Whether or not forecasters

No TV studio on election night will be complete without its academic. Huw Richards reports on a select breed

Whatever the proclaimed disillusionment with party politics, general elections remain massive collective events. More than three-quarters of the adult population voted in the general election of 1992. It would be a shock if this year's turnout were much smaller. And a fair proportion of those who have voted will be tuned to television or radio after the polls close at 10pm. This is a time of tension-ridden enforced idleness for candidates.

But for one group, the election broadcasters, it is the most demanding period of the entire election. Among them will be a small group of academics, the voice of independent, informed authority reaching into the deluge of information and making sense of it.

Five weeks to go

What do the broadcasters expect of their pundits? "Availability" is the first item on political consultant and communications academic Ivor Gaber's list. It is not just a matter of turning up on the night, but also of coming to rehearsals and providing ideas and advice in the months and weeks before.

As a professor of politics, Ivor Crewe could find the time for rehearsals. As vice chancellor of Essex University, he cannot. So after inheriting David Butler's BBC TV role from 1983 to 1992, he will be confined this time to World Service Television and Breakfast Time.

Equally important is the ability to express oneself clearly and economically under the pressure of events. Dr Butler says: "A lot of election night comments are only worth making for a minute or so after you think of them."

Mike Thrasher, of the University of Plymouth, notes the importance of remembering your audience: "You don't talk down, but it is no good talking as if your audience were other political scientists. When I use a word like 'swing' I do my best to ensure that, every so often, I explain what it means."

Straightforward stamina is important, and it helps if, like Dr Thrasher, you obviously enjoy the job. "I get the most tremendous kick out of it, a great adrenalin buzz. Last election I started at 9pm and was in the studio until 5.30am, with only one break. One worry is that your concentration might lapse some time in the early hours."

Dr Butler, drawing on his vast experience, points out: "It is not so much one programme as a succession of slightly different ones, with the size and nature of the audience changing at different times during the night and the pace of events varying. You have to be able to adjust to that."

Being prepared is an important part of the battle. Colin Rallings of Plymouth University says: "I always feel that as long as I have done my homework properly, there should be nothing I can't cope with." Analysts arrive loaded with reference material - Professor Crewe is armed for his appearances with a laptop computer and two lists of constituencies: "One is a list of key seats, which I tick off using a colour-coding system as results come in, and the other is an alphabetical list."

The analysts have also often been involved in longer-term planning, and in the preparation of the exit poll, a key element in election-night coverage. Strathclyde University's John Curtice has been advising the BBC on the exit poll and looking at ways of getting it closer to the final outcome than it did in 1992, when it predicted that there would be no majority. "One of the problems we had was that the projected result changed in the last 40 minutes of polling day. We still had Labour ahead at 9.15pm and the data came through too late. That's one problem we have to solve."

The exit poll is most important in the first couple of hours, before the results start to flood through. It is, as Professor Crewe notes, not invariably reliable - the difficulties in 1992 follow an understating of the Conservative majority in 1987 and a spectacular misfire in October 1974, when the exit polls projected a Labour majority of more than 100. It finished up at three.

He and Professor Gaber both point to the tension which rises when the exit polls and early results are in conflict. Professor Gaber says: "The exit poll covers 100 seats, but is based on what people say they have voted. You may only have two or three results - but those are real votes. Which to believe is a terribly difficult decision".

The pressure is on to predict the final outcome as rapidly as possible. Professor Crewe warns against "being over-confident early on". Aside from possible exit poll-early result tensions, there are other constraints: "Politicians and producers frequently do not understand, or do not want to understand, the law of probability. They will be looking for an exact prediction when all you can offer is a range of possibilities between, say a Labour majority of 40 and a Conservative majority of ten."

Dr Butler argues: "Experts are no more likely to make accurate predictions than anyone else, they just make inaccurate ones for more sophisticated reasons." He is similarly wary of prediction: "You don't make predictions, but conditional statements of how the election would turn out if the remaining seats behave the same way as those that have already declared".

