The NBER Business Cycle Dating Committee issued a statement today declaring that the bottom of the most recent recession was reached in June of 2009, with the economy in the expansion phase of the business cycle during the 15 months since then. This confirms the announcement issued by the Econbrowser Business Cycle Dating Committee last April.
Said Econbrowser Business Cycle Dating Committee is of course not a committee at all, but a soulless, purely data-driven algorithm whose pronouncements we’ve been sharing with readers since the inception of Econbrowser in 2005. Today’s announcement from the NBER allows us to update the table comparing the recession announcements made by the NBER with those of the algorithm. The NBER today declared that the last quarter of the recession was 2009:Q2, exactly what we reported to Econbrowser readers last April.
| Start of recessions | ||||
|---|---|---|---|---|
| Peak as determined by NBER | Date NBER made declaration | Recession start as determined by algorithm | Date algorithm made declaration | Algorithm announcement lead (-) or lag (+) in months |
| 1969:Q4 | N.A. | 1969:Q2 | May 1970* | N.A. |
| 1973:Q4 | N.A. | 1973:Q4 | May 1974* | N.A. |
| 1980:Q1 | Jun 1980 | 1979:Q2 | Nov 1979* | -7 |
| 1981:Q3 | Jan 1982 | 1981:Q2 | Feb 1982* | +1 |
| 1990:Q3 | Apr 1991 | 1989:Q4 | Feb 1991* | -2 |
| 2001:Q1 | Nov 2001 | 2001:Q1 | Feb 2002* | +3 |
| 2007:Q4 | Dec 2008 | 2007:Q4 | Jan 2009 | +1 |
| Start of expansions | ||||
| Trough as determined by NBER | Date NBER made declaration | Recession end as determined by algorithm | Date algorithm made declaration | Algorithm announcement lead (-) or lag (+) in months |
| 1970:Q4 | N.A. | 1970:Q4 | Aug 1971* | N.A. |
| 1975:Q1 | N.A. | 1975:Q1 | Feb 1976* | N.A. |
| 1980:Q3 | Jul 1981 | 1980:Q2 | May 1981* | -2 |
| 1982:Q4 | Jul 1983 | 1982:Q4 | Aug 1983* | +1 |
| 1991:Q1 | Dec 1992 | 1991:Q4 | Feb 1993* | +2 |
| 2001:Q4 | Jul 2003 | 2001:Q3 | Aug 2002* | -12 |
| 2009:Q2 | Sep 2010 | 2009:Q2 | Apr 2010 | -5 |
Let me reiterate that I am not criticizing the NBER in any way for delaying their announcement until today, nor am I proposing to replace their committee with my computer. But today’s announcement serves as a reminder of why it can be helpful to supplement human inference with objective algorithms. There are always those who may seize on the political dimension of announcements like this. For example, if you were attempting to assist one party or the other, you might prefer to have the announcement come before or after the election. I maintain that there is value in setting out in advance exactly what one means statistically by the statement “the recession is over”, and sticking to that definition.
So what does the statement really tell us? Simply that the economy, as measured by a variety of real economic indicators, has been growing rather than contracting for the last 15 months. Given that historically that condition of economic expansion tends to be highly persistent, in the absence of strong contrary indication, the most likely outcome is that we’ll continue to see further economic growth in the months ahead.
There has been some growth in employment, but it has been far too
slow to bring the unemployment rate down. And we need stronger GDP
growth than I think we are likely to see in the second half of 2010
to make any more progress with the unemployment rate over the rest of
this year. I do not think we are currently entering a new recession, though the
growth is sufficiently slow, and the lack of more progress on
employment sufficiently painful, that it probably still feels very
much like a recession to many people.
In the mean time, we’ll of course continue to report our recession indicator index (currently at 5.4% for the first quarter of 2010) to see how the algorithm compares with the NBER Business Cycle Dating Committee for the next go-round.
Let’s hope we won’t have another row for comparison in the above table any time soon.
|
“Given that historically that condition of economic expansion tends to be highly persistent, in the absence of strong contrary indication, the most likely outcome is that we’ll continue to see further economic growth in the months ahead.”
Based on what grounds do you assume that there is an absence of strong contrary indication to your statement further economic growth and no new recession was most likely?
