Click to enlarge the chart.
Real GDP grew at a rate of 3.4% in the 2nd quarter of 2007, according to the Bureau of Economic Analysis. The Wall Street Journal has an article as well as reaction on their economics blog. Behind the headline number, which sounds pretty good (most of us would regard 3 to 3.5 percent as being consistent with the growth of potential GDP), there are a couple of areas of concern. Foremost in the minds of many economists is the fact that personal consumption, which accounts for 70% of GDP only grew at a 1.3% rate. In the last 3 years, only the 4th quarter of 2005 was lower (1.2%). The weakness in the housing market likely accounts for some of the drop in the growth of consumption, which causes many to worry that soft consumption growth could continue for the rest of 2007.
Also, for the first time since 2003, imports decreased for the quarter. At a very basic level, imports can be an important indicator. In the 3rd quarter of 2000, imports began to fall and did not turn up again until the first quarter of 2002. The 2003 drop seemed to be a one off event that did not presage a recession. Hence, this drop does not immediately set the recession alarms ringing. However, if we had two consecutive quarters of imports falling, I would be much more concerned.
Government contributed 0.82% to GDP growth, which is on the high side of where it has been for the last few quarters.
The strength of the economy right now is clearly in the non-residential investment market. I see evidence of this anecdotally wherever I go as well. Spending on non-residential structures increased at a 22% rate--the highest in years. That pushed overall investment spending to its highest growth rate since the first quarter of 2006--even while the housing slump was putting downward pressure on the number. Neither the government spending, nor non-residential investment spending are likely to continue at this rate for an extended time. Unless there is a tremendous bounce-back in consumption, GDP growth for the rest of 2007 is likely to be below 3%, probably in the high 1's or low 2's.
I always find the chart at the top of this post useful. It shows the contributions of the various components of GDP to the overall growth rate. For the first time in a long time, every major component was on the plus side. That doesn't happen often, and some aspects of it do not bode well. My subjective probability for a rate cut by the end of 2007 just increased considerably. More on that later.
UPDATE: James Hamilton is thinking along the same lines. He also does a chart of the contributions to GDP. His includes residential investment as a category. His recession probability index is up to 26.2%, even with the relatively strong overall GDP number. All of this just goes to show that it's not just the overall number. You've got to look at the individual components. It's 3.4%, but it's a pretty ugly 3.4%.
To be bearish, note the revision that reduced the growth reported over the past year. We had thought the y/y change in real gdp at the end of the 1st Q was 1.9%. But that was reduced to 1.5%, and even after the pop in the 2nd Q the y/y change is still just 1.8%.
If consumption doesn't strengthen it will be hard to get over 2% growth in the 2nd half.
You're right to note the revisions. This post was getting long enough as it is, so I left it out. It deserves a post of its own. I'll see if I have time this weekend.
On a more general point I see a growing disconnect between rhetoric on the Bush economy particularly about productivity and GDP and the data series as revised. The plunge of productivity to near zero in Q4 2005 coupled with the revisions to productivity reaching back to 2002 that came with the Q2 2006 BLS release, and then again with this BEA revision in this report:
" For 2003-2006, real GDP grew at an average annual rate of 3.2 percent, 0.3 percentage point less than in the previously published estimates."
all combine to wipe out huge amounts of reported growth. Yet much of the language in use suggests that most commentators (and particularly tax cut defenders) are 'Still partying like it was (pre-revision) 2003'. The numbers they typically use to 'prove' tax cuts work have been systematically undermined over the last five quarters of reporting, the BLS and the BEA alternatively taking sticks to Supply Side models.
Unfortunately they are also creating havoc for my Social Security argument. Right through Q3 2005 my case that productivity would come in ahead of both Intermediate and Low Cost projections was bulletproof, growth numbers were coming in right under 3.0% in a model where I only needed 2.1% and nobody seemed to see a cloud on the horizon. Suddenly not only did the worm turn (but to mix a metaphor) but got sliced shorter and shorter with effects reaching back to 2002. Suddenly 1.7% ultimate stopped looking ridiculous. While I still believe it is pessimistic I have learned to keep my head down a little more.
That being said I still remain puzzled about why this 40% crash in productivity is not being reflected in the unemployment number, or based on 2005 and 2006 actual receipts by Social Security on wages. (In both years despite troubling productivity numbers receipts came in well ahead of projections. Which would seem to suggest that the breakdown between productivity and real wage gains seen earlier in the decade has to some degree reversed.) The June report on the Trust Fund balances is due Monday or Tuesday. Given the dismal Q1 number it will be interesting to compare actual six month receipts against projections. Unless someone can explain to me why Social Security receipts are not a pretty good proxy for average wage worker earnings then a better than projected result just begs for an explanation.
http://www.treasurydirect.gov/govt/reports/tfmp/tfmp.htm
(The preliminary report for June is out already and is spitting out of my printer)
I'll be crunching some numbers over a cold beer and a hot steak, but it looks like receipts are holding up. The 2007 Report projected a combined OASDI balance increase from $2.048 trillion to $2.236 trillion under Intermediate Cost or $186 billion in assets added which divides out to $15.5 billion per average month. In June it looks like OAS added $17.9 billion and DI $944 million for a total of $18.8 billion. I'll be checking year to date as well as comparables for Low Cost.
So whatever is up with these slumping productivity and GDP numbers it isn't hitting wages as such (though equity withdrawel would paint a little different picture).
That's interesting. I don't have an answer right now either. It's worth talking about though.
2007 Social Security Report projections for 2007:
Intermediate Cost: $2.048 trillion balance to $2.236 trillion for an annual gain of $188 billion.
LowCost: $2.048 trillion to $2.241 trillion for a gain of $193 billion
Trust Fund balances as of June 30: $2.170 trillion for a half year gain of $122 billion which projects to a gain of $244 billion.
http://www.treasurydirect.gov/govt/reports/tfmp/tfmp.htm
Somehow in the face of a Q1 Productivity number that came in at 0.6% we have already compiled 64% of the projected growth in the Trust Fund. Now on the surface it would appear that Social Security receipts are a perfect proxy for wage earnings up to $97500, that is these numbers either suggest that a productivity slump doesn't translate to lower wages, or that there is some flaw in the Social Security model that results in this underestimation, or that we need to take a hard look at the survey data that give us the productivity number to start with.
As noted above we had the same effect in 2005 and 2006: disappointing productivity numbers and receipts above projections. Something is amiss in either the theory, the calculations or the survey. 12.4% of something is flowing into Treasury and is producing an average monthly gain of $4.6 billion more than even the optimistic Low Cost projection (which itself returns fully funded Trust Fund through the 75 year window).
I'd suggest that the government gather statistics on just how many are working more than one job.
patrick
They do. It's in Table A-13 of the monthly employment report. For the last decade the number has fluctuated between 7 and 8 million workers. Presently it is right in the middle of that range. It is a rather noisy, but fairly cyclical indicator.
As a percentage of the employed it is currently 5.1%, which is about as low as it has been in the last decade. The majority of them work a primary full time job and a secondary part time job. Less than 2 million work more than one part time job.
Periodically, the BLS publishes more detailed data in the Monthly Labor Review. Here is an article with some more information from back when they first started publishing the data monthly. Before 1994, this data was only collected yearly (in the month of May) and sometimes less frequently.
http://www.bls.gov/opub/mlr/1997/03/art1full.pdf