How to Evaluate Employment Numbers

I have been tracking the jobs information since last August. Three categories are paramount for me: Non-farm payrolls (the total), private and government. What interests me most of the three is not the total number each month of job declines, but the percentage change that number represents from the total employed the month before.

Why? If we have a firm that employs 10 people and lays off one, we have laid off 10% of the workforce. If we double employment to 20, then lay off 2, we have still laid off 10%, BUT, people would also be accurate in saying "layoffs have increased 100% (from 1 to 2)". In this case, simply looking at the total number skews reality. For that reason I want to know what percent of the total who had a job the month before lost it. Then I want to know how that percentage is trending. I also want to know this for the private sector and government.

Here is the data from the BLS (all numbers seasonally adjusted where applicable):

The numbers do show an improvement in the rate of decline since the Dec./Jan. highs.

The key point to note is the April numbers. Without the highly abnormal hiring done by the government (72k jobs for temporary census workers), the improvement in April would have been roughly halved. That does place a large dose of doubt as the the "improvement" in the job situation. May estimates are for -532k non-farm job loses in May (to be released 6/5).

My thought is that this number is too low. I think -575k or greater is more likely unless the government hired another 72k people like it did in April. Assuming the government employment changes are normalized, then I think a spike is in order...

Again, what is of more importance is how government hiring affects this number and does the rate of decline in private sector jobs abate. True growth will only come from the private sector. Government can only grow payrolls and pay for them either through increased borrowings or increase takings of private capital (taxes). Neither is conducive to growth. We cannot keep losing a greater percentage of private jobs that the overall picture would lead us to believe and hope the economy improves soon.

That also goes to earnings. Q2's earnings were better than many of the reduced expectations. BUT, it was not due to top line sales growth but increase cost cutting (jobs). If we have sales declining but cut costs ahead of them, then we can improve the earnings picture. Now, this is not to say this is a bad thing, but it is reality.

Comments

[twitee:]

What happens to those that their unemployment time ran out after one year? Are they counted as unemployed?

Jun 04 11:54 AM

[conceptwizard:]

These are only the posted numbers not the real numbers. Go to "Shadow Stats" for the real calculation. Its hovering around 20%.
These numbers do show a positive trend, but are not including people taking part time and lower qualified jobs. It is interesting to note that the average hourly work week dropped, likely due to job sharing, and positions going part-time.

Jun 04 01:06 PM

[toobad41:]

My biggest gripe about all this statistical data and the seasonal adjustments that are made is that it is too broad as an indicator. With all of the super computer power we have today I think the government could do a better job. I would like to see these figures broken down by geographical area to see the real trends. I think we would see this data would be more applicable to the upper midwest, northeast, Florida and California. That would be significant. But I guess that would be politically incorrect.

Jun 05 09:05 AM