Digital Media
The acceleration of TV to Online. Far too many people have been saying “TV is Dead” for far too long – ignoring the actual amount of television people watch. Americans watch, on average, an astounding nearly 8 hours of television a day. (Nielsen). We watch as much TV now as we did before Facebook, Netflix, and YouTube. However, we are observing accelerated changes to HOW people are watching television. In the most recent Nielsen Total Audience Report – we observe adults 18+ spending 50% of their daily time on digital devices and only 40% on Live+Time Shifted TV. One of the accelerators of this trend appears to be people shifting to work from home models, and with a number of workers preferring to continue to work from home, this will have profound effects on reaching political audiences.
Download the report:
https://www.nielsen.com/us/en/insights/report/2020/the-nielsen-total-audience-report-august-2020/
Political Research
It is no secret that some in the polling industry had a rough couple of cycles. While we are still collecting data, we are finding evidence that the conventional wisdom that the polling was catastrophically off is wrong. One reason conservatives may be so down on the polling industry and quick to disregard it – there is growing evidence that our political biases are leading to a biased evaluation of polls and pollsters.
“Respondents viewed polls as more credible when majority opinion matched their opinion.”, a study in Political Behavior (2020) by researchers Gabriel Mason and Sunshine Hillygus of Duke University finds “evaluations of polls are biased by motivated reasoning.”
In an online survey experiment, participants were shown polling results from one of three conditions: a roughly even split, one with a clear majority (61% to 34%), or one with the opposite support (34% to 61%). The question was on a registration for immigrants from Muslim countries. Respondents were then asked questions about perceived accuract, trustworthiness, and informativeness.
In addition, there was a pre-treatment of the participants asking them their opinions prior to seeing the “polling results.” The expectation was the “perceieved polling credibility depends on the respondent’s prior opinion towards the issue.”
In addition, the authors conducted a second study using a horse race question between Trump and Clinton.
Using these two experiments, they find “evidence that the American public evaluates the credibility of a poll based on the extent to which the poll’s results offers support to the predispositions.”
Simply put the starting point matters. We understand this in political communication – the more you care and the stronger your opinion – the more difficult to change your mind – again, the strength of priors matter. What this is demonstrating is the same effects at work on receiving information.
For me this is a difficult read, because we pride ourselves on using polling to get us out of our bubbles and to test our critical assumptions. The day will allow polling and research to be soley used to confirm our biases is the day polling ceases to be a worthy tool.
Madson, G.J., Hillygus, D.S. All the Best Polls Agree with Me: Bias in Evaluations of Political Polling. Polit Behav 42, 1055–1072 (2020). https://doi.org/10.1007/s11109-019-09532-1
Political Data
In my 2020 year in review, I spoke of the death of third party cookies and what that may mean for political adversting. Where is ad tech going? Monopoly cases! As of now, several antitrust cases are working their way through the court system – against Google, Facebook, and likely soon to be Amazon and Apple. Nothing is likely to be decided this year – the gears of justice turn slowly; however, we are likely to see effects this year as big tech begins to change in response to the suits. For this year, while important, it is not how they respond in court it is how they respond in privacy and ad markets.
