Paul Krugman says that the reduction in long-term unemployment has been rather disappointing. This is important, because 2014 was a sort of “natural test” of the hypothesis that the extended unemployment insurance program led to higher unemployment. The program ended at the beginning of 2014, although it wasn’t clear until the spring that the program would not be renewed, with workers being given back benefits missed in early 2014. So it wasn’t a perfect natural experiment. Here’s Krugman discussing the issue:
Ben Casselman points out that we’ve had a sort of natural experiment in the alleged effects of unemployment benefits in reducing employment. Extended benefits were cancelled at the beginning of this year; have the long-term unemployed shown any tendency to find jobs faster? And the answer is no.
Krugman wrote this post in April, right after a difficult winter and negative RGDP growth in Q1. Today we see one liberal blogger after another hyping the great economy, citing fast GDP growth in the past two quarters. Here’s Krugman:
What the report should do, however, is further discredit the “Ma, he’s looking at me funny!” theory of the Obama economy. Remember, we were supposed to be having the worst recovery ever because Obama was a Kenyan socialist who scared businessmen. Actually, it’s a better recovery than the alleged Bush boom – and what’s really striking, as you can see from the chart, is how strong nonresidential investment — essentially, business investment — has been; all the weakness has been in housing.
Of course, you can count on hearing, any minute now, from people claiming that the numbers are cooked — we really have plunging output and double-digit inflation, plus they’re stealing our precious bodily fluids.
I wonder if Krugman realizes that we aren’t in the 1950s anymore, today it’s “progressive” groups that see fluoridation as a big conspiracy.
What most struck me about this post was how upbeat Krugman is about the investment data. And yet doesn’t investment always show bigger declines in recessions and steeper advances during booms? So is 2014 a great year? Or is it a lousy year if we are evaluating the end of unemployment insurance, and a great year if we are considering whether Obama’s policies have adverse supply-side effects? Here’s what we do know:
In 2013, the year of “austerity,” RGDP grew 3.13%; almost double the rate of 2012. So far this year RGDP is growing at a 2.43% rate. Nothing special, because the first quarter was horrible. Let’s count the overall pattern as ambiguous–I have no ax to grind.
In 2014 the growth in employment is going to come in at close to 3 million, well above the 2.3 million average for the previous 2 years. So it’s the job growth that looks really good, much more impressive than the GDP numbers. If we just focus on the long term unemployed then the first three months were poor, with only a small reduction in those unemployed for 27 weeks or more. But since then the rate has fallen very fast. It looks like 2014 will be better than the previous few years in terms of reduction in long term unemployed. That’s especially impressive given that the absolute numbers are now far lower than in 2013, so it’s tougher to continue reducing long-term unemployment at the same pace. In percentage terms, the reduction sped up significantly in 2014. So if this was a “natural test,” then Krugman lost again in 2014, just as he lost in 2013. And I’d argue that the long-term numbers are not the right variable; you want to look at job creation. Extended unemployment benefits also raise the unemployment rate in the under 26-week group. And again, the job creation numbers in 2014 have been much better than the previous few years.
Update: The original version mistakenly said UI reduced unemployment.
Overall, the extended unemployment benefits were not the major cause of unemployment during the recession, it was inadequate aggregate demand. But Krugman’s wrong in claiming supply-side factors are not an issue in a recession—the supply-side always matters. My best guess is that when unemployment was in the 8% to 10% range a few years ago, 5% was the pre-2008 natural rate, about 0.5% was due to extended benefits, a few tenths of a percent was due to the 40% jump in the minimum wage, and the rest was falling NGDP. But I’d be the first to admit that this is a crude guesstimate.
When trying to decide whether the economy is doing poorly or well, you should look at the data and blot out any political issues that hinge on your appraisal of the economy. It’s not easy to overcome the natural tendency to highlight data points that seem to support your policy preferences. We can see this from looking at Krugman’s changing views on the health of the US economy in 2014, switching from pessimism when thinking about ending extended UI, to optimism when thinking about right-wing claims of negative effects from Obamacare.
Fortunately there is “a lot of ruin in a nation.” But that makes it hard to isolate the effect of individual government policies when looking at macro data. Everyone has an issue they are interested in, but don’t assume that you can see the effects of your favorite issue in the macro data (unless, of course, monetary policy is your favorite issue.)
