The findings of NewsLabTurkey Research Hub’s new report entitled “Ahead According to the Polls”: An Inquiry into the Media, Opinion Polls and Election Reporting are highly insightful. They range from the masculine language that dominates elections to the impact of opinion polls on the voters. The most vital finding is that the media is faced with a huge ethical responsibility at all stages of the election process. Other main findings of the report include:

• “Horserace reporting” dominates: Election reporting that focuses on the candidates’ likelihood of losing or winning rather than on their policies has been described as “horserace reporting”. This type of reporting is heavily shaped by opinion polls. The first commercial opinion polls focusing on elections were conducted in the 1930s in the USA by George Gallup. Since then, opinion polling has turned into a large industry especially in several countries. The existence of a competitive polling industry is a prerequisite for horserace reporting. 

• Horserace reporting is more prevalent in commercial TV stations: Research from across the world has demonstrated that horserace reporting is favored by commercial TV stations, especially 24-hour news networks, rather than by public broadcasters. The main motivation of commercial broadcasters in selecting this format is to create excitement among the viewers. Horserace becomes the dominant format where a competitive polling industry and a competitive media industry co-exist.

• Two male candidates offer the ideal excitement: One of the most ideal conditions to turn the election into an exciting horserace is the presence of two candidates who are competing “head-to-head” according to the polls. Elections where more than two candidates or political parties are competing are too complicated for the horserace format: it becomes more conducive when the chances of winning/losing are split equally between two candidates on a 50 per cent basis. The masculine language that often characterizes horserace reporting overshadows female candidates. Excitement peaks when the horserace is between two male candidates.

• Turkey is closer to the USA in structural terms, to Europe in ideational terms: Following the recent transition into a presidential system, two candidates and two party alliances have started to dominate the political scene in Turkey which is a structurally favorable condition for horserace reporting. The polling industry has started to develop after 2002 as the newly-elected Justice and Development Party and Recep Tayyip Erdoğan have started sponsoring polls regularly. A competitive media industry, however, does not exist in Turkey. The Turkish media system is closer to its European counterpart in terms of partisanship but it is under heavy government pressure in contrast. The format that dominates election reporting in Turkey is debate programs with the participation of guests and experts selected from an “appropriate” guest list pool.

• Opinion polls create a “bandwagon effect” among the voters: As several social psychology experiments have vindicated, the “bandwagon effect” is the most common effect of opinion polls on voters where they tend to vote for the candidate who is ahead in the polls. The bandwagon effect is particularly strong among first time and young voters, and among voters who do not hold strong political views. The bandwagon effect can turn opinion polls into self-fulfilling prophecies. The media thus play a critical role in the election outcome by highlighting particular polls over others. Partisan media which keep reporting biased and unscientific polls may create cynicism, disillusionment and election denialism among their own followers when partisan forecasts mislead them. One particular finding that concerns Turkey is that most opinion polls are unscientific when examined under close scrutiny.

• “Nate Silver wins the election”: The fact that opinion polls are frequently incorrect has opened up a new debate that points to big data and algorithmic forecasting of elections. Nate Silver, who has since 2008 been forecasting the USA elections with nearly a hundred percent accuracy is the face of this debate. In 2012, Silver’s forecasts were fully accurate for all 50 states. The phrase “Nate Silver wins” has entered the election reporting jargon in the USA following his 2012 success. However, a new debate around the question of whether this level of precision is harmful to the election reporting profession has started. The plain language that journalists use to report the news is an advantage as terms associated with algorithms may be too complex for most individuals to comprehend. This debate also concerns the future of election betting markets which are illegal in Turkey but popular in the United Kingdom and elsewhere. The potential impact of algorithmic forecasting on the election betting business needs a fresh debate.
• Focus on policies and the specialist opinion poll reporter position: The main alternative to horserace reporting is to focus on candidates’ policies on important topics. The creation of a “opinion poll reporter” post can also be a step in the right direction. This specialist opinion poll reporter can carefully select which polls to report. Well-trained in scientific methodology, the opinion poll reporter can ensure that the media only report polls that are of a high scientific quality. The scientific skills expected from the opinion poll reporter ties in with the debate on what kind of training journalists should be given. Formal journalism training started in the late 19th and early 20th centuries with an emphasis on language skills and essay writing. From the 1970s onwards, the weighting of social science has progressively increased in the journalism curriculum and there are calls today for including computational social scientific methods as well.