How alarming is it that 200% of 2007:Q1 GDP growth came from consumption spending?
When one looks at BEA’s breakdown of the 1.3% 2007:Q1 GDP growth, one finds that, if consumption spending had been as observed, and all other components of real GDP had simply held constant between 2006:Q4 and 2007:Q1, first quarter real GDP growth would have been 2.66%. The observed 1.3% growth of GDP is due to the fact that several other components, particularly residential fixed investment (housing) were not holding steady but were instead experiencing significant declines.
Unlike Kash Mansori, I do not see the fact that strong consumption growth alone was holding things up as being in and of itself alarming. If income growth is temporarily low, standard consumer theory suggests that consumption spending would continue on something close to its original trajectory, as consumers borrow to get past the temporary problems. Strong consumer confidence suggests that is how consumers currently see things. So, if you thought the other new negative contributions to 2007:Q1 growth (specifically the inventory drawdown and weak exports) were strictly temporary, this pattern might be exactly what you’d want and expect to be seeing.
My view is the same as Calculated Risk [1], [2]— the key thing to watch is nonresidential investment. I was very worried when this had made a negative contribution to 2006:Q4 GDP growth, and relieved that its contribution to 2007:Q1 is back into positive territory.
I feel the latest data leave the situation much as it has been. If all that happens is that housing continues at its current level of problems, that will continue to subtract 1% from each quarter’s GDP growth, as it has each quarter for the last year now. But unless that spills over into other sectors, the economy will continue to grow. We’re all watching for more indications of that spillover. But it hasn’t happened yet. 
Professor Hamilton, you say “We’re all watching for more indications of that spillover. But it hasn’t happened yet.” Barry Rittholtz at the Big Picture recently had a couple of posts about the current spillover into areas such as auto sales.
More importantly, I would please like to know your opinion of this paper by Brandeis University Professor S.G. Cecchetti called “Measuring the Macroeconomic Risks Posed by Asset Price Booms”. This paper shows that housing booms (but not equity booms) lead to declines in output and increases in prices, and the bigger the boom the bigger the fall in output and the bigger the increase in prices:
http://papers.ssrn.com/sol3/papers.cfm?abstract_id=932899
Actually, auto sales have been flat for 6 years and there are no significant signs that they are breaking out of this flat trend one way or the other.
Jim: I’m glad you’re feeling more optimistic than I am. One thing, though. You write: “If income growth is temporarily low, standard consumer theory suggests that consumption spending would continue on something close to its original trajectory, as consumers borrow to get past the temporary problems.” But what if we have consumers with a very high level of debt, and credit market imperfections that mean that credit standards are currently tightening? That might make consumption smoothing difficult in this case.
spencer, Ford says this month is awful and GM’s Lutz says that mortgage problems are hurting car sales:
http://money.cnn.com/2007/04/27/news/companies/bc.ford.sales.us.reut/index.htm?source=yahoo_quote
Also, the CEO of Autonation sees weakness ahead:
http://online.wsj.com/article/SB117759598063783428.html
No spillover? Ford CEO just said Friday that the only thing that could explain the dramatic decline of car sales in April is “spillover from housing slump”.
I was already told once that the pop of dot-com bubble will not spill over. And, looking back, you should admit that dot-com bubble was a small bubble that popped gently, comparing with huge real estate and credit bubbles, which just only starting to deflate, not even popped, but it already lead to congressional hearings, subtracted 1% of GDP and so one.
So what will happen if real estate will subtract 3% of GDP? Will it spill over then?
Based on the level of starts in the recent quarters, it’s unlikely that residential investment will subtract 1% from growth in Q2. There would have to be a big decline in starts for that to happen. While that’s possible, the level of permits suggests that’s unlikely.
KM, that’s a fair question. Lending standards are tightening on mortgages, but not on other consumer loans. The Fed’s household financial obligations ratio is historically high at 19.4% of disposable personal income, but only 0.6% above recent peaks in the early ’00s.
