Let us observe that the promised 72 seats are exactly the two thirds majority of seats in the Federal Parliament allocated to Serbia. The SPS election campaign headquarters relied, justifiably so, on the assumption that the coalition partner of SPS in Montenegro, Democratic Party of Socialists (DPS) would easily win at least 20 out of 30 seats allocated to Montenegro in the Federal Parliament. Therefore they predicted that DPS would be able to win at least the two thirds majority of the seats allocated to Montenegro.
Trying to further improve the president's already good mood, his election campaign headquarters informed him that the SPS-DPS coalition could expect to win the two thirds majority in the Federal Parliament, which is sufficient to change the federal constitution. This two thirds majority allows, among other, a modification of the, until now, largely ceremonial duties of the Federal president. As could be heard in the SPS circles, that was very important for the election campaign headquarters because the increase in the authority of the comrade president who will in 1997 complete his last mandate as the president of Serbia allowed by the constitution would influence his good will with respect to the members of the election campaign headquarters. The same circles didn't mention the Federal president Zoran Lilic. It seems the election campaign headquarters didn't worry about his mood.
At the same time, predictions of two very important state institutions, Federal and Republican statistics institutes, according to well informed sources, were significantly less optimistic regarding SPS. Later, it turned out that those predictions were very realistic, even precise. Their predictions said that SPS could not expect to win more than 68 seats in the Federal Parliament, and that it probably would not win more than 64. According to the predictions from these two institutes, coalition "Zajedno" could expect to win at most 30 seats in the Federal Parliament. Since the parties which formed this coalition won together 28 seats in the previous Federal elections, it was expected that this number wouldn't significantly change.
What was the basis of the prediction from the SPS election campaign headquarters? We will look for an answer in the preparations for the elections.
It is known that SPS, as a very well organized political party, does not leave anything to chance in the electoral process; therefore, it uses state infrastructure which is under its total control, among other, to conduct very expensive opinion polls. We can illustrate the cost of such an endeavor by pointing out that it was conducted on a very large (and very well prepared) sample of 6500 persons, which allows very precise predictions. Let us mention that, in order to properly conduct such a poll, besides professional preparation which requires long term planning of the implementation, the organizer of the poll must educate poll takers, perform the control of all collected data, and conduct the polling on a large territory and on a representative and stable sample. Finally, after the conducted poll, it is necessary to analyze the data, which requires tremendous expertise. All this requires time, money and involvement of institutions under state control, such as Radio Television Serbia (RTS) and Federal and Serbian statistics institutes. Of course, SPS could conduct such a poll since it controls the state funds and resources as its own, and the institutions which did most of the work are owned by the state, so that the experts employed by them did everything during their normal work hours and on the taxpayers expense. Such a poll would cost any opposition party a lot, and it is questionable whether any one of them would be capable of paying about 150,000 dinars [5 dinars=$1] (and this should be done at least twice a year in order to asses the changes of the mood among the electorate) which only approximately covers costs of such an endeavor, or would be capable of conducting it since opposition parties lack access to the data necessary to conduct such a poll.
Of course, the results of this poll, besides several less important exceptions, which were reported by RTS, were planned for the exclusive use by the SPS leadership. It seems that the analysis of the poll wasn't done properly, otherwise there wouldn't have been differences between the predictions of the election campaign headquarters and the Federal and state statistics institutes. It is also possible that the election campaign headquarters deliberately "tweaked" the results for the president.
SPS also conducted a thorough analysis of the December 1993 Serbian elections, December 1992 Federal elections, and 1992 elections for the Montenegrin parliament. In accordance with the results of that analysis, it was decided to increase the number of electoral districts in Yugoslavia to 36 (from 10). In Serbia the number of electoral districts was increased from 9 to 29 and in Montenegro from 1 to 7.
In practice this meant the increase of census [percentage of votes in a district needed to win a seat in the parliament] to 15%-25% of votes, depending on the number of voters and the number of representatives elected in the district. Let us mention that the law stipulates that "only those lists which receive more than 5% of votes in a electoral district will participate in the division of the mandates from that district" (article 87 of the Law about the election of representatives to the lower chamber of the Federal Parliament, FRY Official Paper, issue 57, September 24 1993). The percentage mentioned in the article 87, which is usually referred to as census, was in practice dramatically increased by the changes in the Law about electoral districts (Official paper FRY, issue 28 from 1996); the consequence was that all those lists which failed to win over 15% of votes in districts with 5 federal representatives (electoral district Podgorica was the only one with 7 representatives, while the other districts have at most 5 representatives), or close to 25% of votes in districts with 3 federal representatives, as was the case in 13 electoral districts in Serbia. Consequently, our proportional electoral systems became very similar to the first-past-the-post (single round majority) electoral system. It was predicted that such electoral system would favor ruling parties in Serbia and Montenegro.
