‘I can see how at a moment in time 49 percent of the public would think impeachment proceedings are at least worth exploring…’
(Ben Sellers, Liberty Headlines) A survey conducted jointly by The Washington Post and ABC News released on Friday showed a growing number of respondents supported impeachment of President Donald Trump and disapproved of his job performance.
However, tell-tale signs of sampling bias and other questions of survey methodology undermined the confidence interval, as many remain skeptical of political polling in the wake of the 2016 election.
According to the survey, which claimed to have a margin of error of 3.5 percent, 49 percent of respondents agreed that Congress should begin impeachment proceedings on the president (including 40 percent who strongly agreed), while only 46 percent disagreed (33 percent strongly).
The survey says it randomly sampled 1003 adults nationally, both by landline and cell phone. For a confidence level of 99 percent with the provided margin of error, however, it would have needed to survey about 350 more adults given the U.S. adult population of approximately 250 million.
Instead, the left-leaning news organizations chose to accept a significantly weaker confidence level (opening the possibility that the entire survey could be bunk) to fit their margin of error within the spread.
But assuming the survey results were an accurate statistical reflection of the population sampled, the bigger concern would be whether the sample itself accurately reflected the U.S. population.
As the investigative blog Zero Hedge explained in October 2016, many polling agencies, including the WaPo/ABC ones, systematically build in bias to their polling by oversampling certain segments of the population, even as they claim to be presenting a random sample.
Zero Hedge showed, based on the leaked John Podesta emails, that Hillary Clinton agents actively encouraged oversampling, resulting in such headlines as “Clinton Vaults to a Double Digit Lead, Boosted by Broad Disapproval of Trump” two weeks before the November 2016 election.
One clear indicator of oversampling on the latest survey may be reactions to the economy. Even if respondents were reacting on ethical questions to the double-whammy conviction of former Trump associate Paul Manafort and the guilty plea of Trump’s former lawyer, Michael Cohen, few could complain about the stock market surging to historic highs during the survey period.
However, the WaPo/ABC results found 45 percent approval to 47 percent disapproval (36 percent strongly) in Trump’s handling of the economy.
Despite conflicting data from a CNBC economic survey in late June, which showed a majority approval, the WaPo/ABC poll had Trump’s economic approval in decline since it last asked the question in April.
The survey also suggested a precipitous decline of about 4 percent (or about 10 million U.S. adults) in Trump’s overall approval rating since April, with those approving at 36 percent and disapproving at 60 percent.
Despite the questions surrounding the survey methodology, as writer Dan Calabrese said, the results are nothing to scoff at with the midterm elections approaching.
“When you start with the fact that 40 percent of the electorate hates Trump no matter what he does, and you pound away at those themes day after day, sure, I can see how at a moment in time 49 percent of the public would think impeachment proceedings are at least worth exploring,” Calabrese wrote.