Tuesday, August 28, 2012

Labor Force Metrics 101

So the stats out today say the region's unemployment rate is up 3/10ths of a percent.  I wonder a bit. While all data out there has issues, I don't like poking at methodology too much since it lends itself for folks coming to doubt the data.  You just need to understand the strengths and weaknesses of virtually all data you look at other than maybe the spot temperature outside as reported.  I guess we can trust that for the most part... or maybe not?

Still, I don't quite believe it.  There is an unspoken truth that what we know of the monthly metro-area unemployment rate is less than what you might infer given all the news coverage of that one factoid.  There is no time for the full seminar one could have on the subject, but the labor force data you see monthy is from the Local Area Unemployment Statistics (LAUS) program.

Read just one snippet from the LAUS methodology. Skip down to the section subtitled "Estimates for substate labor market areas" and skip the first paragraph which wil not apply to us.  The whole think will explain certain issues, but in particular read this paragraph on counting the number of folks unemployed:

The second category, "new entrants and reentrants into the labor force," cannot be estimated directly from UI statistics, because unemployment for these persons is not immediately preceded by the period of employment required to receive UI benefits. In addition, there is no uniform source of new entrants and reentrants data for States available at the LMA level; the only existing source available is from the CPS at the State level. Separate estimates for new entrants and for reentrants are derived from econometric models based on current and historical state entrants data from the CPS. These model estimates are then allocated to all Labor Market Areas (LMAs) based on the age population distribution of each LMA. For new entrants, the area’s proportion of 16-19 years population group to the State total of 16-19 years old population is used, and for reentrants, the handbook area’s proportion of 20 years and older population to the State total of 20 years and older population is used  (emphasis added)
That paragraph and a few earlier parts get into this big issue that what is going on in the metro area is pretty much presumed to be the same as is happening across the entire state.  It is never true, especially for Pennsylvania.  It is not a flaw, you do what you have to do given the limited data, but at times like this it is important to understand the methodology.

(and mostly for frequent commenter BrianTH here)  This also is why the reported labor force data for the City of Pittsburgh proper is worth taking with a grain of salt.  The methodology for reporting municipal area labor force data is in the following section of the methodology.  If the metro area data has issues, those extrapolations (warranted or not) become ever more acute in the small areas with labor force data being reported on.  I mean, the state actually reports a change in unemployment rate each month for Ross Township.  Rest assured neither they nor any of us have any meaningful idea what happened in Ross Township last month, up or down to any precision worth reporting on. 

Where would there be uncertainty in the unemployment rate?  Most folks believe the unemployment rate is counted by just adding up folks getting $$ benefits from the state and dividing.  Not true.  All the folks looking for work are counted as unemployed including those who either ran out of benefits or lets say you don't qualify for benefits at all because you are just entering, or reentering the labor force.   We can count the folks getting benefits, but the number on others being counted as unemployed have to come from other data sources which if you read the methodology you will see are often extrapolated from state patterns.  Actual benefits-receiving unemployed are well less than half the total 'unemployed' in Pennsylvania, so it really matters how you add count them all.   If you really dig into it, go look up the new (initial) unemployment claims in Pennsylvania for the week ended July 28.  I see a number that is the lowest it has been in just about 4 years. Yet the state's unemployment rate is moving up at a decent clip.

So what does that all mean?  Probably the harder number of what is happening in the metro area economy each month is the non-farms jobs number coming from the Current Employer Survey (CES) which has no major extrapolation problem.  Sample error in that, but not such a big assumption that what is happening locally mirros the state.  What does the CES say these days?  Pretty much we are pushing to new employment highs here.

Still.. what is odd in the data just out is what it says about the region's labor force.   If the labor force really went up by 7,700 over the month then we are well into record territory for the size of the region's labor force.  It is really odd when you consider most would say it is one of those quasi rules that labor force participation should go down if the unemployment rate is shooting up... yet the opposite would nominally appear to be the case in Pittsburgh.  Of course it is not the case.  I will repeat that labor force participation rates move much more glacially than most assume when trying to explain these monthly shifts in the numbers.. 