The borders of new electoral systems were adjusted based on the election results from 1993, with the goal to guarantee the largest possible number of seats to the two ruling parties. Various parameters which were taken into account in this process included the overall number of electoral districts, number of electoral districts with odd number of representatives and the winning margin which guarantees SPS victory in particular districts. Namely, the number of electoral districts was supposed to be as large as possible, with the condition that each one of them has at least two representatives; otherwise, the electoral system would be majority electoral system instead of proportional, as is prescribed by the law. SPS demanded that the number of electoral districts with odd number of representatives be as large as possible as well: such districts provide for up to two thirds majority with the smallest possible victory margin [for example if all electoral districts had 3 representatives and in all of them SPS won with 30% of votes against 28% for the opposition, it would win 2 seats in each district and insure two thirds majority in the Parliament]. Of course, the solution of such a problem is very difficult and requires human expertise and computational resources.
As far as we know, the best experts in Yugoslavia were not recruited for this job. It is likely that the analysis was performed abroad and, naturally, paid for by the state; another possibility is that a foreign expert performed the analysis using the support from the local institutions. The latter solution improves the accuracy of the analysis, since the foreigners are not contaminated by local political predictions and can focus on numbers only.
The opposition didn't have any control over the voter lists. According to the article 19 of the law about the election of federal representatives to the lower chamber of the Federal Parliament "Federal Electoral Commission performs general control of voter lists". Since there were 7,597,504 registered voters in the federal elections it is clear that "general control of the voter lists" was only possible in the same way in which these lists are compiled: with the help of computers.
As the member of the Federal Electoral Commission representing Peoples Party of Montenegro (Narodna Stranka, NS), I asked whether it was possible to install a computer in the parliament which could be used by the commission members to access and check the lists of registered voters from Serbia and Montenegro; I was told, with a smile, that that was not possible. According to the president of the Federal Electoral Commission, Mr. Rajko Nisavic, a judge in the federal Court, the commission had discussed the "general control of the voter lists" several times in the past and the demands of the members representing the opposition parties to provide ways to check these lists were rejected with the justification that the law does not specify how the "general control" should be carried out. Fantastic!
Why is it so important for the Federal Electoral Commission, whose majority represents the ruling establishment, that no one but the authorities which compile the lists of registered voters has access to those lists? The only plausible answer is that the regime wants to make sure it can manipulate the lists if necessary. Why keep the lists secret, unless to provide opportunity for vote rigging? Whether and how those lists are used against the opposition parties will become clear once they become accessible at least to the members of the electoral commission which, according to the law, have the duty to control those lists.
We shall illustrate the validity of the doubts regarding the rigging of the lists of registered voters by one example: between 1993 and 1996 something strange happened with the lists of registered voters from electoral districts in Kosovo. The number of registered voters sharply increased. According to the data provided by the Federal statistics institute, the number of voters in the electoral unit number 29, Pec, increased between 1993 and 1996 from 333,537 voters to 392,967, therefore by 17.82%; in electoral unit number 27, Kosovska Mitrovica, the number of voters increased in the same period from 159,481 to 202,999, by 27.29%; in electoral unit 28, Pristina the number of voters increased from 228,879 to 317,311, by incredible 38.64%. Such short term increases in the number of voters cannot be explained by any sort of demographical changes.
The opposition also couldn't control the printing of ballot papers. The Federal Electoral Commission was very frugal and chose the [state owned] publishing company "Politika", which submitted the best offer to the commission, to print the ballot papers. The ballots were printer on cheap, imported, low quality paper.
The control of the ballot printing process boiled down to one visit by the commission members to the Politika print works in Belgrade. All members of the commission, except those representing the opposition parties, received passes in advance. Nevertheless, Politika's representative showed good will and allowed everyone to enter the building. In our presence, in less then 45 minutes one printing machine (out of three in the room we visited) printed all ballots necessary for one electoral unit. Our impression was that the print work had an almost unlimited capability for this job and that in the two weeks left before the election it could easily print 30 times more than 8 million ballots needed for the Federal elections. The opposition representatives had no way to perform the control of the printing process, and the government representatives testified that there could be no doubt in the good intentions and honesty of the people in "Politika". Touching!
Election material was packed and transported in trucks, again without an opportunity for control by the opposition parties. The whole process of printing and distribution of the election materials offered unlimited opportunities for fraud. It occurred to me that, having in mind the nature of the printing and distribution process, the only effective control of the ballots would have been to have every member of the Federal Electoral Commission sign all 8 million ballot papers and later to check the signatures on all of them.
