Was 2007 Q4 GDP Positive — or Negative ?

Many breathed a sigh of relief over the final revision of 2007 Q4 GDP. However, we took a closer look to at some of the data to see what was happening beneath the surface.

0328bizwebeconrealeconOur advice to those who think we escaped recession in Q4 2007: Not so fast.

As you might have guessed, actual below-the-headline data was less encouraging than even that weak 0.6% final number.

Under Gross Domestic Purchases, the BEA wrote:

"Real gross domestic purchases — purchases by U.S. residents of goods and services wherever produced — decreased 0.4 percent in the fourth quarter, in contrast to an increase of 3.3 percent in the third."

Real gross domestic purchases are purchases made by U.S. residents of goods and services wherever they are produced (domestic and imports). They decreased 0.4% in the Q4, very significant drop when compared to the 3.3% increase in Q3. Add to that the Gross private domestic investment decline of 2.2% in Q4.

Given those huge swings, how was it possible that GDP in Q4 was still positive? 

It all comes down to the Current-dollar GDP (and the implied implied price deflator). Current dollar GDP was lowered by a significant 0.3% more than was expected.


Why does this matter? Real GDP (after inflation) is obtained by dividing nominal GDP by the GDP deflator (x 100).  The smaller the deflator is, the less of GDP gains can be attributed to inflation. Had the change to above not occurred, Real GDP would very likely have been 0.0% — or worse.

UPDATE: March 31, 2008, 1:30pm

I just got off the phone with BEA — in the current GDP release, the changes in Q4 Current-dollar GDP were due to lowered "Imputed Financial Services" prices.

Let me also emphasize that I am not in the "books-got-cooked" camp. I am merely trying to wrap my head around how Real GDP was the same, but current dollar GDP fell so precipitously when compared to the last revision . . .


Given the impact of Inflation on GDP, is there another measure that might provide a clearer picture of the economy’s direction without the pernicious impact of rising prices?

It turns out there is: Last year, Fed economist Jeremy Nalewaik suggested a different measure: GDI, or Gross Domestic Income. Nalewaik argued in a 2007 paper that GDI "has done a substantially better job recognizing the start of the last several recessions than has real-time GDP."

According to Nalewaik, GDP-based models did much worse at forecasting recessions than did GDI: The past four
recession odds at their actual starting points were only of 52%, 40%, 45% and, for the 2001
recession, just 23% according to GDP data. The alternative measure of GDI did much better, signaling odds of
a recession of 78%, 44%, 72% and, for 2001, 70%.

And what of today? Recent data shows an annualized GDI decline of 1% — its largest drop since the 2001 recession.

While many people are debating whether or not the economy will fall into recession, the GDI data suggest that we are already in one — and have been for several months.


MARCH 27, 2008

Estimating Probabilities of Recession in Real Time Using GDP and GDI
Jeremy J. Nalewaik
Federal Reserve, December 19, 2006

Did Economy Really Escape Fourth Quarter Drop? 
Brian Blackstone
WSJ Real Time Economics, March 27, 2008, 2:02 pm

4th-Quarter Data Confirms Frailty of the Broad Economy
Published: March 28, 2008


Print Friendly, PDF & Email

What's been said:

Discussions found on the web:
  1. John Bruso commented on Mar 31

    That’s really interesting. Your blog is so helpful to us folks who are not economists. When .6% GDP came out, it did seem awfully convenient.

  2. michael schumacher commented on Mar 31

    Does it really matter??

    If one uses these numbers as a true gauge you (along with the Fed) are horribly behind any realism in any sense of the word.


  3. pr commented on Mar 31

    Putting aside your detailed analysis and using just the “official” numbers: simple math dictates that if the economy had a decent head of steam in Q3 (though obviously not as high as the 4% or more as reported) and going into Q4, then to produce an average 0.6% in Q4 most likely the first half of the quarter was around +1.5% and second half was around -.5%. i.e. GDP went negative in mid-November, which is probably what the gurus will conclude in a year’s time.

