Opinion / Columnist
Zimbabwe Election 2018 Predictive Model-It's not even close!
25 Jul 2018 at 13:58hrs | Views
In less than a week Zimbabweans vote in a watershed election. Not surprisingly the impending elections evoke a myriad of emotions for different reasons. Some are hyper-ventilating; some are quacking in their shoes; and many are hyper-vigilant in defending anything perceived as unkind to their preferred candidates. Much is at stake. But who is going to win this election? This piece presents and discusses a few factors to consider, and then hazards a bold forecast.
Voter Turnout
If one distills the data from the elections since 1980, some key winning factors emerge. The first of these is voter turnout. Zimbabwe experienced a decrease in voter turnout from a high of 94% in 1980 to a low of 32% in 1996. After the formation of the MDC the turnout ranged between 26% in the February 2000 referendum and 59% in 2013. Actual numbers can be helpful in revealing an important secret here. In the referendum, the turnout was 26%. The NO vote which was sponsored by players that are now associated with the MDC won with 54.31% to 45.69%. If this had been a presidential election the MDC would have won in the first round. In the parliamentary elections in 2000, the voter turnout was 48%. ZANU PF had a small majority with 48.6% of the vote to MDC's 47%. To date this is the closest election outcome. If it had been a presidential election, a run off would have been called.
In 2008 the voter turnout was 42.4%. The MDC candidate Morgan Tsvangirai won with 47.9% of the vote to ZANU PF's Mugabe (43.2%). This necessitated a run off. In the 2005 parliamentary elections the turnout was 47.7% and ZANU PF won with 59.6% to MDC's (39.5%). In the 2002 presidential elections the turnout was 55.4%. ZANU PF won 56% to 42%. In 2013 the turn out was 59% and ZANU PF prevailed (61.09% to33.94%). Reading figures can be a dreary and boring exercise, yet it is often very informative. From these figures a few salient trends can be noted. First; the MDC HAS NOT WON an election in which the voter turnout was more than 50%. Second; when the turnout is below 50% ZANU PF has tended to struggle. Why is this important? To date all polls indicate that the voter turnout will be well above 50%. This is also supported by the sheer multitudes attending rallies of both major parties. However, given the trend since the formation of the MDC the turnout is not likely to exceed 65%. Decipher what you may; but there is a trend here which cannot be easily wished away.
PARTY COHESIVENESS
For simplicity party cohesiveness as used here is the extent to which a party is pulling in the same direction. To better assess party cohesiveness another term called situational proximity will be introduced. This answers the question-how close to the election did the internal contradictions occur. As already indicated ZANU PF failed to garner 50% of the vote on three occasions. The loss in the 2000 referendum threw ZANU PF into a tailspin. In the election that followed the direction in which the party was headed was openly questioned in some provinces. In all the Matebeleland provinces the base was literally burnt. Even the once mighty oaks and untouchables like Dumiso Dabengwa were reduced to mundane shrinking shrubs. Some have never recovered. In Masvingo the Zvogbo-Mavhaire combination morphed into scheming grandmasters. Manicaland was not spared either. This continued right into the election. In 2008 the Simba Makoni factor and the Bhora Musango rebellion prevailed right into the election. How did this affect the party? To answer this question, the actual numbers will be consulted. The numbers for ZANU PF were as follows: 2000 (1 211 284); 2002(1 685 212); 2005(1 569 867); 2008(1 079 730) and 2013(2 110 434). The average is 1 531 300. The elections that ZANU PF lost were characterized by high dis-cohesion that continued right into the election period, and the numbers were well below the average. In 2013 the ZANU PF numbers were above the average by about 580 000; a whopping 38% above average. This is an outlier that raises eyebrows. What can possibly explain this? The MDC has alleged vote rigging without providing proof. However, there is one phenomenon that was widely reported. After the 2008 run off imbroglio the ZANU PF campaign team did not demobilize. Reports from that day allege that the war veterans led by Jabulani Sibanda were camped in the villages and campaigning daily. Many reports, which the army has denied, have alleged that throughout this period the army was actively deploying both personnel and resources on behalf of ZANU PF both overtly and covertly. Why is this important? In the current election period, that has not been the case. ZANU PF's challenge is to repeat or exceed the 2013 figures. That would seem to be a tall order.