I know at least two strong contrary indications:
1. The data collected by the Consumer Metrics Institute:
http://www.consumerindexes.com/
According to these data, which measure online purchases of discretionary consumer goods, we have already been in a demand recession for a number of months again. The quarterly GDP-change, as a variable of the supply side of the economy, seems to follow these demand side data quite well, but with a lag up to two or three quarters.
2. Hussman issued a recession warning at the end of June 2010, which is based on a combination of data sets:
http://www.hussmanfunds.com/wmc/wmc100628.htm
The data from these sets have fulfilled certain criteria since then, which have only been present in a combination during or shortly before a recession; there hasn’t been any false signal for the available statistical sample going back to the early 1960ies so far. Using Hussman’s recession signal, I would say the probability of an imminent or already ongoing recession is very high. Even if this signal fails this time for the first time, this would be a fail rate of 1 out of 8 and put the probability for a recession still at almost 90%.
I also could mention that new home sales dropped to the lowest value since data have been available just in July 2010.
http://calculatedriskimages.blogspot.com/search/label/New%20Home%20Sales
In no other case of the sample it has happened that new home sales bottomed after the end of a recession. So either this time is the first time that they did, or the recession has never ended, whatever NBER and your program say, or there has already been a new recession again.
There is also an early indication that the Fed Manufacturing Surveys and the Purchasing Manager’s Index, i.e., indicators for the supply side of the economy, have already peaked and are trending toward contraction again, although it’s not conclusive yet.
How do you reconcile all these contrary indicators with your probability statement and your statement about the absence of those indicators?
JDH wrote:
“So what does the statement really tell us? Simply that the economy, as measured by a variety of real economic indicators, has been growing rather than contracting for the last 15 months.”
Not so sure this is accurate. what the statement “really tell(s) us” is that the contraction has slowed. It tells us nothing about recovery. The Great Depression lasted around 18 years but the decline slowed after 1933 only to persist and return to decline in 1937.
With the NBER finally making their statement we can not expect a double dip, but from a much lower level.
Reading the post drove the thought that the essential is to keep reading habits,benchmarks even though not as useful as they were.
The G7 economies have nowdays, particularities that may not fit any longer with former criteria. How relvant is a GDP when most of the markets are nationalized
In the USA there is no longer a market for financial and commercial papers
Series: COMPOUT, Commercial Paper Outstanding
http://research.stlouisfed.org/fred2/series/COMPOUT
Series: FINCPN, Financial Commercial Paper Outstanding
http://research.stlouisfed.org/fred2/series/FINCPN
In France and in other European countries they are no longer real estates markets,no longer sovereign bonds markets,no longer interest rates markets.
One question that comes to my mind as regards the GDP when expressed in currencies amount,how much “quantity of works” is incorporated in the Gross domestic product.
It leads to the same author writing,” at the end of the capitalist system,all corporations will merge in one entity under the state umbrella”.
Actual governments in Europe have messed up the capitalist system and I hope they will not do the same with the Marxist principles.
The NBER announcement is a beautiful set up for the November elections. Don’t get me wrong, I don’t dispute the NBER’s data, but if one throws a $trillion or so at an economy of course GDP will rise. However, what happens in the absense of a QE2, particularly as nothing has been fixed systemically or structurally?
The worst is not behind us. QE2 will determine how far forwards the final bottom lies.
the NBER and economists should explain that we mean only direction by recovery.
we date it when ONE MONTH up from the bottom.
From down this low, the good times are years away
I vividly remember yours ‘we are most likely already turning around’ statement based on your model based on the initial unemployment claims data in the midst of the whole uncertainty at what everyone perceived as the worst of the recession.
Was it in June 2009 already?
It was good to see rational, purely quantitative, scientific, and also very much real time assessments with some small and careful commentaries then. It is also good to see they have shown their power.
Personally, as a layman, I suspect this must have been an exciting time to be an practicing economist 🙂
Looking forward to your next analyses
Impressive. I have to trade in my caribou shoulder blades for some of JDH’s medicine.
Simplicity is good and usually cost-effective. Incidentally, what data point or political consideration was delaying the release of the announcement?
