Monday, January 11, 2010

Another look at the economy

Just trying something new.  Here is a Google gadget that tracks how Pittsburgh has been performing over time in the recession and compares it to other regions.  You should be able to 'play' the dynamic graphic below.  Each bubble here is one of the 40 largest MSA's in the US.  One axis tracks the unemployment rate and the other tracks the year over year employment growth or decline. The size ot the bubble represents the size of each region's labor force. The bottom line is the farther a region is toward the top-left = good, the bottom-right = bad.  The animation tracks the changes each month.  Kind of cool:


Anonymous Anonymous said...

Two conclusions. (1) Google is cool. (2) As usual, glad we aren't Detroit.

Monday, January 11, 2010 12:45:00 PM  
Anonymous johnnyg said...

Hmm. How come we keep trending down at the end of 2009 when everyone else seems to be bouncing back up? And, really, we've lost more jobs in this recession (as a percentage) than Cleveland? Are there bad local job numbers right around the corner?

Monday, January 11, 2010 3:19:00 PM  
Anonymous Queensryche said...

can you explain the difference between "Employment" under "Labor Force Data" (which you used for this graph... -3.6% for PGH in Nov 09) and "Total Nonfarm" under "Nonfarm Wage & Salary Employment" (which is -2.3% for PGH vs. -4.3% for Cleveland)?

Monday, January 11, 2010 5:13:00 PM  
Blogger C. Briem said...

Queensryche as data wonk.. I like that.

Could I explain it.. mostly.

Could I do it briefly, probably not.

A short answer. Some of the differences are methodological, but fundamentally the two numbers measure different things. I know people treat them much the same, but they shouldn’t.

The non-farm employment number is a count of jobs. The labor force employment number is a count of people. Roughly 5% of folks have multiple jobs so you have to take that into account. One series counts self-employed the other doesn't. One series counts certain unpaid family workers, the other doesn't. There are some farm workers in the mix. One comes from a survey of employers, the other comes from a survey of people. An establishment survey is by definition a count of jobs by place of work. The other counts employment by place of residence. Commuting can cause a big difference between employment by place of residence vs. place of word.

What is going on here.. with respect to Cleveland by itself or vis a vis Pittsburgh. I really wonder about Cleveland labor force data over the last year. There is this period of completely flat unemployment rate during the peak of the recession. Hard to imagine that is really what was going on.

Monday, January 11, 2010 6:17:00 PM  
Anonymous Anonymous said...

Why is annual change important? I think a region with low unemployment rate but a worse employment change would be better off than the other way round (though the two are obviously correlated).

In particular, Pittsburgh's precipitous drop in 2009 (relatively bad annual change) "looks bad" but is probably meaningless(?)

Monday, January 11, 2010 8:41:00 PM  
Anonymous Queensryche said...

I imagine there's some geographic differences in those two metrics as well...

say you live in Burgettstown but your job is located in Weirton... you're counted in the "Employment" total for Pittsburgh MSA... but your job would be counted in the Weirton-Steubenville MSA in terms of "Non Farm Employment", right?

Monday, January 11, 2010 9:17:00 PM  
Blogger C. Briem said...

@Queeensryche.. yes. But that is not the only distinction between the two time series.

on the characterization of the unemployment rate vs. employment change: big issue.

Monday, January 11, 2010 11:13:00 PM  

Post a Comment

Links to this post:

Create a Link

<< Home