There’s a chart flying around the blogosphere that seems to show that American families were wealthier in 1983 than in 2010, and the figures in it are attributed to me. I never made that claim, because I have no idea whether or not it’s accurate. What I can say is that American families were wealthier in 1989 than in 2010. There is no solid information for years before that, at least from the Federal Reserve.
The story of the chart that took on a life of its own gives some insight into how memes are created and circulated on the Internet. It started life at Reuters with a sensible disclaimer, but as it circulated those disclaimers have been sandpapered off. It has popped up on the Washington Post website, the PBS NewsHour’s Rundown blog, Brad DeLong’s tumblr feed, and probably some other places I haven’t seen yet.
Let me explain what happened. When the Federal Reserve issued its Survey of Consumer Finances on June 11, it presented data extending from 2001 to 2010 but added an intriguing note. It said, ““median net worth was close to levels not seen since the 1992 survey.” It just didn’t provide the data to support that assertion.
So I dug through old surveys online and updated the figures in them into 2010 dollars, using the same price index the Fed said it uses itself. I created a table going back to 1989, which was the first year for which the Fed conducted the survey using the same methodology it uses now.
The problem I faced was how to deal with surveys the Fed did using other methodologies in 1962 and 1983 (there were no surveys between those two years). I mentioned my calculations for 1962 and 1983 at the bottom of my article—in a paragraph, not an official-looking table—explaining pretty clearly that they were not comparable to the 1989-2010 figures because of a change in methodology. My article on businessweek.com was headlined, “American Families Are Poorer Than in 1989.” (It shouldn’t have said “are,” by the way, because the data only went through 2010.)
Once I finished the article, the meme machine started cranking up. I plugged my businessweek.com article on Twitter (@petercoy), and I sent messages via Twitter to people I know at other news outlets, calling it to their attention. Felix Salmon, the super-smart blogger at Reuters, picked up my numbers and made a chart out of them. He included the 1962 and 1983 numbers, which gave them more credibility than they deserved, but to his credit he put them in a different color and added the qualifier that they could not be compared directly with the rest. Actually, those two years can’t even be compared with each other. In any case, the chart was born.
The PBS NewsHour linked to the businessweek.com article but threw in the Reuters chart without explanation. The Washington Post’s “The Fix” blog left off the caveat about methodologies and made the even stronger claim that the data “suggest net worth may well be at its lowest point since the 1960s.” DeLong, a top economist at the University of California at Berkeley, posted the Reuters chart on the tumblr microblog without any context at all.
By the way, Bloomberg Businessweek‘s own chart, which will be appearing in the print magazine later today, goes back only to 1989.
One bad thing about the Internet is that subtleties can be quickly lost. One good thing is that it can self-correct. I’m hoping this article will begin that process.