Friday, February 3, 2012

January Unemployment Data

You know the drill, so here are the numbers for January, 2012:

Population: +1,685,000
Employed: -737,000
Unemployed: +849,000
Labor Force: +112,000

So while we added 1.6 million people to the population (that almost seems like a typo, given that we only added 1.9 million people in all of 2011! See update below.), we lost -737,000 jobs and 112,000 people started looking for work again, resulting in a total increase in unemployment of 849,000 for the month of January.  This brings unemployment up to 8.8% and the employed-to-population ratio down to 57.8%.

While this may seem like a huge loss of jobs compared to previous months, we have to keep in mind that the majority of temporary seasonal jobs created around Christmas were going to end in January, so we would expect this unemployment number to be somewhat high.  It might be more cheery to see how employment fares when split into part-time and full-time workers.  According to the these numbers:

Full Time Jobs:  -1,171,000
Part Time Jobs: +435,000

Looking at it that way actually makes the situation worse.  It isn't just that we lost 737k jobs in total, we actually lost 1.17 million full time jobs, and only made some of that back by increasing part time jobs by 435k. That's not a positive development if we're looking for long-term growth.  But employment numbers are reported in the media largely as a matter of whether or not we beat expectations, and expectations can always be lowered, so this will probably be trumpeted as an indicator of strong economic growth.  In an election year, can we really expect any different?

Following Trends

To cap this off, we'll look at some more trend data to see if it follows previous months. Dividing the employment data by gender reveals a reversal this month:

Male Employment: -1,065,000
Female Employment: +329,000

Dividing employment by age is also interesting: 

16-19: -303,000
20-24: -88,000
25-34: -805,000
35-44: +44,000
45-54: -36,000
55+: +351,000

You might say that in the first month of 2012, the US workforce got a lot older, and a lot more female.


UPDATE: The BLS explains that the apparent huge spike in US population in January is actually an adjustment due to 2010 census data (which means that the employed-to-population ratio numbers for 2010 were actually worse than was presented).  They also said that they won't be going back to fix the 2010 records with the real data.

Friday, January 13, 2012

Consolidating the Bureaucracy

President Obama announced today that he is seeking some special powers from Congress in order to streamline several federal agencies with overlapping areas of responsibility by consolidating them into one agency.  The savings, at the federal level, would amount to an estimated $3 billion over ten years.  $3 billion seems pretty good to most of us (for which even a small fraction would set us up for life), but how does that stack up to government spending?

First, we'll note that last year's projected federal deficit was $1.65 trillion dollars.  Note that this is not the total amount the government spent in 2011, but only the amount it spent over what it took in (i.e. the amount it needed to borrow).  Also note that the $3 billion in savings is over a ten year period, so the average yearly savings is "only" $300 million.  A simple division of the two will tell us how much these savings can be expected to offset our borrowing this year:

300 million / 1,650,000 million = 0.000181, or 0.0181%.

Almost two hundredths of a percent of deficit reduction is not exactly inspiring on the face of it.  Still, these figures only take into account what we can see (the alleged savings by the government), but if Bastiat teaches us anything, we must also take into account what is unseen.  To the extent that businesses will be able to breathe easier because they don't have six different (and often contradictory) bureaucracies to deal with, the business climate will improve in ways that are not easy to measure in dollars.

It's a start, anyway.

Friday, January 6, 2012

December Unemployment Data Is In

The BLS reports that unemployment dropped to 8.5%, from 8.7% (November's unemployment numbers were "revised" upwards to 8.7%).  Based on the issues raised in the last post, what can we make of the numbers reported?

Not Seasonally Adjusted

Population: +143,000
Employed: -389,000
Unemployed: +79,000
Labor Force: -310,000

So, we had a net loss of 389,000 jobs in December (not unexpected, due to temporary holiday employment ending), of which 79,000 are still looking for work, and 310,000 are no longer looking for work.  Meanwhile, we tacked on 143,000 new people.  According to the "unadjusted" numbers, unemployment actually went up by 0.1%.

Note that the above numbers represent the real statistics as collected by the BLS.  As we noted previously, however, employment throughout the year is cyclical.  For example, people are hired for temporary jobs during the holidays and are then let go after.  That is, unemployment naturally decreases during the holidays and then increases afterwards.  Some people find this cycle to be unsatisfactory for reporting purposes, and so the BLS "seasonally adjusts" the data by comparing it to historical trends.  Here are this month's "adjusted" numbers:

Seasonally Adjusted

Population: +143,000
Employed: +176,000
Unemployed: -226,000

Labor Force: -50,000

Quite the swing!  You might say that these adjusted numbers tell us that we lost 176,000 fewer jobs than we expected to lose at this time of year.  While that's valuable to know, I would argue against the headlines that we will no doubt read saying "Unemployment has gone down!".  Instead, they should say "Unemployment has gone up, but less than would be expected based on historical trends!".

I guess that isn't as easy to fit on a page, though.

Following Trends

Even in the seasonally adjusted data, the trend is still towards a declining labor force.  If the economy is going to truly pick up, this number has turn around dramatically. The numbers show us that since December of 2010, the labor force has only increased by 217,000 people.  If January follows the trend of the last two months, we could actually have less people in the labor force now than a year ago, even though we have since added 1.695 million people to our population!

