Jan Herzmann: surveying the Czech political scene

Jan Herzmann is a managing partner in one of the top Czech polling agencies, Factum Invenio. While most of the work is marketing surveys for companies, the current workload is spiced up by polls and surveys ahead of lower house elections at the end of next week. I asked Mr. Herzmann about the special characteristics of polling in the Czech Republic and how he came to be in the forefront of moves to re-launch political opinion polls after more than 40 years of Communism.

“I was lucky enough to be the first one to have the chance to start opinion polling on election issues early in 1990. Before the Velvet Revolution, there was no polling because the results of the elections was clear in advance. At the time I was one of those people working at the statistical office and it was my idea to restart polling, pre as well as post election polling. So we tried that and it worked.”

There was a lot of interest from whom, the newspapers, the media or the parties themselves?

“Everyone was curious about this. And there was great interest from the whole population and the media. At that time Czechoslovak Television was extremely interested in the results. Everybody was curious. It was a nice time.”

And in the last 20 years, has the market for political polling and maybe other polling, changed?

“In the beginning, there was no market. Step by step it started to develop and the unusual tendency in the Czech Republic is that the political parties are interested in buying opinion polls but the media is not. So the market is a bit biased. It is huge as many political parties order polls on different issues but the media do not. And I would say there is a lack of an independent, corrective mechanism to balance the party organised polls and surveys.”

Do the parties conduct polls throughout the year or is it mostly in the pre-election months or weeks?

“In my opinion, we are dealing with two different decades. In the 1990’s the parties were interested in polling during the last three or four months before the elections. It changed at the beginning of this century. Now the big parties, and in the first place the Social Democrats, spend a lot of money on polling throughout the whole election term.”

And as a result do you see a fairly quick change in policies when there is a poll showing a policy is not popular or hitting the mark?

“I should say there are big differences. Some big parties react fairly slowly and this is the case of the right-wing party, the ODS (Civic Democrats), and some of the smaller parties who are also rather cautious. On the other hand, the Social Democrats are well known for a very swift reaction and sometimes they are called populist because they react to changes in public opinion very quickly and then change some of their positions and policies fairly quickly.”

Can you give some examples of where polls said you are out of tune with public opinion and the policy changed swiftly?

“This applied to some of the issues concerning the reform of healthcare and the situation on the labour market. So there are different, rather small, issues that were changed as a result of opinion polls, but not the strategy or policy of course.”

And the media are still not interested in carrying out polls ahead of elections?

“This year there has been a big change. Public television and radio asked for a series of polls that they used for organising their broadcasts. This was not the case in previous years. On the other hand, the commercial media, both electronic and print, wait for polls to be printed so that they can use them for free. They do not order polls. Otherwise, they prepare rather peculiar estimates based on, say, 100 clever people, things like that. It is not very typical that the media do not order their own pre-election studies.”

Would you say that the Czech polling market after 20 years is now a mature market or does it still have some things to learn and develop?

“As to the frequency of polling and modalities, it is a mature market now. There are at least five polling agencies that can deliver high quality pre-election studies. The difference is that the media do not distinguish between agencies working for political parties and the results of those agencies and results of independent information services. It makes a bit of a mess of the message you get from the electronic and printed media.”

Does it make a difference if the company is carrying out work for a party and carrying out public work or polling? Is there an influence or just a suspicion of influence?

“There is no evidence that any of these agencies who are working for political parties influence the results. But in my humble opinion the reader or public should know whether the results are based on a study that was ordered or paid for by a political party or not.”

One problem I find when I am out on the streets doing interviews is that people seem to run away or do not want to answer questions. There still seems to be a latent suspicion of people asking questions, perhaps a hangover from the communist times. Is this a problem for Czech surveys and are they accurate if people do not want to answer questions?

“Definitely this is not a hangover from the ancien regime, the communist era, because in the 1990’s people were willing to give their opinions. They were very open and the refusal rate was rather low, about 10 percent. This has changed. We now have a response rate at the western level, which means between 60-70 percent. Of course, corrections are necessary if we want to prepare an election model that makes a forecast of the probable election result. I mentioned five agencies that can deliver rather good results. All these agencies have their correction procedures, statistical procedures, to take away the bias of non-response.”