Roger Southall
Right to Reply
It is precisely because of the deep respect which is widely accorded to the Mail & Guardian’s reportage that one simply has to object in the strongest possible terms to the criminally ignorant and politically irresponsible coverage of the Lesotho election by your reporter William Boot. His constant, uncritical repetition of opposition accusations that the election was “rigged” is founded not only upon a dismal appreciation of Lesotho’s politics, but upon a quite alarming failure to understand the mechanics of first-past-the-post electoral systems.
Bluntly, Boot has totally misread the situation. If we rely upon careful, informed analysis rather than a recycling of bar talk and tittle- tattle, it becomes very evident that even if the conduct of the Independent Electoral Commission (IEC) leaves much to be desired, the outcome of the election as presented by the IEC, and as endorsed by 550 election monitors, was a fair reflection of popular opinion in Lesotho on the day of the election.
In stating this, I must make it quite clear: I have absolutely no brief on behalf of the Lesotho Congress for Democracy (LCD). Indeed, I am co-author of a report to the Truth and Reconciliation Commission which details the extent of collaboration between Ntsu Mokhehle and his Lesotho Liberation Army in the 1970s and 1980s with the apartheid regime. So, in short, I am unable to enthuse about the LCD as model democrats.
But despite all my reservations, I have no hesitation at all in defending their election victory in 1998.
Boot and the opposition repeat constantly that the election was “rigged”. Africa is littered with electoral processes deliberately designed to produce victories for ruling parties. But the manners in which election results can be “rigged” in first-past-the-post electoral systems vary greatly. In the case of Lesotho 1998, none of the ways in which the election is said by Boot to have been rigged would have made a serious impact upon the overall result. And if the Langa commission had been more forthright, it would have said so, instead of coming up with its limp conclusion that the election result could not be said to be invalid.
The typical ways of structuring a result under first past the post are as follows:
l Appointing an electoral commission which is unabashedly biased in favour of the government.
We may concede that the LCD took good care to appoint “reliable” people to top positions in the IEC. As the Langa commission has pointed out, the IEC was far from faultless. But IEC personnel were drawn from all political persuasions, and the IEC’s performance before the election was kept relatively clean by lively media criticism and opposition objections.
l Delimitation of constituency boundaries to “load” constituencies; that is, the devaluation of the weight of the opposition by unfairly concentrating opposition voters in single-member constituencies of above- average size.
This method is used extensively by Daniel arap Moi in Kenya, where the number of votes it takes to elect Kenya Africa National Union MPs is on average far less than it takes to elect opposition members. It was also used by the National Party before 1994, which deliberately over-valued rural votes. But this method was not used in Lesotho 1998.
The delimitation which preceded the election aimed for roughly 10 000 voters per constituency. In the event, when voters did register, Maseru constituencies did exceed the 10 000 – because many migrant workers appear to have registered there instead of returning to their home constituencies. But precisely because the 1993 election results demonstrated that political support for the different political supporters was fairly evenly distributed around the country, there was no obvious way to “load” the delimitation of constituencies in favour of the LCD.
l Manipulation of the voters’ list to exclude opposition supporters.
Much has been made by Boot and the opposition about the way in which the election register in Lesotho 1998 records a “scientifically impossible concentration” of birth dates. There is no denying that it does. This does need explanation (and there could well be an above-the-board reason). But even if actual birthdays are wrongly recorded, this does not predetermine how such voters will vote. Statistically, all that matters is that any potential voter fulfils the qualifications to vote, in this case, that voters were over 18 and were Lesotho citizens.
Even if – as Boot has suggested – various under-age voters were registered, their impact would have been limited. Based upon the 1986 population census, we may estimate a potential electorate of 900 000. In the event, only 800 000 were registered. No statistical evidence whatsoever has been marshalled by the opposition to demonstrate that a significant cohort of voters was under age. Anyway, even if we do presume that some under-age voters were registered, Boot must demonstrate that they voted disproportionately for the LCD.
l Refusal by an electoral commission to register opposition candidates, and use of intimidation by the ruling party against its opponents to minimise their vote.
This did not happen in Lesotho 1998. More candidates stood than in any previous contest. The Basotho Congress Party (BCP)put up candidates in all 80 constituencies, the Basotho National Party (BNP) in 79, and so on. There were no reports by the opposition or election monitors that any force was deployed by the government whatsoever.
l Fraudulent counting of votes, stuffing of ballot boxes, statistical manipulation of results and so on.
There is no body of worthwhile evidence that this happened in Lesotho. Votes were counted at constituency level, in the presence of opposition representatives and election monitors. The results were published at constituency level as well as being despatched to the central IEC. This rendered any attempt to manipulate constituency results at central level nigh impossible.
It is possible that individual constituency results could have been fiddled. But to assume systematic fiddling assumes that the IEC, party representatives and electoral monitoring staff in all 80 constituencies were in a massive conspiracy to produce a fraudulent result. Had there been such a conspiracy, someone by now would have squealed. They haven’t. But we have had protests from local IEC staff about the way the opposition has assailed their integrity.
