Thursday, January 12, 2017

Pigskin Mythos

Maybe this blog will live on with only meta commentary looking back on past thoughts. But on this day the Pittsburgh Steelers began an era. Defeating the Minnesota Vikings (16-6) on January 12, 1975, the victory was the first of 4 Steelers Superbowls over just 6 years.

There continues to be this persistent mythos that the Steelers fans during those first Superbowl years were largely made up of vast number of unemployed workers who were cheering the team on as their jobs were being eliminated en masse. It remains to me an odd memory that only lacks in any economic data to support it.  So strong is this whole idea that whole books (by respectable authors)have been written on the topic. I have written about this here many times (for example in 2010: The Economics of Cliff Stoudt), but have had this conversation with innumerable folks who all, almost universally, disagree with me. Often quite vehemently.

Let’s start with this factoid.  Here is a short table of the unemployment rates in Pittsburgh and the United States during the months of each of Pittsburgh’s Superbowl victories.

Superbowl – Pittsburgh Unemployment Rate – US Unemployment rate

Superbowl IX (January 1975)      6.4%      8.1%
Superbowl X (January 1976)      6.7%      7.9%
Superbowl XIII (January 1979)      5.8%     5.9%
Superbowl XIV (January 1980)      6.8%      6.3%

So during 3 of those 4 games, the local unemployment rate was well below the national rate, and in two cases far below the national average.  If you want to take the last game as an exception, here is something to think about.  Pittsburgh MSA employment reached 1,099,600 jobs in February 1980, or just a few days after the Superbowl was played.  That was actually an all-time high in the region’s employment count.  By that I mean an all-time high going back as far as there is consistent metropolitan data reported in 1946. 

Not convinced or think, as some have tried to argue with me, manufacturing employment was collapsing at the time even if overall labor force stats were more positive?  I once posted here the manufacturing employment levels across the 10 counties of southwesternPennsylvania during the 1970s. In 1974 (when those games leading up to Superbowl IX were played), SWPA averaged almost 316 thousand manufacturing jobs, an increase over the most recent years and likely a peak going back over a long time before then.  In 1979 the same region still averaged over 300 thousand manufacturing jobs, about the number the region had at the end of the 1950s if not higher. There had been no big plunge of factory jobs here (yet). 

Still, there are few things I’ve been less able to convince people of than this.  Numbers don’t matter and I’ve had very thoughtful people just say it does not add up to them. What they remember is something very different and contradictory. It's not a question of memory persisting, its a memory that only came later. This may all mean something to recent political news as well, but I will leave that to other pseudo-pundits. 

So why the disconnect? I think the 1970s were a traumatic economic time with multiple oil embargoes making gas prices jump, high inflation and the dreaded ‘stagflation’ or inflation without growth.  Certainly those things impacted everyone in Pittsburgh, but they impacted all Americans.  Did manufacturing have slow times during the two national recessions of the 1970s? Yes, but even that has to be taken into context.  Manufacturing employment had always been highly cyclical and fluctuated with the national economy.  When national unemployment rates went up, unemployment in manufacturing industries, and manufacturing regions, went up even higher. Pittsburgh experienced multiple periods of double digit unemployment rates many times during the 1950s and 1960s, yet never even came close to a a double digit rate during the 1970s. Even in the two national recessions of the 1970s, US Steel - then still the region's largest employer - did not fare badly during the decade (again, strongly against popular belief).  Don't take that from me, here is a contemporary quote from then US Steel Chairman in 1975 saying the recession the nation was then in the middle of was "Steel's best recession ever."   

But history has many ironies.  In the very month the very last of those superbowls, Superbowl XIV in January 1980, the NBER later recorded that month as the beginning of a new recession.  Though it was a short recession, ending officially after just 7 months, there was only a brief respite before another recession started in 1981.  Some consider the two just part of one big economic downturn, but one which see the palpable collapse of so many local steel jobs.  But another Superbowl victory did not come for decades. 

Anyways…  playoff time is a funny time in pundit land.  The merging of sports and general opining can get kind of funny this time of year.   Who remembers our friend Bill Johnson from Denver who parachuted in to write a story criticizing Pittsburgh aesthetics during the championship game in 2006.  I am sure we have a few alien reporters among us from Kansas City right now.  

