Public discussion and the liberal shift

Martin Roiser examines different ways of estimating public opinion and asks how views change after group discussions.
Early one October morning at the time of the war in Afghanistan, I was listening to the radio. The newscaster wondered why opinion polls revealed a public in favour of British intervention in Afghanistan while ‘all those who phone this programme are against’. This woke me with a start, wondering what would happen next. More listeners phoned in. They spoke of self-selected groups and biased samples, but offered no explanation for the direction of the discrepancy. A few days later, in a televised debate, two well-known commentators spoke in favour of British participation in the war, and two against. A lively debate ensued in front of a studio audience of 100 people, who subsequently voted. The audience was selected by National Opinion Polls to be a representative sample of the British electorate. At the end of the debate the vote for continued British participation in the war was carried by a narrow 51 per cent to 49 per cent (Channel 4: War on Trial, 27 October 2001). Though opinion measures were shifting during this period, the studio audience was remarkable for its close result, given that a poll carried out by MORI on 9 October found that 71 per cent thought Britain was right to have joined the military strikes, 20 per cent thought it was wrong and 8 per cent did not know (MORI, 2001). Was the method of sounding opinion affecting the outcome?

BPS Members can discuss this article

Already a member? Or Create an account

Not a member? Find out about becoming a member or subscriber