office, there's no way anyone in your position gets paid more than $Xk per year." Or even paying employees wages within a band determined by job classification. Face it: if you're making the company $500k/yr in profit, you're not going to get your fair share.

So, similar to how a monopsony buyer's market power means they can pay significantly less in wages than the employee's marginal product, an oligopsony that successfully colludes will also pay workers far less than their marginal product. So the demand curve doesn't look like the above. In fact, the labour buyer is a price-maker.

That means there is no demand curve.

So the tl;dr is that NDD killed a bunch of helpless electrons for nothing?

Well, he wrote a basic criticism of the labour supply-demand model, which is maybe enough to get von Mises-worshipping clowns thinking. Thing is, there is even more bullshit to the model than NDD let on. It really is a worthless piece of trash, which is why anyone with any sense should be worried this is all that engineering and finance and business majors learn about labour markets in their 2 or 3 intro econ classes.

In fact, the reason there's so much paleolithic, incompetent, truly dangerous pseudo-economics out there is because universities like teaching this garbage to all their first and second year students.

And that's why countries like the USA are as fucked-up and economically crippled as they are. Because economics departments are purposefully destroying the country.

No comments:

Post a Comment