Humans Are Awful at Predicting Their Own Behavior

This morning, Steve Liesman on CNBC showed all of the National Retail Federation numbers for spending plans this holiday season. Increases of 22%, big numbers.

Unfortunately, Steve omitted to show how the actual data of the past years have lined up with the NRF forecasts. As was discussed this past Sunday in the Washington Post, the answer is not particularly well.

Indeed, when we look at how far off these surveys have been in the past, it makes one wonder why anyone should pay any attention to them at all.

In 2005, based on a survey on Black Friday and Saturday, the NRF forecast a 22% percent increase in holiday shopping based on surveys conducted over the Thanksgiving weekend. The actual results? Up just 1 percent.

The same foolishness resurfaced again in 2006, with an 18.9% sales increase forecast. Sales were actually up almost 5%.

In 2007, just as the recession was getting underway, the NRF forecast a 4% gain in sales. U.S. retailers “unexpectedly” dropped 0.4% in December 2007, the weakest holiday season since 2002.

Given the broad scale of the economic collapse in 2008, expectations were for a 2.2% sales gain; they actually fell 6%

They don’t just get it wrong to the upside; during severe weakness, they exaggerate losses to the downside. With memories of the economic collapse still  fresh, NRF’s 2009 Holiday Consumer Intentions and Actions Survey for holiday shopping reflected an awe-inspiring drop of 43% versus 2008. 2009 holiday sales actually rose ~3%.

Last year, 2010 Black Friday weekend sales rise were estimated at 9.2%; They actually gained 5.5%.

Anyone quoting NRF surveys but failing to show the actual results are doing their viewers a disservice.

Year NRF Survey Forecast Actual Holiday Sales
2005 22% 1%
2006 18.9% 5%
2007 4% -0.4%
2008 2.2% -6.0%
2009 -43% 3%
2010 9.2% 5.5%
2011 16.6% ??

What's been said:

Discussions found on the web:

Posted Under