Sunday, October 02, 2011

Peak Gas... or The metrification of Washington County

So I've been thinking of this story on the data showing Washington County, PA as one of the biggest job growth stories in the US over a recent 12 month period.  NPR's Stateimpact and Jim R. caught it, then the Trib followed up on the factoid that popped up of Washington counting ranking 3rd in percentage of job growth in the nation. 

A good story? Sure.  Related to Marcellus Shale.  Yup.   But still not quite the story it seems for at least two reasons.  Using the exact same data source as the stories are talking about, here is the monthly trend in Washington County employment going back a few years.  You can see the big jump in the last year that resulted in the headlines.  Yet part of that connects to the really sizable and successive drops in the immediate years before. I thought we were a few years into the Marcellus bump? I don't recall any stories on those big drops. Looks like there was a long term growth trend that pretty much stopped in recent years and the very recent gains may be seen as some catching up.  Those who like to infer causality a bit too quickly might even say the arresting of growth in the county seems to coincide with the arrival of Marcellus-related employment.  Of course there was this little recession along the way. 

But there is something else. Despite the claims that the data showed Washington County to have one of the biggest gains in employment among all counties in the US.... it is a bit narrower than that.  The actual data in question only ranks the 323 largest counties in the US.  When you rank US counties by employment, you get that Washington County and Butler County, the very same two stories mentioned in the local coverage for their rapid job growth of late, are literally tied for 311th.  So they are very barely making the cut to even be included in this data at all.  Since the metric being talked about is percentage growth it makes a big difference when you realize we are comparing some very small counties (80K employment each in those two counties) with some very large counties (Los Angeles at the top has over 3 million workers).  Consider that if you added every worker in Washington County to Los Angeles it would not generate the percentage growth there that is being reported for Washington County here.

Actually..  I just realized.  Washington County gained 4K or so workers over the year per this data.  If it had not gained that much employment it would still be down at 75-76K workers.  75,100 is the current cutoff for inclusion in this data.  So they really are barely making it on to the list.

Finally..   Where are all the other counties being impacted by shale gas development in Pennsylvania? No mention of them in this data or related headlines.   Consider that virtually none of the Marcellus impacted counties in PA are large enough to even come close to being on the list of the top 300 or so counties in the US.   The only other one I see in the list is Luzerne County which is at the crossroads of a lot of Marcellus development.  Did it have big job growth in the data?  +1%, or less than that of Philadelphia County if you want a benchmark for a region minimally impacted by shale gas employment. 

Wait, I see Lackawanna County there in the data. That is deep into Marcellus activity. What was their job growth over the same period? 0.4%.... decline! Some folks missed that. Sure isn't mentioned in the Marcellus Shale Coalition's PR on this data. Funny that. Must not be the place with the zero % unemployment rate.  That by the way is the worst job number reported for a Pennsylvania County in this data and one of the worst performances nationwide in the data being reported on.  Seems like that might be worth a headline in a local paper up there.  I may need to check the Rolodex,

Given all the talk, let alone the zero unemployment rates....  How can there possibly be an employment decline of any kind in the heart of Marcellus Land in Northeast PA?  Hmmmmmmmm.   That would be my longest hmm yet.


Anonymous The Wiz said...

Much of the area in N Central Pa where the Marcellus rush is on also has other industries, a couple of which has been impacted by the recession. There is a substantial logging industry for hardwood logs....oak, cherry, walnut, and more. This has been severely hurt by the collapse of the housing market. There is also a major(for them) powdered metal industry in the Dubois, St Marys, Emporium area that supplies the auto industry and it too has been hit hard.

If I were the Congressman's PR dude, I would try to spin his "virtual 0%" comment. In my Econ 101 class many moons ago, the prof said that there will always be around 4% unemployment even in the best of times as companies, products, and jobs naturally come and go. So he said that a 4% unemployment rate is considered "full employment" Thus I would spin that as what the Honorable One meant.

Of course even that's not true as no county is anywhere near 4% but at least I could cash my paycheck knowing I did my best.

Interesting how your graph looks like my last EKG. The annual rise and dip is nearly identical throughout the timespan. But the timeline shows a consistent increase in total employment until the 2008 Big Bang. And is just now about back where it was prior.