Steve, I am puzzled by your comment. Year-over-year starts have been down 20-30% over the last three quarters. How do you work the math so that residential investment does not again subtract 1% from real GDP? Have you checked out the homeowner vacancy rate?
I am a little more nervous about the economy than our host. Consumption seems healthy, but to my mind consumption follows income, income follows employment, and employment lags. However, “soft landing” includes the word “landing”. Recession is uglier than this.
wcw,
See this chart first.
It has to do with how residential investment is calculated as a component of GDP. From what I’ve read on the BEA website, when a new home is started the estimated cost of new units is taken and spread over the estimated time of completion, which I believe is 6-9 months depending on the type of unit. So new units that were started last year are still impacting GDP in Q1.
If you go back to the chart above you can see the rate of decline in starts was the sharpest in the first three quarters of the year. We’ve now leveled off a bit in the 1.5 million starts range.
In Q3 and Q4 of 2006, RI subtracted 1.2% from GDP. The reason is because not only were starts now at a lower level, but the starts that began in 2005 and the beginning of 2006 were completed and no longer a factor in terms of GDP.
In Q1 of 2007, RI subtracted 1% from GDP. That’s still a lot but less than the two previous quarters. It was less because the level of the drop off in activity from Q3/Q4 ’06 to Q1 ’07 is much less than, say Q1 ’06 to Q3 ’06. The drop off in Q2 is going to be even less compared to previous quarters.
Now, this all assumes that starts don’t fall off not only significantly but also rapidly. I suspect we’ll see another down leg in starts this, but probably not too much further. Given that march new permits for starts came in at 1.544 (and actually increased over February), so far it looks like residential investment should remain at current levels in Q2.
This is another chart that shows how we’ve leveled off (also notice permits). Residential investment will continue to be a drag on GDP for the year, but most likely progressively less each quarter.
There’s a really good Goldman Sachs chart from their housing conference at the start of the year that shows how they expect it to play out, but I can’t find it at the moment. If I do, I’ll post a link here. So far it is right on track.
Well JDH uses a pretty provocative opening with “200% of 2007:Q1 GDP growth came from consumption spending”, but the number was 2.6 of the total GDP growth…down from the 3.0 of the previous quarter’s contribution to total GDP, no?
So consumption appears to be on the wane to me (I have such a way with pacification when provoked, no?)
Why would it be on the wane? Could be Kash and CR are right: declining help from MEW.
It is not the Texas Janitors that Yellen appears to see with her “gangbuster” labor remarks.
It’s not that Chrysler workers are looking forward to a generous attitude from management with contract negotiations this fall.
It’s not that house prices have receded to the point where new buyers not only feel confident to find a mortgage, but also fill the box with furniture and consumer goods…stimulating business investment and domestic plant build-outs.
Thanks theoroxy for those counter notes about “spillover”..Nattering Nabob carries a daily tally of all those companies failing to meet expectations and the tune from those companies is almost always, “spillover”.
Steve, thanks for the details.
I can see you would expect a lesser drag if you think homebuilding has bottomed. Me, I pointed you at vacancies for a reason. I don’t think homebuilding can bottom and start growing with the economy until people move into those empty units. That’s not going to happen until starts stay well below levels of household formation. I don’t think a 1.5m start rate does it.
KM, great points/hypothetical questions. I think that consumption smoothing will not occur given record high personal debt loads, tightening credit standards (actually, returning to something approaching normal), and increasing interest rates (again, returning to something approaching normal).
Oh, by the way, employment will be taking a big-time hit, given the huge overhang of vacant homes (2.8% nationwide; anecdotally, 30% of the homes for sale in San Diego are vacant).
It’s going to be ugly, folks.
wcw,
Even if it hasn’t bottomed, it’s unlikely we’ll see residential investment subtract 1% from GDP next quarter or beyond. I expect starts to decline further, but not as drastically from current levels nor as fast as they did last year. I’m not suggesting that residential investment will be positive anytime soon, just that it won’t be as much of a drag on growth.
