The results of the latest Marquette Law School poll were made public in two releases on February 12 and 13. Each release began with an enticing title and highlights of the survey. Charles Franklin, Director of the poll, and his team present this survey, as they do all of their work, as a snapshot of what Americans are thinking and how they feel about a wide range of political issues. What can we take away from this report, which is loaded with numbers, tables, and verbatim quotes from the participants? Absolutely nothing. It is meaningless. Data is not information.
This poll was primarily conducted in the second week of Donald Trump’s presidency. Last November, the voters spoke clearly in support of his agenda with a substantial majority of the electoral college, a sweep of the swing states and the majority of the popular vote. Any meaningful evaluation of the public’s support for Trump’s actual performance will have to wait longer than 2 weeks.
The poll is deeply flawed in a number of ways that make it a waste of time and money. This was a national survey with a sample of 1,018 people. The margin of error is identified as 3.5%, which means the scores provided could be 3.5% higher or lower, a swing of 7 points. More than 156 million Americans voted in 2024. Common sense tells us the likelihood that just over 1,000 people spread across the country will come close to accurately representing the views of the voting bloc is close to zero. The goal of representative views is further hampered by the fact this poll drew its sample entirely from what is called the SSRS Opinion Panel. This is a panel of people who are invitation-only, willing to take surveys on a variety of topics every month, willing and able to complete surveys online, and paid for their participation. These attributes disqualify them as a random, representative sample of the American electorate.
Survey limitations go well beyond sample size and method of selection. The team at Marquette decides what questions to ask. A thinking person may wonder why public opinion of the United States Supreme Court ranked high on their list of what the public currently cares about; or why the team believes the size of the penalty for Trump’s felony convictions or a Texas law that would require an ID to access internet sex sites are occupying people’s minds.
The ultimate goal of serious investigators is discovering truth. The process begins with data collection and progresses to a rigorous analysis of what the data may mean. They do not dump raw data on their audience without a thorough discussion of findings including clearly defining the limitations of the study. Credible investigators guard against anyone using their data to draw erroneous conclusions. Whether these omissions in the Marquette report are by design or just sloppy practice, the outcome is the same. The readers are inclined to draw conclusions that support a particular political agenda and the table is set for the dishonest media to take the data and run with it.
Case in point. Without the constraints of explicit limitations of the study, The Milwaukee Journal Sentinel (JS) was free to immediately publish a story entitled “New Marquette poll: strong feelings on Trump’s moves. Early actions receive either high support or fierce backlash.” Authored by Craig Gilbert, who just happens to be a Lubar Fellow at Marquette Law and is “recently retired Washington Bureau Chief and national political reporter for the Milwaukee Journal Sentinel”, the article lives up to the extreme liberal bias of the JS, completely bypasses any cautions about what can reasonably be concluded from this survey and is laced with editorial comments that do not belong in any legitimate news reporting. For example: “Trump’s early moves have attracted wildly divergent levels of support and opposition…” “there is nothing ‘neutral’ about the way they are responding to the blast of activity and audacious claims of power coming from the White House.” In discussing Trump’s initial job approval rating of 48% in the poll Gilbert opines, “If this turns out to be Trump’s ceiling of popularity as president, as often happens with a president’s early job ratings, then these numbers are pretty underwhelming for him.” Gilbert goes on to quote Charles Franklin, “He is coming into this in a better position than he was on the eve of losing the 2020 election, but he has not realigned public opinion to the extent that he is coming in with 55% or 58% approval.” Always reliably dishonest in its political reporting, this JS rendition of the latest Marquette poll is clearly intended to give readers a false impression of what they can conclude from it.
The Marquette Law School Poll is not alone in presenting survey data in a manner that shapes public opinion rather than reflects it. It is common in political polling. Marquette’s poll is often touted as one of the best in the country and is routinely used by dishonest members of the media to advance a political agenda. Last week, the Marquette poll was just one of a handful of newly released polls that were immediately cited by the liberal media nationwide to proclaim that Trump is losing the support of the people. This predictable dance between the pollsters and the media is contributing to the corruption of our political process.
The time has come to reject political polls as a legitimate component of our political process and discard them on the trash heap of irrelevance alongside the so-called legacy American media. Through our total disregard for what they have to say, we have the power to put an end to the dance of the deceivers.
