Geographers met for their annual conference in Surrey this week, The THES reports.
God has lost to Mammon in academics' priorities for questions in the 2001 population census.
Philip Rees of Leeds University's school of geography has conducted a survey on what data researchers want from the forthcoming census. The information will go into the Census Office's development strategy.
Professor Rees presented his research, for the Economic and Social Research Council and the funding councils' joint information systems committee, to the Royal Geographical Society and Institute of British Geographers annual conference held at Surrey University, Guildford, this week.
He found that 94 per cent of respondents wanted a question on income and saw this as essential or desirable research.
"This question hasn't been asked in a British census before, and academics are very, very keen on it," said Professor Rees.
"It's essential for many forms of social science analysis, whether you're taking a welfare or marketing approach, looking at poverty problems and exclusion or wealth and patterns of consumption. It provides much sharper differentiation of people, since occupation is a bit too blurry, and housing questions aren't really discriminating."
But although a broad coalition has lobbied strongly for a question on what religious group people belong to, there has been much less enthusiasm among academics. Professor Rees said: "Only 43 per cent in the interim analysis saw this question as essential.
"The question was asked in the 19th century and in Northern Ireland, but always on a voluntary basis. A large number of people don't want to answer, which makes it very difficult to interpret."
Some researchers feared that a question on religion would prove contentious, endangering the way people answered other questions.
"You could say that's also true of an income question, but it has been very carefully broadbanded, so that people don't have to reveal too much," Professor Rees said.
There is also academic support to count students at their term-time address and not the parental home. But most researchers felt there should also be tables giving the parental address, so that figures can be compared with the 1991 census. It is not yet clear whether this will be feasible.
The impact of students' spending patterns and housing demands is felt most acutely at their term-time address, said Professor Rees. The data could help funding proposals for student housing.
Professor Rees said that researchers really wanted continuity and comparability with earlier censuses. "The census is often caricatured as being for 'geographers', with the observation that other social scientists could make do with a large household survey. However, this stereotype needs extending to include 'historians' as well. Census users need comparable social information on people over space and time."