Uncertainty Isn't Killing the Recovery

Actually, more uncertainty has been correlated with more jobs since 2008


Sorry, Ben Franklin, but it turns out there are actually three certainties in this world: death, taxes, and Serious People blaming a bad economy on uncertainty.

Big businesses like to complain. Sometimes, they complain about taxes. Sometimes, they complain about rules. But mostly, they like to complain about uncertainty. More clarity about future taxes and regulations might make some businesses more likely to invest. But there is little reason to think that too little clarity is what's keeping them from investing today.

Why do we know the slow recovery isn't about uncertainty? For one, investment other than residential investment has already recovered from the crash. It's the long shadow of the housing bust, not regulation, that's the problem. For another, as Jim Tankersley of the Washington Post points out, uncertainty has actually fallen a lot the past few months, but hiring hasn't picked up.

Okay, but how do you measure uncertainty? That seems, well, uncertain. And it is. But Scott Baker and Nick Bloom of Stanford and Stephen Davis of the University of Chicago have tried to put a number on it. Their Economic Policy Uncertainty index looks at how many expiring taxes and new regulations there are, how often major newspapers talk about "uncertainty," and how much professional forecasters disagree about future inflation and spending to quantify it all. It's a simple enough idea, and there should be a simple enough relationship if it really does tell us anything useful: The more uncertainty there is, the less hiring there should be. And that's exactly what their data show going back to 1985 -- but not now. It's admittedly a tiny sample size, but as you can see below, there's actually been a very weak, but positive, relationship between the two since 2008. More uncertainty, more jobs? (Note: The yellow dots show the last six months since the fiscal cliff was sorted out).


Not exactly. As I said the relationship is very weak. Nonexistent even. Indeed, there wasn't a statistically significant relationship between uncertainty and hiring (the P>t value was 0.262), and what relationship there was explained almost nothing (the R-squared value was 0.019). Now, it was a bit better when I tried lagging payrolls by three months -- the idea being that shocks could take awhile to filter into the economy.


This time there was a statistically significant relationship (the P>t value was 0.023), and it did explain a little (the R-squared was 0.08). But more uncertainty was still associated with more hiring. So next I tried looking at just the recovery by itself, and not the recession too, to see if that changed things. It didn't. Once again, there wasn't a statistically significant relationship if I didn't lag the data, there was one if I did, and neither explained much (the P>t values were 0.121 and 0.03, and the R-squared values were 0.05 and 0.103).

And once again, the sign was still wrong for both. More uncertainty correlated with more jobs.

It's almost as if uncertainty isn't what's holding the recovery back. That's clear enough the closer you look at Baker, Bloom, and Davis's (BBD) index. It interprets every problem as a supply problem -- which isn't useful when our problem is a lack of demand. But it's actually worse than that. It's misleading. As Mike Konczal points out, their methodology perversely tells us that up-is-down and saying something enough makes it true. For one, their focus on soon-to-expire tax cuts and credits means that any temporary stimulus shows up as uncertainty. So the 2008 tax rebate, the 2009 expanded tax credits, the 2010 payroll tax cut, and its 2012 extension -- all demand-side measures that put cash in people's pockets -- are "bad." In other words, certain austerity is supposedly better than uncertain stimulus. (Never mind that until recently, poor sales had topped small business concerns going back to 2008). The same applies to the Fed's unconventional easing. Questions about how and when the Fed will step on or off the gas also show up as uncertainty -- and ignore how those policies help the recovery. So it's not the small sample size that's made more uncertainty associated with more hiring the past five years. It's the way their index works. (Or doesn't). It interprets every demand-side policy that boosts employment as boosting uncertainty. In other words, it can't make sense of a depression.

And then there's the derp. Derp, you ask? Yes. As Noah Smith defines it, derp just means loudly repeating what you believe even after reality disproves it over and over again. There's been a lot of it lately -- reality has been busy disproving a lot of bad ideas -- and the BBD index rather unfortunately counts this as evidence of uncertainty. Let's try a thought experiment to see how. Imagine what John Boehner, or any other generic Republican, would say about the economy today. They'd probably ask where the jobs are, say the stimulus didn't work like Obama promised, and then proclaim the real problem is uncertainty about taxes and spending. As Mike Konczal points out, that's pretty much exactly what you'll find if you search newspapers for terms like "economy" and "uncertainty" and "spending" -- which are the terms BBD use when they trawl through newspapers. In other words, the BBD index picks up Republican talking points about uncertainty as proof that uncertainty is a problem. Neat trick. And it's the same one the index uses when it looks at economic forecasts. Predictions about the future are always hard, but they're especially hard if you're not using the right model. And people who think "Keynes" is actually a four-letter word are not. Now, Paul Krugman has spent the past five years trying to explain how liquidity traps work, but many people (including famous economists) don't want to listen. They keep cowering before mythical bond vigilantes, warning that 1970s-style stagflation is just around the corner. It never is. But the BBD index takes these bad predictions not as a sign of a bad analytical framework, but of uncertainty. It's derp all the way down.

Depressions are stupid things, and we shouldn't have them. We wouldn't if we had more spending or more inflation or more of both today. But that seems too easy for economists who only want to talk about the long run. So they talk about 30-year budget projections and Obamacare and, yes, uncertainty instead. Of course, there's no evidence that these are actually causing the slump now, but why would there be? These are rationalizations from people who don't like the actual solutions to our actual slump.

That is certain.