Unfortunately I had a 'good' jobs forecast this AM as I was looking for 125K; we got 115K.
The previous two months were revised up by a net of just over 50K, and that is something. Revisions matter. But I do not like looking at the number 'net of revisions' and saying they are now as expected. They are not as expected... In fact we learn by looking at the numbers and looking a the revisions, as separate events. For example one thing the revision means is that the drop off in April was WORSE than we expected, from +154K to +115K . Had we 'known' March was 154K what do you think we would have 'expected for April? 115K? I don't think so. This NOT as expected...
For now one thing the revisions do mean is the the estimation techniques of the BLS are being overshot by the actual data trends. That implies that there is still some sort of unexpected acceleration going on by the late reporters. And that is good news.
In some ways having a low headline and an upward revision is a separate bit of good news.
DIVERGENCE - dealing with it....
Despite weak PAYROLL growth the unemployment rate fell. It is now 8.1%, down from 8.2%.
No one will like how it happened as the labor force contracted and overall household survey (HHS) employment levels fell. But before going bonkers on this month's results remember that the HHS produced -31K jobs in March, +429K in Feb and +847K in Jan etc
Over six months the HHS job growth is up by 1.255% compared to 0.983% for payrolls... Had the payroll report jobs been growing at the same speed as employment in the household report we would have had an extra 360K jobs over that six month period or an average of about 60K more per month
Over six months payroll job gains average 197K; if they had grown as fast (in % terms) as the HHS the gains they would have averaged nearly 260K per month. So that is why the unemployment rate is dropping. Household jobs are being created at a 'payroll equivalent pace' of 260K per month...
So the household report has been picking up more job growth than the headline-grabbing payroll report and that is why the unemployment rate is falling. Month-to-month the household report is volatile so step back from this volatile jobs number. The trends of the two employment surveys are much more similar over longer periods. So what is really happening?
We still do not know, what we do not know...but we can piece things together
Right now here is what we do not know:
we do not know if the better weather pushed up job growth early this year robbing us of growth in subsequent months. Now, as we get into the period when some of that hiring was supposed to happen since, some of it is already 'on the books', the seasonal factors are reducing growth because they do not see the increase that they expect month-to-month. So the seasonal factors may simply be bringing the employment LEVELS back to where they belong. ..
Is all that true to some extent, but is there a real slowing going on, on top of that? Are we past the seasonal pay-back slowdown and is the economy really slowing and is job growth next month going to be even slower?
This is a key question.
What we KNOW is that the unemployment rate still fell in April. What we also know is that while the part of the HHS job report tha focuses on levels for employment, unemployment and labor force are unstable, that same report tends to get the unemployment rate right. That implies that the errors in the numerator and denominator (U= number unemployed Divided by number in the labor force) offset. You may doubt this but you can easily verify this as the rate of unemployment is very steady while the data on which it is based (ie number unemployed, number in the labor force) on jump all over the place. That should tend to reassure us that whatever slowing is in train, it is not so bad that it has pushed the unemployment RATE UP. In fact the RATE has FALLEN. That fact argues that much of what we are seeing is a reaction to a data compilation process that has not been able to cope with seasonal abnormalities.
That would also suggest that some of the economic weakness seen in economic indicators is real in so far as the levels of activity had picked up previously and simply did not continue to pick up as much as seasonal factors suggested because growth was shifted ahead. Because of that monthly changes were weaker than expected.
For example, jobless claims after a large Easter bump up are back down in the 360K range. they suggest that we might be through the adjustment period...
While we do not know what is going on precisely, the drop in the unemployment rate is one very important and steady signal that continues to point the way to continued economic progress. The continuing upward revisions to jobs suggests that there is still some unexpected acceleration in place. That does not necessarily mean that we are though the period of seeing all our weather abnormalities wash though the system.
But construction job trends are illustrative: They rose by 26K in Dec and by a further 18K in Jan; they are -1k, -3K and -2K over the past three months, averaging just about 5K per month over that five month period. It suggests to me that we have gone through a period of payback and might expect to see job growth picking up in May.
That's all speculation of course, but that is what the numbers suggest to me.
Also April showers bring May flowers.