Comment by dalbasal | Hacker News Reader

Comment by dalbasal | Hacker News Reader

I really think trying to reason about labour markets using pure theory in a Hayek-Friedman-esque way is a dead end.

It's the 2nd time in two days I made the recommendation, but throwing Ronald Coase into the Neoliberal canon would help a lot. He was a "chicago school" academic from the same intellectual family, so it shouldn't be too much of a culture shock.

He wan't like "progressives" in the "evidence based" sense but he also objected to first principles theory like Friedman. Instead, he tried to find persistent phenomenon and theorized about why they exist.There was always a link to the real world.

Anyway, the pure theory approach leads to a general conclusion/assumption that markets are the same. The market for labour, barbie dolls, whatever.

IRL, labour markets are obviously very different than most other markets. It's inflexible. People stay in jobs a long time and most employers actively try to lower their average turnover. Why?

If flexibility is so wonderfully efficient, why do almost all companies have such a big inflexible workforce? Why is the market for labour so different in practice than the market for oral hygiene products and services.

Why isn't Wall street staffed by day labourers or SV products built by quarterly contractors? You have to look past pure theory to answer these questions. I think you probably have to look beyond economics, or at least to its fringes (like behavioral economics).

Transaction costs.

Both the employee (search costs, hours in the day) and employer (job training, search costs, process knowledge) incur far greater transaction costs than you see in commodity markets. Business models that lower transaction costs (e.g. "gig economy" middlemen) tend to lead to the emergence of highly flexible labor markets. This currently only applies to jobs that don't require a high level of nontransferrable knowledge, because nobody has yet invented an "I know Kung-Fu" machine that can transplant a company's 5 year marketing strategy into the head of a day-laborer CMO.

None of this requires going outside pretty basic economic theory.

Edit: I just realized this sounds like I'm explaining Coase theorem to the person who cited Coase. It's not for you, it's for others reading your comment. I'm disagreeing with your conclusion, not your premises.

There are labor market overhead costs that are often easily overlooked/under-appreciated too: tying one's health care, as one obvious example, to one's current employer necessarily undermines liquidity in the labor market.

But more importantly: the overall inability in current culture to have "not working" as a legitimate and survivable labor position (for the majority of the labor force, at least) also threatens the liquidity of the labor market. Employers may fire at will, certainly, but as a laborer you can much more rarely quit at will without threatening your establishment in the labor market (Americans often question holes in resumes and may see them as moral failings), and your possible (even short-term) survival (health care, shelter, food).

Yes, the labor market is distorted precisely because it's not a negotiation between equals. Costs and benefits are disproportionately pushed onto the party with less leverage.

Transaction costs.

Of course. And I'm surprised that the post you're replying to didn't come out and say that, since he brought Ron Coase into the conversation. I mean, for Coase, the effect on markets of non-trivial transaction costs was really the core of his work, at least as I understand it.

:)

That's definitely a very Ronald Coasian answer! We should start that lobby to get Ronald Coase accepted into neoliberalism or novoliberalism or whatever comes next.

I purposely didn't give the answer because I think there are possibly other explanations besides transaction costs and regulation (probably the more friedman-esque go-to). The reason I like his papers a lot is because of his questions. He has a Darwinian sort of approach. Why does the moth have such a strange beak. Lets look for the matching orchid.

I think (don't know, speculation) labour markets are special in other ways, not just high transaction costs. A company is a society. People have loyalties and identity tied into it. We are a social animal in very fundamental ways. Our psyche handles the intricate nuances of human cooperation by default. Things happen in a team of people that know each-other, like each other or just think of themselves as part of a group that I think are fundamental to this question. I don't think transaction costs sums it up. I suspect group-oriented work plays a bigger role. Also employer-specific skillsets, though that probably can be lumped into transaction costs.

Incidentally, I feel that "basic economic theory" tends to acknowledge transaction costs mostly as a caveat to basic models: "assuming no/low transaction costs," more often that actually accounting for it. I think this is also true for "economist hat" thinking. Coase complained about this a lot. Coase theorem is (at least to me) taught as a theory about bargaining and contracts solving problems like externalities. For Coase, the caveat was the "theorem." In the absence of transactions costs blah blah pareto optimals. externalities... He thought it was obvious that the pareto efficiency doesn't actually happen in reality, externalities continue. Therefore transaction costs must be high.

