The Institute of Fiscal Studies dropped an analytical bombshell into the midst of the few remaining economic optimists with its recent report.[i] It estimated that median household incomes will fall by 7% between 2009-12 – the worst situation since the string of crises during the mid-1970s. It then goes on to discuss the resulting huge numbers of people who will be facing poverty.
So far, so clear – and the BBC[ii] helpfully explains, “the technical term of “absolute” poverty [is] defined as being below 60% of the median income, adjusted for inflation.”
But wait a moment! The Daily Mail also provides a definition, claiming the IFS as its source, as it notes that, “An individual is in….. absolute poverty if a household’s real-terms income is below 60% of the 2010/11 average.”[iii] This sentence even comes directly below a reproduced table bearing the caption, ‘Poverty line is 60% of median before – housing –costs’ income’.
In a very short space of time – just 24 hours – we have gone from median to average to median. And one other newspaper’s article on the IFS report that I have failed to track down after an initial viewing used “median average” at least twice…That piece of writing is journalism’s equivalent to the each way bet. (If anyone can find this article for me I will be enduringly grateful. No prizes but a fuzzy warm feeling inside will be yours.)
So what is really meant – what is the difference between median and average, and why does it matter? Actually, one of the biggest applications of all this discussion is indeed to wages, big and small. So now your attention is no longer drifting, let’s first consider the maths terms involved in very basic form:
Mean = add up all values and divide total by number of values. This is what is meant by ‘average’
Median = line up all values in order from smallest to largest and pick the one in the middle . (If you have an even number of values in line, work out the mean for the two values either side of the middle. That figure is your median.)
Here is a fact to contemplate: on average, a member of the human race possesses one ovary, one fallopian tube and one testicle. Already you might be suspecting that ‘average’ or ‘mean’ may be less helpful than might first appear……
The mean is not in the middle of the range with almost everything else grouped closely around it and just a few values at either extreme, as in a classic bell-shaped curve. In the skewed curve, the majority of values are significantly below the mean and far fewer are above it. The distribution is a bump to the left in the graph and a tailing off line towards the right. The mean therefore does not reflect where most values actually are. The average, so beloved of politicians, trade union leaders, employers’ organisations and journalists, is not where most people actually find themselves but gives a far higher value.
In most Western economies this skewed graph with most values bunched well to the left side has direct and hugely significant applications to salaries and house prices (both sales and rents). Such graphs are important to businesses because salaries can easily be 60% – 80% of the costs of a company or commercial sector. House prices are one of the wealth measures which drive consumer confidence and willingness to spend, boosting money circulation in regional and national economies. Rentals may be a significant factor for restricting savings for house deposits and commitment to mortgages similarly can limit construction, DIY and allied supplier industries.
Let’s take a quick look then at skewed graphs, putting in some contemporary values. Although what follows applies to the UK, any reader living elsewhere can insert their own country’s figures and the story will be the same. In the UK national average salary is £26 000 per year. Yet 50% of full time salaries in the UK are below £20 000 per year and around two thirds are less than £26 000. So the mean actually reflects the dividing line between the wealthier 1/3 and the poorer 2/3. It is not the divider between two equal segments of the full time salaried population.
For a skewed graph, if the mean can prove so misleading, what may help? The median value for a skewed graph falls some distance to the left of the mean (ie a lower value) so it represents a downwards correction towards where the graph ‘bulges’. It is not ideal but it is a great improvement for realistic understanding and discussion about the salaries situation.[iv]
Back to our definition of absolute poverty – ‘less than 60% of median income’ means people are worse off than ‘less than 60% of mean income’ by some significant degree. I finish with the words of Barnardo’s CEO, Anne Marie Carrie: “…This isn’t just about statistics as every day thousands of families are being forced into making choices between heating or eating.”[v] Absolutely, distressingly true – but we won’t see much change for the better if we can’t even get our heads around the concept of ‘median’, let alone act on what it is telling us.
[iv] Part of this discussion is adapted from Section 5 of my upcoming book on critical thinking for business, “LIFTING THE BLINDFOLD”