Unemployment measures & what they mean.

[Kber]

There are a bunch of reports on unemployment and they all measure different things with different measures.

For starters, the government releases a report each Thursday of "new claims" and "continuing claims". New Claims are people filing for UI for the first time. This represents job destruction. Continuing claims measures how many people who have previously applied continue to receive benefits. If the number goes down, it can be for two reasons: one is that more people are finding jobs and the other is that people are exhausting their 6 months of regular insurance. The weekly claims report does not report on why people are no longer receiving benefits, just that the continuing claim number is going up or down.

This report also includes a 4 week moving average that is supposed to smooth out statical bumps caused by holidays, snow storms and other similar events. The 4 week average is considered a more reliable number.

Monthly, ADP, a private payroll company releases its estimate of the number of jobs lost or created in the past month. They release this report the Wednesday prior to the release of the official government unemployment. The ADP report is based on their own payroll numbers, does not include any public jobs (including teachers, soldiers, etc) and has only been in use since the early 1990's. However, their has been a correlation between their report and the Government numbers released two days later.

The first Friday of every month, the government releases two reports. The first, like the ADP report, attempts to measure the number of jobs in the total economy and compares the difference to prior months. They arrive at the total jobs via a combination of surveys and estimates. The small business economy is apparently difficult to survey and the estimates for small business creation are considered the least reliable part of the report.

The second report that the government releases at the same time is the unemployment report. The unemployment figure is based on a survey of random households. You are considered "unemployed" if you indicate that you are out of work, but are actively seeking employment. The other thing this survey captures is the size of the labor market. For example, if your son recently graduated from high school and is now starting to look for a job, both the labor market and the unemployed population grow by 1.

There is another report that captures not just people actively seeking work, but previously employed individuals who have become so discouraged that they are no longer looking as well as people who are under employed (i.e. working as a part time consultant because full time work is not available). Both unemployment reports measure survey participants status (working, looking, discouraged, etc) based on how participants self identify. Benefits have nothing to do with this particular measure.

I think that's all, but if I've forgotten anything, please fill in the gaps.

Thank you.

[jtuck004]

1. I don't see anything to add. But you may already know

There is a chart of the U6 here

http://portalseven.com/employment/unemployment_rate_u6.jsp?fromYear=1994&toYear=2010

Mish's Economic Analysis Blog sometimes has an interesting look at what is under the numbers.

It's also interesting to read the reports of revisions the BLS throws in later...

[Kber]

2. Thanks for the link.

[dmallind]

3. I'd clarify two things

None of what you say is wrong, but sometimes people misinterpret the first two numbers you mention.

The initial claims represents GROSS job destruction. It is accompanied, even in the worst of times, with hundreds of thousands of people who GOT jobs. That number is almost impossible to find and almost never reported even as an aside, but as you can imagine over the last few months has been slightly higher than the number of jobs lost, since we have gained NET jobs. All too often people assuming, say, 446,000 new claims = 446,000 fewer jobs. Not the case - not even early last year when we were losing hundreds of thousands net a month.

Second - and my personal pet peeve - is the idea that new claims are in any way connected to eligibility for benefits or affected by UI benefits running out. Initial claims means just that - the number of people who started claiming. Eligibility is not involved at all in this number.

Again nothing here contradicts what you said - just more detail.

[Kber]

4. Excellent points

and absolutely correct.

Thanks for adding!

[daa]

5. The correction is the CES Birth/Death Model

"The small business economy is apparently difficult to survey and the estimates for small business creation are considered the least reliable part of the report."

This month there were supposedly 162,000 jobs created and they are all going gaga. But, 48,000 were two or three month census jobs. Of the remaining 123,000 private sector jobs, 126,000 were made up by the CES model. We all no that small business is NOT hiring from the NFIB survey.

The important thing to remember about these "made up" jobs is that last year they made up 902,000 MORE jobs and had to make an adjustment removing 902,000 jobs from last years total because they didn't materialize.

Take these phony numbers with a grain of salt.

[Kber]

8. The NFIB is new to me

Do they publish their survey results regularly. I tried googling them and found some news summaries, but not the underlying survey results.

This would be a valuable data point to add for anyone trying to figure out the labor market future.

Updated to include the link - again, no actual numbers, just a quote, so it's hard to evaluate.

"The unemployment rate has probably peaked, but it will not fall quickly," said William C. Dunkelberg, chief economist at the National Federation of Independent Business, a small-business organization. New jobs often are created by small businesses, and he said members of his association are starting to hire again, but they are showing "little enthusiasm" about the recovery because sales continue to be weak."

http://www.washingtontimes.com/news/2010/apr/05/trickle-of-jobs-gives-glimmer-of-hope-for-economy/

[daa]

9. Yes monthly from their survey.

Here is this months's survey results.

http://www.nfib.com/Portals/0/PDF/sbet/sbet201003.pdf

[Statistical]

6. One thing I would clarify is that initial claims are not necessarily job destruction.

That figure is used as a proxy but it is incomplete.

Sometimes people are terminated and the position remains open eventually someone else is hired to fill the position thus no job destruction.

By looking at historical initial claims when unemployment remained steady and comparing it to most current week (normalized for population, and season) we can "guestimate" if jobs are being created or destroyed.

Say "normal" initial claims (normalized for season, and population) is 400,000. Initial UI claims > 400,000 indicate job destruction, while initial UI claims < 400,000 indicate job creation.

People will always be fired. Initial claims does not need to go to 0 to indicate net job creation.

[Kber]

7. Absolutely correct.

Thanks for the thoughtfull comments everyone.