Where Are or Were The Jobs?

With the all the alleged concern in Washington now from both parties about job creation, there’s something that’s missing in much of the debate: facts.  So let’s take a look at some.  I really like graphics like the one below.  They’re complex and take quite some time to read and fully absorb what’s there, but they pack a lot of information into a small space.  They’re info-dense.

We hear from the left a lot of talk about “good” vs. “bad” jobs.  Often what they are referring to is the relative wage level of the jobs.  In general, manufacturing and government jobs are “good” because they tend to have slightly higher than average wages*.  Education and health services jobs are a mixed bag with a lot of variation.  Doctors, nurses, and admins do very well.  Home health aides and assisted living workers not so much.  Teachers are either good or bad depending on the state. Leisure and hospitality are generally panned as below-average.

From the right we hear claims that heavily unionized sectors like motor vehicles, parts and manufacturing are holding down growth and killing jobs.  We also hear political conservatives claiming that excessive growth of the government sector has somehow prevented the private sector from adding jobs.

We also hear from the left that it’s lack of demand that is keeping unemployment high.  The right like to claim the unemployment is structural – we have the wrong workforce with the wrong skills.

But what’s really happened?  How have the different policies of Bush and Obama (to the degree they’re different – they aren’t as different as some think) affected the employment picture?  Let’s look a this graph from David Altig, Senior VP at the Atlanta Federal Reserve Bank as posted at macroblog.  It helps to click and enlarge the graph in a new window/tab.

Click to Enlarge

First, let’s examine how the graph is structured.  As always, it’s important to make sure we understand a graph’s axes first.  Horizontally, we have the average monthly change in employment in percentage between Dec. 2001 and Oct. 2007.  This period covers all of the non-recession portion of the G.W. Bush administration.  Industries to the reader’s right grew strongly and thrived under the Bush administration’s policies.  Industries to the reader’s left shrunk. No growth is the zero or mid-line. Next, the vertical axis shows a similar measure, average monthly percent increase in employment, but it’s for the period of July 2009 through Aug. 2011.  This is the non-recessionary months of the Obama administration.  Industries located high up grew under the Obama recovery. Industries low on the scale shrunk and cut jobs during Obama’s recovery.  There’s no tricks here of cherry-picking time periods – both axes cover only the “recovery” portion of each president’s respective time in office.

So looking overall, we have the four quadrants.  The upper right shows industries that have added jobs under both presidents’ recoveries. The lower left are industries that have been cutting jobs under both presidents. Upper left would be winners under Obama but not Bush. Lower right are those sectors that have been cutting employment under Obama but were big growth sectors under Bush.  Finally, the size of each bubble indicates the relative importance of the industry in terms of jobs.

So what can we conclude?  First there are few items that aren’t so surprising.

  • Under Bush, a lot of the employment growth involved construction and financial activities.  Not surprising. This is the Wall Street driven housing and mortgage bubble. Frankly we don’t need that big of construction sector, at least not if it’s focused on housing as it was.  We have too much housing already.  We do have needs for more construction of infrastructure and to the degree that housing construction workers are either in the wrong location or don’t have the skills for infrastructure construction (I don’t know – it’s not my expertise), then the low employment growth under Obama here represents a  structural unemployment problem.  But notice that industry isn’t that big.  Also, we probably don’t want to have Financial Activities come back as big as they were before.
  • The big winners under Bush were Education and Health Services and Professional/Business Services.  In education and health, health dominated.  Not surprising, health care spending has been growing and the population is becoming older and/or sicker.   The growth in professional/business services is probably not really very productive stuff.  A very, very large part of the increase in that area was the huge increase in security personnel and related-security contracting that has arisen from an increasing paranoid insecure society since 9/11.
But there are some items here which are surprising, or at least surprising if you’re believe the normal political rhetoric.
  • First, it was Bush who grew government employment.  Under Obama, government employment has been negative since the recession ended. Shrinking government employment is clearly the single largest drag on the economy. That’s not ideology or belief talking. It’s facts and data.
  • Second, the big reason why the Bush recovery was such a slow recovery for employment, considering the 2001 recession was mild, is that throughout the Bush administration manufacturing shrunk dramatically.  This was the result of globalization policies that provided incentives for U.S. manufacturing firms to locate production overseas or to buy from overseas manufacturers instead of making their own.  Fast growing companies like Apple and other computer companies prefer to design it themselves but to contract with foreign firms for manufacture. Obama has not turned the corner on manufacturing employment, but he has stopped the bleeding. For the U.S. to recover, this sector needs to have positive growth.  Given it’s size, it’s not necessary to rise to the top in percentage terms, but it needs to be positive which it isn’t now.
  • “Manufacturing” does not mean “autos”.  Manufacturing is much worse than Motor Vehicles and Parts.  Too often when politicians talk “manufacturing” they conjure a stereotypical image of auto manufacturing.  In reality, motor vehicles and parts, while not being a source of growth under either, has essentially held it’s own as neutral.
  • The Information industry is the one industry that has shrunk under both recoveries, although it’s not that large.  This largely represents true sectoral, innovation-driven change as the World Wide Web changes information technology.
Finally, let’s see what this graph says about the controversy over is unemployment structural (in which case we need training and incentives to work) or is it a lack of aggregate demand (in which case we need more stimulus spending).  I think the graph is relatively clear in this regard.  We have three very, very large sectors where there is no increase in employment under the current recovery: Manufacturing, Retail Trade, and Wholesale Trade.  These are the three that represent basic total spending.  Retail and wholesale trade are driven by total consumer spending. Period. Retail and wholesale also are very flexible without widespread specialized skills requirements.  When demand exists, they hire. When demand doesn’t exist they don’t hire and may layoff.  To me, the data indicates it’s clearly a lack of demand story that is hurting jobs in this so-called recovery. Reducing government employment right now, like this graph shows is being done, has repercussions in stopping employment growth in retail, wholesale, and manufacturing.