How We Got to 1.3% Growth, pt. 1 (F) (K) (SPX)

This is essentially the same post as I put up on July 29th when the first cut at the GDP numbers for the second quarter came out. I have put the new, revised figures in bold. The original numbers (first release) are shown in regular font and the second look is in italics in parenthesis immediately after the new numbers. This will allow you to see just where the changes are coming from, as well as the differences from prior quarters. New commentary will also be in bold.

The composition of GDP growth is as significant as the overall level of growth.

Second quarter GDP growth came in at 1.3% (1.0%, 1.3%). That is slightly above consensus expectations of 1.2% (1.1%, 1.7%) growth. The downward revision coming in slightly above expectations is good news.

The quality of the growth also improved somewhat. The return of the level to that seen in the first look was also relatively good news. The upward revision came primarily from two areas: Personal Consumption expenditures on Services, and more strength in Net Exports. This was offset by downward revisions to Goods Consumption, particularly of Durable Goods. The change in Inventories was also slightly more negative, which increases the quality of the growth.

…From Extremely Humble Beginnings

The initial below expectations print does not even start to describe how bad the initial report was. In its annual benchmark revision, the numbers were revised all the way back to 2003. Most notably, there were massive downgrades to the growth in both the first quarter of this year and the fourth quarter of last year.

If you are looking for a silver lining, the best I can come up with is that growth actually accelerated in the second quarter relative to the first. That, however, is only because the first quarter was revised down to just 0.4% from the last published 1.9%. The fourth quarter was not nearly as good as we thought is was, with growth of just 2.3%, not 3.1%. This is the sort of report that literally made me feel sick to my stomach.

While in this post I am going to focus on the second-quarter data and how it looks relative to the prior two quarters, there are big and very significant revisions that go much further back than that (unchanged in this release). For all of 2010, growth was actually 3.0% not 2.9%, as the second quarter in particular was much stronger than thought, coming in at 3.8% growth, not 1.7%. For 2009, the economy fell 3.5% not 2.6%.

The recession was also much worse than the previous numbers let on. In the heart of the downturn, GDP shrank at an annual rate of 8.9% in the fourth quarter of 2008, not the already awful 6.8%, and in the first quarter of 2009, the decline was 6.7%, not 4.9%.

The graph below (from http://www.calculatedriskblog.com/) shows the path of real GDP both before and after the revisions. We had thought that real GDP had finally surpassed its old high in the fourth quarter of last year, but after the revisions we have still yet to do so.

The nest graph shows just how anemic the growth in the first half (after revisions) has been relative to history.

There is a lot of important data in this release, and let me just give a note on the format of presentation.

Since the different parts of the economy are of very different sizes — and some tend to be relatively stable while others can be very volatile — I will focus on the contributions to growth. In other words, growth points, not the percentage growth rates. After all a small percentage change in a very big part of the economy can have more impact than a big percentage change in a small part of the economy.

To do this I will follow the familiar Y = C + I + G + (X – M) framework, where Y = GDP, C= Consumption, I = Investment, G= Government, X = exports and M = imports. Since the revisions to the fourth and first quarters are just as big a story as what happened in the second quarter, I will use the current numbers (released on 7/29) when making the comparison, but those numbers will be followed in parenthesis by the previously published number.

So, how did we get to 1.3% (1.0%, 1.3%) growth in the second quarter?

The biggest part of the economy by far is the Consumer, or consumption, or to be more specific, Personal Consumption Expenditures (PCE). It represented 71.0% of the overall economy in the second quarter. Initially it looked like it simply did not show up in the second quarter, but the revised figures show it in a more positive light — still far from great, but not as bad as it originally looked. PCE contributed 0.49 (0.30, 0.07) growth points, down from 1.47 (1.52) points in the first quarter and from 2.48 (2.79) points in the fourth quarter.

Over the long term, our economy is already weighted far too much towards C, and that contribution has been rising over the years. Back in the 1960’s it represented more like 64% of the overall economy. Our consumption share is also far higher than most other economies in the world.

Still, we need consumers to be opening their wallets for the economy to grow at least in the short term. This is a high quality growth area, and it was mostly missing in action this quarter. If 71% of the economy is not participating, it is hard to have robust growth.