 


Create a Link

 

Monday, January 02, 2017

Population past - population future: migration and Pittsburgh

Reading the PG piece parsing some migration data from American Community Survey: Philadelphia to Pittsburgh, a moving trend.  The punch line: the data shows a net gain of population due to migration from Philadelphia to Pittsburgh (metropolitan areas, respectively)

I figure I should update my graphic tracking the relative difference between the unemployment rates in Pittsburgh and the United States.  Why do I care?   Mostly because when it comes to economic migration within the United States, this is a pretty solid predictor of net migration rates across metro areas, or at the very least it has been for Pittsburgh.  '

The PG piece used some data the Census gnomes have put out looking at migration between metropolitan areas over the period 2010-2014. That happens to be at the end of a unprecedented period when Pittsburgh maintained a streak of monthly unemployment rates below the national average.  Basically for 99 continuous months, the local unemployment rate was below the national average. For 109 months the local unemployment rate did not exceed the national average (there were a few moths in there where the two rates were the same.)   Few really noticed, but I do not there was any comparable period that was true for a longer period of time since metropolitan region unemployment metrics were regularly reported in the late 1940s.  Think about that. 

Unsurprisingly for most of those years the Pittsburgh MSA showed positive net in-migration, also an unprecedented demographic trend to be sustained for Pittsburgh since the 1940s. 

Alas, it probably isn't true right now.  See the  trend really was broken in February 2014, and since then the local unemployment rate has been above the national average.  That isn't the end of the world, but what is problematic is how far above the national average Pittsburgh's unemployment rate has been.  For both October and November, Pittsburgh unemployment rate has been 1.2 percentage points above the national average. That also is unprecedented in that you have to go back to the mid 1980s (not a good period for Pittsburgh) when the local rate was so far above the national rate.  As long as you have that type of differential, you will have folks finding better opportunities elsewhere and you will see working age population migration follow.



The future??   Pittsburgh's unemployment rate dropped 0.3 percentage points between October and November.  That, by the way, was the biggest month over month drop in 16 years one of the biggest month over month drops in decades. So a good sign for the local economy and a sign the local job market is tightening.  But the national unemployment rate dropped by the same 0.3 percentage points at the same time.  As long as you see such tight labor markets across the nation, and so much tighter than here, you just can't expect there to be net migration flows like what we saw over those 109 months where Pittsburgh fared at least we well as the nation.  Pittsburgh may do better than it has in the past, but there are just too many places with too many opportunities to expect Pittsburgh to pull folks given the relatively higher unemployment rate here. 

For now and for the next couple years at least - remembering most data like the Philadelphia migration story today, are pretty much backward-looking due to the latency of data coming out - the story will be of population loss due to net out-migration and continuing natural population decline (deaths exceeding births) endemic to Pittsburgh since the 1990s.  Lots of other local stories follow from all of this... Allegheny County population trends... city of Pittsburgh population trends and all sorts of real estate trends.  So....   watch this space?  Maybe,    




 


Create a Link

 

Monday, December 12, 2016

Uninsurance trends across Pennsylvania

I saw this data journalism story out of Michigan, Mlive.com: See how uninsured numbers have dropped in your Michigan county, and didn’t quickly see anything like that for Pennsylvania. So just if you are interested here are the comparable numbers for Pennsylvania counties. As noted for Michigan as well, this data (American Community Survey 1-year estimates) only reports annual data for larger counties. I will leave for others to punditfy.

