The Obstacle To UPND’s Path To Victory – The Electoral Demographic Map

By Muhabi Lungu

In recent weeks the Movement for Multiparty Democracy (MMD) has come under vicious attack, particularly from the United Party for National Development (UPND), claiming that the MMD is composed of “mad people”, and that anyone who goes back to the MMD or stays has a few screws missing in their head. According to them, the MMD can never win an election again and all of us must migrate to the UPND and “live happily ever after.” This is so natural and appropriate for UPND because supposedly, fate has favorably looked upon them as they are the ones who will definitely take over government in 2016.

Screen Shot 2014-08-08 at 08.35.03 Screen Shot 2014-08-08 at 08.35.15The “mad people” statement was issued at the highest levels of the UPND. Of course, the veracity of the accusation is not in itself offensive, but its implications are quite clear to me and ought to be disturbing to everyone who believes in democratic norms. The implicit assumption here is that people must always belong to the winning team, regardless of what that winning team believes in, stands for, or the methods it uses.
For example, in answer to my press conference on 5th August 2014, UPND Deputy Spokesperson Mr. Cornelius Mweetwa proudly proclaims that my problem is that I am “always joining dying parties”. This is yet another clear indication of how the UNPD views the MMD. In my view, such a philosophy which is devoid of right and wrong, or a moral justification of existence is quite provocative. A philosophy of life or competition which is only predicated on winning by any means necessary or belonging to a winning team regardless of their belief system is of grave concern to me. I for one belong to an organization because I believe in its values, principles and methods, even if it means that adhering to those values means sometimes losing. That is me and I cannot compel others to believe as I believe or do as I do.

It appears to me that the commonly received wisdom is that belonging to the winning side is the smartest and most clever thing to do. This is the misconceived political philosophy practiced by those who join politics for personal gain. They have convinced themselves and others that every one of us in politics is in it for the “eating” and the pursuit of power. In accordance with this philosophy, political opportunism must be applauded, encouraged and even rewarded. In fact, such behavior is so totally acceptable that it justifies doing any unchristian or immoral thing under the earth.

Assuming that the above is the commonly held view, let us then put the assertion made by the three main political parties to the test. The assertion is that PF, MMD or the UPND will definitely win the next presidential polls. We will evaluate the chances of the PF, MMD and the UPND on this test question. I will write three letters to my country men and women, articles assessing the chances of each one of the political parties winning the 2016 elections. I will begin with the UPND since they are clearly the most certain of being victorious. Granted that I am human, fallible, biased and could be totally wrong with my critical assessment. I accept that I may turn out to be wrong.

But also consider, just for a moment, that I could be right. After all, right and wrong does not really matter anymore since it is all just about winning. The UPND’s assertion is that they have a clear path to victory in 2016. This is because the PF have become so unpopular and have lamentably failed to govern. On the hand, the UPND’s belief is that the electorate will never give the MMD another chance to govern; hence the MMD must just go away and commit suicide. In fact, they are even prepared to help us carry out this most desirable task.

The primary reason for their conviction appears to emanate from the good results obtained by the UPND in Parliamentary and Local Government by-elections held in the last three years. In fact, they have stated that the soon to be held Mangango by-election (which in their estimation will be won by them) is a good indication of their popularity. First and foremost, let me just indicate that by-elections are a poor indicator of popularity on a national level. By-election results may be useful for democracy, but are not a reliable barometer for General Election performance because the evidence whichever way is inconclusive and often contrary to expectations.

For example, the MMD won the Mpulungu by-election in October 2010 but subsequently lost it to PF less than a year later in 2011. The UPND won the Chilanga, Mufumbwe and Solwezi Central by-elections but subsequently lost all three to MMD in 2011. A By-Election win (or loss) proves absolutely nothing, and it would be folly for any party to derive a false sense of confidence from the results. Therefore, Mangango or any other by-election result has absolutely no predictive value for 2016. (A separate in-depth article will be unveiled on the predictive value of by elections).

Let us now get to the substantive case and assess the probability of a victory by the UPND in 2016. This is going to be done from a mathematical point of view and using the current Voter Register, past performance and consequent projections can be made on the basis of trending data. Please pay attention to the charts at the end which have kindly been provided by

The current 2011 voter’s roll is composed of 9 electoral provinces with a total number of voters registered standing at 5.2 million people (the exact figure being 5,167,154). The largest two electoral blocks, Copperbelt and Lusaka have a combined total of 1.6 million, accounting for 31% of the total registered voters. Compare this to the 31% of the combined total of the four electorally smallest blocks, these being (from smallest to largest), North-Western (6.0%), Western (7.6%), Luapula (7.9%) and Central at (9.3%) percent. One can do very well in these four voting blocks but still lose the national election if one does not perform well in the two urban blocks.

