When I first started working in the government, I was often confused by the many different ways in which economic data is reported. Now, thirty years into teaching economics – -particularly international economics – – I still have to help out my students when they venture outside of the safe world of textbooks to read official reports from different governments and institutions. So, here’s an incomplete (but hopefully helpful) primer on some conventions used in relevant government releases:
Quantities/Rates actual or annualized?
GDP level (US BEA): “Seasonally Adjusted at Annual Rate” SAAR
GDP growth rate (US BEA): SAAR
GDP level (Europe): quarterly, not at annual rate, SA
GDP growth rate (Europe): quarterly, not at annual rate, SA
GDP level (China): quarterly, NSA, sometimes cumulative by year; more recently quarterly is SA
Exports, imports level (US BEA/Census, Int’l Trade release) monthly or quarterly, SA; and NSA for bilateral ***
Balance of Payments components (US BEA/Census, Int’l Trans release): quarterly, SA
Balance of Payments components growth rates (US BEA/Census, Int’l Trans release): quarterly, SA
Flow of Funds component flows (US Federal Reserve Board): quarterly, SA
CPI Inflation (US BLS): both monthly and at annual rate
Industrial production growth rates (US Federal Reserve Board): monthly
Exchange rate changes: Usually depends on sampling frequency
Data pertain to what time?
GDP level (US BEA): over the quarter or year, depending on frequency
Exports, imports level (US BEA/Census, Int’l Trade release): over the month or quarter, depending on frequency
Balance of Payments components (US BEA/Census, Int’l Trans release): over the quarter
Flow of Funds component flows (US Federal Reserve Board): over the quarter
Flow of Funds components levels (US Federal Reserve Board): end of quarter
CPI level (US BLS): throughout the month
Nonfarm payroll employment (US BLS): payroll week including the 12th
Civilian employment (US BLS): calendar week including the 12th
Industrial production growth rates (US Federal Reserve Board): over the month
Money stock (US Federal Reserve board): end of month
Exchange rates: Often end of period, sometimes average of period
Summing Up:
Expressing levels or growth rates at annual rates is common, but not universal, even in US government statistical releases. Hence, care is warranted in the use and interpretation of the data.
*** For maximal entertainment, try to keep straight difference between trade flows on a Census vs. Balance of Payments basis in the International Trade release. Note these will not typically match NIPA figures even when converting to the same (e.g., annual) basis.
“Exchange rate changes: Usually depends on sampling frequency”
I was trying to explain something regarding an old Russian transfer pricing issue involving oil prices where the publicly available information was in rubles but the person I was talking to for some reason wanted to see the figures in dollars. It took me a while to figure out that FRED had monthly data on the exchange rate (yea I was having a bad day) but my goofball colleague wanted to see the exact daily quotes. I invited to do a Google search on his own time.
That colleague is just lazy. How many time have you (we all) been “assigned” a research task by someone who can jolly well look it up themselves?
Your colleague almost certainly missed the point if he/she wanted daily rates in the context you presented. Spot prices and daily exchange rates represent a pretty small fraction of oil trade for delivery. Financial market money is made in daily gyrations. Money in the physical market is made on delivery contracts, with a good bit of FX risk handled through hedging. There is not much new economic information in daily spreadsheet cells that is not available in monthly cells. But you knew that.
Menzie,
Thanks for this clarification. Indeed different data have different time periods and labels. It is not easy to keep them all straight (I have been off on some of them at times, for sure). It would be nice if media would be clear about what they are reporting precisely when they do so, as they often do not.
Happy New Year, you all.
JBR
@ Barkley Junior
You did notice BEA’s headline number for GDP is SAAR right?? I was afraid after the media had quoted that number for the last 40+ years or so you might have gotten confused. Let us know if your Quora website contradicts Menzie’s outline.
Yes, Moses Senior. But it is a minority of the data series listed that do that. Did you notice that some data reported from end of quarter to end of quarter? That was the mistake I was making when I made wildly inaccurate statements about GDP rates, statements you insist on repeatedly putting up here without any recognition that in fact I had the numbers right, the dates wrong.
This dredging over who said what early in the year is getting old. Heck this year is about to end, so it really is time to move on. But I note that your one correct statement about all this was to go along with the main herd on what would be the second quarter number. You did that correctly. Congratulations.
However, you repeatedly denied what I was the first here to point out who was not a Trump follower: that the bounceback from the bottom was sharp and looking more like a V than anything else, even if that V would stall out before getting all the way back, something I accurately predicted. You considered it ridiculous that I was noting data showing such a rapid bounceback, but it was accurate, and Menzie had to point that out to you.
I note that this involved me moving off what I had been saying eaelier, which had been in line with the herd here, including Menzie. That was a forecast along the lines of what Menzie’s source describes as a “deep” nonlinear pattern. That one starts out slowly then picks up. We have seen just the opposite of that, a rapid initial bounceback that then slows down or flattens. Given that the flattening has hit before it got all the way back, Not Trampis and Menzie say that means it is not truly a V, but it was certainly V-like until then, more V-like than anything else. In any case, I was the first non-Trumper here to se what was going on, that the bounceback was rapid and V-like, something you still have not accepted and attempt to ridicule.
