[jordanb]

Yes. The recession began in late 2007. Unemployment didn't start spiking until mid-2008.

What happens is that companies are not anxious to start laying off employees when the economy goes south, for a variety of reasons. So they start spreading less work among their existing workforce. Worker productivity and hours worked both go down.

This makes these leading indicators. On the other end, when the economy starts winding up at the end of a recession, the now-lean companies aren't anxious to start hiring again. So instead they start giving their existing workforce more work, asking people to work more hours, etc. So you have the opposite: unemployment stays high for a while while productivity and hours spike.

Worker productivity is currently "surging": http://news.google.com/news/search?aq=f&um=1&cf=all&...

However, hours worked is "holding steady": http://news.google.com/news/search?aq=f&um=1&cf=all&...

What that means, I'm not qualified to say. And it looks like real economists are (as always) of differing opinions on it. Ideally you'd want to see both of those indicators on the up-tick, but given that one is increasing and the other is not getting worse, it's probably a sign that unemployment is going to start going down here pretty soon.

[SamAtt]

The problem I have with the "Late 2007" start date is it doesn't seem to be backed up with facts. GDP grew in Q1 2008 (http://useconomy.about.com/b/2008/04/30/still-not-a-recessio...) and Unemployment was flat from Dec. '07 to Jan '08 and actually went down a tick in Feb '08 (http://data.bls.gov/PDQ/servlet/SurveyOutputServlet?series_i...)

In fact, GDP continued to grow in Q2 of 2008 (albeit slower) and in Apr. of '08 the Unemployment rate was at 5% (up a mere .1% from Dec. '07).

I was always taught that a recession is 2 quarters of negative GDP growth. Lately that definition has changed to "when economists say the recession began based on rules they made up". The problem with that measuring stick is you are defining behavior based on rules that are supposed to measure behavior. If you've decided the unemployment rate is a lagging factor than you define the Recession as having started earlier so that it fits the rule.

The reality is GDP didn't start to fall until Q3 of '08 and Unemployment had been going up steadily since Apr. of '08 (a full 3 months before the start of Q3). In my book that isn't a lagging sign.

[jordanb]

"Two consecutive quarters of negative GDP growth" is just a "rule ... economists made up" too. And a totally arbitrary (albeit seductively simple) one at that. The NBER look at when the economy actually starts slowing down, not when some arbitrary amount of time has passed after the sign on national GDP growth flipped over. It is more complicated, yes, but also more sophisticated and rigorous than the old rule of thumb.

When it comes to indicators, the fact is that the definition of 'recession' is pretty irrelevant. Your rule of thumb is backward-looking and thus pretty much useless for figuring out how the economy is going to move anyway. You can't know the economy is in trouble until it's already been there for two quarters, eh?

Likewise with unemployment, if you ignore the other indicators you won't know something is going wrong until you're already deep into it (and it starts getting reflected in unemployment data). Whereas if you pay attention to leading indicators like productivity, you'll be able to better predict what unemployment is going to do.

[Agathos]

6 months sounds great. Since the recession is probably over (most recent GDP change was positive, although the NBER hasn't ruled yet), we should have recruiters kicking our doors down by April or May. And then we'll hire carpenters to replace the doors, and everyone wins.

Unfortunately, I think it's fairer to say that the first derivative of the unemployment rate is what actually lags by 6 months. April or May might be when it starts falling (fingers crossed), but it will have to fall for a couple of years before it reaches its pre-recession level. Every economist or seer I've heard of says it will still be around 9% in late 2010.

And that's six months if we're lucky. Consider the last one: recession ends in November 2001; unemployment doesn't really start dropping until early 2004!