By Dirk van Dijk
Initial Claims for Unemployment Insurance jumped by 43,000 last week to 474,000. However, last week was revised up by 2,000, so one could see it as a 45,000 increase. This was far worse than the expected level of 400,000.
The trend in initial claims had been downward, though there had been some bumps along the way. Recently, though, they have started to shoot back up again. That is a very worrisome trend. The weekly numbers, after an erratic few months, seemed to have broken decisively to the downside. This week did a lot of damage and makes it hard to see the trend still being in the right direction.
Let us hope the decline resumes next week. I don’t want to get too alarmist about a single data point, but combine it with the weaker-than-expected ADP jobs report, a very soft ISM services number and first quarter growth of just 1.8%, one has to come to the conclusion tha the economy is softening significantly over the last month. This is the fourth straight week we have been above the 400,000 level. It is also the highest weekly level since last August.
It looks like we are headed back to the “trading range” that initial claims were in for almost all of 2010. Initial jobless claims had been generally trending down since they hit a secondary peak of 504,000 (after revisions) on Aug. 14, 2010.
4-Week Moving Average
Since claims can be volatile from week to week, it is better to track the four-week moving average to get a better sense of the trend. Even smoothing it out this way does not make the picture that much more appealing. It rose by 22,250 to 431,250. The four-week average climbing above the psychologically important 400,000 level is a very bad sign.
As far as the domestic economy is concerned, robust job creation has been the last big part of the puzzle to fall into place. It looks like the combined pressures from government budget cuts, the disaster in Japan and the increased oil prices are taking their toll. Things had been starting to look better on the jobs front, but the last few weeks' numbers are a bit of a wet towel.
Still, relative to a year ago, the four week average is down by 24,500 or 7.4%. On the other hand, the raw weekly number was 14,000, or 3.0% above the year-ago level.
The economy is growing, but is only starting to add a significant number of jobs and to put a dent in the huge army of the unemployed. The recent trend in initial claims suggests that that progress is rapidly slowing.
The March employment report was encouraging, but we still have a very long way to go. We added a total of 216,000 jobs according to the establishment survey, as the private sector total of 230,000 was offset by the loss of 14,000 jobs in state and local government. The unemployment rate dipped to 8.8% from 8.9%. January and February both had net positive revisions, with private sectors jobs revised up more than public sector jobs were revised down.
Gloomy Outlook to BLS Report
These numbers make it likely that when the March jobs report is released tomorrow, the number of jobs added will be lower than in March, perhaps down in the area of 150,000 total jobs added (about 170,000 adds in the private sector, offset by the loss of 20,000 government jobs). We might even see a slight uptick in the unemployment rate.
The unemployment numbers are based on mid-month conditions, so this report does not directly impact tomorrow’s numbers, but they are not a good sign. The consensus is more optimistic about tomorrow’s numbers looking for a total of 185,000 with 200,000 private sector jobs added offset by 15,000 lost government jobs.
The first graph (from this source) shows the long-term history of the four-week average of initial claims.
The Importance of 400,000
So why is the 400,000 level so significant? The next graph shows why. Historically that has been the inflection point where the economy starts to add a lot of jobs. It layers over the monthly gain or loss in private sector jobs (red line, right hand scale) and total jobs (green line). Unfortunately the automatic scaling did not put a line at zero for the job growth line, so you will have to eyeball it a bit.
Notice the strong inverse correlation between job growth and initial claims, and how when the blue initial claims number is below the 400,000 level that job growth is strong. OK, it is not that an increase from 399,000 to 400,000 is all that much difference than from 397,000 to 398,000 or from 402,000 to 403,000, but big round numbers are psychologically important, especially when that round number is near a historical inflection point.
The data on regular continuing claims was also disappointing this week. Regular continuing claims for unemployment insurance rose by 74,000 to 3.733 million. However, the overall trend is strongly in the right direction. They are down by 915,000 million or 19.7% from a year ago. Regular claims are paid by the state governments, and run out after just 26 weeks (although several states have lowered the number of weeks they are going to pay in the future recently).
