Findings influenced by ‘margin of terror’

By Charles Mangongera

Public opinion surveys can be useful planning tools for political strategists and policy makers. This is especially so in well-established democracies where public policy is informed and instructed by citizens’ voices.

In times of elections, political strategists use opinion polls to craft electoral messages that will strike a chord with the electorate on key public policy issues. Opinion polls are also a useful indicator of the likely outcome of an election.

But opinion polls can also get it wrong. Take for instance the recent election in the UK. Opinion pollsters had predicted a hung parliament, which would have meant another coalition government akin to the Conservative-Liberal Democrats power sharing government of 2010-2015.

This however, did not happen. David Cameron’s Conservative party cruised to a comfortable win and is now governing alone. This has obviously caused some discomfort in the UK market research industry, which is estimated to be worth over £3 billion a year.

The recent Afrobarometer survey on Zimbabwean citizens’ attitudes on democracy and governance, the economy, civil society and other issues conducted by the Mass Public Opinion Institute (MPOI) has ignited fierce debate, not least because of its findings, which many have wrongly interpreted as meaning rising support for President Robert Mugabe and his party Zanu PF and declining support for the opposition MDC-T and its leader Morgan Tsvangirai.

Many commentators and so-called “analysts” have interpreted the findings of the survey at face value and have failed to contextualise them in the broader political economy of the country. As a result, most of the analysis has been flawed and the conclusions reached misleading.

Some have gone as far as questioning the veracity of the survey on the basis of the research methodology used. Because I had the privilege of heading research teams at MPOI on Afrobarometer surveys, I know their methods are thorough and their field interviewers are rigorously trained.

For me there are two key determinant variables I think influenced the outcome of the survey and most of the analysis proffered so far has missed them.

First is the issue of fear among respondents. Because Zimbabwe has a well-documented history of state-sanctioned political violence, most communities, especially in rural areas, live in constant fear and to them anything political is considered as highly risky.

The ruling party has maintained a sophisticated and very efficient intelligence system in which it has informants at the lowest level of Zimbabwean society. Asking people in such communities whether they trust Mugabe or Tsvangirai and expecting them to answer truthfully is illogical. They do not even trust that the person who is interviewing them is just a researcher who has not been sent by the state.

A careful analysis of past MPOI surveys will show a high percentage of “I don’t know/Refused to say” and “my vote is my secret” responses to questions on candidate or party preference in elections.

The late founder and director of MPOI Masipula Sithole used to jokingly say that our surveys had a “margin of terror” and not a margin of error, as is the standard in research surveys.

Let me illustrate my point on fear. The survey asks a series of questions to measure the incidence of lived poverty and its findings paint a gloomy picture.

Of the 2400 randomly selected respondents, 33% of urban residents said they had gone without food at least once, while 56% of their rural counterparts said the same. On medical care 52% of urban respondents said they had gone without, while 59% of rural respondents said the same. Nearly six in 10 urban respondents (59%) said they had gone without water while about four in 10 rural respondents (42%) said the same.

A majority (76%) of urban respondents said they had gone without cooking fuel while 40% of rural dwellers said the same. The most fascinating statistic is that on access to income. An overwhelming majority in both urban (86%) and rural (94%) areas said they had gone without a cash income.

Overall, close to two thirds of the respondents said the country was going in the wrong direction and that unemployment is the biggest problem government should address and that foreign direct investment is a better option compared to indigenisation of the economy in creating jobs.

It is almost a contradiction that the same people would say they trust the man who is presiding over the system that is responsible for these grim conditions. But they will say so if they know saying otherwise will invite the wrath of the party militia that burnt down their homes and confiscated their cows and goats not so long ago.

To buttress my point, over two thirds of the people surveyed said corruption had increased over the years and they rate various government departments such as police, senior government officials, tax authorities, judges and magistrates etc, as highly corrupt. But they also said they will not report incidents of corruption because they fear the possible consequences of doing so.

The second factor that I consider key in influencing the findings of the survey is what I would call the power of propaganda. The media plays an important role in shaping mass public opinion.

The survey finds out that a majority of Zimbabweans relies on radio as a source of news and information. What it does not tell us is that there are half a dozen or radio stations that are state-owned and two private radio stations that are owned by entities and individuals with close ties to the ruling party.

While there has been a notable rise in the uptake and usage of social media, which has given citizens new platforms to generate and share content with minimum state interference, the survey results show that social media remains the least used source of news and information with only 10% of respondents saying they use it to get news every day.

Mobile technology presents opportunities for the free flow of information as a building block for democracy but its efficacy has been hampered by poor connectivity and high cost of usage, especially in rural communities that offer low market value for technology companies. In addition, government has realised the liberating power of social media and senior apparatchiks have publicly grumbled about it.

It is therefore conceivable that citizens are unlikely to trust an opposition leader whose political mission, they are reminded everyday, is “to bring back white rule” and are likely to trust a president whose party, they are reminded everyday, “liberated them from white rule”.

I agree that the opposition has a long list of sins that it needs to attend to if it is to offer formidable opposition to Zanu PF hegemony. However, the results of the Afrobarometer survey are not a measure of those sins. They are a reflection of a deeply fractured society in which citizens fear authority and are manipulated by propaganda.

In the end, what we should be questioning is not the validity of the survey’s findings but the efficacy of opinion polling in an authoritarian context.

Charles Mangongera is a Zimbabwean researcher. He is a former Chief Research Fellow at MPOI and headed research teams on previous Afrobarometer surveys.