Why Looking Carefully at Data Matters

Our good friend Ed Weick continues to email lots of (always gloomy!) data on the current economy.   This morning, he sent along this graph on jobless claims:

This, of course, makes the current situation look like it is notably worse than the recessions of the mid-70s and early-80s.  But sharp readers will note that both axes are the right-hand axis is measured in absolute numbers of jobless claims, and not jobless claims as a percentage of the total workforce.  Sure there are a larger number of absolute jobless claims right now, but that's because there's a lot more workers than there were 25 or 35 years ago.

I emailed Ed to point this out and he promptly provided a graph to reflect that criticism.  It provides a different story:

Things are hardly good, but in percentage terms, we are still only just over "half as bad" as the peaks of those earlier recessions.  In fact, we're not that much worse than the Bush I recession of the early 90s.

My point is not to say "all is well."  Rather it's to say that we need to make sure we have an accurate picture of just how much of a "crisis" we're really in so as to try to head off the talk of catastrophe that inevitably feeds Leviathan.

Comments

[Roger Koppl]

This is a really important point, Steve. We are in a crisis, but it is a political crisis, not an economic crisis. We are in an unextraordinary economic downturn. That can be very difficult indeed if you are among those cast out of work or house or both. It's a serious matter indeed. It is not, however, sufficient cause to gut the basic system and move to greatly increased political control of markets. I can hear Pete Boettke asking, "Why is this so hard for people to understand?"

[bill]

It appears on the surface that there are jumps around the presidential election cycle where the incumbent was defeated or likely to be defeated. A big swing from Ford to Carter, Carter to Reagan. Steady in 88 where Reagan was re-elected. Jump after the election of Bush 1. Jump again with Clinton in 92. Steady in 96 where Clinton was re-elected. Jump again in 2000 w/ Bush 2, declining in 04 with his re-election. Jump again in 08 to today.

[Dave Prychitko]

Those cast out of work, etc. This is a growing social calamity in the sense that we are likely to see more alcohol problems, depression, spouse abuse, etc. Those in social work see it coming.

What bothers me is that one of my social work colleagues keeps hounding me about this fact -- that we must do something about it by policy programs to get people back to work and so on. That argument, however, assumes that good intentions lead to the desirable results, which is an *economic* claim.

Everybody's an economist today. Everybody claims to know how and why we've gotten into this mess, and how to fix it.

It goes back to Pete's nice phrase: "Ought doesn't imply can, and can doesn't imply ought."

[Steve Horwitz]

Uh, PETE'S nice phrase?! Once again, Boettke gets credit for my work. ;) Even Pete knows better than that: http://austrianeconomists.typepad.com/weblog/2008/09/please-just-say.html

Watch the May 2009 Freeman for the full version of "Ought Implies Can".

[Steve Horwitz]

The original is here: http://myslu.stlawu.edu/~shorwitz/mackay06.htm

[None]

However, both the median and mean duration (in weeks) of unemployment are approaching all time highs (or at least all times highs for the existing data - mean since Jan 1948, median since 1967 - see the St. Louis Fed.) It always rises during downturns, but really prolonged unemployment can really become a serious issue.

Extending unemployment insurance from 26 to 39 weeks will only worsen this issue.

[Dave Prychitko]

As Johnny Carson used to say, Steve, "I didn't know that."

Also, Pete never gives me credit for what he's turned into a classic statement:

"Economics puts parameters around our utopias."

Oh well...

[Dave Prychitko]

Parameters.

He's said it so much, Pete has, that I almost have to credit him with the original phrase -- and insight!

[tom]

If you are going to title your post "why looking carefully at data matter" you should probably look at the data carefully yourself. UE claims on their own aren't nearly as instructive as statistics that include underemployment, labor force participation rates and I would even argue savings to debt ratios.

[Ed Weick]

Steve,

Your point is well made.

Please send me an email with the good news that I am blind to.

Each of my emails is essentially one of many brush strokes that have been painting a developing Misesian catastrophe.

He wrote the book; I am just reporting facts as I see them that fit his theoretical framework.

ED

[Vangel]

The point is valid but there is a broader problem with the reporting of unemployment data; the methodology used to define the term unemployed has been changed. John Williams points out that the U6 report shows an unemployment rate of 13.5% and when he adds in the discouraged worker as they were defined before the methodology was changed the number goes up to 17.5%.