Percentage of Population with no health insurance coverage
Large Pennsylvania Counties, 2010 and 2015
PA County 2010 2015 Change
Adams 9.5% 5.3% -4.2
Allegheny 8.0% 4.8% -3.2
Armstrong 8.5% 5.9% -2.6
Beaver 7.4% 3.6% -3.8
Berks 9.8% 6.9% -2.9
Blair 10.8% 4.3% -6.5
Bucks 6.7% 4.8% -1.9
Butler 7.6% 3.2% -4.4
Cambria 9.1% 5.0% -4.1
Carbon 14.5% 4.8% -9.7
Centre 8.6% 5.7% -2.9
Chester 8.6% 5.2% -3.4
Clearfield 11.1% 5.3% -5.8
Columbia 8.2% 3.7% -4.5
Crawford 14.3% 9.8% -4.5
Cumberland 7.1% 6.1% -1.0
Dauphin 10.3% 5.5% -4.8
Delaware 9.2% 5.2% -4.0
Erie 9.6% 4.9% -4.7
Fayette 12.5% 5.3% -7.2
Franklin 12.7% 9.9% -2.8
Indiana 11.6% 8.5% -3.1
Lackawanna 10.4% 5.7% -4.7
Lancaster 13.3% 11.4% -1.9
Lawrence 11.0% 5.2% -5.8
Lebanon 10.3% 9.5% -0.8
Lehigh 10.7% 7.0% -3.7
Luzerne 11.0% 6.0% -5.0
Lycoming 10.9% 5.0% -5.9
Mercer 11.3% 6.6% -4.7
Monroe 11.6% 7.3% -4.3
Montgomery 6.5% 4.3% -2.2
Northampton 9.3% 4.9% -4.4
Northumberland 11.5% 5.4% -6.1
Philadelphia 14.9% 9.7% -5.2
Schuylkill 9.4% 7.8% -1.6
Somerset 12.8% 8.1% -4.7
Washington 8.6% 4.1% -4.5
Westmoreland 7.9% 4.2% -3.7
York 9.1% 6.1% -3.0
Compiled from:
Census Bureau - American Community Survey (ACS)
1-year estimates data

Create a Link

 

Friday, December 02, 2016

unemployment trends (how is that for a boring title)



This is the first Friday of the month and all economic data geeks know that monthly labor force data was released by the Bureau of Labor Statistics.  This includes the headline unemployment rate number that most of the business news keys off of.  So today the national unemployment rate dropped to 4.6% which is the lowest it has been since in 2007.

Something I’ve always wanted to do, and I am sure others have done this – though I can’t find quite the same graphic out there.  How has the national unemployment changed over the course of recent presidential administrations.  Since 1948 there have been 12 presidents. Truman was of course already in office in 1948, and Obama’s term is not quite over.  Of those 12, if you look at literally the month they took office (usually January data of the respective year) to the January they left office (or other months were history intervened) I have graphed the trends.  6 presidents saw the unemployment rate decrease over the course of their administration. 5 presidents say the unemployment rate increase.  6 + 5 = 11 so the 12th.   President Carter’s administration both began and ended at 7.5% unemployment, so no change. 

Here is what I get, and I've split gainers and decliners into two graphs for presentation:




Caveats as always.  I've included the inauguration, or other form of transition, month in both the beginning or ending of each time series.  The bigger question is whether it is fair to attribute - implicit in this type of graphic - unemployment rate changes to a particular president in the first few months of their administration.  They say the American economy is more cruise ship than speed boat, which means it take a long time to turn.  Probably fairer to redo these graphs based on a 6 month lag to the month of taking office.  Maybe for another day.

Create a Link

 

Wednesday, November 30, 2016

When lack of data is data - Coal employment in Pennsylvania

A few months ago I threw up here (Coal dropping even faster than you think) this graphic of the rapid decline in Pennsylvania's coal industry. This was the time series of coal employment in Pennsylvania through the first quarter of this year.




The latest quarter of data is out (for Q2 of 2016).  But as is shown in the graphic below, there is no data being released.  Data is suppressed for confidentiality reasons for a couple of different reasons.  If the number of firms is small, or the concentration of employment in any one firm is so high, then the release of this aggregate data may reveal proprietary data about an individual firm.






So I can't update that graph with the latest employment data, but something must have materially changed from the previous quarter for the routine suppression to kick in. I guess a continuing, but slower decline.... maybe 3% employment loss for the quarter, but I am really interested in the trend for the number of establishments. The decline in number of establishments is being reported for Q2 and is down over 5% over the quarter. At 172 establishments that is probably the lowest count for coal mining establishments in Pennsylvania in the 'modern era'.   For the truly wonky labor analyst/historian out there this is all something new.  Never before has statewide coal mining employment (NAICS 2121) been so small or concentrated to ever require statewide payroll employment data to be suppressed. That right there tells you a lot and is a seminal - if incredibly wonky - point in economic history.