Today’s personal consumption release for March shows real consumption down in March, and slower in February than in January. Front-loading growth like that means PCE starts out at a relative disadvantage in Q2. The way the math works out, you need more spending growth in each month of Q2 to get the same quarterly pace of growth as in Q1. As pointed out above, vehicle sales were probably weak in April. Target says chain store sales were weak. Gasoline prices probably put a ding in restaurant spending.
So having started in the hole at the end of Q1, PCE did not get off to a strong start in Q2. I expect we’ll start seeing downward revisions from the 2.4% median estimate for Q2 GDP growth that Bloomber published back on April 10.
kharris,
Right on cue…
Economists at Morgan Stanley lowered their forecast for growth this quarter to an annual rate of 2.2 percent from 2.4 percent because of the slowdown in spending. They also projected growth last quarter will be revised up to 1.5 percent from 1.3 percent following the revision to February construction figures.
Great timing, kh/Steve!
The details for personal consumption (real/deseasonalized) are available, now: Mar., compared to Feb., had declines in purchases of autos and parts; furniture and household furnishings; food; gasoline; electricity/natural gas. These five components comprise 30% of personal consumption expenditures.
Feb. (revised), compared to Jan. (revised), had declines in six categories; some of those declines in Feb. — in clothing, transportation, and recreation — did NOT continue into Mar. But, my guess is that they’ll resume their downward path in April.
The month of May may prove ‘fun’ for the stock market, as the bad news (April auto sales, released tomorrow) continues to come out.
My view is that residential investment will tank the next 2 quarters as builders reduce. Starts will fall to .800-1000 range by the winter of 08.
Residential investment will have its worst 2 quarters the 2nd and 3rd of 2007 then more modest drops through the first quarter of 2008. My guess is bottoms as a whole then, beginning a slow sluggish rise afterwords.
If consumer spending weakens while this “peak bust” phase occurs, a recession probably will develope. As I said in my previous post, I am not happy.
The revisions we pay attention to are the ones that issue from the BEA and not Morgan Stanley who are no different from than any of the other economists who were predicting 1.8 GDP.
I don’t know why I am not getting the kharris message:
So having started in the hole at the end of Q1, PCE did not get off to a strong start in Q2.
PCE is declining through the last 3 months (and possibly I should check if this extends through the past few quarters) and therefore in terms of growth (not bulk) should be in an increasingly favorable position to post a gain from such a miserable hole, no? Sorta more obvious with the delcining RI: after 3 quarters of serious declination, the stabilization should be increasingly more probable. [Hark! I hear a hammer! Someone is building something somewhere. Post a positive for RI.]
Calmo,
Your view seems to stem from thinking about household finance, and is correct in that sense. Mine has to do with monthly and quarterly GDP math. If growth slows progressively through a quarter, PCE growth did in Q1, then the final month’s figure is low relative to the quarterly average. In fact, March PCE was low relative to the Q1 average. Growth is needed in April merely to get back to the Q1 average.
Assuming 0.3% real PCE growth in each month of Q2 (that is the average over Q4 and Q1), real PCE will grow at a 2.3% annualized pace in the quarter, down from 3.8% in Q1 and 4.2% in Q4.
“The revisions we pay attention to are the ones that issue from the BEA and not Morgan Stanley who are no different from than any of the other economists who were predicting 1.8 GDP.”
I’d just like to point out that not all forecasters should be held responsible for the median estimate. My Q1 GDP growth forecast back in December was 1.3%. When I ran the inputs a week ahead of the GDP release, I revised to 1.4%. Some of us blind squirrels do come pretty close to the acorn, once in a while.
I’m sure there’s a message about group responsibility bein disintermediated as soon as the surveyed economists throw their darts at it…not as if this bunch is weeded on the basis of past performance; not as if we have MS and Goldman economists predictions sorted out from the mere Lehman and Citi economists.
Kudos to you for finding the “advance” acorn and thanks for cluing me in to monthly assessments of PCE rather than just quarterly reports and revisions.