"I Didn't Come Here to Explain"

12/15/2014 - 11h32

"Oh my God! It seems to me that the voter ID cards should be replaced by a voting license, which would only be given to those who pass the test to weed out the worst applicants," joked reader Carlos de Marchi, commenting on the result of a Datafolha survey published on Sunday, Dec. 7 ("Brazilians blame Dilma for Petrobras case").

Marchi did not marvel at the main revelation, within the headline, that seven out of every 10 Brazilians attribute to President Dilma Rousseff some responsibility for the corruption scheme at oil company Petrobras.

I don't even believe that it was anything to marvel at. Brazil's biggest state-owned company is going through a scandal of fraud which spits out new accusations every day. Marveling would be if this omnipresent news did not spill out onto the federal government, tied by the umbilical cord to the oil giant.

What is surprising, and not only by itself, is the apparent contradiction of this number with other revelations in the survey.

If 68% hold the president responsible in some way for irregularities at the company, how is it possible that 42% consider her government great or good, the same rate as in October at the peak of the electoral campaign?

Dilma also does not appear on the list of beneficiaries of the corruption scheme, and 40% said that corrupt people have been punished the most during her term.

It's not unusual to have something erratic in the collective perception. But I doubt that the problem in this case is the quality of respondents, but rather the generic question conceived by the newspaper.

The term "responsibility" is an umbrella too wide to be used as it was, without complementary questions that help not only in the interview, but also the reader, to understand what the concept could mean to each one.

A caveat which lacks the clearest parameters is not the news from this questionnaire. The question is a standard one, identical to those which were made in surveys during previous administrations, also affected by accusations of corruption.

The problem is that repetition of a model created decades ago has the ability to respect history, but brings up a defect again and again, in this case, the ambiguousness. At that time, like now, instead of illuminating the situation, it triggers more doubts.

For example: what do the 43% of those interviewed who ascribed the president a lot of responsibility in this case think? Do they believe that she knew about the irregularities and did nothing? That she had some active participation in the scheme or gained benefits from it?

What is co-participation by spreadsheet, why did she choose and/or approve the managers of the oil company and make or go along with the agreements with political parties which put thieves in charge of watching the money?

And if responsibility comes only as a result of the job that is occupied, would that be "much" or "a little" responsibility? (To explain to those who did not see it: "a little" was the option chosen by 25%; another 20% said that Dilma Rousseff has no blame, and 12% were undecided).

The managing editor's office agrees that additional questions could qualify the perception given, but it asserts that there was no reason to believe that this would change the main result. I also believe it would not, and don't put this in doubt. My point is that the survey could go beyond how much and try to clarify why.

Surveys are valuable instruments to try to understand a society's mood, as much as can be done to reduce their limitations and determine their findings. If not, as the reader defined, the results are diffuse, vague and become material for gossip.

Translated by JOHN WRIGHT

Read the article in the original language