Create a Link

 

Thursday, November 17, 2016

Benchmarking Metro Grantmaking

Just passing this on. I am surprised there is little notice of this locally given how active the local foundation community is.  But recently the Federal Reserve Bank of Atlanta has put together a neat piece of data analysis:  Following the Money: An Analysis of Foundation Grantmaking for Community and EconomicDevelopment.

I admit I have wanted to do some of that type of metro benchmarking on foundation grantmaking, but never got around to it.  Their interactive online data is worth exploring, but is limited to showing only a few areas at the same time.  So I put together their data for the 30 largest MSAs to see how Pittsburgh compares.





Create a Link

 

Friday, November 04, 2016

Post Soylent Burgh update

I have been conducting a natural experiment of sorts here.   I was wondering if this random corner of the 'Burghosphere (remember that term? Others do) remained quiescent, would others pick up commenting on such obscure things as I have obsessed on here?  The results are mixed, but lean toward no.

So here is one of those stats that is an awfully lot more important than you might think given it has had so little notice.  The latest update (with data for 2015) on natural population change for Pennsylvania municipalities has been available.  Following up on past posts here, this is the trend in births and deaths for City of Pittsburgh residents.





Take for what you want.  The shift back to natural population gains (births > deaths) from natural population decline (deaths > births) happened about a decade ago and continues. In itself that has been a positive for population trends in the city of Pittsburgh, but the scale of  the net gain (births minus deaths) is narrowing. I believe that in the long run city of Pittsburgh population will depend more on migration trends, but still the natural population trends are important.

The trend in number of births to city of Pittsburgh residents continues a slow but steady decline.  Not immensely below where births were a decade ago, before the number bumped up a bit, but not showing any sign of trending up at the very least.  The implications for that number span across things like future school enrollments and projections for the number of families living in the city proper. Also the fact that there has been steady natural population gains is important context for interpreting overall city population trends, which have shown more stability this decade than in decades previous.

Caveats as always. 2015 data is clearly labeled preliminary by the state.  There has been some upward revisions in past preliminary data on births, so the 2015 data may show up as higher in the end.

Create a Link

 

Sunday, October 16, 2016

Credit-worthy Yunzers

So here is a factoid of note.  Completely incidental to a recently released report from the Urban Institute focused on Detroit is a table of data on the credit worthiness of residents of 'Rust Belt' cities.  While it is interesting to see what others define as the list of places that constitute the 'Rust Belt,' what jumps out is how Pittsburgh compares.  Here is a figure made from their data:




The metric there is the percentage of credit scores that rank in what is considered 'prime' or above, so higher numbers are better. Prime credit scores representing better credit worthiness.  The source is a the recently released Urban Institute report: The Fiscal Health of Detroit Residents.  Pittsburgh has the highest percentage of credit scores that have "Prime" credit scores.  Conversely Pittsburgh has the lowest percentage of residents with "Subprime" or lowest ranking of credit scores.  But in almost all measures Pittsburgh compares favorably to all the other cities. Take a look at the report to see some other factoids worth thinking about more. For all the attention to student debt, the percentage of student debt in collections is pretty low (2.8% for Pittsburgh).  Pittsburghers also seem to have the lowest credit card utilization ratio and lowest delinquency rates and what may be even more significant, compares favorably to the national average in almost all metrics identified.

I guess my biggest caveat to interpreting the numbers there is that the cities grouped together might all be "Rust Belt" cities in some sense of history, but differ markedly in size and any recent history.  For example, the City of Detroit is over 680K, while the city of Pittsburgh hovers, arguably, around 300K. To be clear, the report was focused on Detroit and not intended to be a comprehensive benchmarking exercise.  But I do wonder a bit how the cities would compare if the geographies were redefined a bit more consistently, but that is harder to do than to conceptualize. I also have to wonder about how age and student populations impact these numbers. Might be worth someone to look into more? Nonetheless, (City of) Pittsburghers seem to be far more able to access credit markets than folks in Detroit and have better credit than all those specified regions and that has to have a lot of implications across the board, not the least of which is in real estate markets. 