Goods & Services

Consumption can be broken down into two main categories, goods and services. Goods can be further broken down into durable goods, which tend to be big ticket items that will last more than 3 years, and Non-Durable goods, which tend to be consumed right away. (For some reason clothing is categorized as a non-durable good. Clearly the people making those decisions have never looked into my closet.)

Services are by far the biggest part of consumption at 66.0% of PCE and 46.9% of overall GDP. It chipped in 0.87 (0.64, 0.40) growth points. That is up from the new 0.36 (0.70) contribution in the first quarter. In the fourth quarter, services added 0.61 (0.70) points to growth. The 0.23 upward revision in PCE services account for almost all the upward revision to overall GDP. The contribution is now more than twice what it was in the first read.

Services tend to be “produced” domestically, not in China, and also tend to be more labor-intensive than goods producing jobs. Goods subtracted 0.38 (0.34, 0.33) points from growth in the second quarter, down from adding 1.10 (0.91) points in the first quarter and 1.87 points in the fourth (2.10). Normally demand for services is more stable than demand for goods, especially durable goods.

Within the consumption of goods, consumption of non-durable goods is about twice as large as the consumption of durable goods. However, since people can defer purchase of durable goods like an Auto from Ford (F) more easily than they can defer purchase of a box of corn flakes from Kellogg’s (K), durable goods demand is very volatile. As a result, durable goods tend to “punch above their weight” in determining is the economy is booming or slumping.

Durable goods consumption subtracted 0.42 (0.40, 0.35) points from growth. That is down sharply from an addition of 0.85 (0.68) points in the first quarter and from 1.20 (1.45) points in the fourth quarter. All of that slowdown and then some was due to the supply chain disruptions caused by the disaster in Japan. The auto industry was the one most affected by the disaster, and it subtracted 0.74 (0.73, 0.65) points from growth, versus adding 0.85 points in the first quarter, and 1.26 points in the fourth quarter.

Early in recoveries it tends to have a bigger positive impact, but it also has a big negative impact when the economy falls into recession. The headwind from the Japanese disaster will abate later in the year, but a swing of 1.59 growth points sure leaves a mark.

All of Durable Goods make up only 7.64% of GDP, and 10.75% of PCE. The easing of the supply chain constraints from the Japan disaster is a key reason to hope that third quarter GDP growth will come in better than in the first half. On the other hand, while that headwind is easing, others are growing.

Non-durable goods are 23.25% of PCE and 16.51% of overall GDP. The sector’s contribution to growth fell to 0.04 (0.07, 0.02) points in the second quarter from 0.25 (0.23) points in the first quarter and 0.67 (0.65) points in the fourth quarter. While one does not expect big flashy moves in this part of the economy’s contribution to growth, it is stall a major part of the overall economy. It is as if it didn’t show up to play at all for the second quarter of the game.

Revised Up to "Bad" from "Awful"

It was still a fairly weak performance by the Consumer, by far the biggest part of the economy. The revised numbers move it up to being simply a poor performance rather than an awful one. Then again, with unemployment high and wage growth stagnant one can understand why they are not opening up their wallets. The massive loss in wealth from the popping of the housing bubble means that the middle class is desperately trying to rebuild their balance sheets. That means using available cash to pay down debts and build back savings, not going out and spending on goods and services now.

To the extent that personal income has increased of late, it has been coming from increases in things like dividend income and rental income. In other words, sources that tend to go to the wealthiest in the society. At the margin those consumers are not going to spend that much more.

The poor and the middle class simply don’t have the ability to spend more, not a lack of desire to do so. While I do think that over the long term it would be good to see consumption shrink as a share of GDP back to where it was in the 1970’s and 1980’s, one would hope that it would come through faster growth of other sectors, not an actual shrinkage in consumption.

You have to remember that the popping of the housing bubble resulted in a huge destruction of wealth — about $7 Trillion worth. That was widely-distributed wealth, unlike stock market wealth which is far more concentrated. The fall in housing asset prices decimated the balance sheets to the middle class.

Many people had been counting on using that wealth for things like financing retirement or sending the kids to college. Now, assuming that people still want to retire or send their kids to school, they have to rebuild their savings the old fashioned way, by consuming less of their income. This will be a substantial headwind to consumption not just for this quarter and next, but for years to come.

