A Dream

Life is better when you have pleasant dreams.  You are one weird dude when you dream about the employment situation report!

So imagine this improbable scenario -- that someone in MSM (maybe even a real star like Steve Liesman) wants a different perspective.....

You only have a few minutes in this interview, Jeff, so take your best shot.

Steve:  I am putting this politely, Jeff.  You basically say that all of the employment pundits are off base.  Why?

Jeff:  There are three big mistakes.

• First, people confuse job creation with net job creation.  If you are interested in sound analysis, the starting point is good measurement. Try this test on CNBC.  Start asking your guests how many new jobs were created last month.  You will expose a lot of ignorance.  They are all focused on the net change.

Steve:  Why is that bad?  We all care about the net change.

Jeff:  It is bad because it cripples our ability to evaluate what is happening.  In labor dynamics, there are job losses and job gains.  You publicize all of the information about losses.  These come in clumps from major companies and show up in weekly reports on initial claims.  The new jobs come in small increments with no publicity.  It is the difference between anecdotes and data.

Steve:  How can you be confident of this?

Jeff:  The actual data are available from state employment offices.  No one pays employment taxes on a phantom job, so this is solid information.  The problem is that it is nine months old, so you chose not to cover it.  Do a search on business dynamics and employment.  You will see that I almost alone in looking at actual solid data.  May I continue to my other two points?

Steve:  Oh.  Yes.  You said you had three.

Jeff:   Here is the second mistake

• Confusing pre-conceptions about government with what is actually happening.  The popular notion is that the team at the BLS cooks the books, leaks advance information to the President, creates fictional jobs and generally does whatever is needed to make things look good.  In addition, these civil servants instantly change allegiance after an election!  The people advancing this viewpoint do not know anything about government, the civil service, or the difficulty in actually developing and publishing a report.  Most of them get big-time credibility because they are singing to a receptive choir on the Internet and financial TV and you all lap it up.  Here is a challenge.  Take the biggest public critics of the BLS approach.  Now see what actual research they did -- comparing the BLS with the actual results.  Even better, their own forecasts with results.  You will not find a thing.  They do not forecast; they just criticize.

• Third mistake.  The world treats the BLS report as the ultimate answer.  I advocate a different approach, and I say this as someone who has deep respect for the BLS analysis.  There are a number of different research methodologies.  Some look at tax withholding.  Some look at online ads.  I look at correlations with other economic indicators.  There are different researchers all trying to reach the TRUTH.  We do not know the truth until nine months later.  By then, no one cares, but it actually tells us which methods are working.

Steve:  So we should dismiss the official government report?

Jeff:  Not exactly.  The BLS uses the state unemployment reports for benchmark revisions.  Unlike their critics, they are completely open to analyzing old predictions.  Smart investors should do the same.

Steve:  What do you mean?

Jeff: Investors should know what approaches actually worked.  We all know that people can be bamboozled for a time, but eventually, we want to know what really happened.

Steve:  What do we need to see in employment gains if we are to have an economic recovery?

Jeff: Monthly job growth of 125-150K is just absorbing demographic changes.  Quite frankly, this number might be even higher since people are delaying retirement.  We need to see job growth significantly in excess of the 250K mark to reduce unemployment  Even that is a very low hurdle.

Steve:  Your work has been a bit skeptical on job growth.  Your estimates have been below those of the BLS for many months.  Meanwhile,  you are a long-term bull on stocks.  How do you explain that?

Jeff:  Many companies are showing great profit growth.  Even if you torture test the results there are many good buys based upon earnings..  Employment growth is trailing other economic indicators including business investment and the technology cycle.  Having said that, it is a very important indicator, and I am watching it closely.  We need to do much better.

Steve:  Didn't Obama and Biden already spill the beans on this month's report?

Jeff:  There are specific rules about the release of data.  The Fed gets a few of the data series on Wednesday morning.  The CEA gets the overview on Thursday afternoon, after the market close.  I understand that it is part of trader lore to create these conspiracy theories, but I do not believe that the President or the VP revealed actual information this week. Most likely, they were coached by their economic team that it would be a good number, especially knowing about the temporary census jobs.

Steve:  You said something about a bet with Mish..... Mike Shedlock...

Jeff:  Mish is a leading blogger writing on economics.  He has many specific and dour forecasts on the economic recovery.  At the Kauffman Conference for economic bloggers I suggested that he was far too pessimistic on employment gains, and I proposed a small wager.  I cited several points I have made here.  Mish was a bit elusive, but agreed that the bet would be no monthly employment gain over 250K in "this cycle."  Obviously, he has already lost if we accept the BLS job growth, but in the spirit of fair play, I am open to a different definition.

It just needs to be based upon real data, not opinion.

.....and then something happens in my dream.....

A ghostly figure appears.......and reality returns.

Steve is not interviewing me.  Instead, he has the regular cast of characters who are describing their conception of "the new normal."  Since these people have lots of money under management, the viewer is supposed to believe that they are providing useful information rather than talking their books.  It is the same report we see every month -- everyone trying to guess what the BLS will find, and then criticizing the result.

I jolt awake....

I am back to the regular monthly report where everyone confuses seasonal and non-seasonal data, blames everything on the birth/death adjustment, and substitutes stories for data.

A Note to Readers --

I understand that my audience expects me to uphold a tradition of providing plenty of links and data.  For those who have this hunger -- -and you should -- just check out last month's report, or use the google search on the blog for birth death or business dynamics.

Meanwhile, back to sleep.......

Comments

[oldprof]

MO -- Thanks for the words of encouragement. I have written about the imputation step a number of times, but perhaps the best description is here: http://oldprof.typepad.com/a_dash_of_insight/2009/12/whats-wrong-with-the-november-employment-numbers.html

I have been considering the problem of improving pointers to various resources and themes. Thanks for your question, which is nudging me along in that direction.

Jeff

[MO]

Hi Jeff,

Love your blog. Especially your detailed analysis of the BLS methodology is much appreciated. However, I would very much like if you could expand upon the following sentence from one of your earlier posts:

"I am concerned about the first part of the BLS estimation process, the imputation step"

Could you please elaborate? And perhaps also go into more detail about this apparently crucial step of BLS's nonfarm calculation?

Keep up the fantastic work!

Regards,
MO