More Bad Data from the NRF?

I’m kinda dumfounded to see this issue come up time and again, but — there they go again: The National Retail Federation is once again putting out data which has gotten misinterpreted by most of the MSM:

“Retailers kicked off the holiday selling season in style as shoppers across the country set their alarms for the wee hours of the morning to catch doorbuster specials. According to the National Retail Federation’s 2006 Black Friday Weekend Survey, conducted by BIGresearch, more than 140 million shoppers hit the stores on Black Friday weekend, spending an average of $360.15, up 18.9 percent from last year’s $302.81.*”

Or so said the NRF’s press release. Note that little asterisk?* Follow it, and we see

“*Spending data includes Thursday, Friday, Saturday and projected spending for Sunday.”

Even that disclaimer is inaccurate:  First, this is not based upon actual sales data, but rather, is a survey of consumers. Not only that, but much of the survey results are self-reported projections of spending expectations — not reciepts. The survey dates are 11/23-11/25. This means survey interviews done on Thursday 11/23 are almost all forecasts of future behavior; depending what time of the day they are done on Friday 11/24, between 2 and 3 days of data are predictions, and perhaps one day is self-reported data.

To get an accurate read of retail sales, we need actual DATA — like company sales reciepts or Credit Card purchase totals. Asking people how much they are planning to spend or have already spent is a surefire methodology of getting bad dope.

Note that this has become an annual rite of error by NRF. In my opinion, based on my read of how they present this data, I suspect they are purposefully attempting to mislead the media. We saw the exact same issue last year  in the way they report their survey: In 2005, the NATIONAL RETAIL FEDERATION said holiday Retail sales rose 22%; We later found the actual sales data was nowhere near that statement. Taking a survey forecast and reporting it as actual sales is not honest.

Given last year’s debacle, as well as the more recent cheerleading back-to-school forecast, one would have hoped the NRF would make it clearer that this isn’t actual sales data — but rather, is a survey asking people how much they intend to spend, and on what items.

It is quite unfortunate that most of the media reports this survey as if they were actual sales. They are not.

Why? Surveys are very subjective; People are notoriously inaccurate at reporting their own behavior: Their own Egos can get in the way (Yeah, I’m a big spender!); How questions are phrased impacts answers; Even the immediate prior question on a survey can and will change a respondents answer tot he next question. Phrasing, grammer, buzzwords — all impact results.

Given that the NRF is an industry spokesgroup, we cannot expect any degree of objectivity — but should instead recognize them for what they they are: A industry trade association promoting the goals of their membership. There’s nothing wrong with this cheerleading, as many industry groups have their own trade associations who spin the data (like the National Association of Realtors do with Housing Data). The problem arises when the mass media reports their marketing releases as a form of objective economic data. They are decidely not.

Consider how widely survey results can and do vary. For example, CNBC also conducted a survey of Holiday shoppers — about on-line
sales, specialty stores, big box discounters, and the factors that might affect their
overall spending. Their results? They found that spending this year is likely to be flat. Here’s an excerpt of the CNBC survey result:

“Despite a huge turnout of shoppers on Black Friday and the rest of the Thanksgiving weekend, most Americans will spend less or the same amount they spent last year on gifts, according to CNBC’s exclusive Holiday Central Survey. Surprisingly, overall falling gas and energy prices had little or no effect on overall expenditures, according to the survey…

Specifically, 46% of those surveyed said they will spend “about the same amount” they spent on gifts last year and 32% said they plan to spend less than they spent last year. The average amount consumers plan to spend this year on gifts is $764. Only 20% said they plan to spend more than they did last holiday season.”

I’m not suggesting the CNBC survey accurate,while the NRF survey is not. But it shows how widely varying survey results can be based on who asks what questions.  However, in the spectrum of likely sales gains at this stage of the economic cycle, between a gain of 18.9% versus flat spending, my expectations — a +2 to 4% range — are much closer to flat than a huge increase over last year.

And, its hard to imagine a 19% increase in spending would lead to triple digit loss for the Dow . . .

Last year, numerous media mistakenly trumpeted misleading data as an actual measure of sales, rather than a mere survey. Once again this year, we saw similar errors showing up in various Holiday Retail articles — some more or less erroneous than last year:

WSJ:  “In a survey of 3,090 consumers, the trade association found that shoppers spent an average of $360.15 this weekend, up nearly 19% from last year’s $302.81. Discounters were still the most popular shopping destination, but their share dropped significantly from last year; about 50% of those surveyed said they visited a discounter over the weekend, compared with 61% last year.”

They didn’t spend that much, they said they spent that much — and that’s huge difference. Bloomberg is even more explicit:

Bloomberg: “U.S. consumers shopping over the Thanksgiving holiday weekend spent 19 percent more than a year earlier, outpacing the advance of 2005, after retailers slashed prices to attract customers.

Consumers spent an average $360.15 from Nov. 23 through yesterday, up from $302.81 a year earlier, the National Retail Federation said yesterday in a statement. Fewer people shopped, with about 140 million visiting stores during the four days including Thanksgiving, down from 145 million last year.”