Investment

Investment tends to be the most volatile part of the economy, and thus is the major reason why the economy either booms or busts, even though it is a relatively small part of the overall economic picture. Overall Gross Domestic Private Investment (GDPI) is just 12.65% of the overall economy.

Overall, GDPI added 0.79 (0.78, 0.87) growth points in the second quarter — up from the new estimate of a 0.47 point addition in the first quarter, but way down from the 1.46 point addition previously estimated. However, in the fourth quarter GDPI “only” subtracted 0.91 points from growth, not 2.61 points. Those are some pretty drastic revisions. Fortunately, they mostly came from the lowest quality area, inventory building.

Fixed vs. Un-fixed Investment

Investment is the key to future growth and as a share of the economy, it is much lower than most other economies. However, not all investment is of the same quality. Fixed investment, particularly investment in equipment and software, is investment that tends to have a positive return on investment and which then drives future growth.

But not all investment is fixed. If companies build up their inventories, that too is counted as investment, and it tends to be of very low quality. If companies are simply adding to store shelves and those goods just sit there, then the investment in inventories will be reversed in later quarters. The inventory cycle is a powerful driver of booms and busts (recessions from 1946 through the early 1980’s were mostly due to the inventory cycle, or at least had the inventory cycle as one of the major components).

The increase in inventories accounted for a drag of 0.28 (0.23, an addition of 0.18) growth points in the second quarter. The bigger drag from inventory is actually a good sign, although I would not get to0 excited about a change of just 0.05 growth points in this revision. However, relative to the first read, it is a swing of 0.46 growth points, and that is significant.

Thus even though both the first read and the final read both came in at 1.3% growth overall, the final read is showing a much better economy than the first read did. Inventory investment added 0.32 (1.31) points to growth in the first quarter. The full point reduction in contribution from inventories thus accounts for two thirds of the downward revision to first quarter growth (in the first release of the data).

If we had to have a downward revision, inventory investment is exactly where you would most want to see it. In the fourth quarter, lower inventories subtracted 0.91 points from growth, not the massive 2.61 points we thought they had subtracted. In other words, even though much lower in magnitude, the quality of growth was much better in the first quarter. The same is not true for the fourth quarter, where both the level and the quality were revised down substantially.

On the other hand, while a 0.28 point drag is not massive (0.18 points is not a huge contribution) for the second quarter, it is still a drag. Had it not been a factor, growth would have been 1.58%, not 1.3%. As for the first quarter, with both over all growth and the inventory contribution revised down, inventories represented 80.0% of the overall growth.

In other words, without inventory investment, the economy would have been dead in the water, growing at less than 0.1%. Clearly inventories can have a big influence on overall growth, but it tends to be very low-quality growth. There does however seem to be somewhat of an inverse correlation with net exports, which I discuss below.

Residential vs. Non-Residential Investment

Fixed investment can be broken down into Residential investment (mostly homebuilding) and Non-residential (or business) investment. Residential investment has been the major thorn in the side of the economy for a long time now. That changed a bit in the fourth quarter, and we appear to be slowly forming a bottom in residential investment, it being up in two of the last three quarters.

In the fourth quarter, residential investment added just 0.06 (0.07) points to growth. In the first quarter, it subtracted 0.06 (0.05) points from growth. This quarter it was a slight positive, adding 0.09 (0.08, 0.08) points to growth. Residential investment is now just 2.23% of the overall economy, down from 6.34% of the economy at the peak of the housing bubble. Normal is about 4.4% of the economy.

While the levels are still just plain awful, it has shrunk so much that it is hard for it to have that much of an effect on the overall rate of growth. Residential investment has been a drag on GDP growth in 15 of the last 19 quarters. We still have a massive overhang of existing homes for sale (including those in foreclosure, and those which are likely to be foreclosed on). Most estimates of the amount of excess housing available today put it at about 1.5 million housing units.

Normally, residential investment is one of the biggest positive contributors to growth in the early stages of a recovery. It is the locomotive that pulls the economy out of a slump. This time around, the locomotive has been derailed.