How about the dis-cohesion caused by the Tsholotsho declaration in 2005. The 2005 election was a parliamentary election in which candidates contested at a local level. Hence the impact was local rather than national. The rebellion was quickly subdued. The main actors were either expelled or contained. Those that chose to run as independents and had the resources such as Jonathan Moyo did well. This demonstrates that ZANU PF is not immune from independent candidates who are well resourced. Kereke and Samkange in 2013 are cases to remember. The "enfant terribles" Agrippa Bopela Masiyakurima in Mutare North, and Mliswa in Norton can cause major localized upsets with no bearing on the national scale in this election cycle. Since the last election ZANU PF has suffered two major dis-cohesion incidents-the Mujuru purges and Operation Restore Legacy. Given this backdrop can ZANU PF repeat the 2013 performance?
The MDC figures for the same period were as follows: 2000(1 171 096); 2002(1 258 401); 2005(1 041 292); 2008(1 195 562) and 2013(1 172 349). The average was 1 167 740. The MDC figures are very consistent. Why is this important? A few salient observations can be made. First, Tsvangirai seemingly had plateaued and was unable to grow the party beyond 1.2 Million votes. This indicates a need for change at the top. Second, this consistency in voter numbers runs contrary to the MDC theory of migrating X in the 2013 ballot. If Xs had indeed migrated the 2013 number would be expected to be an outlier lower than the average. Third the core MDC base stands at about 1.2 Million votes while the ZANU PF base is 1.5 Million. This supports the observation made earlier that when the bases are properly canvassed, and the voter turnout is high ZANU PF tends to win. Fourth ZANU PF's challenge is to replicate the 2013 numbers, while the MDC challenge is to catch up. To achieve this MDC needs to add at least 300 000 more votes if ZANU PF has its average performance, but 1.1 Million if ZANU PF replicates the 2013 performance. This represents a growth of 18% on the lower end, and 100% on the high end. The question is not whether this is in the realm of possibility but whether it is probable. To address this one needs to consult the BVR statistics province by province
Voter Registration Statistics
Zimbabwe has 10 electoral provinces. The BVR registration figures are as follows: Bulawayo (258 690), Harare (900 300), Manicaland (733 293), Mash central (531 864), Mash east (633 126), Mash west (644 974) Masvingo (617 204), Mat North (338 851) Mat South (264 160), Midlands (761 474). Previous elections produced a well-defined pattern. There are three zones. The Red Zone consisting of provinces that went to the MDC is in all elections. These are Harare, Bulawayo and Mat South. The Green-Orange zone consisting of provinces that have voted ZANU PF in all elections. These are The three Mash provinces and Midlands. Finally, there are three swing provinces that have changed hands between the two parties, which are Manicaland, Midlands and Mat South to a lesser extent. No party has won the election without winning both Midlands and Manicaland. Although Mat South has changed hands between MDC-T, MDC-N and ZANU PF the lower voter numbers in that province minimizes its impact. One statistic stands out in these numbers. The three Matebeleland provinces combined have less registered voters than Harare on its own. This demonstrates that in terms of numbers alone Lovemore Moyo's argument for a regional party is viable. However, viability and desirability are two different issues. But this is a digression. What is clearly discernible is that the MDC has one voter rich province, and two with much less numbers. ZANU PF has four voter rich provinces on its side.
In the absence of an extraneous shock, the best predictor of the future is the relevant past. In this case the relevant past are the voter trends from previous elections. Taking an average of all the elections since the MDC formation produces the following pattern for ZANU PF vrs MDC province by province. Bulawayo: 20%:80%. Harare: 35:65; Manicaland: 52:48; Mash Cent: 75:25; Mash East: 75:25; Mash west:70:30; Masvingo:60:40; Midlands: 65:35; Mat South: 48:52 and Mat North: 35:65. Given these historical trends, the voter statistics and party dis-cohesion, is it probable that the MDC will increase its support by more than 300 000 in case ZANU PF support remains constant; and by 1 million in case ZANU PF repeats the 2013 performance? Given how engaged the voters from both parties are, is it probable that ZANU PF support will plummet below 1.3 million, which would make the MDC task much easier? Either way, it is a tall order.
Opinion Polls
The next factor to explore is opinion polls. Opinion polls when properly done can be an indication of the strengths of different parties. A wide spectrum of polls has been published; ranging from very dubious to methodologically robust polls. A poll of polls indicates that in all polls to date ZANU PF is leading, howbeit with small margins in other polls.
One poll has attracted a lot of attention because of its robust methodology. This is the Afrobarometer poll. The first poll gave ZANU PF 42% and MDC 31% with a 2% margin of error. What this means is that the ZANU PF support could be anywhere in the range 40-44% compared to 29-33% for MDC. The second poll gave ZANU PF 40% and MDC 37% with also a 2% error margin. This means the ZANU PF range is 38-42% compared to 35-40% for MDC. This suggests that the race is tightening. If the extremes are considered ZANU PF's lowest (38%) is actually lower than MDC'S highest (40%), which means that according to this poll MDC can actually be leading.
The ZANU PF lead has narrowed from 11% to 3% in two months or so. If the trend continues MDC will surge ahead by election day. This has caused a lot of excitement in the media. Although the data and the methodology are unimpeachable, there are two details that need attention. First, at the time of the first survey the MDC leadership issue had barely been resolved. What is being termed a surge is actually a coalescing and consolidation of MDC base behind Chamisa. These are not new voters attracted to the MDC, but most likely MDC supporters who were previously ambivalent because of the leadership wrangles who have now thrown their lot with Chamisa. Why is this the likely explanation? Consider two details. First, statistically ZANU PF's share of the vote has not changed (range-40-44% vrs 38-42%). Second, the percentage of "undecided voters" is constant at about 20%. Likewise, the votes for the other parties is also statistically constant at around 3%.
What about the 20% "undeclared intention"? This is not likely to be undecided. This is what is here referred to as "scripted responding". Consider this. If for 18 years it is drummed into people's minds that ‘YOUR VOTE IS SECRET" and then in the 18th year you inquire into their voting intentions, what do you expect? A good number of respondents will give an ambivalent answer. This should be construed as undecided. It simply means that the voter education has been effective, and respondents are giving what they believe to be the smart answer. Going back to the latest Afrobarometer data, disregarding the 20% undeclared and then recalculating the percentages without the undeclared reveals the usual pattern that was discussed under Voter Registration statistics above. If you calculate the data this way, and then factor in the voter registration statics above, you will suddenly realize that the MDC is likely to get a maximum of 1.7 million votes while ZANU PF will get a minimum of 1.9 million votes.
When all factors are considered, this race is not even close. If all the factors discussed in this piece are taken together ZANU PF is likely to win with anything between 53% and 60% of the total vote cast.
Voter Turnout
If one distills the data from the elections since 1980, some key winning factors emerge. The first of these is voter turnout. Zimbabwe experienced a decrease in voter turnout from a high of 94% in 1980 to a low of 32% in 1996. After the formation of the MDC the turnout ranged between 26% in the February 2000 referendum and 59% in 2013. Actual numbers can be helpful in revealing an important secret here. In the referendum, the turnout was 26%. The NO vote which was sponsored by players that are now associated with the MDC won with 54.31% to 45.69%. If this had been a presidential election the MDC would have won in the first round. In the parliamentary elections in 2000, the voter turnout was 48%. ZANU PF had a small majority with 48.6% of the vote to MDC's 47%. To date this is the closest election outcome. If it had been a presidential election, a run off would have been called.
In 2008 the voter turnout was 42.4%. The MDC candidate Morgan Tsvangirai won with 47.9% of the vote to ZANU PF's Mugabe (43.2%). This necessitated a run off. In the 2005 parliamentary elections the turnout was 47.7% and ZANU PF won with 59.6% to MDC's (39.5%). In the 2002 presidential elections the turnout was 55.4%. ZANU PF won 56% to 42%. In 2013 the turn out was 59% and ZANU PF prevailed (61.09% to33.94%). Reading figures can be a dreary and boring exercise, yet it is often very informative. From these figures a few salient trends can be noted. First; the MDC HAS NOT WON an election in which the voter turnout was more than 50%. Second; when the turnout is below 50% ZANU PF has tended to struggle. Why is this important? To date all polls indicate that the voter turnout will be well above 50%. This is also supported by the sheer multitudes attending rallies of both major parties. However, given the trend since the formation of the MDC the turnout is not likely to exceed 65%. Decipher what you may; but there is a trend here which cannot be easily wished away.
