Unemployment Insurance and Unemployment

Quick finance thought exercise: Let’s say I’m going to buy 100,000 shares of a stock. Leaving C-3P0, Hal-9000, skynet and all the other flash-trading supercomputers that would have an orgy front-running this large market order out of it for a minute, we expect the stock price to rise rationally. More demand for the stock, price has to rise. It may rise, say, 20%, depending on market factors.

Now let’s say instead I just buy 1 share. What happens? Price has to rise; it won’t go down, and I can show you all the equations that say it has to rise. However it is going to “rise” a hundredth of a penny. The rise as a result of this action is negligible and indistinguishable from zero.

So there is a lot of debate this past week as to what the effect is of means-tested unemployment benefits on our unemployment rate is. Paul Krugman calls out Casey Mulligan for “saying that unemployment is high not because employers have become less willing to hire, but because workers have become less willing to work.”

Let’s go to Casey Mulligan. Here is a quote from him: “employees face financial incentives that encourage them not to work…. [T]he decreased employment is explained more by reductions in the supply of labor (the willingness of people to work) and less by the demand for labor (the number of workers that employers need to hire).” Is that a weird thing to think nowadays? Well, he said that in December 2008 (!), before Obama’s stimulus plan and back when the unemployment rate was 7% (remember those days?). Unemployment rising during a housing crash and financial collapse is the result of government meddling in labor markets. We’ll have to agree-to-disagree there. But I want to go on to some other points.

Two more quotes. Free Exchange rightfully points out, in an excellent entry:

Unemployment benefits clearly do provide an incentive to stay out of work longer. Holding other things constant, we would expect an increase in the generosity of unemployment benefits to lead to more joblessness…

In the absence of unemployment benefits, workers facing job loss would have been forced to liquidate assets to pay for basic necessities (adding to downward pressure on the value of assets of all sorts and leading to an increase in personal and financial institution defaults) and reduce spending sharply, magnifying the initial demand shock.

In addition to a significant increase in human suffering, this would have led to a much steeper and more rapid rise in the unemployment rate…

This is just daft. Yes, without unemployment insurance the newly jobless would have quickly gone back to work—selling possessions for a pittance, working off the books at sub-minimum wage rates, scraping by on charity (nasty incentive effects there, too), begging, stealing, and scavenging—and the country would have been far worse off. For millions of Americans, there was simply no real work to be had at positive wage rates during the worst months of the downturn. Forcing those labourers to get by with no assistance would have been economically catastrophic, and cruel.

This seems entirely correct to me. Will Ambrosini takes issue:

Strange for a magazine with such a name that they don’t know about Alfred Marshall and his marginalist revolution. See, when someone uses marginal analysis to show, unequivocally, that increases (*ahem* marginal increases *ahem*) in unemployment insurance benefits depresses employment, its not ok call them names for not doing inframarginal analysis…

The first $1838 of UI (and food stamps and TANF) keeps families from starving and gives them shelter for a month or so while the bread winners look for new jobs and it prevents them from having to sell off their assets, which, as The Economist mentions, would exasperate the demand shock. Every dollar above that encourages them to search longer and thus increases unemployment….

Free Exchange does bring up this marginal analysis and depressed employment at the beginning, and argues that it is too small to risk a fire sale of household assets. Will says that it must increase at the margin, and that is that.

Fair enough. My questions: how much does the increase in unemployment benefits increase unemployment in this economy? A lot? Is it responsible for half the increase in unemployment? My guess is that it increases it in the same way me buying 1 stock increases the stock price – a negligible and indistinguishable from zero amount.

Why do I believe this? As far as I understand the models, it’s conditional on a “value function”, the chance that you’ll receive a good job offer. You go look for a job, see if you get an offer, compare it to your unemployment benefits, and make a choice. Now what do we know about how looking for a job has changed in the past year? From the big picture:

So when you go and look for a job, in the same way you pick a marble out of a jar, the chances that you’ll find one in any period has fallen off a cliff. Cut in half? a third? It’s a terrible market for job seekers, regardless of whether or not get you a little extra spending cash in a government check. The fact that it is so much harder to find a job makes me think any marginal effect from a little bit more of an unemployment check is negligible and indistinguishable from zero.

I’m willing to be proven wrong, and love back of the envelope estimates. The U3 unemployment rate has gone from 5.6% to 9.5% over the past year, an increase of 70%. How much of that is the result of means testing? .1%? .3%?

The issue that people have with this debate, and you saw it when the panelists at The Week blinkered at Malkin bringing it up, is the presumption that it is a major effect. You get the impression that some people think, with all the issues on the table, that this is one of the major drivers of unemployment – which strikes me as, to use Free Exchange’s term, daft.

Here’s Mulligan’s current view of the unemployed: “Sensible people will recognize that public policies have dramatically reduced the costs to them of searching further for the job they’d like, or making this the year they coach junior’s baseball team, or do some work on their house, or refrain from “coming out of retirement,” or take a trip.” It’s a big country, with all kinds of interesting characters in it, and I have no doubt somewhere someone unemployed thinks right now is the best time to kick back and coach little league on the Government’s dole instead of taking a job offer. But is this a constituency that is anything other than negligible and indistinguishable from zero?

Since the effect may be small, Free Exchange’s point is even more important – we have to worry about fire sales of assets, aggregate demand and widespread misery, and if extending unemployment benefits has a negligible effect it must be true that unemployment is going to be very bad. I want to point out three more things:

1) I believe Will knows his Raj Chetty, so the idea of higher unemployment benefits in periods where work is the most difficult to find is not just about moral hazard of people on the dole but also about people being able to overcome imperfect markets, liquidity and credit constraints. If you aren’t familiar with Raj Chetty on unemployment insurance, I highly recommend you read this slate piece on his research.

2) Means testing mortgage modification, another thing that has caused the unemployment rate to move to 7% and beyond in Mulligan’s world, is less about redistributive justice and more about solving a massive Principal-Agent problems the quants have known about for years but have been unable to hurdle.

3) Higher unemployment spending is a feature, not a bug, of the recovery. Chief Moody’s Economist, and former McCain economic advisor, Mark Zandi concluded that extending unemployment benefits had one of the best bang-for-the-buck ratios for stimulus, second only to food stamps. Consumer savings rate has skyrocketed past 5%, and that is good for them but bad for the economy’s aggregate demand. Tax cuts would automatically be saved, unemployment benefits spent. Why isn’t this a fantastic deal?