Zimbabwe provides an interesting case study of media coverage of elections because this coverage has been so extensively monitored in recent years. The Media Monitoring Project Zimbabwe (MMPZ), a non-governmental organization, was established in 1999 and monitored a series of controversial polls, starting with a referendum on constitutional reform in 2000.
Although several subsequent elections were marked by heavily unbalanced coverage in the government-controlled media, the 2000 referendum provides a particularly clear example because international standards on allocation of time in referenda are so clear. Each proposition – acceptance or rejection of a new government-sponsored draft constitution – should have received equal direct access air time. News coverage in the government-controlled media should also have reported the positions of each campaign roughly equally.
In 2000, broadcasting in Zimbabwe remained a state monopoly. Both radio and television were run by the Zimbabwe Broadcasting Corporation, formally an independent public corporation. The main daily newspaper, the Herald, was owned by a company, Zimpapers, that was ostensibly controlled by a public trust. In reality, it was well documented that editors were hired and fired by the Ministry of Information. The monopoly enjoyed by the Herald and its sister paper, the Chronicle, over the daily newspaper market had been breached in 1999 by the launch of the privately-owned Daily News. This rapidly acquired a mass readership. There were a number of other private quality weekly newspapers with smaller, predominantly urban readerships.
Quantitative analysis of coverage by ZBC and Zimpapers – media outlets that directly or indirectly used public funds – provided telling evidence of their overwhelming bias in favour of the Yes proposition, acceptance of the draft Constitution. For example, television current affairs programmes devoted 16.12 hours of coverage to the Yes campaign and its arguments, against just 1.33 hours for No (and 1.28 hours of general information on the issue). There were 17 opinion or editorial articles in the Herald favouring the Yes vote and not a single one favouring No. Of the 38 opinion articles across the Zimpapers titles, all favoured a Yes vote.
MMPZ’s methodology also places considerable stress on the importance of the sources of information used by news media. Here is the breakdown (typical of the government-controlled media) of sources in stories on the constitution in the run-up to the referendum. It refers to two of the ZBC radio stations:
- Ruling party and government: 53%
- Constitutional commissioners: 18%
- Other Yes voices: 18%
- Newsreader: 6%
- Members of the public: 4%
- Opposition political parties: 1%
These are examples of purely quantitative methods, which seem to tell the story of media imbalance quite clearly. Yet even quantitative analysis requires further explanation. Take, for example, this striking statistic. In the month before the referendum, television ran 139 advertisements for the Yes campaign and just 14 for the No campaign. Clinching evidence of bias – or is it? The figures only have any meaning if the reasons for the imbalance are known. Here are some possible explanations:
- The No campaign may have decided not to place many TV advertisements.
- The No campaign may have had little money to spend on advertising.
- ZBC may have applied different advertising charges for the two campaigns.
- ZBC may have refused to accept advertisements for the No campaign.
In reality, the first of these explanations was not true. The second was true but irrelevant – it was not in fact the reason why so few No advertisements were broadcast. The third of these factors may have been true but was not relevant. (It is unclear that the Constitutional Commission, campaigning for a Yes vote actually paid for its advertising at all.) The reason was the fourth factor: a refusal of ZBC to run No material. The National Constitutional Assembly, the main No campaigner, obtained a High Court order requiring ZBC to run its advertisements, but the corporation still refused to comply. Indeed, its news programmes failed even to report the court’s decision. ZBC stated that the No material was “unbalanced” and of low technical quality. (Yet there was no requirement of balance – these were campaigning advertisements.) Later ZBC issued a statement saying that it was legally barred from showing “pornographic material”. The implication was that the NCA’s material was pornographic, although no evidence was ever presented for this claim.
The methodological point is this: each quantitative finding only makes sense if it is given context and explained. In other words, media analysis is required, not just media monitoring.
Other failings of ZBC’s coverage cannot be depicted in quantitative terms at all. For example, the broadcasters and Zimpapers repeatedly reported the case of a young man in Harare allegedly beaten to death by supporters of the No campaign. This was adduced as evidence of the ill intentions of those campaigning against the draft Constitution. Yet it was clearly documented (in a police statement) that the dead man had died in a traffic accident.
Voter education was another area where the government-controlled media fell short of proper standards. One of the most elementary shortcomings was that it almost entirely failed to explain what the outcome of the referendum would be. It was assumed that the vote would be binding – that if the Yes campaign succeeded the new Constitution would automatically become law. Yet this was not the case. The effect of a positive vote would simply have been that a Constitution of Zimbabwe Bill would have been placed before Parliament for a vote.
However, there was an even greater failure in the “voter education” material prepared by the Constitutional Commission. In an animated slot purportedly telling voters how to complete their ballot, the box next to the word Yes was shown being filled with a tick. This was a flagrant breach of the principles that voter education should be impartial. But it may have been that the No campaign had the last laugh: voters were actually required to place a cross in the box. A tick would have constituted a spoilt ballot.
However, there was a certain irony in all this. When votes were cast on 12-13 February 2000, Zimbabwean voters rejected the draft constitution by a large margin. MMPZ, in its report on the referendum, remarked on this irony and speculated on the effect of the media coverage. Perhaps voters ignored biased coverage. Perhaps they were repelled by it and voted against the draft for that reason. Perhaps biased coverage reduced the size of the No campaign’s victory. Or perhaps media coverage was irrelevant to voters’ decision-making. MMPZ accepted that its monitoring methodology gave it no basis for reaching any of these conclusions. The question would simply have to remain open. So MMPZ ended with a more modest conclusion. The draft Constitution on which the public voted had scarcely been distributed: “So, to the extent that people voted on the draft and not according to some other factor, they will have done so on the basis of information in the media. And that information was lamentably inaccurate and biased.” [1]
[1] Media Monitoring Project Zimbabwe, A Question of Balance: The Zimbabwean Media and the Constitutional Referendum, Harare, 2000.