Create a Link

 

Tuesday, October 11, 2016

Saying Goodbye to the Greatest Generation

It is one month until Veterans Day 2016.  I just thought it worth noting just how few World War II veterans are still with us.  In Allegheny County the number of World War II veterans once exceeded 200K, but only small fraction are still with us.  I know it has been a long goodbye in a sense, but maybe it is worth noting there are WW2 vets around. The estimates available from the Census are a little above 6K just in Allegheny County as of 2015.  As 2016 draws to the close, I am sure we are down well below that number.

A more amazing statistic.  This is a number so small that survey error makes it hard to say for sure it is not actually zero, but the estimate for 2015 is that there were 41 Allegheny County veterans who served in WW II, Korea AND Vietnam.  I think for Veterans Day this year - and the ulterior motive for posting this a month early - we should find any local vets who served in all three of those wars.



Create a Link

 

Sunday, September 25, 2016

Lawrenceville Redux

The PG appears to have a short series on the past and present of Lawrenceville, a Pittsburgh city neighborhood written about even in the Financial Times (April 24, 2009: Diamond in the Rust) and many other media venues in recent years.  Trust me more will be written on the causes, impacts and eventual end of the transition of Lawrenceville before it is over.  For me one key point is that very little of the history of Lawrenceville of late can be separated from what once was, and what became of St. Francis Hospital. With full disclosure it was also my place of birth.

Just a few years ago, the solicitations I myself received attest to the plans that at least some companies wanted to frack the land underneath the neighborhood, which I found amazing because the neighborhood still had some of the highest population densities in the region. Would the Lawrenceville story be the same as it is today if any form of that drilling had proceeded? A counterfactual that will be left to conjecture.

Still, the biggest change in Lawrenceville over the last 2 decades has been the demographics.  Just a decade ago, some still looked to Central Lawrenceville as one of the oldest neighborhoods in the nation. So old that the area was identified as a Naturally Occurring Retirement Community (which has its own acronym now: NORC).

But now the story is about the youth inhabiting the neighborhood, and indeed that is the crux of it all.  From the Nullspace archives(2015: All things Larryville - long past inflection), here is the time series that says it all:



Of course the other story of Lawrenceville is real estate, which is the same story as the demographics really.  People do live in houses and all of that.  But for one story of property appreciation in the neighborhood just take a look at the price history of 5309 Duncan St which was sold, looks like via a foreclosure, for a pittance in 2010, then sold for more than an order of magnitude higher (but still only for $26K) in 2011 and then for $225K in 2013 and I can only speculate that its current market value is somewhat higher.  Makes you want to go back and look at all the complaining about the values set in the last countywide property assessment in Allegheny County.  The map showed big jumps in property values in Lawrenceville for sure, but I will bet the values (which remain in place today under the county's base year system) are a fraction of most market value transations. 
 
The bottom line is something else. As much as the story seems one of the here and now, be assured that the redevelopment story in Lawrenceville has built upon a lot of work in the past.

Create a Link

 

Wednesday, September 21, 2016

fracking Pennsylvania update

So here is something you don't always see, but an economic time series running into a wall. Below is an update of my contrived time series of mining employment in Pennsylvania net of the Pittsburgh MSA, or a measure of what is happening across the vast Pennsylvania "T" (since Philly has little mining employment to begin with).



In short, the collapse in this time series came to an abrupt end in August. Absolutely no change from July.  We kind of skipped the inflection point altogether.  I do admit I was expecting the trend to continue downward, though it could never have continued at the same pace. But you rarely see a time series just come to a halt so abruptly. What happens next?  A big turnaround?  Dead cat bounce? Or this was just a blip in a longer term trend?

As I said last month, the time series showed that pretty much all employment gains in mining employment from before the expansion of alternative gas development were gone, or at least down 98% from the employment average for 2005. It is true that the Pittsburgh MSA has retained some of its gains in mining employment since before the big frack began, but that series has been on its own massive collapse as well.

