Opinion poll Politics

By eriasa mukiibi sserunjigi

• How accurate is Museveni’s 66% poll score?

• IPC survey says Besigye has 60%. Does he?

The heated debate over the credibility of the recent pre-2011 election survey by Afrobarometer that gave the incumbent, President Yoweri Museveni, a commanding lead of 66 percent has obscured assessment of its value.

The survey, based on responses of the 2000 people it sampled across the country, found that Museveni’s nearest challenger Kizza Besigye would have got only 12 percent of the vote if the election had been held in November, barely three months to the February 18, 2011 general elections. The remaining six candidates would have garnered a paltry 8 percent of the vote.

Observers say claims and counterclaims over the Afrobarometer survey show that outcome of the 2011 will not be accepted by the loser.

In fact, the Afrobarometer survey shows that up to 73 percent of those surveyed said the losers will refuse to accept the official election results.

Besigye’s team, which maintains its support is somewhere near 60 percent, have refused to blame the pollsters and its funder, Deepening Democracy Programme, who they say are victims of manipulation by the ruling party.

“What Ugandans don’t know is that the questionnaire, the research design and areas where the poll was to be conducted were first submitted to the National Council for Science and Technology and the President’s Office,” said FDC’s Head of National Mobilisation, Maj. Gen. Mugisha Muntu, who spoke on Dec. 17 on behalf of the Inter Party Cooperation (IPC) that is backing Besigye.

He said this enabled the Gombolola Internal Security Officers (GISO) and other intelligence operatives to intimidate the people into answering the poll questions in favour of the ruling party.

But Robert Sentamu, the Executive Director of Centre for Democratic Governance and Wilsken Agencies Ltd which conducted the poll, told The Independent that the requirement to submit their questionnaire and research design to the National Council of Science and Technology exists but that he did not comply with it.

He said some of the questions in the survey might not have pleased the President or his agents.

Muntu also pointed out that most of the respondents, 63 percent, said “one always or often has to be careful about what they say about politics in this country”.  Muntu said they answered the way they did out of fear.

Besigye’s team intends to announce its full provisional results 24 hours after closure of voting despite the Electoral Commission’s claim that it holds exclusive mandate of announcing results. According to the law, the Electoral Commission is the only body that can announce final results, but parties and other bodies can relay provisional results. But this could create tension.

According to the Afrobarometer survey, 59 percent believe there will be post-election violence and 70 percent say the security forces will step in to prevent large-scale violence.

This is the second survey that Afrobarometer has released on the 2011 election. In May 2009 it released findings of a survey conducted in September 2008 that showed that if the election had been held at the time, Museveni’s NRM would have got only 41 percent of the vote. Besigye’s FDC would have got 21 percent, DP 4 percent, and UPC 4 percent.

Political opinion polls can be inaccurate – sometimes disastrously. However, the Afrobarometer survey shows that since then Museveni has been able to gain 25 percentage points while Besigye has lost 13 points. What could be the cause of this?

Part of the answer could be that in the 2009 survey 13 percent of respondents said they would not vote, 8 percent refused to answer, and 7 percent said they “don’t know”. That is a total of 28 percent of the vote that was still up for grabs.

These “undecided voters” are important if the Afrobarometer surveys are treated as “brushfire polls” designed to give an indication of how voters feel at a particular moment in time. They can be won over if the candidate or party uses the survey results to re-adjust their message. This means that Museveni and the opposition should have used the survey released in May 2009 to check themselves. Did they?

Wrong message?

Where the opposition’s message centres on the issues people seem to be preoccupied with, the people seem to believe that the current government is doing well on them.

hen respondents in the Afrobarometer survey were asked what they consider to be the most important election issues, they led with health and education (28%), fighting corruption (21%) and maintaining order (20%). Improving social services and fighting corruption are key cornerstones of opposition’s message, but not so much maintaining order in the country.