The next graph shows the long-term history of continuing claims for unemployment, as well as the percentage of the covered workforce that is receiving regular state benefits. It does a good job of showing just how nasty that the Great Recession was for the job market. It also shows how things at the regular state unemployment benefit level have been getting much better over the last year (but still well above the peaks of the last two recessions.
Note that the insured unemployment rate generally follows the direction of the number of claims, but has been gradually diverging over time. That is a function of the overall growth of the labor force, and of tighter eligibility standards for getting unemployment insurance over time. It also reflects the fact that this time around a very large proportion of those getting unemployment benefits are getting them from the extended Federal programs, not from the regular state programs.
Length of Unemployment
In March, half of all the unemployed had been out of work for 21.7 weeks (down from a record high of 25.5 weeks in June, but up from 21.2 weeks in February), and 45.5% had been out of work for more than 26 weeks. Just for a point of perspective, prior to the Great Recession, the highest the median duration of unemployment had ever reached was 12.3 weeks near the bottom of the '82-83 downturn.
Clearly a measure of unemployment that by definition excludes 45.5% of the unemployed paints a very incomplete picture. The number of short-term unemployed (less than 5 weeks) was actually exceptionally low. The problem in terms of employment is not a lot of firing, but a lack of hiring. This has been the case for some time now.
After the 26 weeks are up, people move over to extended benefits, which are paid for by the Federal government. While regular claims are down, it is in large part due to people aging out of the regular benefits and “graduating” to extended benefits. Unfortunately, the data on extended claims in prior recessions is not available at the St. Louis Fed database. However, given the extraordinary duration of unemployment, it is a safe bet that they are higher than in previous downturns.
The extended claims have also been trending down, but it has been a bumpy decline. They (the two largest programs combined) fell by 43,000 to 4.122 million this week. Relative to a year ago they are down 1.507 million or 26.8% A much better measure is the total number of people getting benefits, regardless of which level of government pays for them. This is particularly true when looking at the longer term, not the week-to-week changes.
Combined, regular claims and extended claims (including a few much smaller programs) fell by 172,000 to 8.015 million on the week and are down 2.510 million or 23.8% over the last year. (The extended claims numbers are not seasonally adjusted, while the initial and continuing claims are, so there is always little bit of apples to oranges. In addition, the continuing claims data are a week behind the initial claims, and extended claims are a week behind the extended claims data.)
Even with the deal that extended unemployment benefits for this year in exchange for continuing the top end of the Bush tax cuts, people will still “graduate” from the system after 99 weeks, but people will continue to be able to move to the next tier up to the 99 week limit. Extended benefits are in four different tiers, so if benefits had not been extended, some people would have lost their benefits after just 39 weeks of being out of work. The key point though is that the year long extension does not mean that some people will be getting benefits for a total of 151 weeks, as is sometimes thought.
Employment Picture Remains Awful
While the employment picture has improved a bit this year, in any absolute sense it is still just plain awful. The pace of improvement appears to be slowing significantly based on the initial claims data over the last few weeks. The Federal Reserve is doing its part by keeping rates low and by using quantitative easing. While that helps a little bit, it is also much less effective than fiscal stimulus would be.
Right now monetary policy is sort of “pushing on a string." In any case, QE2 is going to end in June, and the likelihood of it being followed by QE3 is very low right now.
Some claim that the long duration of unemployment benefits has actually discouraged people from looking for work. That is, people are content to live forever on 60% of their previous income, or $400 per week, whichever is lower. The average benefit is only about $300 a week. Ask yourself how well could you live on $300 per week.
Right now (well, as of the February JOLTS data) there are almost 4.4 people out of work for each job opening. Just telling all of them to “get a job” isn’t going to work. The extension of benefits is one of the key reasons that initial claims are falling. While that may sound counter-intuitive, it is because extended benefits are a very effective form of economic stimulus. The March JOLTS data should come out next week.
The graph below shows the ratio of jobless to job openings since the JOLTS data first started tracking job openings. As with so many other things, we are making progress and headed in the right direction, but the absolute level is still awful.
