(Fwd) Zimbabwe voting stats (pre-count)

Patrick Bond pbond at wn.apc.org
Tue Mar 12 10:31:28 MST 2002

>From ZWNEWS, 12 March

Voting analysis
An analysis of voting data from the presidential election indicates a
definitive voting victory for the MDC. Based on the government's INITIAL
turnout figures, together with polling percentages from the June 2000
parliamentary election and recent opinion polling data, the electoral
arithmetic shows that the way voters are likely to have decided in this
watershed ballot leaves Zanu PF and their presidential candidate in trouble.

The initial overall turnout for the whole country - by the end of voting on
Sunday evening as announced by ZBC - was around 2 980 000, or 53 percent of
the registered electorate. Monday's voting in Chitungwiza and Harare -
disrupted and cut short by administrative obstructions - is not included in
this total. By province, the turnout varied widely, according to the
official release, from 46 percent in Bulawayo and Matabeleland North, to 69
percent in Mashonaland Central. Overall, the turnout was almost eight
percentage points higher than in June 2000. Again, the figures vary - from
under one percentage lower in Bulawayo compared with the parliamentary
elections, to around 14 percentage points higher in Manicaland and Mavingo,
and 17 percentage points higher in Mashonaland Central. The Harare turnout
by Sunday evening, at 47 per cent of those registered, was around 2
percentage points higher than in June 2000. These initial turnout figures
have been subsequently revised - considerably - by the government. But on
the INITIAL figures, the calculations show that an announced Mugabe victory
will be totally at odds with the likely course of voting on the ground.

In order to calculate how these turnout figures translate into support for
the two main presidential candidates, we have used two sources of data.
Overall actual support for the two candidates, by common agreement amongst
political observers, would have largely been decided in the swing provinces
of Masvingo and the Midlands, and also, crucially, how voters have changed
their preferences in the three Mashonaland provinces since June 2000. For
these - overwhelmingly rural - constituencies, we have taken the relative
support for the two main candidates by those rural voters who were willing
to reveal their intentions to researchers from the Mass Public Opinion
Institute at the University of Zimbabwe. For the other areas, where the
swing in the vote from June 2000 is likely to have been much lower, we have
taken the relative shares of that actual vote in 2000. The arithmetic yields
a total of just short of 1.7 million votes for Morgan Tsvangirai, compared
with just over 1.2 million for Robert Mugabe - a victory of over 58 per cent
as against 41 percent for the MDC.

As with everything else connected with this election, however, the data
comes together with a number of health warnings. The February opinion poll
data recorded a huge percentage of voters who were unwilling to say who they
supported. The figures for Harare's extra day of voting are not included in
the calculations, and they could be crucial at the margin. And, most
importantly of course there is, no allowance made for the distinct
possibility of fraud between vote and count. The initial turnout figures
upon which we have based our calculations have subsequently been revised by
ZBC, further inflating the rural versus the urban vote. The numbers for
Harare later showed a fall in votes cast versus June 2000. The initial
figures showed an increase. And those for votes cast in Mashonaland Central
were later raised enormously.

But of the vote itself, there is little doubt. Those who declined to reveal
their voting preferences in the February opinion poll are more likely to
have been MDC than Zanu PF supporters. And although we have assumed that the
shares of the vote in the non-swing provinces stayed static compared with
June 2000, it may have in fact increased in favour of the MDC since then.
Masvingo is a particular case in point. The devil, as always, is in the
detail. Why the large changes in the reports of voting totals on Monday?
Those monitoring the count today will have to have their wits - and
calculators - about them.

                                        Registered         Votes cast by
Turnout     Turnout         MDC vote

                                        Voters 2002         Sunday night
% 2002     % 2000         share 2000

Bulawayo                             368 028                 169 000
45.9         46.7                 83.6

Harare                                 882 176                  415 000
47.0         44.7                 75.9

Manicaland                          658 694                  363 000
55.1         41.4                 47.2

Mash Central                       480 092                  331 000
69.0         51.7                 19.8

Mash East                           589 185                   328 000
55.7         47.2                 24.0

Mash West                          572 677                   293 000
51.2         43.1                 32.8

Masvingo                            655 122                   380 000
58.0         43.8                 37.0

Mat. North                          338 186                   157 000
46.4         43.7                  73.6

Mat. South                         343 993                   165 000
48.0         46.4                 59.3

Midlands                            724 659                    379 000
52.3         48.4                  33.2

Total                               5 612 814                 2 980 000
53.1         45.5                 46.7

PLEASE clip all extraneous text before replying to a message.

More information about the Marxism mailing list