Imagine an alternative universe where the NBER announced the end of the recession in Q1 2010 and around about the same period, the US federal reserve had given the first federal funds rate target hike to 0.50%. By now, the FF rate might have been 0.75 or maybe 1.00. Confidence would have been partially restored, people would not be so spooky and so suspect of the ability of public institutions to manage for the long-term.
“The Great Depression lasted around 18 years but the decline slowed after 1933 only to persist and return to decline in 1937”
Incorrect. The decline of 1937-38 was a piddle to the 29-33 contraction. By the end of 1938 the recession was over and by 1940 the economy was booming.
The real end of the depression was in 1941. The depression lasted roughly 12 years.
Do they have double-headers in opera?
The Rage,
Not to quibble but the Depression lasted through WWII. The unemployed did not find productive work. They were sent to Europe and the South Pacific to be shot to pieces. Production of consumer goods declined during the war and had to be rationed. Keynesians warned of a return to the Depression after the war but it did not happen because Roosevelt was dead and Truman replaced his crack-pots, not to mention that Republicans passed tax cuts.
In the Keynesian world, a war is good perhaps the greatest thing economically because there is virtually nothing but consumption, WWII’s consumption is still scattered all over Europe, but to those of us in the real world war and war’s destruction is hell.
All major wars were preceded by domestic social crisis,one may have to detect first the place of birth of a social crisis.Throughout hereunder samples social crisis have symptoms but were not predictable in their timing.
The make up of revolutions (Comment naissent les revolutions “Edition Perrin’)
A template of short essays on the past revolutions covering:
France 1788 and 1847 and 1871 and arguably 1968
USA 1773 sugar act followed by the stamp act survived by the tea tax,
Mexican 1810
Russia 1905 and 1917
Italy 1922
China 1949
The common features (symptoms) of these revolutions:
Unexpected in their timing.
Unexpected in their cities of birth (Grenoble 1788, Buzancais 1848,Boston 1773,Dolores (Mexican village) 1810,1905,1917 Petrograd, Fiume 1919,Shangai 1949)
Long lasting financial issues leading to social frictions and frustrations with aborted attempts to be taken care through either parliamentary or regal legislations,depleted public finance,rocketing unemployment,hidden or recognized prices inflation.
Distrust in the political instances and parties,doubts on the government ability.
They are no recipe for the make up on time of a revolution.The French revolution of 1848 was a surprise to K Marx himself.
Ricardo –
To the rage’s point, by 1941, GDP had returned to the long range trend line. In other words, the output gap had been completely filled in. That is a pretty conservative measure of ending the downturn.
http://macromarketmusings.blogspot.com/2009/01/great-depression-debate-in-one-picture.html
Ramp up to the expected war was no doubt a part of it, but far, far from the whole story.
Further, I defy you to find one word in Keynes that says war is an appropriate economic solution to anything.
BTW, the ’38 recession was a direct result of FDR’s mistaken attempt to balance the budget in ’37. Things were looking pretty much up by then, and he thought it was safe to go back to being a conservative. This earned him a chiding letter from Keynes, which you can easily find on the inter-tubes, if you are so inclined.
To your other non-point, there is not one shred of evidence that cutting taxes helps the economy in any way. As I commented here in another thread, taxes are the lowest they have been in at least 80 years. The tax cuts from about 1980 on have done bupkus for the economy. Bush’s were ruinous and insane. GDP growth is grinding to a bumpy halt.
Nothing in your comment is supportable by facts or data.
Cheers!
JzB
I think the question is really whether we relive the early eighties (and not just in fashion, which is – oh horror – already a fact) or whether we can actually best Carter-Reagan’s 5 years of economic wasteland. If we’re going with Carter/Reagan, we’re in for another two or three years of economic turmoil. If this crisis is going to be worse than the C/R one, it will be impressive to watch, and who knows when it will be over.
jazzbumpa wrote:
As I commented here in another thread, taxes are the lowest they have been in at least 80 years.
I was taking you seriously until I read this. Now I see you are just a comedian. (check tax rates when Reagan left office)
The NBER Business Cycle Dating Committee issued a statement today declaring that the bottom of the most recent recession was reached in June of 2009, with the economy in the expansion phase of the business cycle during the 15 months since then.
If the recession ended in June of 2009 what explains ZIRP?