  4. DL commented on Mar 31

    Given that the government has a tendency to “cook the books” on CPI data, it stands to reason that they would do the same for the “implied price deflator”. And it’s not just the government. Most economists, for whatever reason, “gloss over” the contribution of inflation to the reported GDP numbers. There is far too little discussion of the contribution of inflation to the official economic growth numbers.

  5. Eric Davis commented on Mar 31

    I was Psyched, if we arn’t in one Yet, it means we won’t come out of it for a minimmum of 2 more months…. as a “shallow” recession.

  6. Estragon commented on Mar 31

    Three caveats to using GDI as a “better” recession indicator, both relating to the corp profits component:

    1. Corporate profits tend to have more real time lags because of reporting cycles, and revisions can be large.

    2. In the current environment, where export sales have picked up at least some of the slack in domestic sales, issues of transfer pricing (i.e. where profits are reported) are potentially significant. Dell, for example, may transfer US made systems or components to an Irish subsidiary at a price that puts a larger share of the overall profit in the subsidiary for tax reasons. The effect of this would be to overstate domestic production value (a GDP component) and understate domestic profit (a GDI component).

    3. Financial and housing sector profits were impaired by writedowns relating to past activity. Current activity is clearly slowing in both areas, but the income approach may overstate the rate of slowing.

    That said, there are lots of indicators flashing slowdown. Whether we were slightly negative or slightly positive is really just semantics.

  7. John commented on Mar 31

    It doesn’t matter to the markets whether we are or are not presently in a recession, nor does it matter whether the recession will have started in November or December or April or May. What matters to the markets are future cash flows and how much they must be revised from current expectations, as well as whether or not current prices have discounted the actual future outcomes.

    In other words, cash flows determine the presence or absence of recession, not the other way around.


  8. The Financial Philosopher commented on Mar 31

    Perhaps I am over-simplifying this:

    1. It would seem that those individuals with the applicable “knowledge” to understand that The Fed numbers are, at a minimum, “massaged,” will make up their own minds as to the status of the US economy and whethor or not we are in a “recession.”

    2. Those without this knowledge are ignorant of it and either do nothing, don’t care, believe we are in a recession anyway, or some combination of those…

    3. It is my observation and assumption, therefore, that to “argue” the numbers would be simply to assert our “knowledge” in an attempt to quantitatively prove we are in a “recession” and/or a Bear market.

    In other words, its classic Bear vs. Bull arguments that lead to knowhere but a “smart-talk trap.”

    With that said, I must say that Barry’s arguments are by far more intelligent, well-presented, and entertaining than the vast majority of information media sources out there… Including the “Bulls.”

  9. John commented on Mar 31

    Quarterly GDP numbers don’t determine the beginning or ending of recessions.

  10. Estragon commented on Mar 31

    TFP – “in an attempt to quantitatively prove we are in a “recession” and/or a Bear market”

    Speaking only for myself, I’m not out to “prove” anything. For me, the discussion improves my understanding of the world as it is, not as I would have it.

  11. mock turtle commented on Mar 31

    i’m not an expert or even close…so please correct my thinking here

    have always been told that consumer spending is plus two/thirds us economy.

    how is it that pre Christmas season, where traditionally more than one/third of all retail spending occurs, the 4th quarter numbers consistently look so low?

  12. Stefan Karlsson commented on Mar 31

    Estragon, the fact is that writedowns are not considered in the national accounts profit statements. Writedowns are considered balance sheet items and so do not affect profits.

    Barry is right BTW that there was negative growth during Q4 2007. However you actually don’t need to resort to GDI to reach that conclusion. As I pointed out already after the advance estimate back in January, the more relevant terms of trade adjusted GDP was -0.6%, not +0.6%.

    This number is derived by deflating nominal GDP growth of then 3.2% with the gross domestic purchases deflator which rose 3.8% in that estimate. Using the latest number, real GDP fell 0.7% as nominal GDP rose 3.0% and the gross domestic purchases deflator rose 3.7%. As it is the gross domestic purchased deflator which measure how much purchasing power production generates for Americans, it is more relevant than the GDP deflator, which instead measure to a large extent purchasing power for foreigners.