Film exists of a youthful Dr Butler hailing the uniformity of swing across the country at a 1950s election as a symptom of underlying national unity. How far that still applies is a matter for debate - while Labour's strong performance in target seats was a factor in the failure of the 1992 exit poll, Dr Butler argues that "uniformity is still far more important than the differences. The overwhelming majority of seats are within three points of the national swing."

With experience of election night punditry in Australia and India, he points to the British peculiarity of only declaring results by constituency rather than polling district: "In a British election you have 659 largish bits of information which take a time to arrive. In Australia you have 12,000 small bits which start to arrive very rapidly and give you a much quicker sense of what is really happening."

After the unsettling fluctuations of election night 1992, the analysts might not mind too much if the result is clearcut this time. But that too will set challenges. Dr Thrasher notes: "If the result is clear from early on, you have to find a way of keeping up interest - for instance spotting key seats which might indicate, say, whether Labour can make a real breakthrough in the South."

Heavily crunched numbers, elaborate graphics and computer projections will abound on election night. But through it all, as Professor Gaber points out, will come the human element: "The whole business is an art not a science, and you have to remember that."

THE EARLY PUNDITS

Short of in-house expertise in the early days, broadcasters looked to academics - in particular David Butler of Nuffield College, Oxford. The 1950 election, won narrowly by Labour, saw his debut both as senior co-author of the Nuffield Election Studies and as the BBC's on-screen analyst.

"I hadn't even watched television before. R. B. McCallum, who co-authored the 1945 election study, was not very numerate and insisted that I accompany him to deal with the numbers," he says.

There were no computer graphics and projections in those days: "Computers came in in the early 1960s, but there was a manual back-up as they weren't entirely reliable". As results came in, they were mounted on cards, giving Dr Butler, wielding a slide rule, 45 seconds to calculate the swing in that seat and its impact on the national swing.

From 1955 to 1979 he formed a well-remembered partnership with the Canadian Robert MacKenzie, inventor of the "swingometer".

Ivor Gaber, who combines the post of professor of broadcast journalism at Goldsmiths College with working as a political producer and consultant at ITN and Channel Four, notes that Dr Butler defined a distinctive style: "There's a slight touch of eccentricity, a suggestion of the mad professor, that grabs an audience."

This style has its antipodean variation in Keith Jackson of the University of Canterbury, New Zealand, who journalist Tom Scott said "has a moustacheless beard, modern wire-framed spectacles that somehow look horn-rimmed and a high-domed head. The total effect is one of vague but prodigious intelligence. His enthusiasm never seems to flag, though I suspect that if you chained him to a bus shelter for three days without food and water and with only lolly wrappers to read, he would still describe the experience as absolutely fascinating."

TODAY'S PUNDITS

The touch of eccentricity lives on in today's pundits. Professor Gaber points to John Curtice of Strathclyde, previously of Nuffield, as an exponent of the style.

A more conservative approach has come out of Essex in the form of Anthony King and Ivor Crewe, and out of Plymouth in the form of Colin Rallings and Mike Thrasher. "They do the business very professionally and very straight, but there is less star quality about them," says Professor Gaber.

Not that the broadcasters mind this. Professor Gaber notes that a recent trend is to develop in-house experts: "People like Alistair Stewart and Peter Snow know their psephology and have real authority as front-men. The broadcasters want their people rather than the academics to be the authority figures."

The academics' role has changed slightly in response: "They are to provide a change of pace, a bit of a break from the main presenters and to spot the points of real detail. They are an extremely important back-up and safety check."

Of the major broadcasters, only BBC radio o has opted to do without an academic - using instead its in-house political expert, Bill Bush. BBC television will have Professor King as their on-screen academic analyst, with Mr Curtice as backroom back-up. "If you see me on screen, the chances are that something has gone wrong," says Mr Curtice.

ITV will combine Dr Butler with Dr Rallings and in-house expert David Cowling, while Dr Thrasher will be working for Sky. "We're pretty good at preserving Chinese walls," say both Dr Rallings and Dr Thrasher, once dubbed "the Starsky and Hutch of election studies".

Justin Fisher of London Guildhall University was on Live TV's pre-election programme, but on election night, he will be helping LBC radio.