Thread by @Noahpinion

Now at a session about employer power, with @mioana and @RyanDNunn!

First paper is by Evan Starr, with J.J. Prescott, about noncompete agreements.

The question is: What do workers BELIEVE about their noncompetes and their enforceability in court?

It's all about the CHILLING EFFECT.

California "bans" noncompetes because they're not enforceable. But tons of California workers have these un-enforceable agreements anyway!!

The authors randomly inform workers about the enforceability of their noncompetes, and observe how worker outcomes change.

Workers either don't know that non-competes are un-enforceable, or respect non-enforceable contracts out of integrity and honor.

Authors find that simply telling workers that their non-competes are un-enforceable makes them more likely to say they're considering switching jobs.

(But is this just resentment after realizing they were tricked? Or would they really leave?)

Basic message of this paper: We need institutions that constantly remind people their non-competes are un-enforceable.

(Personally I'd also consider some legal penalties for companies that knowingly write contracts with un-enforceable non-compete provisions.)

Next paper, by Ben Lipsius, alleges that monopsony power does NOT explain the falling labor share:
ideas.repec.org/p/jmp/jm2018/p…

Lipsius reproduces the now-common result that labor market concentration DOES reduce wages.

In other words, market power in labor markets is real and substantial (which is probably why minimum wages don't reduce employment much, or at all).

Employer power matters - a lot!

But has it been increasing over time?

Can it explain why wages have been stagnating as profits increase?

Lipsius says: NO.

Lipsius calculates EMPLOYMENT-WEIGHTED labor market concentration and finds that it's falling.

In other words, compared to previous decades, more Americans now work in labor markets that have more employers.

But within some industries, like finance/real estate, manufacturing, retail, and wholesale trade, employment-weighted concentration has actually been increasing.

Lipsius notes that the result persists when you use a different definition of a local labor market. (See other paper by @kevinrinz).

Lipsius combines his estimates of the importance of labor market power with his estimates of employment-weighted concentration, and finds that essentially nothing has changed since 1990.

In other words we need some other theory to explain the falling labor share.

Discussant Ryan Nunn: we ought to look at data on OCCUPATIONS, not just industries, in order to properly measure concentration.

Nunn: Even if labor market concentration hasn't increased over time, it's still hugely important! And our antitrust and wage policies ignore it!

So we still need to worry about concentration a lot more than we have been.

(Nunn is right.)

Ironically and tragically I had to leave for a meeting before @mioana could present, even though it was her paper I came to see, so I'll have to buy her lunch to make amends...

(end)