So as the big Shale Insights conference kicks off here in Pittsburgh, the news is good or bad depending on how you want to take it.  The collapse has ended, quite definitively at least for this month, but still all past employment gains remain eviscerated.  I do think there has been some gains in natural gas employment across the state, but just barely enough to offset the resulting collapse in coal employment in Pennsylvania, now at levels not seen since before the civil war. 

Funny that there are no longer any economic impact reports on fracking being produced, despite the reports were originally promised to come annually.  Before the reports were all a bit conjectural about the future, but now we have data to work with.  I know local interest has switched focused to the new $7 billion dollar ethylene refinery (i.e. cracker) being planned for Beaver County.  Some in the region are aware of what Chinese steel production did to steel prices and the steel industry here.  Well, take a look at $24 billion ethylene refinery being built in China.  But we won't talk about that. 


Create a Link

 

Tuesday, September 20, 2016

Pittsburgh GDP up 3.3% (sort of)

So here is a data conundrum.   Headline this morning from a dump of data from the BEA shows that Pittsburgh's GDP jumped 3.3% in 2015, a decent bump over the national jump of 2.5%, both as measured in real (inflation adjusted) dollars.   Here is the BEA's time series of GDP data for Pittsburgh as they show for both current and inflation adjusted (real) $.



So for a reference number if you want to know how big the Pittsburgh economy is, the current dollar estimate for metropolitan GDP is $138.9 billion annually. GDP is, broadly speaking, intended to be a measure of value added production. A good number to keep in mind when comparing the size of things you read about in the news.

But I need to look at something closer here.  The real GDP estimate for Pittsburgh shows a 3.3% jump, but the current dollar increase between 2014 and 2015 is only showing a 0.9% increase.  Inflation is low, but typically even with low price inflation current dollar increases appear larger than the inflation adjusted amount in almost all time series. Looking at this data, for most all regions current GDP growth is higher than real GDP growth, but for Pittsburgh it isn't, by a palpable difference.  So I calculated the ratio of the current dollar GDP to inflation-adjusted GDP across all metropolitan areas in the BEA data.  The ratio of the two is different across all MSAs, but within a narrow range of around 1.12. For Pittsburgh, the ratio is much lower at 1.06, one of the lowest ratios among all MSAs that I calculated. 

So now we get all wonky, but this is interesting.  Basically if the data is correct, it is saying there has been an abnormal deflation in prices in Pittsburgh based on the mix of goods consumed here.  The methodology used by the BEA does not assume different regional prices per se, but it does de facto calculate different regional price deflators based on different mix of goods produced here. So if we produced more of some good that had abnormal price changes, that results in regional differences in deflators and the result observed.  If I had to speculate, I think the strange result for Pittsburgh (much lower current dollar GDP growth compared to 'real' GDP growth) is coming from price deflation in energy markets. The price collapse in coal and natural gas over the time period for this data would give this result for regions that have concentrated output in those products.  Makes for some different interpretations of the data for surer.

So yeah, methodology matters, no matter how easy it is to quote a headline number.


Create a Link

 

Sunday, September 18, 2016

Still a magnet for educated immigrants

Your Wonk Quotient is high if you know that last week the Census Bureau released 2015 data from the American Community Survey (ACS).   More may have noticed various media stories looking to pull some interesting factoid from the latest data.  The problem is that the ACS has a panoply of data and finding one interesting datapoint to comment on is hard.  There is a bigger problem that the ACS survey data, especially the 1-year estimates data just released, has sampling error and related issues that really make most media stories an exercise in misoverinterpretation at best. You really want to exercise extreme caution when commenting on something in the 1-year ACS data if the latest data looks to be a significant change from the past.

That being said, I thought it safer to look at what hasn't changed in the ACS data for Pittsburgh. I thought I would start with a look at immigration for Pittsburgh. Not just immigration in general, but the characteristics of the immigrants who currently reside in Pittsburgh. This came to mind since I see yet another immigration attraction effort has started in Pittsburgh, the last in a long long line of similar efforts.  The new programs are always written about as if they are a new thing no matter how much immigration here has been written about worldwide.

