The Index of Consumer Sentiment is misleading: but there is a survey-based measure of consumer psychology which is strongly correlated with retail sales growth

Since mid to late 2010 retail sales growth has been weaker than would have been expected given economic conditions.  For example, standard mortgage interest rates were more than 2% lower than their 2008 peak and are still nearly 2% lower in 2011.  Petrol prices are also still below their 2008 peak.

Many reasons have been put forward over the past 12 months to explain this change in consumer behaviour, including:

  • Increased online shopping;
  • Higher interest rates;
  • Abnormal weather;
  • Fears of a carbon tax;
  • Political instability;
  • Higher prices of services such as electricity;
  • Higher household saving.

There has been little in the way of compelling evidence to support most of the claimed causes.

David Jones’ chief executive, Paul Zahra, recently said “we see ourselves as being in the eye of a perfect storm.  We have the strong $A, we have leakage to offshore online retail … we have new taxes which predominantly target our customer base, a high savings ratio, international uncertainty, and higher outbound tourism.  Customers are choosing not to shop – they’re saving their money and paying down debt.  People don’t know what else is about to hit them”.
Australian Financial Review, 15 July 2011, page 45.

Westpac senior economist Matthew Hassan said the improvement in consumer spending was surprising, given the fall in consumer confidence to levels only ever seen during downturns.
The Australian, 6 October 2011 (consumer confidence is the frequently used synonym for the Index of Consumer Sentiment).

Retailers and economists often talk about the Index of Consumer Sentiment when describing the state of discretionary consumer spending.

We have consistently found that the Index of Consumer Sentiment (ICS) has a very low correlation with retail spending growth – telling us very little about the present and nothing about the future.  Our conclusion was mentioned by Australian Financial Review on 20 February 2001, page 29.  Indeed, the ICS is very often a misleading indicator.

Researchers at the Reserve Bank of Australia have reached much the same conclusion (their research report on sentiment surveys in January 2002 concludes “there is little evidence that the surveys tell us anything we didn't already know.” Research Discussion Paper 2001-09, “What do sentiment surveys measure” by Ivan Roberts and John Simon).

The main problem with the ICS is that it is much more strongly influenced by variations in the Australian dollar exchange rate than is consumer spending.  Its strongest driver is a weak driver of discretionary spending.

Chart 1 illustrates the problem.

Chart 1

In 2009/10, the ICS rebounded strongly after weakening during the global financial crisis.  But retail sales growth slowed.  Indeed, during 2008/09 the ICS slumped - prompting talk of a recession that Australia never had - and retail sales growth was stronger than many who study the ICS expected.

The Consumer Pulse® tracking surveys conducted by foreseechange are much more strongly correlated with retail sales growth (Chart 2).

Chart 2

The Consumer Pulse measures provide a reliable diagnosis of the health of retail sales - in total and for the broad categories.  Being survey based, they can be analysed by demographic and geographic region.  And they can be used for developing scenarios of future growth rates.

We no longer have to list a range of possible factors which may or may not be affecting spending or comment on a spurious index.  We know the primary reason for the downturn and what needs to happen before retail sales growth can recover.

For more information, see Consumer psychology driving the retail spending downturn - Foreseechange Forecast and Analysis Reports

Charlie Nelson
October 2011