No, spending was not up 19%. People said they expected to spend 19% more. These were not actual spent dollars, but rather, were expectations of spending. There’s an enormous difference, especially when we consider that people tend to be very poor judges of their own behavior.


Newspaper readership has dropped significantly over the past few years. Budget cuts have reduced the number of reporters, fact checkers, assistant staffers. And, it may only get worse, as the internet pulls even more readers from traditional print outlets, pressuring news gathering budgets even further.

Its easy for trade organizations to crank out official sounding statements; their adeptness at manipulating the media is apparent. We should brace ourselves for more of this sort of marketing spin being reported as fact in the future.

And thats a shame . . .


UPDATE: November 28, 2006 11:09 am

So, now that the data is out — how did the NFR “forecast” square with reality? The WSJ’s  Holiday Sales News Tracker noted:

Chain-store sales slipped 0.4% in the week ending Saturday from the prior week, the International Council of Shopping Centers reported Tuesday. Not too impressive given all the reports of jam-packed malls over the holiday weekend. However, when you compare last week’s performance with the same week a year earlier, chain-store sales increased 2.6%.

Redbook puts out a similar survey but got slightly different results. Chain-store sales rose 0.1% in the first four weeks of November from October, and 3.2% from November 2005, according to Redbook’s gauge.

2.5% . . . 3.2% — thats very consistent with my pre-Black-Friday forecast of 2.5-3.5% range for the holidays.



UPDATE: November 28, 2006 4:37 pm

I spoke with Scott Krugman of the NRF; He pointed out that the full press release specified that this was the result of a survey and not actual sales reciepts; The NRF retains the services of “Big Research,” an independent marketing research firm — and then reports back the results of their survey.

Further, the NRF economic forecast back in September was fairly circumspect:  NRF Sees Subdued Holiday Gains in 2006.
So to be fair, while I have issues with their first paragraph, its the Media’s responsibility to do their homework and report the news accurately — not a Trade Association’s.




Early Birds Catch Plenty of Worms as Retailers Have Lucrative Black Friday


Black Friday survey

CNBC, November 27, 2006

CNBC Video Report on Black Friday

Holiday Sales Get Off to Solid Start, But Wal-Mart Doesn’t Share Cheer
November 27, 2006; Page A1

U.S. Thanksgiving Spending Increases 19% on Discounts
Mary Jane Credeur
Bloomberg, Nov. 2006


Category: Consumer Spending, Data Analysis, Financial Press, Retail

Have we already enjoyed the Year End Rally?

Category: Currency, Investing, Markets, Psychology Gift Certificate

Category: Retail

Massive Retail Roundup

Category: Consumer Spending, Economy, Financial Press, Retail

Recession: The Stage Is Set

Fascinating interview With Richard Arvedlund, Founder, Cypress Capital Management, who is not particularly optimistic on the economy going forward:

WE CAN ALWAYS COUNT ON RICHARD ARVEDLUND to take a different tack. Independent and bold calls on the economy come easy to this longtime money manager, who’s seen it all in his 30-plus-year career. But his balanced investment approach, with a focus on high-yielding, big-cap stocks combined with some bets on bonds, helps his clients preserve their capital as much as build it. The founder of Wilmington, Del.-based Cypress Capital Management, which has $450 million in assets and is now a unit of WSFS Financial, is at his best in troubled times. Trouble, the way he sees it, is straight ahead.

Barron’s: It took a year, but the calls you made when we last spoke are looking pretty good now.

Arvedlund: Well, until midyear the economy was running much stronger than I had thought it would. However, a GDP [growth domestic product] slowdown has clearly begun. The GDP growth rate dropped to 1.6% in the third quarter from 2.6% in the prior quarter and 5.6% in the first quarter. We have not seen GDP growth below 2% for four or five years. We now have preconditions in place for a recession.


The preconditions would be the following: Whenever housing starts and permits drop by the rates of decline that have been exhibited — 10% to 20% — it has always preceded a recession. What is remarkably different in housing than just about any other sector of the economy is that whenever housing cycles turn down, and that’s happened twice in the last 30 years, once in the late ‘Seventies and once in the late ‘Eighties, the downturn tends to last much longer than people dream. The average cycle is three to four years.

Continue reading "Recession: The Stage Is Set"

Recession: The Stage Is Set
Interview With Richard Arvedlund, Founder, Cypress Capital Management
Barron’s, Monday, November 13, 2006

Read More

Category: Earnings, Economy, Investing, Markets

Goldilocks Word Count

Category: Economy, Financial Press, Psychology, Web/Tech, Weblogs

New WSJ Blog: Holiday Sales Tracker

Category: Financial Press, Retail, Weblogs

Retail Sector Market Performance

Category: Consumer Spending, Markets, Retail

iPod versus ?

Category: Music, Technology

Best Buy Doorbuster

Category: Retail