Be Careful Using Consumer Confidence Measures

futurelab default header

by: Dick Stroud

Marketing Week has an article about the age and regional variation in the TNS/Nationwide consumer confidence measure (Nationwide Consumer Confidence Index). This is the measure most often quoted in the UK – something similar to the Conference Board in the US.

The article focuses on the changes in this index during the past 3 quarters and what it tells us about the way that different age groups react to our present financial Armageddon.

I really don’t know what I make of these numbers. Obviously, youthful enthusiasm (some would say naivety) makes them the group with the highest confidence level, although it has dropped by the highest amount. The 45-59 year olds seem to have plateaued out. The 60-75 year olds have had a sedately decline into gloom. I feel sorry for the 75+ crowd who are perceived as worth measuring!

The regional analysis is interesting but difficult to explain, other than in gross archetypes of inhabitants of each region. I better keep my theory on this one to myself.

I decided to dig a bit further to find out about the way this index is calculated. This is the latest press release from Nationwide showing the figures for October. Interestingly, there has been a rise in confidence.

If you want to look at the data and the research methodology then Nationwide have created an excellent site with all the information you could require (well done Nationwide).

Here are my concerns with the index.

It is not clear from the research methodology how the results are collected. My bet is it via the Web. To what extent does this bias the sample?

The research sample is said to be:” nationally representative of all adults in term of age, sex and socio-economic group”. This sample of 1000 people is being used to draw conclusions over 13 UK regions and four UK age groups. I am no statistician but I wonder what degree of accuracy can be attributed to the numbers

I wish there the data was reported by gender. How much more/less confident are men than women by age?

My major concern with the index is the conclusions that are drawn between quarterly measures. I suspect that consumers become gloom immune the longer time goes on. A fall of the FT index or the announcement of job losses when the background news is neutral will have a much larger impact than when there is a background of doom laden pronouncements. If there is then the slightest ray of hope it will have a disproportional impact on consumer attitudes.

I wonder about the sensitivity of the index to the news stories at the time of measurement. Confidence is an instinctive thing. If I have just listened to 2 days of unremitting bad news I will respond differently to when the sun has been shining and the news has been ‘neutral’.

So what am I saying? Beware reading too much into these index numbers.

Original Post: