Sputnik: What's behind the stark difference in estimations of Joe Biden's approval rating between Gallup and Democracy Institute in your opinion?
Patrick Basham: Several media- and university-sponsored polls are showing Biden’s approval rating in a different light to our polling. Most of the differences in current poll numbers can be explained in methodological terms.
Some of these polls employ an “adult” voter screen, which is almost no screen at all. These are basically surveys of American consumers (citizens and non-citizens alike), not voters. Such surveys always skew heavily Democratic. Other polls are employing a “registered voter” screen, which in a midterm election overstates the actual ballot-casting electorate by approximately 30 percent. Registered voter polls also skew Democratic.
The remaining polls are using a “likely voter” screen, which is the method used by the Democracy Institute. This is normally a far more accurate method. However, media and college likely voter polls still overstate Biden’s approval rating for two reasons.
The first is that they tend, largely for efficiency (that is, cost) reasons, to allow unrepresentative Republicans to dilute the Republican sample. Moderately conservative, establishment-supporting, urban and suburban Republicans are far more reachable (that is, willing to respond), from a polling standpoint, than are populist conservative, Trump-voting, exurban, small town, and rural Republicans. Not only are these polls not finding the now-legendary Shy Trump voters, but they are still not making a concerted effort to look for them. In contrast, Democracy Institute polling continues to make a particular effort to identify and measure these voters.
So, in practice, many pollsters (Democrat and Republican) will fill their required quota of “Republican” respondents, and move on to Democrats and independents, without filtering out the establishment Republicans (a sizeable chunk of whom now vote Democrat) from the Republican quota, and without divining the populist Republicans, who are today’s most anti-Biden Americans. Consequently, these polls lower Biden’s disapproval total in a way that is largely accidental, yet crucial in misrepresenting the actual state of play.
The second reason applies to most of the higher Biden approval polls regardless of their respective voter screen. They simply overweight, by accident or by design, their samples in favor of Democrats.
Such samples automatically bump Biden’s approval rating up by anything from the low to the high single digits. This year, our Democracy Institute polls use a sample that is only very slightly weighted in the Democrats favor. This is the sample weighting we also used in 2020, and that was the eventual partisan breakdown of the 2020 electorate.
Sputnik: In the past couple of months, Biden cancelled student loans, denied that the US has regressed into recession, inked the Inflation Reduction Act, and branded MAGA Americans as semi-fascists. Meanwhile, Biden's DoJ dispatched 30 FBI agents to raid Trump's house. Could one assume that these actions are part of the Dems' effort to improve their election odds at the eleventh hour? Could it work, judging from the trend exposed in the recent Democracy Institute's poll?
Patrick Basham: Each one of the actions initiated and executed during the summer was intended to improve the Democrats chances in the November election. Although in politics any outcome is possible, there is no hard evidence to suggest that any of these moves are working, or eventually will work to the Democrats advantage.
Each action is popular with the left-wing, anti-Trump base of the Democratic party, which does not want to repay its student debt, wants the government to continue to print, borrow, and spend money, and wants to see Donald Trump prevented from seeking the presidency in 2024. But, each of those actions is unpopular with a majority of [other] voters. While Republicans predictably oppose these actions by massive margins, it is also the case that most independents and moderates dislike them, too.
No political issue, economic news, or cultural topic is working in the Democrats’ favor. Quite the opposite, in fact. Democratic candidates in politically competitive and swing districts continue to run away from their party label and especially from the White House.
The bottom line is that the Democrats will continue to throw political and policy mud at the wall in the hope that something, anything, eventually sticks to that wall and, by doing so, minimizes the Republicans inherent advantages in the run-in to the election. The proof that nothing has yet stuck to the wall for the Democrats is that every few days they alter the focus of the election narrative as they rapidly move from one issue to another.
Seven weeks is still enough time for the Democrats to hit on a winning message, but they have yet to find one.
Sputnik: Could one assume that some surveys and mainstream media enthusiasm over Biden and Dems' successes are misleading? Could one suspect some sort of manipulation of public opinion in these reports?
Patrick Basham: Many media and academic polls have poor track records over the past several elections.
These polls continue to be plagued by the same methodological problems regarding partisan weighting and voter screening detailed in my answer to an earlier question.
As these easily identifiable problems have not been addressed, the only logical conclusion is that the sponsors of these respective polls may be more interested in furthering pro-Democrat, and more recently pro-Biden, narratives than ensuring their polls’ accuracy.