PARTY COHESIVENESS
For simplicity party cohesiveness as used here is the extent to which a party is pulling in the same direction. To better assess party cohesiveness another term called situational proximity will be introduced. This answers the question-how close to the election did the internal contradictions occur. As already indicated ZANU PF failed to garner 50% of the vote on three occasions. The loss in the 2000 referendum threw ZANU PF into a tailspin. In the election that followed the direction in which the party was headed was openly questioned in some provinces. In all the Matebeleland provinces the base was literally burnt. Even the once mighty oaks and untouchables like Dumiso Dabengwa were reduced to mundane shrinking shrubs. Some have never recovered. In Masvingo the Zvogbo-Mavhaire combination morphed into scheming grandmasters. Manicaland was not spared either. This continued right into the election. In 2008 the Simba Makoni factor and the Bhora Musango rebellion prevailed right into the election. How did this affect the party? To answer this question, the actual numbers will be consulted. The numbers for ZANU PF were as follows: 2000 (1 211 284); 2002(1 685 212); 2005(1 569 867); 2008(1 079 730) and 2013(2 110 434). The average is 1 531 300. The elections that ZANU PF lost were characterized by high dis-cohesion that continued right into the election period, and the numbers were well below the average. In 2013 the ZANU PF numbers were above the average by about 580 000; a whopping 38% above average. This is an outlier that raises eyebrows. What can possibly explain this? The MDC has alleged vote rigging without providing proof. However, there is one phenomenon that was widely reported. After the 2008 run off imbroglio the ZANU PF campaign team did not demobilize. Reports from that day allege that the war veterans led by Jabulani Sibanda were camped in the villages and campaigning daily. Many reports, which the army has denied, have alleged that throughout this period the army was actively deploying both personnel and resources on behalf of ZANU PF both overtly and covertly. Why is this important? In the current election period, that has not been the case. ZANU PF's challenge is to repeat or exceed the 2013 figures. That would seem to be a tall order.
How about the dis-cohesion caused by the Tsholotsho declaration in 2005. The 2005 election was a parliamentary election in which candidates contested at a local level. Hence the impact was local rather than national. The rebellion was quickly subdued. The main actors were either expelled or contained. Those that chose to run as independents and had the resources such as Jonathan Moyo did well. This demonstrates that ZANU PF is not immune from independent candidates who are well resourced. Kereke and Samkange in 2013 are cases to remember. The "enfant terribles" Agrippa Bopela Masiyakurima in Mutare North, and Mliswa in Norton can cause major localized upsets with no bearing on the national scale in this election cycle. Since the last election ZANU PF has suffered two major dis-cohesion incidents-the Mujuru purges and Operation Restore Legacy. Given this backdrop can ZANU PF repeat the 2013 performance?
The MDC figures for the same period were as follows: 2000(1 171 096); 2002(1 258 401); 2005(1 041 292); 2008(1 195 562) and 2013(1 172 349). The average was 1 167 740. The MDC figures are very consistent. Why is this important? A few salient observations can be made. First, Tsvangirai seemingly had plateaued and was unable to grow the party beyond 1.2 Million votes. This indicates a need for change at the top. Second, this consistency in voter numbers runs contrary to the MDC theory of migrating X in the 2013 ballot. If Xs had indeed migrated the 2013 number would be expected to be an outlier lower than the average. Third the core MDC base stands at about 1.2 Million votes while the ZANU PF base is 1.5 Million. This supports the observation made earlier that when the bases are properly canvassed, and the voter turnout is high ZANU PF tends to win. Fourth ZANU PF's challenge is to replicate the 2013 numbers, while the MDC challenge is to catch up. To achieve this MDC needs to add at least 300 000 more votes if ZANU PF has its average performance, but 1.1 Million if ZANU PF replicates the 2013 performance. This represents a growth of 18% on the lower end, and 100% on the high end. The question is not whether this is in the realm of possibility but whether it is probable. To address this one needs to consult the BVR statistics province by province
Voter Registration Statistics
Zimbabwe has 10 electoral provinces. The BVR registration figures are as follows: Bulawayo (258 690), Harare (900 300), Manicaland (733 293), Mash central (531 864), Mash east (633 126), Mash west (644 974) Masvingo (617 204), Mat North (338 851) Mat South (264 160), Midlands (761 474). Previous elections produced a well-defined pattern. There are three zones. The Red Zone consisting of provinces that went to the MDC is in all elections. These are Harare, Bulawayo and Mat South. The Green-Orange zone consisting of provinces that have voted ZANU PF in all elections. These are The three Mash provinces and Midlands. Finally, there are three swing provinces that have changed hands between the two parties, which are Manicaland, Midlands and Mat South to a lesser extent. No party has won the election without winning both Midlands and Manicaland. Although Mat South has changed hands between MDC-T, MDC-N and ZANU PF the lower voter numbers in that province minimizes its impact. One statistic stands out in these numbers. The three Matebeleland provinces combined have less registered voters than Harare on its own. This demonstrates that in terms of numbers alone Lovemore Moyo's argument for a regional party is viable. However, viability and desirability are two different issues. But this is a digression. What is clearly discernible is that the MDC has one voter rich province, and two with much less numbers. ZANU PF has four voter rich provinces on its side.