The extremely weak performance of residential investment is what has made this recovery so weak relative to prior recoveries. Residential investment tends to move first, because it is so sensitive to interest rates. As it does, it pulls other parts of the economy, most notably consumption along with it.

With that much excess supply, building more houses is, on one level, simply a massive misallocation of resources. Residential investment is extremely volatile, and as such tends to “punch far above its weight” when it comes to the overall growth rate of the economy.

Eventually population growth and new household formation will absorb the inventory overhang, and residential investment will pick up. That, however, it not going to happen right away. Still, starting from such a low level, it seems likely that residential investment is likely to be a positive contributor to growth in 2011. Not a very big one — that is more likely a 2012 story (hopefully) — but simply by not being a major drag on the economy is a major turn for the better.

That bump is almost entirely a function of just how small residential investment has become as a share of the overall economy. The overall bottoming process in residential investment is not over, and it will be a long time before it returns to its historical norm of about 4.4% of the overall economy. However as it does, it will set off some very strong economic growth.

We are now at an all-time low for residential investment as a share of the economy, as shown in the next graph (also from this source). Note how at the end of every previous recession, residential investment increased sharply as a share of GDP, in effect leading the economy out of the recession, but how it has persistently declined this time around. That is THE key reason that this recovery seems so sluggish.

Business Investment

Non-residential, or business investment can also be broken into two major parts, investment in structures, such as new office buildings and strip malls, and investment in equipment and software. Investment in structures added 0.54 (0.38, 0.20) points to growth in the second quarter. That is a major positive swing from the first quarter when it subtracted 0.40 (0.41) points from growth. In the fourth quarter it added 0.26 (0.19) points.

After consumption of services, this was the biggest area of upward revisions, and this is a high-quality growth area. However, I am not that confident that it will last. Vacancy rates are still extremely high in almost all areas of the country, and in almost all major types of non-residential real estate.

We simply don’t need to be putting up a lot of new commercial buildings right now. On the other hand, we are starting to see some signs of those vacancies being absorbed, and prices for commercial real estate seem to be starting to firm up. The best leading indicator of future investment in non-residential structures is the amount of activity among architects. That index moved back into positive territory during the winter, but has since slipped back.

The most recent reading though showed a sharp improvement, but only to slightly positive levels. Given the lead time, this suggests that we will have a small positive contribution from non-residential structures starting in the third quarter, and then a bit more in the fourth quarter, before becoming a drag again in the first quarter of 2012. If we continue to see strength in architects’ billings, then we may well see a more positive contribution in the second quarter of 2012.

Equipment & Software

Investment in equipment and software (E&S) is what we really want to see to power future growth, and there the news continues to be good, though not as good as in previous quarters. E&S investment added 0.44 (0.55, 0.41) points to growth, which is solid showing since it is only 7.33% of the overall economy. However, it is down from a 0.60 (0.61) contribution in the first quarter and a 0.56 (0.54) point contribution in the fourth quarter.

This is the ninth quarter in a row that E&S investment has made a positive contribution to growth. A year ago, investment in E&S was just 6.71% of the overall economy.

The downward revision in E&S investment was probably the single worst part of this report (relative to the previous report and expectations). This is probably the highest quality form of growth out there, as it is growth that feeds future growth. While the downward revision here was disappointing, the level is still pretty good.

The next graph, (from this source) shows the contributions to growth from the three components of fixed investment on a rolling four-quarter average basis (unfortunately not updated with today’s revisions, but since it is a rolling four-quarter average, today’s revisions would not change the picture significantly).

Note the very sharp rebound in E&S spending, back up to the levels of the late 1990’s, relative to the persistent drag from investment in structures, both residential and non-residential. However, the contribution looks to be fading, as is the drag from construction.

Here is a different way of looking at the key parts of business investment spending — investment in structures and investment in equipment and software as a percentage of GDP. The rebound in E&S investment has been quite strong since the end of the recession, but it has not yet recovered to the share of the economy it commanded before the recession started.

Then again, in the last economic expansion, it did not even get to the same neighborhood as the peak of the Clinton expansion. Investment in structures (which tends to lag the overall economy) fell to a record low share, and is not showing much in terms of a rebound, offsetting the improvement on the E&S side.

(Editor’s note: Part 2 of this report will be continued in a separate blog.)