New politics, margin of terror, and the 2013 elections

by Tony Reeler

One of the enduring questions raised by the 2013 elections is the magnitude of the win by ZANU PF, and the even more staggering win by Robert Mugabe.  Questions about rigging aside, one relationship that keeps being posed (and answered) is the notion that ZANU PF did so well because the MDCs are so bad. Stephen Chan, a respected Zimbabwe commentator, has raised this recently, pointing out (in his view) that “…Morgan Tsvangirai’s Movement for Democratic Change (MDC) performed appallingly. Outwardly confident, it made the same mistakes it had in previous elections – as if internal reflection, self-criticism and learning from mistakes were impossible”.

A similar view is expressed by the Solidarity Peace Trust (SPT):

All these issues point to a party that has not been able to strengthen its organizational and strategic framework against a repressive regime that has constantly harassed its leadership and structures. However, since 2009 these weaknesses have eroded the support of both MDCs, as was evident from opinion polls carried out in 2012, which showed a drop in support for the MDCs and Tsvangirai and an upsurge in popularity for Mugabe and his party. These weaknesses and, of particular importance to the election campaign, the failure of the two MDC formations to develop an electoral pact in 2013, resulted in the loss of several seats to ZANU PF due to a split vote. For example in Matabeleland South, 8 of the 13 seats were lost to ZANU PF because of the this factor, while in Matabeleland North a united opposition would have won 11 of the 13 seats instead of which ZANU PF won 7 out of the 13. Together these factors meant that the MDCs were a much weaker force in 2013 than they were in 2008.

Now we must also note the arguments about ZANU PF developing a strong social base due to land reform and indigenisation, and also the problems of explaining this social base by reference to Mugabe’s 1 million voter margin over Tsvangirai, and concentrate on another source of “evidence” for the result. This is the evidence deriving from public opinion surveys. Essentially, there is an argument that ZANU PF has been increasing in its popularity with the citizens of Zimbabwe, and it is this popularity that explains the election result.

But before looking at this “evidence”, let us look briefly at the argument that ZANU PF obtained this enormous increase due to a massive increase in its “social base”, and we will ignore the issues of whether this is a social base due to “positive affiliation” (identification and voluntary support) or due to “patronage” (compliance and support based on comparative advantage for supporting ZANU PF).

The basis for this “social base” is generally argued to be three-fold: old Liberation War allegiance (mostly in the rural areas), access to land from land reform, and access to the benefits (or hoped-for benefits) of indigenisation. As regards the first of these, it is indisputable that ZANU PF commands voluntary allegiance from substantial rural supporters, and it was largely this group that gave Robert Mugabe his 43% in March 2008. And, in 2008, most land reform had already been in place for nearly 10 years, and that did not seem to give Robert Mugabe the advantage that is alleged he got in 2013. The big benefit on land seems to be increased tobacco production at the cost of food production.

As for indigenisation, it seems fairly clear that there have been few actual beneficiaries in rural communities, especially in Matabeleland North and South, although there are promises for the future when ZANU PF wins the election. So the benefits of indigenisation on the voters must have been expected in the future and a reason to vote for ZANU PF, which may be a possible explanation. However, it should be pointed out that very little of the US$4 billion supposedly realised from indigenisation seems to have landed in the fiscus, so it can equally be argued that the lack of benefits for the ordinary citizens of Zimbabwe should have counted against a vote for ZANU PF.

However, the assumption is that the “social base” can be inferred mostly from voter turn out, and there seems a missing step in the logic here.

Contrast these two statements:

If ZANU PF has a social base, they will get lots of votes.
ZANU PF got lots of votes.
Therefore, ZANU PF has a social base.

If ZANU PF gets lots of votes, they have a social base.
ZANU PF has a social base.
Therefore, ZANU PF gets lots of votes.

What is obviously missing here is any obvious link between lots of votes and social base. We can empirically verify the votes: they were in the boxes and counted. We can believe that land reform and indigenisation can make people ZANU PF friendly, but there is a missing step in the logic that shows that being ZANU PF friendly comes from land reform and indigenisation independently of the voter turn-out. This is where public opinion surveys and other forms of social research have their application.

Partial support for many has been a 2012 report by Freedom House, which concluded that support for the MDC-T had slipped markedly – from 38% to 20% – and that for ZANU PF had risen by the same amount – from 17% to 31%.  This report (and the SPT report) is being mendaciously used by the ZANU PF media to “explain the victory”. Essentially, the spin is that, in the period from 2010 to 2012, MDC-T’s support was sliding, and, one year on, ZANU PF’s support had increased so dramatically that the result was intelligible to all on the basis of changes in political party support.

