Labor Turnover
One of my favorite jobs-related graphs comes from JOLTS "(Job Openings and Labor Turnover Summary"). It shows how dynamic the economy really is, underneath the net numbers.
Net New Jobs (monthly): First some background. Every month, the government reports how many net new jobs were created. The last month, they reported about 250 K net new jobs. In a month like January, there is a huge seasonality adjustment in these numbers. Still, if we were to average 250K per month, it would be good news. How about if we added 50,000 jobs each month? While it would be better than zero, it would end up raising the unemployment rate.
Level to keep the unemployment rate from rising: Here's a back-of-envelope calculation. The U.S. has 300 million people, but only 155 million are working or want work. (Children, retirees, college-goers, housewives, and a few bums account for the rest). The population growth rate is about 0.85% per year. If we ignore age-distributions etc., the labor force will grow at that rate too. So, about 1.3 million people will enter the labor force each year: a little more than 100,000 new people each month. (The number is a little higher when one adjusts for people who are currently dejected about looking for a job.)
So, next month, when you hear the number of net jobs added, evaluate it this way:
Initial Weekly Unemployment Claims: Now, on to the quandary. Every week, we also get a report of new claims for unemployment insurance. These are around 400K each week! In fact, these rarely fall below 300K even going back to the 1980s. Here's a relevant graph (sharper one at Calculated Risk).
At that rate, there are more than a million people seeking initial unemployment claims each month.
In fact, it is even higher because not all people who are fired or laid off get unemployment compensation. A separate survey called "JOLTS" reports that about 2 million people were laid off or discharged each month, and that number has varied in the 2 - 2.5 million a month range since at least 2001.
So, what gives? How to reconcile the notion that the economy added (say) 250K jobs in the month with the notion that 2 million people were discharged?
New Jobs started: The answer is: people are also finding new jobs. In the past year, we have seen about 4 million new jobs each month. Not net new jobs, just the total gross number. Here are the very rough numbers for a typical month of the last decade:
Net impact = zero + a few net new jobs.
In Jan 2012, as an example, and we see (still pretty roughly):
Once more, I turn to Calculated Risk for the picture (sharper image here).
Labor Turnover: In our global political context, high labor turnover is usually a good thing. It is usually a sign of less government intervention in the labor market. Some studies claim that Europe has a lower labor turnover than the U.S. This means uneconomical jobs are retained rather than having people on something more productive.
In summary: When you hear weekly and monthly job numbers, don't pay too much attention to the unemployment rate itself. The number of net new private sector jobs created is the important number. Be aware that in some months this number has a huge seasonal component that makes it almost worthless unless we're having a very "typical year". More than 200 K net new private sector jobs is good news, if it stays consistently that high. Jobless claims around 300K - 320K are a good sign. (Aside: Another good sign is a high number of "quits". People start to quit jobs when they feel confident about the economy.)
Net New Jobs (monthly): First some background. Every month, the government reports how many net new jobs were created. The last month, they reported about 250 K net new jobs. In a month like January, there is a huge seasonality adjustment in these numbers. Still, if we were to average 250K per month, it would be good news. How about if we added 50,000 jobs each month? While it would be better than zero, it would end up raising the unemployment rate.
Level to keep the unemployment rate from rising: Here's a back-of-envelope calculation. The U.S. has 300 million people, but only 155 million are working or want work. (Children, retirees, college-goers, housewives, and a few bums account for the rest). The population growth rate is about 0.85% per year. If we ignore age-distributions etc., the labor force will grow at that rate too. So, about 1.3 million people will enter the labor force each year: a little more than 100,000 new people each month. (The number is a little higher when one adjusts for people who are currently dejected about looking for a job.)
So, next month, when you hear the number of net jobs added, evaluate it this way:
- below 100 K is bad,
- 100K - 150K is coasting,
- 150K - 200K is good, but we need better for a while, to get out of the hole we're in
At that rate, there are more than a million people seeking initial unemployment claims each month.
In fact, it is even higher because not all people who are fired or laid off get unemployment compensation. A separate survey called "JOLTS" reports that about 2 million people were laid off or discharged each month, and that number has varied in the 2 - 2.5 million a month range since at least 2001.
So, what gives? How to reconcile the notion that the economy added (say) 250K jobs in the month with the notion that 2 million people were discharged?
New Jobs started: The answer is: people are also finding new jobs. In the past year, we have seen about 4 million new jobs each month. Not net new jobs, just the total gross number. Here are the very rough numbers for a typical month of the last decade:
- 2 million people are fired or laid off
- 2 million quit
- 4 million+ start new jobs
Net impact = zero + a few net new jobs.
In Jan 2012, as an example, and we see (still pretty roughly):
- 2 million fired or laid off
- 1.9 million quit
- 4 million started new jobs
Labor Turnover: In our global political context, high labor turnover is usually a good thing. It is usually a sign of less government intervention in the labor market. Some studies claim that Europe has a lower labor turnover than the U.S. This means uneconomical jobs are retained rather than having people on something more productive.
In summary: When you hear weekly and monthly job numbers, don't pay too much attention to the unemployment rate itself. The number of net new private sector jobs created is the important number. Be aware that in some months this number has a huge seasonal component that makes it almost worthless unless we're having a very "typical year". More than 200 K net new private sector jobs is good news, if it stays consistently that high. Jobless claims around 300K - 320K are a good sign. (Aside: Another good sign is a high number of "quits". People start to quit jobs when they feel confident about the economy.)
Darius, you make it all sound so good that I began by writing a little too caustically. I am starting over again.
ReplyDeleteCertainly, the economy does function intelligently, and the reports the government produces do take advantage of that fact. However, the government number crunchers do not have a sound understanding of how to actually connect their product to the real world. What we have now is one set of numbers (new jobs) derived from one place that are seasonally adjusted using one methodology by one group of people (a point you made nicely in a previous blog post) compared to another set of numbers (new claims for unemployment benefits) about something related to the first set that is seasonally adjusted by a different methodology by different group of people. The numbers they gather from the actors in the economy could even be accurate, but the consequent processing and adjusting makes for some serious errors, that are compounded over time. We should not take any specific announcement as meaning anything. But people do. The inference drawn from each announcement by even people with some understanding is that an increase in jobs is good and a reduction in the weekly number of new unemployment benefit application is good. Only occasionally do people try to figure out something more meaningful in the numbers, such as the observation in January that lots of people weren’t looking for work anymore and were no longer counted as unemployed. The “theoretical” justification of the numbers that might have had some truth to it no longer applies.
I don’t mean to say that your first comments are tarnished. The U.S. economy is still very dynamic. It isn’t as dynamic as a capitalist economy would be, but it does have some life in it.
I am just saying that the numbers do not fit together and tell the story of our economy.
Hey, here is a chance to return the favor and the fire. I just posted after a few weeks of silence. http://krazyeconomy.blogspot.com/2012/02/note-on-greek-banks-recapitalization.html.