Extended unemployment benefits are, dollar of dollar, one of the most effective forms of economic stimulus there is. It is a pretty good bet that the people losing their extended benefits have depleted their savings and run up all the debt they can in trying to make ends meet. The maximum unemployment benefit works out to be just $20,800 per year, or less than the poverty line for a family of four. You think any of those people have been able to sock any of that away?
There is a concern that by cushioning the blow of unemployment, people might be more reluctant to take a marginal job opportunity, but a below poverty level income is not that much of a cushion. I’m not sure it is good for the economy for highly skilled people to be taking jobs in other fields that have no use of those skills, and then be unavailable when those skills are needed again.
The people who get extended benefits tend to spend the money quickly on basic needs. This in turn keeps customers coming in the door at Wal-Mart (NYSE:WMT) and Family Dollar (NYSE:FDO). It means that, at the margin, some people are able to continue to pay their mortgages and thus helps keep the foreclosure crisis from getting even worse than it already is.
However, by the time they are well into extended benefits, they might also be spending food stamps as well as the unemployment check at Kroger’s (NYSE:KR). These customers keep the people at Wal-Mart, Family Dollar and Kroger’s -- and of course their competitors -- employed. It also keeps the people who make and transport those goods employed as well, although in that case much of the stimulus is lost overseas if the goods are imported.
Who Imports More - Rich or Poor?
However it is not clear if the marginal propensity to import is higher for poor (or temporarily poor because they are unemployed) or for the rich. Lots of the stuff on the shelves of Wal-Mart comes from China. On the other hand, the poor are not likely to be buying Swiss watches or German autos. They will not be buying frivolous things like water imported from halfway around the world (Fiji water).
What is clear is that they will spend it quicker, increasing the velocity of money, than will the rich who will tend to save more of it, particularly if they see the increased income from say a continued tax cut for the highest income people as temporary. The rich are much more likely, in other words, to fit Milton Friedman’s “Permanent Income Hypothesis” than are the unemployed, since the rich do not face liquidity constraints.
Also, if you remember your Friedman, velocity of money counts, and it counts a great deal. P * Q = M * V. Price times quantity in the aggregate is nominal GDP, and it is equal to the amount of money in circulation times how quickly it changes hands. Money in the hands of the unemployed has a much higher velocity than money in the hands of a multi-millionaire.
Stairway to Heaven, Highway to Hell or Sitting in Limbo?
There really is not good way to tell from this report if the decline in the number of people receiving benefits is due to them getting new jobs, or due to even the extended benefits running out. If it is the former, it is very good news. If it is the latter, it just means more people are falling into absolute destitution.
However, on a back-of-the-envelope basis, if the total number of people getting benefits is down by 2.5 million, and we have only created 1.3 million jobs (as of March) over the last year, one can surmise that at least 1.2 million have left the wrong way. That is not good news for either the economy or for social stability.
The 4-week average rising above the 400,000 level is a discouraging sign. In the last two recoveries, when it got below that threshold, that job creation really started to take off. If it stays there, we will climb the "Stairway to Heaven" of a rapidly falling unemployment rate. If it shoots above 500,000 then we are on the "Highway to Hell." The current level, if we stay here for a prolonged period of time, puts us back "Sitting in Limbo" of a pseudo-recovery.
The trend over the last four weeks makes it clear that we are no longer on that Stairway to Heaven, and seem to at least be looking for the entrance ramp to that highway. The increase this week is troubling, as was the fact that it was much worse than forecast.
This series can be volatile from week to week. It is not yet a reason to despair, but it is getting awfully close to being so. If claims continue to climb over the next few weeks it will be.
The continuing and extended claims numbers look much more promising, but only if people are leaving the rolls for the “right reason.” If they are leaving for the wrong reason, that the benefits have simply run out, the declines are not really good news (even if they do help reduce government spending), they are just a reflection of millions of people slipping into poverty. Hardly a thing to celebrate.
This was an ugly report.