However, basing election outcomes on public opinion surveys is always a dicey affair, especially when so many of the people do not give an opinion. In the Freedom House 2012 survey, 47% were unwilling to state their voting intentions, and this needs to be thought about carefully: nearly half would not say. But presumably, and in the light of 2013, these were interviewees who were worried about declaring their affiliation to ZANU PF for fear of victimisation. Reality suggests that this is not the case!

But there is a more serious problem with reliance on the Freedom House report – that it is not methodologically sound, and its conclusions are probably erroneous. A more empirically sound report, “The Margin of Terror”, was compiled by Michael Bratton and Eldred Masunungure of the Afrobarometer, and this came to rather different and more nuanced conclusions than those of  Freedom House. They also provided a number of reasons why the Freedom House report was methodologically unsound.

Firstly, this report concluded, on simple voter preference, that ZANU PF and MDC-T were in a “statistical dead heat”, with the former getting 32% and the latter 31%, and only 22% were unwilling to state their preference. This did represent a decline for the MDC-T from the post 2008 election heyday to early 2009 where expressed preference for MDC-T was 57% to ZANU PF’s 10%. So the difference here between Freedom House and the Afrobarometer is both over the extent of the decline and the final positions that both parties found themselves in 2012.

Secondly, Bratton and Masunungure posed two hypotheses to account for the decline: one was the positive effect of improved government performance, and the second was the negative effect of political fear. They tested both, and the conclusions, on face value, supported both working in favour of ZANU PF.

After careful statistical analysis, it appears that some voters attributed the improved “right” direction, the “good” management of the economy, and “improved educational services” to ZANU PF’s role in the Inclusive Government. This is a paradoxical finding, given that MDC-T was responsible for fiscal control and MDC for education. This would support a view that the MDCs were not marketing themselves as effectively during the GPA as ZANU PF, and even provides support for the SPT thesis about a growing social base. But there was more to come.

On political fear, the study revealed a negative relationship between an interviewee’s expectations of violence and an intended vote for ZANU PF, and, more interestingly, that silencing opponents was more likely to make an interviewee vote against ZANU PF. Most interesting of all, the more likely an interviewee was to perceive the survey as government-sponsored, the more likely they were to express preference for ZANU PF. Furthermore, this last factor had the strongest effect on whether a person would come out openly in support of ZANU PF, and strongly suggests the operation of political fear.

However, Bratton and Masunungure then tease out the effects of people that perceive the survey to be government-sponsored and hence fearful of being honest. Through a slightly complicated analysis, they then conclude that the probable split of support for the various political parties in 2012 was MDC-T (49%), ZANU PF (45%), and all the others put together (7%). This was almost exactly, as they point out the split that obtained in the March 2008 poll, and, as they say, the analysis implies that, if voting intentions do not change, Zimbabwe can expect another close election in 2013.

If this had been the case, then Tsvangirai should have got 1.7 million votes and Mugabe 1.5 million, but actually Mugabe got 700,000 more votes against this supposed trend, and actually got 1.03 million more votes than he got in 2008, a more than 95.46% increase. But interestingly, Tsvangirai got pretty much the same number of votes as he got in 2008.

Well, opinion surveys certainly cannot argue against ballots in the box, all 3.4 million of them. But then Zimbabwe tends to disconfirm all the more general findings of political science, as we have pointed out before: clearly trends only mean things in countries other than Zimbabwe. However, given the support of analyses of the Voters’ Roll (especially in comparison with the census), it can be argued that the survey is likely to be more accurate than election result, and it still remains to explain the result on grounds more closely related to the process of the election itself.

In Zimbabwe, other trends seem to apply. Zimbabwe sees ZANU PF losing ground, regaining it through violent elections, and then being returned with two-thirds majorities in a subsequent “peaceful” election. 2013 marks the end a second such cycle in Zimbabwe: 2000 was fairly violent, 2002 was at least as violent as 2008, and 2005 was peaceful. Then we have the peaceful poll in March 2008, the horrors of June 2008, and the peaceful poll in July 2013. And obviously citizen’s opinions (and their votes) have very little to do with it. No wonder some commentators, like Dr Mandaza, wonder why we bother with elections: all they do is make a mockery of the best that political science can offer!


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s