Dec. 2 (Bloomberg) — One would hope that after a decade of people lamenting the poor reporting of Thanksgiving weekend retail data, the news media might learn to do better. Yet we see few, if any, signs that reporters understand the misleading press releases put out by the retail industry and its flacks.
The innumeracy of the media has not been greatly exaggerated.
The Wall Street Journal, the New York Times and others reported industry estimates of sales as if they were grounded in fact. As we discussed last week (“Ignore the Black Friday Hype“), these shopper surveys and measures of foot traffic have in the past had no predictive value relative to retail sales. Their track records are fairly awful. The National Retail Federation has been especially terrible in its forecasts. A coin flip — 50/50 chance of getting it right — has much greater accuracy than these surveys.
Let’s look at each of these reports to see if we can identify where they went awry.
“Retail spending over Thanksgiving weekend dropped for the first time in at least seven years, the industry’s main trade group said …”
It may have said so, but we do not have any reliable data yet as to how much was spent over the holiday weekend. What the WSJ was reporting were answers to a survey question as to what shoppers recalled spending last year over Thanksgiving weekend – – $423.55 — versus what they thought they would spend this holiday weekend — $407.02.
Neither shopper number is based on any review of actual spending; these are figures that random shoppers literally make up on the spot. They are made without the benefits of looking at checking account balances or credit card statements. History informs us that neither guess is remotely accurate.
One would imagine that the best-read business publication in America would be able to handle analyses of this sort easily. Over the years, I have witnessed other sorts of innumeracy from the Journal when it pays attention to to industry flacks. The one that stands out most was confusing the annual seasonality of home sales with a housing recovery (see this and this).
Perhaps these stand out because the Journal generally does a good job, and has columns like Carl Bialik’s The Numbers Guy (he’s now going to fivethirtyeight.com) to put data into better context.