Election 2016 Problems with the Current Election Analysis: Summary
Election analysis concerning the 2016 US Presidential race is truly awful. This series of posts will discuss this topic and of course how to improve upon it. There are many issues and problems with the current US presidential analysis. Highlighting some of the most obvious:
• Pre-Race Republican Advantage – judging from coverage, it would seem like Obama and the Democrats had a tremendous electoral advantage as a carry-over from the previous elections but this is not the case. In fact, judging purely from Obama’s election record, he has set the stage for a Democratic loss in 2016. This has not been covered and is frankly shocking. This is the foundation from which the current election has grown and without an understanding of it any analysis becomes questionable.
• Black Box Polls – polls seem to be fairly scientific from the outside but there is a black box element where the data is manipulated which is not transparent between the collection of the data and the presentation of it. This is where ‘weighting’ or ‘re-balancing’ occurs and this stage is very non-transparent. As long as there is such opaque data analysis, the quality of final output remains in question.
• Expected Turnout – this election will likely have an extremely high turnout on both sides. This will be one of the defining factors of this election. But, analysis and forecasts of how turnout will impact the race are sparse and almost all analysis is based on previous election turnout which seems terribly deceptive. Analysts need to make turnout projections even thought it is uncomfortable and risky.
• Demographic Turnout – in addition to overall turnout projections, analysts should also forecast turnout of different demographic groups. This is essential in the 2016 election as the turnout should be record-setting. Polls should be focused on forecasting turnout as well. Without forecasting turnout of different demographic groups, polls are faulty as they become too dependent on previous elections and become backward-looking.
• Bias and Transparency – journalists should be neutral and report the news so that the public can make up its own mind. Debate moderators should be neutral and conduct debates highlighting issues important to the public. Journalists and moderators have proven themselves less than neutral during this election.
• Social Desirability Bias – many in the general populace feel they should publicly agree with socially acceptable norms and will change their survey or poll responses to coincide with such. This goes beyond politics, as it is a phenomenon that permeates many issues including personal finance, health, and sexual practices. There are strong indications that many in the general populace are reticent of publicly supporting Trump due to Social Desirability Bias but likely will vote for him nonetheless.
• Social Media Trump Dominance – there is a quasi-disdain in the media for social media analysis. It is unclear why this would be the case other than it benefits a candidate that they do not support. Social media is the newest form of communication and has marginally displaced media and analysts ignore it at their own expense.
• Search Trump Dominance – Trump dominates search over Clinton. Put another way, the electorate is simply more interested in Trump. This seems extremely relevant to a national election but has to-date been ignored by the media and analysts. This is an error in 2016 as search is how most get their information.