Also, last month it was noted that well over 99% of those dropping out of the labor force were women.  This month, the numbers are -61,000 for men (19.7%), and -249,000 for women (80.3%), so the trend continues, albeit not as strongly.

Friday, December 2, 2011

Stunning Decline in Unemployment, But Is It Real?

Much ado has been made about the Bureau of Labor Statistics' (BLS) just-released unemployment numbers, which showed a stunning decline from 9.0% unemployment in October to 8.6% in November. But what does that number mean, how is it calculated, and does it really imply that things are getting better?

How Is Unemployment Calculated?

Simply put, the unemployment rate is the number of unemployed people divided by the total number of people in the labor force.  I think we all have an idea of what an unemployed person is, but what constitutes the labor force? The labor force consists of people who are currently working (the employed), and people who are not currently working but are actively looking for work (the unemployed).  People who are neither working nor actively looking for work (the retired, the young, etc.) are said to be not in the labor force.

What Factors Influence the Unemployment Numbers?

Obviously, hiring and firing directly influence the unemployment numbers, as would be expected.  But there is another factor here that often gets overlooked.  Let's use a simple scenario to show how it plays out. Imagine a country with 100 total people, of which 50 are either retired or too young to work, 45 of them have jobs, and 5 of them are looking for jobs. Therefore, 50 is the total labor force for this country, of which 45 people are employed, and 5 are unemployed. The unemployment rate for this country is 5 / 50, or 10%.

Now imagine that 10 employed people retire.
The new unemployment rate for this country is 5 / 40, or 12.5%.

Now imagine that the 5 unemployed people get tired of job hunting and give up.
The new unemployment rate is 0 / 40, or 0%

Finally, imagine that 10 young people become old enough to get a job, and start looking.
The new unemployment rate is 10 / 40, or 25%.

As we can see, the unemployment rate fluctuates not just between employed and unemployed people, but also between people entering and exiting the labor force as a whole.  So with just a percentage in mind, there would be no way to see exactly what was going on in the job market.  Luckily, the BLS produces the whole range of numbers for us to look at.

What Do the Current Numbers Look Like?

According to the BLS: 

Oct 2011 Employed: 140,987,000
Nov 2011 Employed: 141,070,000
Change: +83,000

So 83,000 more people had a job in November than in October. That seems pretty good, right? So what about the unemployment numbers?

Oct 2011 Unemployed: 13,102,000
Nov 2011 Unemployed: 12,613,000
Change: -489,000

Wait, so only 83,000 more people are employed in November, but the number of unemployed people decreased by 489,000?  How does that make sense?  The answer lies, as we have seen, in shifts out of the labor force entirely:

Oct 2011 Labor Force: 154,088,000
Nov 2011 Labor Force: 153,683,000
Change: -405,000

This means, for every 100 people who are no longer counted as unemployed, that only 17 of them actually got a job!  Of this 405,000 people no longer in the labor force, it's also possible to infer how many people still want a job, but are no longer actively looking for one (so-called "discouraged workers").

Oct 2011 Not In Labor Force, Want a Job: 5,969,000
Nov 2011 Not In Labor Force, Want a Job: 6,183,000
Change: +214,000

Looked at this way, at least half of the people who moved out of the labor force did so not because they don't want a job, but because they gave up on finding one.  That's two and a half times the number of people who actually gained employment last month, and is not a great indicator of economic recovery.  But at least we gained some jobs, right?  Surely that means things are getting at least a little better?  Well, perhaps not:

Oct 2011 Civilian Population: 240,269,000
Nov 2011 Civilian Population: 240,441,000
Change:  +172,000

So in the same time frame that we added 83,000 jobs, we also added 172,000 people!  Now, most of that population growth no doubt comes in the form of newborns, who obviously won't be entering the labor force for some time.  However, their graduating big brothers and sisters are.  You might say that, if today's newborns are tomorrow's workers, then yesterday's newborns are today's workers. So how many workers are we adding each month?  As it turns out, that number is 125,000.  So, all else being equal, we would need to add approximately 50% more jobs each month than we did in November, just to meet increased job demand from population growth! Even with an increase in jobs, the job deficit is still widening

(Interesting Note: The data shows that 99.7% of people moving out of the labor force were women.  With childcare so expensive and wages coming down, are we perhaps seeing a resurgence in the number of stay-at-home moms?)

What Other Important Factors Do the Unemployment Numbers Hide?

We can also note that the simple unemployment percentage says nothing about what kind of jobs people are employed in, how much they pay, or whether or not they are making a good use of a particular worker's skills.  Thus, the unemployed quantum physicist who acquires a part-time job at Starbucks mopping floors for minimum wage is now employed for the purposes of calculating unemployment.  Such "underemployed" people are thought to comprise a not-insignificant percentage of the labor force these days (the next time you're at Waffle House, the guy cleaning the table could be a former hedge fund employee). These underemployed workers, if present in significant numbers, represent a market inefficiency that restrains economic growth.

In short, don't start celebrating just yet.