Sunday, October 02, 2011 9:38:00 PM  
Anonymous Anonymous said...

Astounding lies always get you better press than sober truths.


Sunday, October 02, 2011 11:00:00 PM  
Blogger C. Briem said...

no way one could infer full employment from that zero employment statement. If that was what he meant then the US hit 'zero unemployment' at the end of the 1990's. Guy must really love Bill Clinton.

let's hope that isn't your EKG.. which if it was ramping up like that you will never make it to the yacht.

Sunday, October 02, 2011 11:02:00 PM  
Blogger C. Briem said...

and for the record, none of these counties are near 4% unemployment.. most are quite high compared to their own historical unemployment rates. Tioga: 8.2%, Lycoming: 8.6%. Are these not some of the places with the most Marcellus pads?

and I don't quite get the point about the other industries in those areas. You are countering your own point about 'full employment'. and anyway is not the job demand being caused by Marcellus so large that it would more than sop up any of the unemployed folks from those other industries? Realize places like Tioga county, which has one the largest numbers of Marcellus pads, is all of 20K workers. So if 200 workers are hired it counts at 1% of the entire jobs count in the county. If those workers were pulled off the unemployment rolls, just 200 jobs would bring down the unemployment rate by almost a full percentage point in current conditions.

Sunday, October 02, 2011 11:35:00 PM  
Anonymous The Wiz said...

I just said I'd spin it that way as a PR dude. And I said "Of course even that's not true as no county is anywhere near 4%"

That anyone would say they had zero unemployment is incredibly stupid. And for other groups like the MSC to repeat it is even stupider because it does serious damage to their credibility on any other statements they put out.

But those areas had a very high unemployment rate in 2009. Their other industries were hit very hard. Can't expect one industry like gas to hire everyone that lost their job.

Other problems with the data; how many of those are out-of-county employees? These guys move around a lot. You could get hired in Washington Co and work in several counties in a years time. And many are from out of state as they have the training and experience. So if 2000 are unemployed and 200 are brought in from other areas, the impact is diluted due to the higher total workforce numbers.

And this a very specialized business. One crew spends three or four weeks drilling and then moves on to the next well pad. There are often two different drilling rigs used that split the time on each pad. Another crew spends a week or two fracing and then moves on. There are crews for surveying and design, wellpad construction, a completion crew, a pipleline crew, and more. How does this get figured into the numbers when they spend such a short time on each pad?

Its a pretty transient business that is highly fractured. Makes it hard to quantify the affects. And when groups like the MSC are willing to distort the picture, it makes it all the harder to nail down.

Monday, October 03, 2011 9:49:00 AM  
Anonymous MN said...

I really just can't believe that stat is being reported and repeated. My very low opinion of journalists is now even lower. Same goes for politicians.

Anyways, I read the BLS releasae last week before anyone even picked this story up and even I didn't catch Washington County listed. I should've. But I guess I missed it because I wouldn't expect Washington County to be included.

But... I suspect some of the Marcellus activity isn't helping employment in some of the smaller counties (even those with high Marcellus activity) because companies and their employees are not really located in those counties. They spend a lot of time traveling all over the state and even into other states, but a lot of "camps" and headquarters are located in SWPA. And some counties are so small they don't even have enough hotel rooms so crews will stay in hotels hours away from well sites. Just a thought.

Monday, October 03, 2011 3:33:00 PM  
Anonymous MN said...

And Butler County is also included in the list of large counties with the highest percent increase in average weekly wages for the 1st quarter 2011 at 9.3 percent. Anyone care to pose a theory as to why?

Monday, October 03, 2011 3:38:00 PM  
Anonymous MN said...

And you know, Katie really should know better.

Monday, October 03, 2011 3:48:00 PM  
Anonymous The Wiz said...

Butler Co; Natgas wells and the Westinghouse nuclear research center

Tuesday, October 04, 2011 10:29:00 AM  
Blogger C. Briem said...

Butler County planning: more Westinghouse than Marcellus -

Tuesday, October 04, 2011 10:34:00 AM  

Post a Comment

Links to this post:

Create a Link

<< Home