I listened to NPR news this morning while stuck in Boston traffic and heard three different reporters speak about how “the U.S. economy grew 3.9 percent in the third quarter [of 2014].” That’s an awesome 16.5 percent annual growth rate (like you might get if you implemented congestion pricing so that Americans didn’t get stuck in traffic and have to listen to NPR news after running out of books on CD…). However it turns out that the real number is closer to 1 percent and 3.9 percent is the annualized rate (BEA press release). As someone who has been reading about America’s educational system a bit lately (see my review of Higher Education?, also Academically Adrift, and the American schools section of The Smartest Kids in the World), I’m kind of curious about how people who spend the time to put together a story to be broadcast nationally could have missed the difference between 3.9 percent and 16.5 percent.
9 thoughts on “How to feel better about your country: Be confused about numbers”
Comments are closed.
The press is not just innumerate but ignorant.
Recently , the NY Times had a (not very good) magazine piece called “A Brief History of Failure”
http://www.nytimes.com/interactive/2014/11/12/magazine/16innovationsfailures.html
about technologies that ultimately failed. One of them was direct current. Here is the articles explanation of AC:
” Using transformers, alternating current could send out a large charge on small wires across great distances, from a single power plant, which was then reduced, or alternated, upon reaching a home or business. ”
The author is (wrongly) using “large charge” to mean “high voltage” and thinks that “alternated” is a synonym for “reduced”. In other words, he actually has no idea how AC really works. He understands electrical power distribution about as well as the late Senator Stevens understood the Internet (“a series of tubes”).
Basically what is going on is that journalists are the English majors who were not smart enough to do the math required for STEM and it shows.
I’ve found over the years for NPR to be the most crudely propegandistic (always in support of whichever administration is in power) of the media outlets. I have no doubt that there is an editor there somewhere who knows perfectly well that the GDP growth figure for the first quarter was really just under 1%, not 3.9%.
Anyway, I suspect that these GDP growth figures are made up completely, or at the least inflation is not taken into account or badly understated. I haven’t paid attention to them in years.
An annuallized figure based on one quarter’s data is simply a projection of what the figure would be for the year if the exact same trend from the past three months were repeated during the next nine months. Sometimes in sports reporting, which is generally more honest than reporting about the economy, they will say that a player who has hit 25 home runs in the first half is “projected” to hit 50 home runs for the season. That is all the annualized figure really is. But they don’t say that the player has hit 50 annualized home runs as of the All Star break, or put more significance on that number than its deserved. “Projections” are generally not made for team won loss totals. Its understood that the actual annual statistics will differ from the projections, sometimes to some extent, despite the home runs or wins or whatever in the first half remaining and counting towards the final total. I’ve done projections of financial data for companies based off of quarterly financial statements, and they are to be taken with big grains of salt, and the technique is only used only if harder data is not available. At any rate, if you want to present the most yearly GDP data, you could take the last four quarters of quarterly data for any quarter and calculate from that, you don’t need to wait until the time the BLS puts out annual data in a press release.
I have no doubt that there is an editor there somewhere who knows perfectly well that the GDP growth figure for the first quarter was really just under 1%, not 3.9%.
It could be that there’s just an assumption that NPR listeners consume a lot of news and thus know what it means when a statistic comes out saying that there was 3.9% growth in a quarter.
Vince,
That’s the assumption that I made when I heard the same news story this morning… that the growth rate in the 3rd quarter was the annual rate.
The fact that it was adjusted up from 3.5% to 3.9% was the more curious part of the story (for me). That 0.4% adjustment is a pretty big swing (11.4% change). I guess that the early numbers must have been missing something pretty huge in the economy.
I listened to NPR this afternoon and they clearly said that the economy grew at a 3.9% “annual rate”. So maybe someone bought them a clue.
Economic statistics have a strong seasonal bias, so it wouldn’t make sense to measure them quarterly. Some indicators are reported quarterly or even monthly, but the measurement period is typically the trailing 12 months, *or* the statistic is an annualized projection.
GDP and CPI are clearly gameable and actively gamed. And consumers are certainly innumerate, but annualized measurements really are the right thing to do.
But I also think financial media reports that happen to be done properly are largely accidental. They just rip and read the stats from the report abstract without any real comprehension. And consumers return the favor.
My ire is raised every time I hear a “stock market report” that only mentions the DJIA and NASDAQ, and of course reports the deltas in raw dollars instead of percentages.
But it’s a bit unfair to call out NPR on this one. Every single general media outlet does the same thing.
Scott,
The usual response when folks are surprised by large up- or down-wards revisions is that some parts of the beast are extremely hard to get timely numbers on. I don’t know how true that is, but it sounds reasonable to me.
http://www.bea.gov/about/pdf/jep_spring2008.pdf explains why these are “estimates” and it isn’t surprising that there are mistakes!
Most people have some degree of illiteracy, innumeracy or otherwise deficient training and skills. It is the norm, not the exception.
At some point we have to be honest with ourselves that not everyone can be or wants to be a professor. We have to be realistic about what people should reasonably know and how extensive that knowledge is.
Being poor with one skill or another is only shameful because we have made it so. We have created a culture of dishonesty where everybody is claimed to be above average. I am not sure how much of the population even understands what is ultimately wrong there.