We'll illustrate the possibilities for misuse of the uncontrolled production of election material by several examples from Kosovo, since the elections there passed without opposition control and the opportunities for fraud were the largest.
In electoral unit 27, Kosovska Mitrovica, the number of valid ballots jumped from 26,464 in 1993 to 34,358 in 1996, by 29.8%. In 1996, 35,548 voters cast their ballots and the number of Serbs, Montenegrins and Yugoslavs in that electoral unit, according to the 1991 census, is 36,531 [ethnic Albanians from Kosovo boycotted both elections].
In electoral unit 28, Pristina, the number of valid ballots jumped from 41,256 in 1993 to 60,008 in 1996, by 45.5%. In 1996, 62,332 voters cast their ballots and the number of Serbs, Montenegrins and Yugoslavs in that electoral unit, according to the 1991 census, is 57,985.
In electoral unit 29, Pec, the number of valid ballots jumped from 46,001 in 1993 to 71,704 in 1996, by unbelievable 55.87%. In 1996, 75,127 voters cast their ballots and the number of Serbs, Montenegrins and Yugoslavs in that electoral unit, according to the 1991 census, is 38,126; even if the number of Muslims living in that electoral unit is added to the previous figure, the total of 73,391 is still less than the total number of valid ballots.
Knowing that the Albanian population didn't vote in the November 1996 elections, and that if some of them did take part their number is statistically negligible, and that on average 67% of eligible voters took part in the elections elsewhere in Serbia (it is most likely that a similar proportion of Serb and Montenegrin population voted in the elections in Kosovo), such election results can only be explained by fraud which was much more blatant than that in the second round of the local election in Nis. If it wasn't for the second round of local elections and shameless fraud committed in it, I admit I wouldn't dare be so explicit in my claim regarding the electoral fraud.
As the press later wrote, certain members of ruling parties, after the end of elections admitted that they had taken part in election fraud which consisted of stuffing of ballot boxes with in advance prepared ballot papers. After the cancellation of the second round of (lost) local elections in some cities, numerous instances of electoral fraud which were the consequence of the refusal of the authorities to give up power, became public. It seems that the regime naively believed that the November 3 [Federal] elections were regular and therefore exhibited lack of caution and too much confidence in its victory. In the end, the authorities demonstrated how far they were prepared to go, and the forged election records from Nis proved that this regime does not flinch from a criminal act in order to preserve its monopoly of power.
The role of "satellite" parties. There are numerous parties which are SPS "satellite parties" and are not represented in the Parliament but receive financial and media assistance in their struggle with the opposition. The are also "so-called parties" which only appear just before an election and disappear as soon as the election is over.
SPS motivation for assistance to "satellite parties" is quite clear and does not require further discussion.
There are two sorts of reasons which affect friendly attitude of the ruling parties with respect to the "so called" parties which take part in elections. As such the "so called" parties have no chance of winning a seat in the parliament, because they have no chance of crossing the legal census (5%), let alone actual census (15-25%), all votes won by them join the mass of votes which at first glance do not affect the outcome of an election. However, every such vote is eventually split by the parties winning seats in the parliament in accordance with the number of votes won in a particular electoral unit. Similarly, invalid and spoiled ballots, have the same effect as those cast for parties which do not cross the census. Also, the increase in invalid and spoiled ballots and those cast for the "so called" parties increases the actual census.
In other words, an increase in invalid ballots and those cast for the "so called" parties increases the difference between the percent of votes received by a party and its representation in the parliament, usually referred to as discrepancy. That can also have a crucial effect on the number of votes needed by a list to win a single parliament seat. For example, SPS-JUL-ND needed 28,885 votes for one seat in the federal parliament, "Zajedno" needed 44,059 votes for one seat, while SRS needed as much as 48,704 seats. Above all such discrepancies are the consequence of the large census and the large number of electoral units, but also the consequence of the deliberately designed chaos on the political scene which is manifested through support of the ruling party to various parties without representation in the parliament.
The other set of reasons for tolerance of participation of such a large number of political parties in elections has to do with the use, better said abuse, of the media by the authorities in an election campaign.
A study done by a group of sociologists (Vreme, issue 323, December 12, 1996) showed that almost 50% of participants in the demonstrations against the annulment of the local elections are college graduates. The other half of protesters, according to the study, have a high school diploma. According to the study, the number of protesters with Masters and Doctoral degrees was higher that the number of those who only graduated from primary school. Namely, 2.7% of participants in the protest have stopped their education after completed primary school, while 3.1% have a Masters or Doctoral degrees. Having in mind that 60% of Yugoslav population stops education after completed primary school, we can conclude that there is a significant discrepancy between the composition of the protesters and the rest of the population.
