Battlepanda: It's Free-Market Kismet


Always trying to figure things out with the minimum of bullshit and the maximum of belligerence.

Wednesday, March 08, 2006

It's Free-Market Kismet

Funny how I just wrote a post about how true-believers mistaking the tenets of their religion for agreed-upon facts and using them as such in arguments, and now I'm going to write another.

Consider this Jane Galt post arguing against the minimum wage. It follows the rubric of what I call the "free-market kismet" post pretty tightly. The idea of free-market Kismet is that The Market is an omnipotent equilibriating force that defys any human attempt to change the economic landscape for the better, except by giving ourselves over more fully to The Market.* Not sure what I mean? Let's take a look at Jane's arguments through the three steps to Free-Market Kismet:

(Step 1) Discredit empirical evidence supporting the other side
Empirical evidence can always be discounted through poking holes at the methodology, sample size, time frame, political leanings of the experimentors or even just musing that the whole experiment is just a fluke.
Many liberals, citing Card and Krueger's paper on changes in the minimum wage, say that any fall in employment is so small that it might as well be zero; it is too small to detect in the constantly fluctuating marketplace. [Snip] This makes conservative economists say "Whoa!" Card and Krueger didn't simply find that effect of a relatively small change in the minimum wage was too negligible to measure; while a conservative might dislike this finding, it wouldn't be obviously, outrageously wrong. But Messrs Card and Krueger actually found that raising the minimum wage increased low-wage employment, a result that simply makes no sense to all but a handful of hard-left activists who used the study to argue that they'd found some sort of magic money machine. No one has posited any plausible reason that an employer whose wage bill has just gone up would respond by . . . taking on more, now more expensive, workers. Without such a mechanism. . . and colour me suspicious that we will find one . . . we are left with two possibilities:
1) There is some other effect, such as a boom in New Jersey's fast food sector, that masked the employment drop

2) There's something wrong with the data

Number 1 seems possible, and number 2 seems very likely, because they relied on survey data, and surveys are notoriously unreliable. Just ask the folks at Coca-Cola, who did the most extensive surveys in history, and unveiled "New Coke" only after every single study had reported that consumers overwhelmingly favoured the new taste. A later study using more-reliable payroll data (but funded by the retail/restaurant industry, and involving a smaller data set) found the effect you'd expect: minimum wage employment in New Jersey went down compared to Pennsylvania. Other criticisms of the study's methodology helped to seriously weaken their assertion that there was no measurable employment effect. Most of the studies that have been done in the past have tended to reinforce the economic conventional wisdom.

Nonetheless, while modest increases in the minimum wage may increase unemployment, the effect doesn't seem to be huge; when the studies tend to point both ways, that's a good sign that whatever change you're looking at is pretty small.

(Emphasis mine, obviously.)

Is it just me or is the above a lengthy exercise in overinterpretation? By Jane's own admission there are two studies on the New Jersey minimum-wage hike: One found that it slightly raised employment levels and one found that it slightly depressed employment levels. How is that not consonant with the liberal hypothesis that Jane herself outlined, that any fall in employment might as well be zero? Attacking Card and Kreuger's methodology when the study she quotes approvingly supports the liberal hypothesis just as well seems nothing short of a bizarre compulsion to waste pixels. I can counter Jane's claims by pointing out that the New Jersey minimum wage hike is by no means the only real-life example of a minimum-wage rise not to result in the confidently-predicted drop in employment, but I have a feeling Jane won't care. She has performed the ritual step of "discrediting" some research on the other side.

(Step 2) Provide "evidence" based not on empirical research, but from Econ 101.
Turn about is fair play, right? So let's see what real-world evidence Jane can muster to support her hypothesis. Except she's too smart to fall for that one. Empirical evidence can always be discredited, but long-winded derivations from principles cannot:
Standard economic theory tells us that if you artificially raise the price of something, suppliers want to sell more of it, while buyers want to purchase less. In the labour market, this means that more people want to work, but employers don't want to hire so many of them, so you end up with too many people chasing too few jobs.
Hey you can't argue with standard economic theory. Believe me, I have no desire to, except when its first tenets are being wrongly applied. Any discussion of the minimum-wage employment dynamic in the real world without mentioning "monopsonies" and "bargaining power" is simplistic at best, and dishonest at worst. This is not some crazed leftist plot. In England it is regarded as high-school level econ.
Another little-considered downside of minimum wage increases is that employers who are forced to pay higher wages often find ways to get it out of their employees in other ways; as Tyler Cowen pointed out:
Gordon notes that the government can make an employer raise nominal money wages, but can't stop him from turning off the air conditioner. [A more optimistic scenario is that the employer invests in creating a higher-productivity job.] Surely just about every job out there can be made worse, one way or another, in a way that saves the employer money.

So the scenario is now simple. The government boosts the minimum wage. Low-wage workers earn more. Few lose their jobs. Workers sweat more too, one way or another. Few are much better off.

Oh, I see. Card and Kreuger can be criticized for using surveys to get their data because surveys were what lead Coke to make the disasterous leap to New Coke, but approvingly quoting Tyler Cowen's fanciful conjectures about how a higher minimum wage might lead employers to take actions to adjust work conditions so that life is exactly just as bad as before for the low-wage worker is OK. Can't argue with logic.
But employers can just raise prices! say advocates. Mmmm, yes . . . unless there are substitutes for their product. McDonalds competes with frozen dinners and 7-11 burritos and Kraft Mac n' Cheese; The Gap competes with the J Crew catalogue. No, McDonalds won't shut its doors, but it doesn't have to in order to reduce employment; if 15% of its customers defect to cheaper alternatives, it will need fewer burger-flippers and order-takers. There's also the fact that workers complete with labour-saving equipment, which may become cost-effective if wages skyrocket. Plus, since poor people are disproportionately likely to shop at places that pay minimum wage, including fast-food outlets, the higher prices often come out of their pockets.
In which Jane exposes herself as an inhabitant of Widgetland where competition is perfect and their ain't no such thing as profits.So much for the stock market.

(Step 3) Congratulations, free-marketeer! You just established the infallibility of Free-Market Kismet to your own satisfaction!
Nothing left do do but pile on some smarm and take a gratuitous dig at the Unions:
it seems at least as plausible that unionization and the high minimum wage were the result of high demand for labour, which also caused income inequality to shrink. If this is true, enacting the minimum wage, or getting the NLRB to beat up on companies, will be no more effective at bringing back those halcyon days of income compression than resurrecting Burma Shave billboards across the land . . . and considerably less entertaining.
Oh, I see. Unions were the result of income equality, not the other way around. That'll explain why the improvements in wages and working conditions preceeded the formation of unions.

*In Galt's defense, she is not against all poverty-fighting measures. For instance, she supposedly supports the EITC. I would give her partial credit for this, except she is obviously wrong even given her Econ 101 logic -- like any commodity, labor has a cost below which it cannot be sold -- if it took me $4 to make this teapot I will not sell it to you for $3.99; if takes $x for me to keep body and soul together I will not work for you for less. But if the government steps in and gives me $y in an attempt to to improve my lot, I will now be able to cut my price for labor all the way down to $x-y, which results in my not having any more money in my pocket than when I started. It's the government providing the employers with subsidized labor.