March 01, 2007
New feature: Econbrowser faces the data
We've added a couple of new features to the sidebar.
The first is in response to longtime reader Dryfly, who complains that he's sometimes confused about what our bottom line is here at Econbrowser. Actually, we seem to have at least two different readers who both call themselves Dryfly, which I find a little confusing myself, but my mission today is to resolve complaints, not raise new ones. Anyway, at least one of our Dryflies doesn't like the fact that one week we seem to say everything's fine and the next week we're all worried.
Part of that is just the nature of the beast, the beast (me) being a multi-handed economist. But I kind of felt that I'd been doing the sort of thing Dryfly is requesting in my rare invocation of emoticons to signal a decisive turn in my outlook. We brought out the happy face last September 13 for the incipient big drop in gasoline prices that followed a significant decline in mortgage rates, and resurrected Sad Sack for the cumulative drop in industrial production revealed February 21, coupled with the calculation that the stimulus from lower mortgage rates has now been played out. So for the benefit of Dryfly and anyone else who wants to know my current bottom line on the implications of all the recent incoming data, I'll keep the most recently used emoticon in the right-hand panel labeled "Econbrowser faces the data."
A second addition is in response to a complaint from UCSD Communications Director Barry Jagoda, who says he couldn't find where in Econbrowser we reported the latest recession probability index. OK, I admit he's the only person in the world ever to raise this issue, but it's near and dear to my heart, and I've got a notion my index may get some actual exercise this year. So just to keep Barry and me happy, we'll be posting the most recent value of the recession probability index (currently, that represents an assessment of the situation of the U.S. economy for the third quarter of 2006) in the right-hand column as well.
I don't mean to get anybody too excited about checking these new fun items every day to see if they've changed, because unless something really extraordinary happens, they won't-- longer term stability is the whole point of Dryfly's complaint, and the recession probability index is constructed from the GDP advance estimate, which is only released once per quarter. But, if you only check in with us once every few months, you can just glance at the right column for a quick summary of all you've missed.
Of course, if you only check in with us every few months, you'll probably never read this explanation. But that's the occupational hazard of blogging, don't you think?
Don't bother answering, that's supposed to be a rhetorical question.
Posted by James Hamilton at March 1, 2007 08:09 PMdigg this | reddit
time-varying parameters stubbornly vary with time...
why should not our opinions?
Posted by: sam at March 2, 2007 06:30 AM
JDH - Can you explain why the recession probability indicator is not currently calculated for 2006:IV? Is it because "Our recommendation is that one should wait until one extra quarter of GDP growth is reported or one extra month of the monthly indicators
released beforemaking a call of a business cycle turning point." If so, does that mean that you are using all current information in estimating the recession status of the previous quarter? How are revisions used in the model? Finally, is there a way to produce some estimate of the size/severity of the recession with some s.e. using your method? I would work through the 74 pages of the ms if tenure weren't breathing down my neck. Apologies for the laziness. Thanks
Posted by: stirthepot at March 2, 2007 08:41 AM
Stirthepot, the index for 2006:Q3 uses the advance estimate of 2006:Q4 GDP growth reported Jan 31 to refine the inference about where the economy likely was in 2006:Q3. The reasons for the delay are (1) data revisions for the most recently reported quarter can be large, as we've just witnessed, and (2) you really need the second quarter of data to be confident about recognizing the trends. The goal of the index is to replicate the calls made by the National Bureau of Economic Research as to when the economy goes into and comes out of a broad business cycle downturn. To do that accurately, you need decent data, but in my opinion you don't need to wait one or two years before making your announcement, as the NBER often has. My research suggests that we can make a pretty reliable call if we wait just for one quarter of new data and data revisions.
Yes, one could use the same principles to try to make an earlier guess, but the index reported here is intended to offer something a little more reliable and authoritative. Based on the historical analysis, anyway, I'm confident that I'll tell you pretty much the same thing that NBER eventually will confirm, only you'll get it here sooner.
Posted by: JDH at March 2, 2007 09:03 AM
I think the usefulness of the Recession Probability Index for the layman will come if and when we get into an "iffy" situation where some people say we're in a recession and others see it as just sluggish growth. Then, if we're willing to wait six months or so, we can get a more objective picture of where we were. Of course, if the RPI reads near 50% that may not shed much light.
But otherwise, if we do run into a serious recession, it seems like the RPI will just confirm what everyone will by then have already known.
Posted by: Hal at March 2, 2007 10:57 AM
Hal, to a certain extent I have to agree with you. On the other hand, I remember an interview I did with a local TV reporter in January 2003. When I told him that the recession was over, and had been over for a year, he looked at me as if I was from outer space, my claim being so completely foreign to everything that he as a journalist was seeing and hearing at the time. Indeed, NBER did not announce the end of that recession until July of 2003. So, I can certainly imagine situations in which it could be helpful to have an authoritative declaration about this earlier than it comes from NBER.
Posted by: JDH at March 2, 2007 11:57 AM
Just to be clear, what the NBER announced in July 2003 was that the recession ended in November 2001.
Posted by: JDH at March 2, 2007 02:17 PM
I think the reason you were met with that "stare" was because the way economists define "recession" differs maredly from how the man on the street defines it. To Everyday Man a recession is when he, or people around him are hurting economically. That is most likely to be the case right after the recession is over (because that's when you are most likely to have maximum unemployment).
Don't you think?
Posted by: bellanson at March 8, 2007 09:14 AM
Perhaps that's part of it, bellanson. But I also think that reporters sometimes have trouble balancing the good news with the bad.
Posted by: JDH at March 8, 2007 09:34 AM