The other issues the opposition has campaigned on were rated far below in importance by the respondents. The economy got 14 percent, infrastructure, 11 percent, and equitable access to resources and opportunities, 6 percent.

How come the respondents do not feel so strongly on the issues the opposition thinks are important?

When Afrobarometer asked them how they rated the government handling of issues, the respondents rated it highly in education (67%), health (57%) and the economy (54%). On the flipside, the government was rated very poorly on its handling of corruption (23%) and job creation (24%).

But Muntu insists the survey indicates that their campaign is spot-on, and “we will not change our message”.

Money in campaigns

Money and vote buying is playing an important part in this election. More voters than ever are being offered money and the majority (52 percent) say it is wrong but “understandable” if voters take it.

“The vast majority of the offers came from the NRM, with FDC, DP and UPC jointly making up the very small remaining share,” the report says.

Although over 70 percent of respondents said they would take the money but vote for their preferred candidate or party anyway, money no doubt is a major influence in favour of the NRM.

Money is being offered most in the poorer regions of eastern and northern Uganda.

A close look at the survey results also shows that Museveni’s campaign could have, in fact, exploited weaknesses in the opposition’s message, campaign strategy, and preparedness.

The Afrobarometer survey reveals two significant points regarding the campaigning strategies.

First, under the topic, `Voter intentions: What voters want and intend to do (early indications)’, the survey has a section on `voting intentions’. It found that 95 percent who say they are affiliated to the NRM said they intend to vote for Museveni. However, among those who said they are affiliated to the opposition, only 57 percent said they would vote Besigye. DP’s Norbert Mao and UPC’s Otunnu got 14 percent each, Betty Kamya 3 percent and Bidandi Ssali 1 percent.

The opposition’s failure to unite behind and sell one candidate could be hurting them.

But could the poll be wrong?

Not a chance, says Prof. Carolyn Logan, who was involved in the polling. She says it matters a lot how the respondents are selected and that they followed scientific methods with a margin of error of plus or minus 2.5.

The poll was based on 2000 respondents that represented the spectrum of the Uganda population based on the 2002 Uganda Bureau of Statistics Census figures and 2010 population projections of Uganda. Respondents were proportionately selected based on regions, religions, sexes and whether they live in urban/rural areas. For example, 86 percent of respondents were rural based, 88 percent were Christians, and 80 percent below 45 years of age.

According to the survey, although Museveni has never won in the north since 1996, the poll shows that if elections had taken place between Nov. 18 and Dec. 6, he would have won 59 percent of the northern vote, followed by Kizza Besigye and Olara Otunnu, both at 11 percent.

“Our people know that we in the opposition ended the Kony war (so) how could they decide to reward (President) Museveni for the same?” wondered Aswa County MP Reagan Okumu, who is one of the leading opposition politicians in northern Uganda.

In Kampala, another area where Museveni has lost to Besigye in the last two elections, Museveni would win 44 percent against Besigye’s 15 percent.

In fact, Okumu has written to the Electoral Commission complaining that FDC agents in Acholi are suffering intimidation and arrest from pro-NRM security agencies.

Okumu doubts the Afrobarometer research methods.

“I think there was a problem with their sampling method,” he said, although he admitted he had not yet read the full report and how the study was conducted.

Generally, however, criticism of the Afrobarometer survey by the opposition has been partly moderated because it was funded by the Deepening Democracy Programme in Uganda, which is a major sponsor of the opposition parties.

But political polls have been known to be inaccurate sometimes.  Eight months before the 2006 elections, Afrobarometer did a survey in which it asked the question: If a presidential election were held tomorrow, which party would you vote for? Up to 62 percent of the respondents said they would vote NRM, 6 percent FDC, 6 percent UPC, and 4 percent DP.

When the election was held, NRM’s Museveni got 59.3 percent of the vote,   FDC’s Besigye got 37.4 percent, and the rest got about 3 percent between them. True, it predicted that Museveni would win, but how can one explain Besigye’s significant gain?

