Imprecise And Infrequent Questioning

Imagine sitting at a desk, wringing your sweaty hands nervously as the teacher waltzes around the room, handing down divine judgment. Well, not biblical, but pretty close-your grade depends on the final term paper.

A hearty slap of paper meeting hardwood makes you almost jump out of your chair. Your instructor has delivered the goods, but without batting an eye, has moved on to the next victim. Your sweaty hands emerge from their sub-desk cave.

You reach out.

You hesitate.

You reach out again, flipping the paper over.

‘85’ is written in bold red pen across the top. You look for comments throughout the paper-there are none. You flip to the back, expecting some analysis. There isn’t any.

You furrow your brow and walk up to the teacher, your nerves replaced by confusion and a slice of irritation.

“Why did I get an 85?” you ask, politely yet firmly.

“It feels like an 85.” Your teacher responds, with a blank look on their face.

“What did I do well? What did I do poorly?” you say, confusion and irritation rising.

“I don’t know.” Your teacher shrugs. “Why don’t you look at your other papers and compare them to this one?”

You return to your desk, spreading out previous papers as if a general on campaign, looking for what should’ve been made explicit.

Do you think that the teacher’s grading is appropriate?

Of course not. Thankfully, most don’t grade that way.

However, the above story is what I often feel when I look up polls designed to get a politician’s approval rating.

Let’s take industry standard Gallup News, who began tracking presidential approval in 1938 and Congressional approval in 1974.

As far as I can tell, Gallup asks poll respondents these questions:

“Do you approve or disapprove of the way Donald Trump (or other president) is handling his job as president?”

“Do you approve or disapprove of the way Congress is handling its job?”

That’s it. A simple yes-no question. While Gallup published “The Metrics That Matter for Election 2020”, key indicators that have proven to be strong predicators of presidential elections, including approval rating and analyzed how President Trump’s approval and disapproval changed among key demographic groups during his presidency, their initial question strikes me like what my hypothetical teacher wanted you to do.

You wanted clear reasons for the paper’s grade. Although you may figure out why you got an 85, comparing past papers is inefficient. You may assume incorrectly-looking at writing style, for instance, rather than the argument.

It is much better for the teacher to explain their position with clear examples.

Take the same reasoning to Gallup’s polling. President Trump’s lower approval ratings are most likely correlated with his defeat, however, so much else happened that a simple look may miss the mark. I don’t want to assume what other voters thought. An explicit, direct question helps me avoid making false conclusions. I imagine campaigns and politicians feel the same!

Gallup should follow-up with something like “Why do you approve or disapprove of President Trump’s/Congress’s job performance?” every couple of weeks or at least monthly, along with asking “If you have changed your opinion of approval or disapproval, why?”

Perhaps some outfits could try allowing responses of a few sentences and trusting the pollsters to categorize/group them, but that seems to be quite tricky when polling 1000 or more people, so allow the respondent to choose between a few options such as policy, demeanor or scandals (allegations such as corruption or extra-martial affairs, for instance). Pollsters could include ‘other’ which would allow respondents, if confident enough, could provide their own reason in their own words.

Some of the polls that RealClearPolitics uses to average presidential job approval ask more precise questions. USA Today/Suffolk University asks which policy area was Mr. Trump’s greatest achievement and failure, how will history assess him and who respondents voted for. Reuters/Ipsos breaks down Mr. Trump’s issue approval into four categories as well as asking what issue was most important for all voters.

Yet, they don’t have a simple “Why do you approve or disapprove?” question. These polls are just outside the bullseye of the truth, but not quite there.

Congress has few tracking polls, and its approval rating is frequently tucked into a larger poll, usually about presidential approval. RealClearPolitics has only four in its average. Not only does Congress need a why question like the president, but it also needs more frequent polling-at least monthly. It’s important to note that Congress rarely has gotten above 50% approval in Gallup’s 46 years of tracking them, usually attaining 30% or lower.

At the state level, this type of polling is even less frequent. For my home state of Wisconsin, Marquette Law School’s poll, conducted about every 4–6 weeks, does ask about state politicians such as Governor Tony Evers, Assembly Speaker Robyn Vos and Senate Majority Leader Scott. L Fitzgerald, who was elected to Congress this past November.

Even for Marquette, a prestigious law school based in the state’s most populous city, it asked about Mr. Vos and Mr. Fitzgerald’s approval ratings three times in 2019. State level politicians need more polling and better questioning to be able to better respond to and inform their citizens.

Marquette’s questions about approval and disapproval are worded similarly to their national level counterparts, again demonstrating a need for the elusive “Why do you….?” question.

The general public has been less trusting of government since the Vietnam War and Watergate. I firmly believe that more frequent polling, especially for Congress and state-level politicians and institutions, along with more “Why do you approve/disapprove….?” type questions will help make politicians more accountable to their voters and the voters trusting them more, producing a better political system. For those who don’t contact their representative, which is the vast majority of voters, polling is one of the few ways they indicate their preferences besides voting.

Yes, I know the polls have been problematic lately. As statistician Nate Silver of FiveThirtyEight pointed out, many polls had a rough 2020 along with everyone else, but not as bad as initially thought. I encourage you to read the piece in-full.

Furthermore, what’s the alternative? To not have them at all? Having less polling means politicians have to rely on historical trends, vocal supporters, active citizens and the good ol’ gut feeling.

In our (democratic) republic, our politicians should champion our needs. If they don’t know what they are, how can champion them? If we are unsatisfied with their performance, how do they improve if they don’t know?

They can’t. Much like our suffering student above, they require consistent, in-depth feedback instead of simple yes-no questions or letter grades/percentages.

With more precise questioning and more frequent polling, they just might do their jobs.

And Mr. Smith won’t have to go to Washington, after all.