It is necessary to emphasize several facts which have to do with the educational level of the population, in order to shed more light at the methods of preparation for the elections and the election results. For example, according to the Federal statistics institute data, in electoral districts Prokuplje and Leskovac more than 15% of population aged more than 15 declared themselves to be illiterate. Of course the true number of illiterate persons is significantly higher. It is estimated that at least 30% of adults in Yugoslavia are functionally illiterate, that is, they are unable to reproduce a single paragraph from a daily paper.
It is not necessary to discuss the significance of the influence of the electronic media, especially television, on this portion of the population. This is the main reason behind regime's attempts to maintain the control of all media outlets. The media under the SPS party control mostly address illiterate or semi-illiterate population. The leadership of the ruling parties is aware that the reliability of their traditional voters depends on their lack of education, and also, but to a much lesser extent, on the corrupt nature of a portion of the adult population. Public opinion polls, conducted twice a year by SPS, confirm this.
The so called election campaign presentations by political parties on national television, RTS, mostly affect the educated portion of the population. Mad ravings which can be heard in these programs can only alienate potential voters from the political scene and reduce a turnout, as is in the interest of the ruling parties.
Space for "fudging". The methodology employed by the Federal statistics institute in collection of results leaves space for manipulation, above all because the process itself goes very slowly and therefore allows additional "fudging" of election results. Software which is used for the processing of results should be tamper proof, but the data arriving to the institute can be forged wherever there is no effective control by the opposition.
One should have in mind that there are 10,078 polling stations in Serbia and 878 in Montenegro, all together 10,956. Ruling parties, which use state bodies as their own, have, naturally, total control over all polling stations. This is not true for the opposition parties, because some of them are not active in all parts of the country or cannot afford to have observers at all polling stations.
For example, all parties received funds from the budget, based on the number of candidates fielded in the elections: SPS received 171,810 dinars [5 dinars=$1]; DPS 41,471 dinars; "Zajedno" 171,810 dinars; SRS 213,285 dinars; NS 41,471 dinars etc.
In comparison with the funds necessary for the election campaign, the portion received from the budget is negligible. If daily expense for a single activist for a day spent at a polling station is only 50 dinars [$10], a party with candidates in all electoral districts in Serbia would have to spend more than 500,000 dinars for this. For two rounds of elections more than a million. All that for only one observer at every station, and at least two are needed for effective control. In present conditions only those parties which control the state budget can afford that.
The Federal Electoral Commission demonstrated very little good will to try to reduce the possibility of fraud in Federal elections. Just the opposite, they rejected the demand that the overall number of voters who voted at a particular station be made public an hour after the station is closed although that could have been done very easily by simply counting the number of circled names on the list of registered voters. The argument was that the data cannot be easily obtained from the mount Jelica. As if that polling station could significantly affect overall results. Out of many ways to prevent or reduce possibility for fraud in the elections, the counting of votes cast in a station is the easiest one to implement. Regardless, the Federal Electoral Commission didn't express the desire to reduce the opportunity for fraud in the elections.
The opposition had no observers at many polling stations. An initial inspection of the publication Elections '96, lower Parliament chamber, published by the federal statistics institute, is sufficient to notice that in electoral districts Nis, Prokuplje, Leskovac, Vranje and all districts in Kosovo there are polling stations in which the coalition SPS-JUL-ND won 100% of votes.
For example, in the electoral district Vranje, municipality Vitina, polling station Djelekare, there were 1217 registered voters, all 1217 voted in the Federal election, there were 1217 valid ballots and all 1217 were cast for SPS-JUL-ND. In the same electoral district, municipality Gnjilane, polling station Malisevo, there were 2102 valid ballots, all 2102 of them cast for SPS-JUL-ND. There are hundreds of similar examples. We can openly state that this can only be explained by electoral fraud.
We can assume that the list of registered voters are perfect and that there are exactly 1217 voters in the village of Djelkare; we can assume that since the compilation of the list there were no changes in the village, that everyone is healthy, no one has died, no one was absent, and that everyone voted; we can even assume that all the voters vote for the same party and think exactly like their neighbors, but knowing that in the electoral district Vranje at least 15% of population is illiterate and another 15% functionally illiterate it is impossible that at least one out of thousand voters who vote at that station didn't get confused and circled a different party. This is a proof that there was fraud in the federal elections.
The author is a university professor of mathematics