The company set up to measure audiences for BBC and independent radio has pledged to continue with its new system of audience measurement - even though it has apparently mislaid about one million listeners each from Radios 1, 2, 3 and 4.
The slump occurred when Radio Joint Audience Research (Rajar) changed its methodology toward the end of last year.
Instead of asking 50,000 representative listeners to write by hand the names of the stations they listened to in the previous week in "listening diaries", it introduced lists of stickers pre-printed with station names.
The move was aimed at simplifying the diaries following the increase in radio stations. But it had a different effect. Unpublished results from Quarter Four - 18 September to 17 December, when the methodology was first used - are understood to show Radios 1, 2, 3 and 4 to have lost one million weekly listeners each.
This would amount to an 8 per cent drop from the previous quarter for Radio 1; 12 per cent for Radio 2, 38 per cent for Radio 3, and 11 per cent for Radio 4. Curiously, Radio 5 was barely affected and retains almost all its 4.8 million audience.
Drastic falls also occurred in the commercial sector. Virgin, Classic FM and Talk Radio are said to have "lost" about 700,000 listeners each - a 19 per cent drop for Virgin, 18 per cent for Classic FM and 30 per cent for Talk Radio. Rajar admits the slump is due to the new system. A statement said: "A key problem has been isolated as a failure by respondents to stick in sufficient labels."
The problem is that the new system apparently fails to acknowledge listeners who tune in only briefly to stations or listen involuntarily. But it has infuriated the radio industry, which relies on audience figures to pull in advertising. Radio 1 even threatened to pull out of Rajar. But, despite pressure to revert to the old system, Rajar refuses to change.Reuse content