Anyways, here is a true example of Pittsburgh exceptionalism.  Yes, immigration here is pretty low, usually measured as low as you would expect among large metro areas.  However it has also been true that among the immigrants who do reside in Pittsburgh, they are clearly the most educated cohort of immigrants among large metro areas in the US. Again, nothing new and something I have been pointing out for years. Still, the latest ACS data is interesting in that it confirms that factoid is still true, but if anything is stronger than ever.

Here are some rankings of how the educational attainment of the foreign born populations compare across the 40 largest Metropolitan Statistical Areas.  Here is what it looks like for the percentage with a bachelor's degree or higher.



I'd point out one important caveat here which gets to definitions.  Strictly speaking the ACS is measuring the characteristics of residents here.  The definition of 'foreign born' is just that. What you will see by many, and myself on occasion is referring to the same group as immigrants which isn't actually correct.  Immigrants are those who have moved permanently to the United States.  Foreign born residents here can include those not intending to stay in the US, such as many foreign students currently enrolled, or on non-permanent visas... although many of those may be intending to convert their visa status eventually.  So an important distinction, though I think the general pattern would hold if it was possible to strictly measure immigrants.

Still, not only is Pittsburgh out in front with over 56% of immigrants here now holding a bachelor's degree or higher.  Unsurprisingly, San Jose, CA, is #2 but well behind Pittsburgh at 50%.  But here is another figure comparing the percentages with a graduate degree.


So Pittsburgh is again out in front, but by an ever larger margin compared to #2.  Over 34% of the foreign-born population in Pittsburgh have a graduate degree of some kind.  #2 again is San Jose, but back at 26%. So when you read about efforts to attract and retain immigrants in Pittsburgh, remember we do pretty well at attracting immigrants with advanced professional skills. We are so far ahead of most anywhere else, you have to ask how much higher is even possible no matter the effort put in. What Pittsburgh lacks almost entirely are immigrants with less than a college degree, or even less than a high school degree. How and why those flows are not coming to Pittsburgh is the issue.

I will put the data that made the tables here, in a 'Footnotes' section on my web page. 

Create a Link

 

Tuesday, September 13, 2016

Coal employment dropping even faster than you may think in Pennsylvania

So last month I posted about the employment trends for coal mining in Pennsylvania.  The latest data at the time I saw was for the 4th quarter of 2015 when Pennsylvania coal mining employment had reached a new all time low of 5,918, or as I pointed out at the time a level Pennsylvania had not seen since before the civil war.

So the next quarter of data is out.  From a level of 5,918 for 2015Q4, the average employment for the first quarter of 2016 dropped to 4,835, or an 18% drop over the course of just 3 months.  Within that data the monthly data actually shows March of 2016 coal employment statewide dropped even further to 4,775.  Again that is total coal mining employment across Pennsylvania. How bad does the trend look like:



So the trend went from bad to worse, right.  Dig deeper and it gets even worse.  The same data also captures total wages by industry.  While coal mining employment dropped 18% from 2015Q4 to 2016Q1, total wages by quarter dropped by just under 30% over the same period.  Take a look at that:



So not just jobs are being cut, but among the jobs folks still have, either hours or wages (or both) are being cut as well. 

Create a Link

 

Tuesday, September 06, 2016

Pittsburgh always fighting the last demographic trend

It was a bit odd reading the recent headline from a few weeks ago: Pittsburgh's Youth Exodus Revereses - Millennials are being drawn to Pittsburgh.

Reverses?  Compared to last year? The last decade?  Maybe it's been lot longer since Pittsburgh was losing young folk. Not that you would know it from the headline.

Still... the headline implies some positive demographic trend is just beginning when it may actually be at the end  For years the message that Pittsburgh was NOT hemorrhaging young folks was met with less than deaf ears.  It was very lonely not being negative. Even after years of the data showing clear net migration into Pittsburgh, and lots of other positive demographic trends, the Post Gazette itself would write editorials ("Reports of a population uptick are unproven") in utter denial of hard data and calling me out by name for having the temerity for not being more negative.  Of course my favorite was the letter to the editor in the Trib calling virtually all non-nabobism, especially mine, as "Unctuous Bafflegab".

