The work by Bob Elliott and Keith Duffus to which he refers first appeared as Treasury Occasional Paper No. 3.
The calculations and assumptions on "wage drift" are highly questionable.
1980 is a very selective starting point. It was a time of high inflation, and various comparability studies, such as the Clegg reports, were playing havoc with statistics on earnings. Moving the starting point by one year has a significant effect on the calculations.
Starting from 1981 reduces the average real earnings increase for teachers by around 30 per cent. As Bob Elliott himself demonstrates, this reduces the "drift" to a negligible amount.
The second factor is the need to take account of the increase in the number of newly-qualified teachers. In 1984, newly-qualified teachers were at the same level as teachers returning to service after a break. By 1994, newly-qualified teachers outnumbered returners by 5,000.
This change is deflating average salary levels and needs to be assessed when discussing "drift".
All statistical conclusions needed to be tested against experiences. This risk must increase if one set of figures is taken in isolation.
M L WALKER National Employers' Organisation for School Teachers 36 Belgrave Square London SW1