Statistics are usually expected to get things wrong but in the October 25th election, all the predictions were way off. As the results finally began to filter in, the country watched open-mouthed as we saw that there would be a runoff. Not only that, but the gap between Victory Front’s (FpV) Daniel Scioli and Cambiemos’ Mauricio Macri was far smaller than anyone would have expected. This not only revealed the difficulties political analysts face when attempting to make predictions, but also reflected the fact that consulting services’ political inclinations sometimes get in the way of objectivity when conducting polls.
Over the weekend and throughout the week, various polls have been published in the tense run up to the runoff set for November 22nd. Although most agree that Macri is in the lead, there are discrepancies as to by how much.
Two polls published yesterday alone gave two very different results: for one consultant, Haime and Associates, the difference would be 3.8 percent and for the other, Poliarquía, 8.5 percent. In addition, for Haime and Associates, not only is there a smaller difference, but that gap is in fact closing from 7.3 percent last week.
There are also discrepancies regarding the distribution of votes that went previously to A New Alternative’s (UNA) Sergio Massa, the second runner up in the October election. For Haime and Associates, half of these key voters would go to Scioli while for Poliarquía, Macri woud count on 60 percent.
What makes this case particularly interesting is that both polls were requested by two different newspapers, Página 12 and La Nación, respectively: the first being sympathetic to the incumbent government, the second not so much. This isn’t to say that the results of one or the other are completely useless, but there are differences that should be taken into consideration. For example, for the October election, the consulting group Equis, led by openly pro-Kirchnerite Artemio López, had stated that Scioli would easily win in the first round with 41.7 percent of the vote.
Now that we’ve seen that the October election predictions were so off, political analysts and consultants are being far more prudent in their statements and forecasts.
“Despite Cambiemos’ advantage that is shown by the poll, one cannot consider that the election is defined”, said Eduardo Finanza, director of the Poliarquía poll.
This was echoed by all the consulting firms involved. Even Artemio López said that the “panorama is too uncertain” to make solid predictions. However, this is not just due to the political inclinations of each analyst (not even Macri’s consultants of choice predicted him going past the 30 percent mark), but inherent problems of polls.
Statistics is an imprecise science. Polls are especially tricky because they are based on the assumption that, as Carlos Gervasoni mentioned in an interview on El Exprimidor, people tell the truth. Which they often don’t. Not trying to have a Dr. House moment here, but the truth is that many who voted for Macri may not have felt comfortable admitting it. Add the volatility of the Argentine electorate in a polarized election and one almost wonders why we do it at all.
A poll is basically a screenshot of what public opinion looked like at the time the surveys were taken. Whether that is congruent with what happened or not is really a case of luck: according to Gervasoni, this is something that happens all over the world, but Argentina showcased the major limitations of polls last month. Hence the current prudence of the analysts and, it should be said, politicians.
For one thing, Scioli has been skeptical of the polls that puts him at quite a disadvantage. That’s not surprising though, as all the polls for the first round had indicated that there wouldn’t even be a second round. In an interview with MDZ radio (Mendoza), Scioli maintained that:
“The people will have the last say on the 22nd, but the consulting firms have had a gross level of error during the last election. What most of the firms have done in the last elections is coarse.”
A little harsh, but understandable. Macri, on the other hand, has defined his ballotage strategy accordingly. Since voters showed a tendency to close the gap between the candidates, i.e. vote for the runner up, not the winner, Macri has been very careful not to brag about his supposed advantage in order to ensure support.
However tempting it is to take people at their word (or numbers), one should take into account the limits of these surveys and where they come from. In addition to political inclinations and the inherent limitations of polls, there are also factors like the upcoming debate. Raúl Aragón, author of yet another poll for Aragón and Associates, mentioned yesterday that there is 10 percent of undecided voter to be taken into account: thus, the debate may have an effect on the controversial margin.
The analysts coincide on two things: that Macri is a little more comfortable after the October elections and that nothing is certain. I know the suspense is killing us, but let’s try not to get too ahead of ourselves.