To me, this is the big difference between Friedman and Coase. Friedman's a pure theory guy. The main opposition to this view today is a theory-less econometrics/evidence based "wonkish" analysis. Coase is a middle ground. Theories make predictions. In some cases they are good. In cases where they're not, we need theoretical expansion. Neoliberals tend to reach for "distortion" as an explanation far too often, and selectively IMO.

(we're really in the weeds here, sorry)

Huh. I think this highlights the fact (that I didn't realize until now, really) that I never really considered Friedman an "economist" in the strictest sense — in my view he's more of a normative political theorist, who uses economic models combined with empirical evidence from history to draw conclusions about policy. Though I'm politically & philosophically simpatico, I've never really looked to Friedman (or Hayek, for that matter) for practical economics.

Friedman was a professional economist all his life. His most famous area was on monetary policy which was very politically important in that era of hyperinflation disasters.

But, economics bleeds into political science readily. Marx was an economist. Hayek (mentioned in the article) is most famous for political pamphleteering, but he was certainly an economist too. Both he and Friedman got the nobel prize for their economics.

Neoliberalism is very much an economics-politics hybrid movement.

I like your "in the weeds" thoughts. There are a lot of intagibles that get wrapped up in employment, and many of them are very personal and contextual. Such things are incredibly resistant to meaningful quantification -- often the individual might even struggle to explain the dynamics that led to certain choices. A lot of food for thought here... thanks for the post!

Would you mind if I told you that you're not even wrong?

First, neither Hayek nor Friedman were well known for contributions to labor market research. Hayek was well known for his critiques of central planning while Friedman was famous for his contributions to monetary theory. But my point is that these guys aren't "labor market economists", so taking an issue with them over labor market issues would be barking up the wrong tree.

Plus, I don't see why "pure theory" should push anyone anywhere: constraints on model-building are actually quite weak. On the other hand, labor market models haven't become more like consumer goods market models over the years. If anything, the opposite has happened: labor markets are unique among other markets because of stuff like human capital, efficiency wages, signalling, gift exchange, searching costs, internal labor markets, non-cognitive abilities, poaching externalities, hedonic benefits of work, etc. This list isn't exhaustive.

There are very fair and well-documented economic reasons why present day firms might want to put more money into their employees than a 19th century landowner hiring day-laborers during harvest would. Firms benefit from having (and keeping!) a skilled and well-motivated labor force, and so they act accordingly. Trying to look "beyond" economics if you haven't heard of any of those is just looking past it.

They were both pretty wide, but I disagree. I think both touched on labour markets quite a bit. For example, "Friedman's" best known policy is negative income tax (earned income tax whatnots).

Have a look at https://en.wikipedia.org/wiki/Natural_rate_of_unemployment http://journal.apee.org/index.php?title=Spring2008_2

Frankly, I think that in order for something to "touch on labor markets quite a bit", they'd have to at least dedicate a book to it. In the current day, David Card and Orley Ashenfelter might be said to have touched on labor markets quite a bit.

Hayek and Friedman? Not so much.

   Anyway, the pure theory approach leads to a general conclusion/assumption that markets are the same. The market for labour, barbie dolls, whatever.

Proof needed. Have you ever read Hayek or Friedman you're talking about ?

Go to your whiteboard and draw supply-and-demand curves for a couple derpy markets of widgets and other goods. Then, draw one for labor. Note how the curves are reversed; if you draw it "backwards", then it goes back to behaving like a normal market. Now, try adding employment policies, like minimum wages, to the market; what happens to the curves?

To be fair, the labor market does have some gotchas. In particular, there's the phenomenon of "price stickiness", which is a big deal in labor (i.e., wage stickiness), but doesn't have nearly so much effect with other goods.

Don't bring Keynes into this! We're way to far in the weeds already.

"Pure theory" or Hayek / neoliberal theories don't seem to be just about perfect S/D curves and reasoning with S/D curves doesn't prove anything, it's just a visualization tool. I didn't understand why the curves are reversed ?

Reply on news.ycombinator.com