Zajedno-Leftists (SPS-JUL-ND) 26.1-23.9

The three polls conducted over the past month and a half show that the Zajedno opposition coalition always had a lead over the SPS-JUL-ND coalition. Partner’s polling team conducted the third poll in Serbia without Kosovo from October 22 to 26, covering 1,000 people in 97 municipalities. Their response:

SPS-JUL-ND 23.9%

Zajedno 26.1%

Serbian Radical Party 10.4%

Other parties 4.9%

Undecided 15.9%

Abstentions 19.5%

That means that the poll commissioned by VREME shows that the leftist coalition is close to percentages the SPS won at the 1993 republican elections (24.3%, 23.9%) and that Zajedno has slightly higher percentages (by 3.1%) than the joint percentages won by DEPOS, the Democratic Party (DS) and Serbian Democratic Party (DSS) (23.5%, 26.1%).

Partners pollsters decided to publish a cautious prediction for the elections saying some four million voters will turn out in Serbia and that Zajedno will win slightly more votes than the leftist coalition. They didn’t specify the percentages because the poll was completed (Sunday, October 27) prior to the end of the campaign and possible surprises but they were less surprised by the results of this poll than the two previous ones.

The third and last poll showed that Zajedno voters have approached the percentages of the last republican elections even though the number of leftist voters is lower. That means about 1.5 million people could vote for the Socialists in Serbia and almost 1.6 million for Zajedno. Those numbers are no prediction but just an interpretation of the results. If that poll had been an election exit poll Zajedno would have kept the number the of votes its members won in 1993 and attracted a group of voters about as half as large as Democratic Serbia -while the leftists would have kept the number of votes won by the SPS in 1993.

To me, the results of the third poll seem closer to reality than the second poll in mid-October which indicated that Zajedno could win 50% more votes than it had earlier.

As this article is completed (Tuesday, October 29), the Center for Psychological Research at the Belgrade Social Sciences Institute (IDN) published its prediction for the 1996 federal elections which says 61% of the electorate will turn out and 39% will abstain. Of the people who turn out, 27.3% will vote SPS, 18.3% for Zajedno, 6.4% for the SRS, 7.1% for other parties and 1.9% of the ballots will be invalid. Those numbers cover the whole electorate. Since election commissions base exit poll results on the number of votes cast, the number of abstentions is played down. IDN researchers predicted the following results:

SPS-JUL-ND 45%

Zajedno 30%

SRS 10.5%

Other parties 11.5%

Invalid ballots 3%

That prediction is drastically different to Partner’s poll. In comparison, the third Partner poll showed that Zajedno will win 39% of the vote and the leftists 36%.

The IDN prediction was based on one poll in 76 municipalities covering 1,520 people in the FRY, including 1,280 in Serbia without Kosovo. The poll was conducted on October 5-15, during Partner’s second and 10 days before its third poll. IDN researchers only predicted the mood of the electorate in Serbia and gave up on a prediction for Montenegro since 250 people are not enough for a reliable prediction. Their results show a huge lead of the Montenegrin DPS over the National Unity opposition coalition (43%-17%). A week ago, IDN published the results of a poll which showed a very low level of trust for system institutions. Researchers could not find a convincing explanation for the discrepancy between disappointment and votes for the party that orchestrated the situation. That poll was conducted in May and it seems researchers discovered a very low level of support for the opposition (up to 12%). The IDN mid-October poll indicates a big rise in votes for the ruling party. The IDN is the most experienced FRY institution in terms of opinion polls (they began their research 33 years ago) and their predictions have to be taken into account although this writer is surprised by them.

This could mean that the IDN expects the leftists to win more votes in Serbia (obviously including Kosovo) than Zajedno by up to 600,000 votes which seems impossible five days before the elections. In 1993, the SPS could count on about 1.5 million votes and the parties now rallied in Zajedno on 1.4 million. Except for Seselj’s SRS in 1992, opposition voters never crossed the red line to date. Parties did.

The IDN has several explanations for their predictions: a) two thirds of the electorate are ideologically to the left of center; b) this is a choice between lesser evils - people are unhappy with the system but fear something even worse; c) federal elections and identification with the federal state are the weakest link in the opposition chain; d) opposition leaders all suffer from the Ante Markovic (the former Yugoslavia’s last federal prime minister) syndrome, enjoying wide but shallow support unlike Milosevic whose support is narrower but stronger; e) the leftist coalition dominates the media and has campaigned all year while the opposition campaigns occasionally.

The IDN said young people still favor the opposition but the left has improved on its percentages of young supporters. The left has strongholds in places with populations of less than 5,000 including ill-educated and older people. Regionally, central Belgrade has stronger opposition support while the suburbs prefer the Socialists. the Sumadija region has the strongest concentration of leftist support.

The IDN did not calculate the number of parliament seats but it’s obvious that the ruling party will get a big majority if the IDN predictions are true.

Partner also didn’t calculate seats.

VREME made another simulation of the elections after the third Partner poll.

The calculation is based on 1993 election results and corrected in electoral districts by the percentages changed in the polls compared to 1993. If the exit polls are anywhere similar to the third poll, the leftist coalition will win 51 seats and Zajedno 46. That is fairly uncertain since a tight race could happen in at least 10 districts with the victory being decided by 600 to 2,000 votes.

The Partner poll gave some indications of votes at local level. Those figures are not too different from the vote at federal level.

Which party candidate would you vote for today (October 22-26) in local elections?

SPS 17.5%

JUL 5.9%

Zajedno 22.5%

DSS 9.6%

Other parties 4.3%

Undecided 18.9%

Abstentions 18.1%

Regionally, the SPS results are worst in Belgrade (13.8%) and about equal in Serbia and Vojvodina (18.3%, 18.5%). JUL did best in central Serbia (6.6%) and worst in Belgrade (5.5%). Zajedno did best in Belgrade (26.3%) and worst in Vojvodina (19.4%).

The DSS did best in Belgrade (6.3%) and worst in Vojvodina (1.2%). The SRS did best in Belgrade (10.9%) and worst in Vojvodina (8.1%).

The other parties did best in Vojvodina (10.9%) and stand little chance of winning anything anywhere else. At local level, the leftist coalition is less than 1% stronger than in 1993 while Zajedno is 1.7% weaker. Zajedno did worse at local level in the third Partner poll than previously.

At local level, JUL is twice as strong as Kostunica’s DSS in central Serbia 6.6:3) and Kostunica is 1.,5 times stronger in Belgrade 6.3:4.4). What that is going to mean on the ground can’t be predicted on the basis of these results except for the general prediction that Zajedno seems to be the strongest player in the first round at local level.

So will these polls and simulations held VREME readers find their way around these elections or at least prevent surprises afterwards?

IDN director Dragomir Pantic said debates were held in the West that showed that pre-election polls do not affect voters’ decisions since just 2-3% of them base their decision on polls by influential agencies but added that polls help keep elections rational, expand political culture and help political parties with their campaign strategies.