BTW, You should immediately discount Krugman’s claim that we are having a better recovery than the “alleged Bush boom.” During the first Bush recession the unemployment rate peaked at 6.3%. During his first 5 years in office, Obama never saw the unemployment rate fall below 6.6%. There were two reasons for this; Obama inherited a horrible economy from Bush, and his own macro policies were inept, producing a weak recovery. To compare the recovery from 2001-2002 and the Great Recession is utter nonsense. No comparison is possible because the recessions were so different. In terms of employment levels the Bush recovery was far better, ditto for RGDP growth. In terms of decline in the unemployment rate Obama’s was better, as he started from a much higher level. It’s apples and oranges. Comparisons between countries are also tricky. Should you focus on Japan’s 3.6% unemployment rate, or its slow RGDP growth (due to a fast falling labor force)?
READER COMMENTS
E. Harding
Dec 27 2014 at 9:37pm
-This is only for the relevant fourth quarters compared to previous fourth quarters. Annually, RGDP growth in 2013 was lower than in 2012. Employment growth was also lower, with more than 2.5 million jobs created in 2012 and less than 1.5 million jobs created in 2013. This was partly due to the gov’t shutdown and fiscal cliff battle in the year.
Also, yes, we are in the 1950s, lewrockwell.com frequently paints fluoridation as a vast statist conspiracy.
-Can you elaborate?
Agreed with having to block out any political prejudices before appraising the economy.
Scott Sumner
Dec 28 2014 at 12:34am
E. Harding, Almost all economists believe the payroll jobs numbers are better, and growth in payroll employment was higher in 2013 than 2012.
When looking at the effect of fiscal austerity that began on January 1st 2013, obviously you want to use Q4 over Q4 RGDP numbers, not annual numbers. I think all economists would agree on that point.
Kevin Erdmann
Dec 28 2014 at 1:05am
By the end of ext. Unemployment insurance, there were only 1.3 million receiving it, but nearly 4 million workers unemployed over 26 weeks. Many of those additional workers had timed out of ext. UI, but their exit rates wouldn’t have been affected by the end of the program. Coincidental with the end of the program, exit rates among all other groups, including shorter durations, took a distinct turn to faster re-employment. Durations over 26 weeks are now down to about 2.8 million, but average durations for this group haven’t declined at all, which suggests that those persistently very long term unemployed are still a sizable portion of the group, and that unemployed workers under 100 weeks have declined basically to recovery levels, given current demographics.
Between February and August we shed nearly 1 million unemployed, and it basically all came from the over 26 week group. The 15-26 week group had been fairly flat at about 1.6 million, then dropped to about 1.4 million from March to May (about 15-26 weeks from the end of ext.ui) where it has remained.
Scott Sumner
Dec 28 2014 at 10:52am
Thanks Kevin. I’ve always thought the extended UI benefits also raised unemployment in the under 26 week category, but much less than in the over 26 weeks group.
william Occam
Dec 28 2014 at 12:38pm
So Krugman thinks the Obama recovery is stronger than the Bush recovery?
So “austerity” led to a stronger recovery than the spending-fueled 1990 version. Interesting claim.
Hmmmmmm
E. Harding
Dec 28 2014 at 2:46pm
Okay, I concede that using fourth quarter numbers makes more sense, and, therefore, on RGDP. But I still don’t see how the civilian employment numbers support your case:
http://research.stlouisfed.org/fred2/series/CE16OV
The nonfarm payroll employment numbers were more stable and were ever so slightly better in December 2013 than in December 2012. But, remember, total nonfarm payroll
“excludes proprietors, private household employees, unpaid volunteers, farm employees, and the unincorporated self-employed”.
These matter, too (except for the unpaid volunteers).
Nick
Dec 28 2014 at 8:19pm
“Extended unemployment benefits also *reduce* the unemployment rate in the under 26-week group.”
Should that be “increase”?
Scott Sumner
Dec 29 2014 at 1:17am
Nick, Thanks, I fixed it.
E. Harding, In principle I agree with you, but the payroll data is the only fairly accurate data we have for month to month changes.
ThomasH
Dec 30 2014 at 2:27pm
Scott:
I’m pretty familiar with the ways you think Fed policy has been inept so as to contribute to the slow recovery and I agree with you, but I’m not familiar with your criticisms of what President Obama has thought or said or done or left undone that has contributed to the slow recovery.
James in London
Jan 4 2015 at 3:42am
And now PK switches to GDP per capita as the measure of success. Obviously, he wouldn’t want to put Japan on that Jan 2nd chart as it would invalidate his claims of Japanese failure.
Equally, he wouldn’t want unemployment or total employment as the UK would then be far better than France.
Talk about picking and choosing.
Daublin
Jan 5 2015 at 12:11pm
I don’t see how we get any natural experiment given the passing of Obamacare. It’s a massive change to labor law, and businesses are sure to have started adjusting to the changes–and fears of changes–as soon as Obama got elected.
Comments are closed.