Sorry.. truth is I'm not really sore about the latter. I'm really a bit envious of Mr. Brock's wordsmithing and wish I could turn a phrase so well.  His is the single reference that turns up in the entire Googleverse of the phrase 'Unctuous Bafflegab"

So now all of a sudden the news coverage says all is well. Our long demographic nightmare is over, or so you would think. The thing is that any trend of young people migrating to Pittsburgh story was clearly a reflection of some remarkable economic trends, but remarkable trends that have clearly ended. A regular meme here was tracking how Pittsburgh's unemployment rate dropped well below the national unemployment rate and remained below the national rate for an extended period of time.  My updated graphic of the region's  unemployment rate (United States minus Pittsburgh MSA rates) shows this:


That big block of green on the right end of the graph was a period of 100 straight months where the Pittsburgh region's unemployment rate was below the national average.  If you include a few months where the two rates were the same you can extend it out to say that it was 109 straight months, literally 9 years and a month, where the local rate was not higher than the national rate.  There has probably not been a longer period where Pittsburgh unemployment rates stayed below the national rates in at least the previous half century, and probably going back as far as unemployment rates were reported at a regional level. So a big deal, and probably resulting in  the net positive migration the Pittsburgh region experienced by Pittsburgh for probably 7 straight years.

So after 7 years of positive net migration for the region, the headlines finally came to believe things have changed.   But already the headlines are lagging data and the most recent shows the migration trend is again negative for the region.  What about the future?  Look at the graphic above again and look at the last year there and the red on the far right. You see that in 2016 the Pittsburgh region's unemployment rate gave up the ghost and shot up well above the national rate. Not just by a marginal amount either.  In June the local unemployment rate was 0.8 percentage points above the national unemployment rate.  You have to go back to 1988.. yes 1988, to find a month where Pittsburgh was faring that far below the nation. That relative migration metric is a great predictor of net migration at the metropolitan level, so the next couple of years at least its unlikely the region will be pulling too many folks from elsewhere.  It is not so much that Pittsburgh is doing horribly, in absolute terms the local unemployment rate is still well below historical levels, but the much lower national unemployment rate reflects economic conditions in other regions that are developing serious labor shortage issues that will limit reasons to move here. Simple and unavoidable economics.


Create a Link

 

Monday, September 05, 2016

New low for Unionization Rate in Pittsburgh

I used to do this calculation from CPS data myself regularly, but the folks at UnionStats do the same thing and I am sure they do it much better so I will just pull their data.   But their latest data on the unionization rate by Metropolitan area show a remarkable thing. The private sector unionization rate within the Pittsburgh MSA has fallen to 6.8% in  2015. This is what their time series looks like.



I'd point out two things.   The annual time series compiled from the CPS shows more variation than you would expect in the real world, and that is to be expected from survey data.  But if you smooth the trend out to remove the sample error you clearly see the downward trend.  I'd suspect the actual rate for 2015 to be more like 8%, which still is pretty much an all time low, and something below that now in 2016.

But the bigger question that comes up when I talk about this data is what exactly is being measured.  Private sector unionization here measures union membership, but only among those currently employed.  So when you look at the greater obvious union presence at the Labor Day parade for example, realize that here in Pittsburgh for sure there are a lot of households with retirees or former union workers as well that would not be captured in this stat.  Note also the overall unionization rate in Pittsburgh is just a little big higher because public sector jobs have a higher unionization rate here.

A couple thoughts on what it means.  Whenever you hear of someone saying - and they still occasionally do  - that somehow unionization impedes economic growth in Pittsburgh...  the unionization here is actually below that of a lot of other metro areas.  Seattle for one example now show a private sector unionization rate of 10.0% in Seattle, compared to Pittsburgh at 6.8%. Seattle has been doing pretty well nonetheless.

Also, the historical trend is still a big deal.  This time series only goes back to 1986, but even then the private sector unionization rate in Pittsburgh was measured at 18.8% - or about 3 times the current rate.  1986 was clearly after so many union jobs went away across Western Pennsylvania. Remember this metric measures unionization only among those currently employed.  So the rate was much higher just a few years before that.  But go back further.  My research tells me that peak manufacturing employment came in the early 1950s when manufacturing jobs alone represented 40% of the local workforce.  Any measure of unionization back then was far upwards of that 40%. 

Create a Link