The largest five voting blocks (3,565,872 voters) account for 69% of the total voters registered. These are the following in order of reducing size: Copperbelt (16%), Lusaka (15%), Northern (13%), Eastern (13%) and Southern (12%). No presidential candidate can win an election without out-rightly winning at least two of these five electoral blocks plus coming second in another. Rupiah Banda, who narrowly won the 2008 election, is the only exception. He won convincingly in one province but came an impressive second in the remaining four. An analysis of the numbers reveals that no presidential candidate can win an election in Zambia unless they can obtain at least thirty-five (35%) percent of the vote as a combined total in this key demographic block.

Fredrick Chiluba won more than 50% in all the five blocks in both his elections (except Eastern Province in 1991). Levy Mwanawasa performed well in Northern, Copperbelt and Eastern in 2006, getting 36% of the combined vote in the five provinces. Rupiah Banda performed well in Eastern, Northern, Copperbelt and Lusaka in 2008, getting 38% in the five provinces. Finally, current President Michael Sata won in Copperbelt, Lusaka and Northern provinces, getting 42% of the vote in the five major provinces.

After the fragmented 2001 elections, all 3 subsequent elections have produced a consolidation of the votes between MMD, PF and UPND. Other parties have produced less than 5% of the total vote. The statistical results from 2006 to 2011 show that to win the presidential elections, a candidate must get at least 40% of the national vote. This closely mirrors the combined 35% in the five largest voting blocks.

Before I delve in to detail on the UPND obstacle, let me emphasize the above point. President Michael Sata failed to win the presidential elections in 2001, 2006 and 2008 because he was unable to perform well enough in the five provinces. In addition, he failed to reach 40% of the national vote until the 2011 election. He eventually got into office by winning three of the five provinces with large margins and coming second in another province.
The founding president of the UPND Anderson Mazoka failed to win the fragmented presidential vote in 2001 because although he won Southern Province and did relatively well in Lusaka, he lamentably underperformed in Copperbelt (12%), Eastern (4%) and Northern Provinces (5%). The failure to win by Anderson Mazoka was regardless of the fact that he performed very well in Western, North-Western and relatively well in Central province. Those who blamed his loss on electoral rigging simply have no empirical evidence. However, the arithmetic can indicate the path of loss quite convincingly.

This is the same reason that the current President of the UPND, Hakainde Hichilema has been unable to win in a presidential poll thus far. He has consistently underperformed in four of the five big provinces. For example, his total tally in the 2011 presidential elections was as follows, in order of provincial size: Northern (0.8%), Eastern (3.3%), Copperbelt (3.6%), Lusaka (11.3%) and Southern (71.4%). The average percentage of the UPND in this key demographic block is 18%, a significant distance from the minimum 35% that is required for you to be in contention in a presidential win. In addition, the national minimum total of 40% is even further.

The required percentage change for the UPND for them to reach the minimum threshold is a massive task, especially if one considers the upward gravitational pull that will be required on the Copperbelt, Northern and Eastern provinces. Taking into consideration current trending data, it is highly unlikely that the UPND will make any significant gains on the Copperbelt and Northern provinces. Even if the UPND were able to double its performance in Lusaka to say 22%, in the absence of significant improvements in the Copperbelt and Northern Provinces, reaching the minimum 35% average threshold in the five provinces will be a tall order.

The capacity for the UPND to double and reach 22% in Lusaka will have to be weighted and compared to its past trend performance in the last four presidential elections in this province: 2001 (31%), 2006 (21%), 2008 (15%) and 2011 (11%). This is a declining trend in Lusaka for the UPND in four consecutive presidential elections, whereas MMD has increased from 16% to 31% in the same period. Even if we were to give UPND the benefit of doubt that they have the capacity to reverse this trend in Lusaka for the 2016 elections, the upward trajectory of Mr. Hichilema would have to significantly surpass all his past performances. It will be a tall order for the UPND to reverse this trend and double their 2011 performance from 11% to 22% in Lusaka.

On the national level, for the UPND to jump from a national performance level of 18% in 2011 to even reach the MMD’s 36% would be a miraculous performance. This would entail that Mr. Hichilema would surpass the 27% national total obtained by Mr. Anderson Mazoka in 2001, the closest that the UPND came to wining a presidential election (Mr. Mazoka was much more popular nationwide than Mr. Hichilema). This leap, in a single five-year cycle would also have to be weighted and compared against the step by step increments the PF were able to achieve from their 29% in 2006, 38% in 2008 and finally 42% in 2011. This was an upward swing of 13% over five years, from a starting base of almost 30% and three elections.

A dramatic change from 18% to the winning number of 40% for the UPND in one five-year election cycle with only one campaign season would be incredibly impressive, considering the vote consolidation factor among the three major parties. This unprecedented 22% upward swing would be nearly three-quarters greater than Mr. Sata’s 13% gain. The magnitude of the swing becomes more unbelievable when you consider that the current presidential election trajectory for the UPND over four elections has been downwards: i.e. 27% in 2001, 25% in 2006, 20% in 2008 and 18% in 2011.

I have never said that the UPND can never win. However, it is extremely unlikely for the UPND to produce this 22% upward swing within five years. Please refer to the following charts, courtesy of

Share this post