Pay rises act 'as an addictive drug'

Huw Richards reports from the annual conference of the Royal Economic Society in Swansea on take-overs, pay and satisfaction.

Research has challenged the economic theory of a direct relationship between income and personal satisfaction.

Andrew Clark, of the Organisation for Economic Cooperation and Development who has worked on the issue with Andrew Oswald of Warwick University, told a session of the RES conference at Swansea that the key factor might be how much income workers have received previously.

His attention was first drawn to the issue by the findings of surveys showing that most people, when offered a choice between a pay progression with a steep advance from a relatively low starting point and a flatter profile with a higher starting point, opted for the steeper progression. In many cases this was even after being told that it was less remunerative in total.

"We don't like people behaving like that," he said, when the standard assumption that the rational way to behave is to maximise income in the early stages and then invest any surplus. But such behaviour ceases to be irrational when considered in terms of personal satisfaction and most people's preference for a rising income.

He has tested his proposition on a recent panel survey of 2,000 British employees, comparing their responses on job satisfaction to their 1992 wage and the pay received the previous year. He found no direct relationship between pay and job satisfaction, but a strong direct link between satisfaction and the pay rise received over the last 12 months.

This suggests that wages operate rather like an addictive drug. "As with a drug, the more you're used to getting, the more you need to get a real high," said Dr Clark. And, as with a drug, increases are rapidly absorbed and treated as normal.

This may be one reason, Dr Clark supposed, for the historic tendency for wages to outstrip increases in productivity and also the general reluctance to accept pay cuts, however high wages are to start with.