To put a point on the labor force story (or lack of story as it were) take a look at the updated graph of where we are going...

Curious trend there over last couple of years for the region... a trend which is even more surprising when you consider it is not really supported by any trend in local demographics.  I went into the longer parse on that last month when we reached a new all-time labor force peak for the region. 

So if you read any of that above.  My personal take is one of two things is happening.. or one of two things almost have to be happening.  Either the non-benefits receiving unemployed count is being overestimated for the region because it is an extrapolation from state data, which would make the unemployment rate look higher than it really is... or....  there really is a sizable flow of new workers moving into the region.  Each would make the headline today a bit different.     If I am wrong and the common wisdom is right that the whole answer here is that the labor force participation rate just in Pittsburgh is shooting up??  Then that is a big story unto itself since someone would have to explain why Pittsburgh of all places is going against national trends, but also against the pattern almost always true that the unemployment rate and the labor force participation are negatively correlated. 

and when I say that more people are moving here, I of course mean in terms of net migration.  So it could be that fewer are leaving.  That may be the nexus of all of this.  Since we are such a prodigious producer of college graduates, and given the semi-anemic state of the economy in a lot of regions folks.. the trends may all be reflecting a larger number of recent grads unable to find work elsewhere and are remaining in place here. That would all be quite consistent with this recent surge in the region's unemployment rate coming right as the school year ended.  If true then Border Guard Bob was successful in the end.  Of course he had to tank the national economy to make it happen.  Who knew he went to work for AIG after he was fired here? 


Anonymous MH said...

Maybe it's just me, but there has been a huge increase in the amount of stupid and stupid-aggressive driving lately. I'm guessing new people are moving here, probably from the greater NYC area.

Tuesday, August 28, 2012 10:32:00 AM  
Anonymous BrianTH said...

If this is turning into a dialogue, I have a bad feeling I am not being cast as the Socrates.

Anyway, I think I understand the issue. Looking back and forth at the state and Pittsburgh components of labor force, it looks like the unemployed count in Pittsburgh is tracking pretty well the unemployed count for the state, and that may be an artifice of the methodology.

Meanwhile, though, the employed count in Pittsburgh is NOT tracking the state count, and that appears to be why overall, the Pittsburgh labor force count is outgrowing the state labor force count.

Which leads me to a question--couldn't both things be true? The unemployment rate may be overstated because they may be overestimating the unemployed, but the employment count may also reflect jobseeking migrants coming here AND actually getting jobs.

In other words, forget the unemployed count, but just look at the record employed counts. Is it plausible that is ALL because the real unemployment rate is fantastically low?

Tuesday, August 28, 2012 5:55:00 PM  
Anonymous MH said...

I have a bad feeling I am not being cast as the Socrates.

You know what happened to Socrates in the end.

Tuesday, August 28, 2012 11:31:00 PM  
Blogger C. Briem said...

No hemlock here...

But to be clear, I do believe the regional labor force surge is more the result of net migration. Otherwise you have to conclude Pittsburgh has a unique surge of labor force participation among in-place residents all while the unemployment rate is jumping up. That would be really odd and noteworthy.

All I wanted to point out is that given the details of the methodology, there is just nothing in the labor force data telling us much about what is happening to the city resident labor force. If one agreed then the inference of population gains of late (for the city proper)would also be suspect. Maybe it is happening, maybe not. Just that the data is really a bit to contrived at how it breaks the data down to the sub-county level to really say much.

Wednesday, August 29, 2012 7:11:00 AM  
Blogger Laura Kathryn said...

Hi.,We looked your post informations,It was really wonderful stuffs to seen in your blogs,Share more informations.Thank you.
laser hair removal bc

Friday, August 31, 2012 4:08:00 AM  

Post a Comment

Links to this post:

Create a Link

<< Home