The timing of direct access slots is clearly of paramount importance. A broadcast when everyone is asleep or at work will be of little use to anyone. As with commercial advertising, everyone will aim for 'prime time'.
All this is obvious, yet it is surprising how often it is overlooked. In the 2000 Zimbabwe referendum campaign, the Yes vote campaign (supported by the government) almost invariably received slots at around the time of the main evening news. The No campaign had to go to court to get its own broadcasts aired - yet the ruling did not specify when these were to be aired, so they received less advantageous times.78
The issue may not only be when a slot is broadcast, but also what is on the other channels. In the 2000 presidential elections, Serbian television tried to reduce audiences for broadcasts by opposition candidate Vojislav Kostunica by scheduling them simultaneously with a popular soap opera.
Yet the issue can be exaggerated. In Chile's 1988 plebiscite, broadcasts were deliberately put out at obscure off-peak times in order to dampen down political enthusiasm. But a population denied any active political debate for 15 years was not to be deterred and watched them eagerly.79
The key point is equality of access to the best slots, whenever these may be. A popular way of achieving total equality is by drawing lots - an approach that is most common when there is also equality in the amount of time allocated.
A mechanism that found favour in the past was the simultaneous broadcast of party election broadcasts on all channels. (In Britain this reputedly created great strain on the national power grid, as everyone took the opportunity of simultaneous political broadcasts to switch on the kettle to make tea.) This approach has something to recommend it, but has been generally abandoned in favour of a philosophy where viewer choice is sovereign. In practice the proliferation of television channels in many countries made it unenforceable.
A second issue is the length of broadcasts. There are two competing trends here. Traditionally the purpose of law and regulations has been to ensure that slots are long enough for parties to get their message across.
But in the age of slick advertising and sound bites it is increasingly felt that the 10-minute election broadcast is a thing of the past. In the United Kingdom, for example, the main parties are allocated five 10-minute slots - but only actually broadcast for five minutes of each of them. If the rules permitted they would no doubt take 10 five-minute slots, but they do not. So the parties prefer to forego half their time allocation in order not to repel the voters by going on at too great length.
In the United States there are moves to ensure a minimum length for political advertisements in order to compel politicians to make appeals to the voters' reason rather than their emotions.
For their regulator there are two alternative approaches. One is to specify precisely the time slot available - say a five-minute broadcast - and then it is up to the party to fill it. If they choose not to then they lose the time not used. The second is to give an overall allocation of time that the party can then use as it chooses. The problem with the second approach is that it makes planning on the part of a public broadcaster almost impossible.
A third approach might represent a compromise between the two. Parties could be given a total allocation of broadcast time in accordance with an agreed system. That time allocation could then be broken into different length time slots, allowing parties a mixture of lengthy and reasoned argument on the one hand and snappy advertising messages on the other.