  13. corrigan commented on Mar 31

    The deflator was depressed because import prices were so high…. another nonsense which drops out of this most Keynesian la-la land of economic numbers…. Imports are a negative on GDP, remember, so a high import price reduces real imports and boosts GDP – arithmetically reducing the deflator at the same time….

    Thus, if oil went to $1000000 a barrel, the material well-being of the average American would plummet, but the GDP deflator would fall and, hence, ‘growth’ might miraculously be assured…

    Similarly, if foreigners would only buy American goods at a 99.9% discount, the export deflator would plummet and again, GDP would rise.

    The Red Queen woudl be truly proud!

    FWIW, tho’ far from perfect, I tend to calculate net private product per capita (subtracting capital consumption and ignoring government) and then deflate the balance with a domestic consumption index. Such foibles then drop out

  14. lurker commented on Mar 31

    Estragon, you sound like a smart dude. Thanks for posting.

    For me, the question is whether the market is done discounting the slowdown, whatever we call and whenever it started. Have stocks bottomed or not is what matters most to me.

  15. jojo (brion) commented on Mar 31

    Recessions are “felt” first….
    Just find a reasonably intelligent “middle classman” of a certain age (say 50 years old who’s lived thru previous econ cycles) and ask them….
    “How is the economy doing?”

    The last person you wanna ask is an analyst sifting thru gov. propaganda….

  16. The Financial Philosopher commented on Mar 31


    I respect and completely understand your logic. In fact, an attempt in finding an “understanding of the world as it is, not as I would have it” is a primary reason for my reading blogs, such as The Big Picture. I would say, however, that my purpose in reading TBP has more to do with my fascination with human behavior and observing it as it occurs…

    I prefer to spend the most energy on understanding myself rather than understanding “the world.” The former will place me on the right “path” while the latter tends to lead me away from it…

    Thanks for provoking thought…

  17. Michael Donnelly commented on Mar 31

    Barry so glad you write about this. GDP prices (general inflation) was lower in 2007 than it was in 2006 or 2005 ? Hard to believe.

    The big reason for prices falling apart in the fourth quarter was non res structures (business building) prices fell apart in that sector, sounds like housing doesn’t it?

  18. pft commented on Mar 31

    The president of NBER, the organization that is the final arbitor for deciding when we are in a recession, said recently he thought we were in one already. His saying so is not official of course, and they no longer use GDP as the sole yardstick.

    If it looks like a duck, acts like a duck and quacks like a duck, it’s probably a duck. Sometime later, as you are halfway through your Beijing Duck course, having correctly concluded it was a duck and acted accordingly, NBER will announce it is a Duck and not to worry, the duck is dead and is being consumed.

    Then you can relax, unless they warn you a Big Bear is heading for you and the duck.

  19. Giovannoni commented on Mar 31

    This is simply not correct.
    Gross domestic income is just GDP without Exports and Imports, so that already includes (the falling) investment values:
    GDPurch = C + I + G
    No need to subtract investment TWICE.

    See definition here:

    And indeed GDPurch does a much better job at predicting recessions; I have a paper on that and according to it, we are in a recession.
    Yet do we even care if we are in a recession or not? This discussion seems so 2007 to me.

  20. Ironman commented on Apr 1

    Barry – keep in mind too that just because the BEA called it the final release for 2007Q4 doesn’t necessarily mean that the BEA is done revising the GDP figures for the quarter.

    Going back to 2000, the BEA ultimately revised the original GDP data of that year significantly downward from what they had previously put out in their “final” revisions, which finally confirmed that the economy began sliding into recession beginning early in that year. It just took several years before they got around to adjusting the data.

    We’ll see what the future brings – I just find it interesting that they seem to have the most trouble getting the numbers right during election years.

Posted Under