To correctly appreciate the survey findings, however, it must be noted that it was conducted just a few weeks after the presidential candidates were nominated on October 25-26. In the preceding period, the opposition candidates remained invisible to the electorate as President Museveni, who has been in power for 24 years, hogged all the visibility and freely campaigned.

Museveni’s challengers have about four months, two of them already gone, to sway voters. Have they made any impact or is Museveni widening the gap?

That is why the late December distillation of the Besigye campaign message to focus on five key issues from the 13-point Manifesto is important.

First of the top five is job creation, where the government has performed badly. But keeping education, healthcare, and infrastructure, and the new one – agriculture, is dicier. Perception on whether the opposition presents a credible alternative to the current government is evenly spread but the nays have a slight edge, 45 percent to 44 percent.

Elsewhere, the idea of using small samples to predict the outcome of an election was popularised by George Gallup in the 1936 US election. Relying on a method called quota sampling; Gallup surveyed some 3,000 people as against the 10 million surveyed by the Literary Digest magazine.

But whereas The Digest predicted Republican Alf Landon would win with 57 percent of the vote to Roosevelt’s 43 percent, Gallup forecast a win for Roosevelt with 54 percent of the vote. Roosevelt was re-elected with 61 percent of the vote.

Gallup’s trick is that even though he had polled a few people, he had taken care of the differences in the American population and selected his respondents in proportion to the groups they represented.

Dr. Yasin Olum of Makerere University Political Science department says probably 66 percent of the respondents in the Afrobarometer survey were NRM supporters. Olum says even in a situation where the selection of respondents is scientifically done, it is possible to come up with a sample that has a disproportionate number of like-minded respondents that could end up skewing the results.

He says it is important to carry out more random studies of this nature to be able to establish the truth.

Gen. Muntu told The Independent that the FDC has done their own surveys but they are for “internal consumption”. The IPC estimates its support at about 60 percent of the vote.

“We urge the NRM also to carry out theirs because we don’t want them to be shocked and behave like Gbagbo (defeated president of Ivory Coast who has refused to hand over power).”

But Olum says in order to understand the actual situation on ground, it is important for parties and candidates to consider what has happened between the last presidential election and the coming one. “The opposition may not have done enough to challenge government, but what has government done to gain so much ground?” he asked.

Although attention has been on the prediction of the 2011 election, the majority of the respondents approved of government’s handling of education (67 percent), basic health services (57 percent) and managing the economy (54 percent). However, only 24 and 23 percent felt government is performing well concerning job creation and fighting corruption respectively.

Improving public services such as education and health, fighting corruption and maintaining order in the country, in descending order, are the issues that voters consider most important.

Strangely, the people of northern Uganda, who have only recently come out of war, are the least preoccupied with the need to maintain order in the country. Improving public services such as education and health and fighting corruption are more urgent issues for them.

The people of central Uganda are the most preoccupied with maintaining order in the country, with the east and west tying on this consideration.

The poll doesn’t explain though, why the people of Kampala are this time round more likely to vote Museveni. They seem to be more interested in issues like job creation and fighting corruption, where Museveni’s government is shown to be doing badly.

Some of the findings of the poll make it hard to understand why the opposition has significantly lost ground. For example, the poll shows that people’s rating of the opposition’s programmes has improved. Approval for the opposition’s alternative programmes is high, at 44 percent as against 45 percent who disapprove of them. Why is it the case that only a few of these respondents are ready to vote for the opposition’s “viable” programmes?

The results of the poll couldn’t have been shocking only to Besigye’s team. Even NRM, which was busy thinking up ways to invigorate its campaign when the poll results came out, could have been both pleasantly surprised and disturbed. The Independent is aware that NRM’s internal projections indicate that Museveni wouldn’t win outright if voting took place today. The state-run daily conducted a survey last August that placed Museveni at 52 percent. The parties have limited time to reconcile the poll results with the objective reality on the ground. Every second matters.