Incidentally, Boot alleged (M&G, August 7 to 13) that a recount conducted by the opposition in half the constituencies indicated that the LCD had won only 12 out of these 40 seats. He has not retracted this allegation, even though when it came to the crunch the opposition did not offer such evidence in its written submissions to the Langa commission.
What the BNP did offer, instead, was a “Review of the 1998 Lesotho General Elections using SPSS/PC+ Statistical Package”, whose conclusions were subsequently dismissed as without foundation by an independent team of electoral and demographic consultants who made a submission to the Langa commission.
Finally, Boot fails to explain the conformity of the 1998 with the 1993 results. There was a 72% turnout of registered voters in 1993, and 74% in 1998; the proportion of votes cast for the BCP in 1993 was 75%, and 71% for the BCP and LCD together in 1998 (calculated together because of the LCD having broken away from the BCP in 1997); the BNP vote in 1993 was 23% and 24% in 1998; the Maremoutlu Freedom Party gained 1,5% in 1993 and 1,3% in 1998.
If some statistical wizard manipulated these results on behalf of the IEC and the LCD, then they were damn clever. In so doing, rather than providing a consistent pattern of votes in every constituency, they would have had to have allowed for credible variations in individual results, such as in Mount Moorosi, where BNP leader Everistus Sekhonyana performed better than most other BNP candidates, and in Mokhotlong district, where the BCP performed better than elsewhere. Frankly, production of such a sophisticated and subtle pattern of results would have been beyond the capacity of an IEC whose functioning has been labelled chaotic by the Langa commission. Anyway, who does Boot think LCD votes were stolen from? The BNP? The BCP? The insignificant MFP?
What Boot completely fails to understand is how first past the post over- represents winning parties. In only one British election this century (1935) has a ruling party gained a majority of the popular vote. The massive majorities obtained by Margaret Thatcher in 1983 and Tony Blair in 1997 were obtained with between 38% and 42% of the vote. The LCD obtained 60%. Given that there was no undue regional concentration of the vote, there is no mystery at all that it should have won 79 out of 80 seats.
What is quite clearly at fault is the choice of the electoral system: Lesotho badly needs a form of proportional representation to ensure an adequate opposition presence in Parliament.
Roger Southall is professor of political studies at Rhodes University
n William Boot replies: First, if the whole process was as transparent as Southall apparently wants to make out, why did the Langa commission refer in its interim report to “gross irregularities” in the way the election was handled? And indeed why were such incisive references absent from the final product?
Second, it is not really good enough to refer to the role of the IEC in the basically indulgent terms that Southall does. With his obvious background in Lesotho’s political life, he will also have some understanding of the role the Tactical Intelligence Bureau (TIB)of the defence force has played in recent times, and he will surely agree that having the TIB handling logistics for the elections is rather like entrusting Vlakplaas with the source documentation for South African elections during the apartheid era. As regards the (frankly feeble) claim that there were always representatives of the broader democracy present when such documentation was handled, I want only to ask where he could have heard such nonsense and why did he believe it?
Responding to the opposition recount issue, I will only note that by the time the Langa commission came on the scene – even with its somewhat forgiving approach to the rigours of documentation – it was itself forced to admit that no real conclusions were possible given the state of the material with which it had to work. To rely only on one set of records as it was forced to do answers fewer questions than it begs, particularly when the commission admits that this documentation was itself flawed.
On the subject of the errant birth dates, I am frankly gobsmacked by the explanation offered (if I understand it) that clerks made mistakes, but so systematically that a clear and sustained statistical pattern emerged throughout the voters’ roll. Hello! And, by the way, the South African auditors did not refute or reject the OF&A findings: the report notes that they were not given access to sufficient documenation by the IEC to make any findings one way or the other. I leave you to ponder the implications at your leisure.
To answer the elaborate arguments about Lesotho demographics would, I think, land us in the kind of scholasticism in which Southall appears to delight and which, despite the kind of statistical gymnastics which give us the LCD share of the vote in 1998 out of the old BCP’s share in 1993, does not go to the point of the matter at all. The real question here is: is there evidence of election tampering?
The answer, I believe, is yes, and by the way, that was also the opinion of a number of both Southern African Development Community and Commonwealth observers with whom I discussed the elections in Maseru in May. However, when it came to certification of the elections, the brief of both observer groups was interpreted as being limited to monitoring whether or not there had been visible crookery, intimidation or violence on voting day itself. This rather limited interpretation of the brief was confirmed at press conferences held by both observer groups in the aftermath of the elections.
A final point, which speaks to Southall and the Langa commission itself: the brief given to the commission was not to discover, as both appear to think, whether or not the LCD probably won (a judgment on which I have no comment), but whether or not there was evidence of irregularities and/or substance in the opposition’s claims, and to manage the situation from there.
Oh, and I do understand the principle of first-past-the-post balloting, but it has nothing whatever to do with the issue, however pertinent the point might be in relation to the projected future of Lesotho political systems.