In the absence of an extraneous shock, the best predictor of the future is the relevant past. In this case the relevant past are the voter trends from previous elections. Taking an average of all the elections since the MDC formation produces the following pattern for ZANU PF vrs MDC province by province. Bulawayo: 20%:80%. Harare: 35:65; Manicaland: 52:48; Mash Cent: 75:25; Mash East: 75:25; Mash west:70:30; Masvingo:60:40; Midlands: 65:35; Mat South: 48:52 and Mat North: 35:65. Given these historical trends, the voter statistics and party dis-cohesion, is it probable that the MDC will increase its support by more than 300 000 in case ZANU PF support remains constant; and by 1 million in case ZANU PF repeats the 2013 performance? Given how engaged the voters from both parties are, is it probable that ZANU PF support will plummet below 1.3 million, which would make the MDC task much easier? Either way, it is a tall order.
Opinion Polls
The next factor to explore is opinion polls. Opinion polls when properly done can be an indication of the strengths of different parties. A wide spectrum of polls has been published; ranging from very dubious to methodologically robust polls. A poll of polls indicates that in all polls to date ZANU PF is leading, howbeit with small margins in other polls.
One poll has attracted a lot of attention because of its robust methodology. This is the Afrobarometer poll. The first poll gave ZANU PF 42% and MDC 31% with a 2% margin of error. What this means is that the ZANU PF support could be anywhere in the range 40-44% compared to 29-33% for MDC. The second poll gave ZANU PF 40% and MDC 37% with also a 2% error margin. This means the ZANU PF range is 38-42% compared to 35-40% for MDC. This suggests that the race is tightening. If the extremes are considered ZANU PF's lowest (38%) is actually lower than MDC'S highest (40%), which means that according to this poll MDC can actually be leading.
The ZANU PF lead has narrowed from 11% to 3% in two months or so. If the trend continues MDC will surge ahead by election day. This has caused a lot of excitement in the media. Although the data and the methodology are unimpeachable, there are two details that need attention. First, at the time of the first survey the MDC leadership issue had barely been resolved. What is being termed a surge is actually a coalescing and consolidation of MDC base behind Chamisa. These are not new voters attracted to the MDC, but most likely MDC supporters who were previously ambivalent because of the leadership wrangles who have now thrown their lot with Chamisa. Why is this the likely explanation? Consider two details. First, statistically ZANU PF's share of the vote has not changed (range-40-44% vrs 38-42%). Second, the percentage of "undecided voters" is constant at about 20%. Likewise, the votes for the other parties is also statistically constant at around 3%.
What about the 20% "undeclared intention"? This is not likely to be undecided. This is what is here referred to as "scripted responding". Consider this. If for 18 years it is drummed into people's minds that ‘YOUR VOTE IS SECRET" and then in the 18th year you inquire into their voting intentions, what do you expect? A good number of respondents will give an ambivalent answer. This should be construed as undecided. It simply means that the voter education has been effective, and respondents are giving what they believe to be the smart answer. Going back to the latest Afrobarometer data, disregarding the 20% undeclared and then recalculating the percentages without the undeclared reveals the usual pattern that was discussed under Voter Registration statistics above. If you calculate the data this way, and then factor in the voter registration statics above, you will suddenly realize that the MDC is likely to get a maximum of 1.7 million votes while ZANU PF will get a minimum of 1.9 million votes.
When all factors are considered, this race is not even close. If all the factors discussed in this piece are taken together ZANU PF is likely to win with anything between 53% and 60% of the total vote cast.
Source - George Chingarande
All articles and letters published on Bulawayo24 have been independently written by members of Bulawayo24's community. The views of users published on Bulawayo24 are therefore their own and do not necessarily represent the views of Bulawayo24. Bulawayo24 editors also reserve the right to edit or delete any and all comments received.