Really time to move on past this running in circles. You were right on some specific numbers, but you were and continue to be massively wrong about the big picture of what happened.
This situation created a division between specialists and non-specialists, a division which is evident in comments here and which Menzie appears to be addressing in this post. The effort required to interpret data often leads to frustration, not just for non-specialists. It also leads to the oft-repeared “what economists don’t understand…” sort of assertion from non-economists. (In some contexts, it is reasonable to suspect that economists lack understanding, but not generally when it comes to using economic data.)
The Saint Louis Fed has done wonders to simplify the use of economic data with FRED: https://fred.stlouisfed.org/. Many regular commenters here are aware of FRED and use it in their discussion. FRED overcomes some of the differences between data series with the phish of a few buttons. Weekly data can be converted to quarterly data to allow easy comparison with GDP or flow of funds components. Differences in scale can be adjusted for. One can even do a bit of data shopping to see just what is included in this 768,000 series, if so inclined.
Too often, the comments section here is polluted with mindlessly repeated right-wing blather, an effort to derail legitimate discussion, but we could do better. Interesting discussion, based on actual data, is not beyond the reach of non-specialists. It would be great to see new names join in, and FRED is there to help.
Apologies to regular (non-troll) commenters and to our hosts. End of editorial.
Good Lord! With the PUSH of a few buttons. Push.
Phish of a button made me chuckle!
How many discussants want to be associated with a place filled with ad hominem invective? No one worth their salt.
FRED is a true treasure.
Now there are certain things it does not report such as financial information for publicly traded companies. I raise this because one of our trolls is saying consumers are not being the incidence of these tariffs as companies like Lowes are taking it on the chin. Now this troll has never wondered over to http://www.sec.gov which will give users handy access to all sorts of useful data noted in the 10-K filings. BTW – Lowes profit margins are doing quite well.
And of course Peak Trader tried to peddle the line that the US is the largest consumer of steel and aluminum. Fortunately for us simple Google searches point to reliable data that shows China dwarfs us in consumption of both commodities.
But our trolls persist with claims never checking the basic data.
I had to search for an example of this but back during the Bush43 days, a National Review troll named Jerry Bowyer constructed “Buzz Charts” which had to be some of the most misleading presentations of economic data ever. Mark Thoma may have stopped blogging but one can find some of his classic blogs, which often went after Bowyer and his “Fuzz Charts”:
https://economistsview.typepad.com/economistsview/2005/07/two_lines_on_a_.html
Oh to Menzie, congrats on your journal being in the top 3.1% of econ journals according RePec recursive discounted factor. Mine is only in the top 12.5%. I know you are doing a good job there.
Off topic but funny:
https://www.youtube.com/watch?v=fn3KWM1kuAw&feature=youtu.be&fbclid=IwAR0o-x-T6OSijmEzwDjkQa9UBX0zWlMfvPz3SdtXBjH5MiveJ0UOOU5A-sk
If anyone plays “Do you love me – now that I can dance” I have to get up and dance. But I do not have the moves these robots display!
My favorite dancing robots is Beck’s video for “Hell Yes”.
Those things built by Cyberdyne? May be less funny in a couple of iterations.
Helpful list, Menzie. Thank you.
At the risk of being ungrateful, one could ask for a table with name of the series, source (link), source (issuing agency), ordinary dates of issue. Maybe on a spreadsheet? Maybe worth putting up on the your academic website with a link as a standing reference?
Happy New Year!
You need help figuring out how to go to http://www.bea.gov or http://www.bls.gov or Census or the Federal Reserve? Just wow!
BTW – those excellent FRED graphs do provide the kind of information you want. You do know how to Google “FRED”.
Oh, c’mon. We talked about this…
The IMF has changed the format of data offered, so that as far as I can tell saved data links extending from 1980 through 2019 now no longer work:
https://www.imf.org/external/pubs/ft/weo/2019/02/weodata/weorept.aspx?pr.x=45&pr.y=2&sy=2000&ey=2019&scsm=1&ssd=1&sort=country&ds=.&br=1&c=924%2C134%2C534%2C158%2C111&s=NID_NGDP%2CNGSD_NGDP&grp=0&a=
October 15, 2019
Total Investment & Gross National Savings as a Percent of GDP for China, Germany, India, Japan and United States, 2000-2019
Would there be a way to get the old links to work (saving my slides)?
ltr: You’re using links to old WEO datasets, which have probably had their URLs changed. Go the most recent WEO, then look for ancillary material.
You’re using links to old WEO datasets, which have probably had their URLs changed. Go the most recent WEO, then look for ancillary material.
[ Yes, I understand now how I can comfortably make the link conversions. I am quite grateful.
Thank you so much. ]