Sceptics Corner
D. Thomas
February 2001

Treat Confidence Indexes with Caution

There have been many doomsday headlines in the Australian press recently about the collapse of business confidence, as measured by several surveys.  But business confidence plummeted in March and June 1998, in the midst of the Asian economic crisis, and yet GDP growth paid no heed and powered ahead.

At the time of the Asian economic crisis, most official and market economists badly misread the implications for Australia, in part on the basis of falling confidence.

When the Reserve Bank noted yesterday that the economy was travelling ‘better than either the bank or other forecasters had foreseen’, it was being too kind to the economics profession.

The experts who last year correctly tipped how strong the economy would be in 1999 could hold a reunion party in a phone box and still hire a band
The Age, 7 May 1999, Business page 1

Dun & Bradstreet’s economic consultant, Dr Duncan Ironmonger, said the rate of economic growth would be weaker than at any time in the past two years and there was a 50% chance of Australia sliding into recession.
The Age, 17 July 1998, Business page 1

 


Of course, at the time of the Asian economic crisis, interest rates had fallen to the lowest level in a generation over the previous two years and the lagged boost to consumer spending was about to kick in.  This time, rates have been rising over the past 15 months and this increases the risk that the confidence indexes may indeed presage a slowdown.

The relationship between consumer confidence and retail spending is very weak and often misleading.

 


 

After adjusting retail growth for interest rate movements, there is a significant but weak contemporaneous correlation.  If confidence rises by 10% (from 100 to 110, for example) then retail turnover will increase by 0.4% in the same quarter, all other factors being equal.

The consumer confidence index does provide some information about the future.  Not about consumer spending but about the behaviour of firms.  For example, imports tend to rise two quarters after a rise in consumer confidence – because importers believe that the consumer confidence index does contain information about the future, even though it doesn’t.  That is, their use of the consumer confidence index leads to sub-optimal inventory decisions.

Research in the UK has reached similar conclusions about consumer confidence providing only information about consumer spending in the current period (“Composite Forecasts.  Non-stationarity and the Role of Survey Information” by Holly and Tebbutt, Journal of Forecasting, April 1993).  A US study found that some consumer confidence measures contained information about the future of consumer durable purchasing, but not consumption goods (“The Indexes of Consumer Sentiment and Confidence: Leading or Misleading Guides to Future Buyer Behavior” by Huth, Eppright, and Taube, Journal of Business Research 29, 1994).  However, the latter research examined only bivariate correlations, whereas the former (and our Australian research) allowed for multiple effects including interest rates.  In other words, once we allow for interest rate movements, consumer confidence measures contain no information about the future of consumer spending.

The penultimate word goes to the Australian firm Access Economics, whose column in the Business Review Magazine that hit the newsstands the day before the 1996 federal election said that it would be unusual for a government to lose an election with consumer confidence so high.  The next day, the government was booted out of office in a landslide!

Treat confidence indexes with great caution.  Evaluate the impact of other factors before relying on confidence as an indicator.

Postscript

The Reserve Bank of Australia issued a research report on sentiment surveys in January 2002. It concludes "there is little evidence that the surveys tell us anything we didn't already know."

Despite the overwhelming evidence, economic commentators still persist with their bankrupt commentary on consumer sentiment/confidence surveys.  For example:

"Consumers say they were more cautious in February, according to the latest reading of consumer confidence.  But that hasn't stopped them from shopping.  Store sales remain healthy, and car sales are holding up better than anyone thought possible after the late-2001 boost from dealer incentives."
Business Week Asian Edition March 11 2002, page 21.