Unrepresentative Policy of the Week: Is Polling Accurate?
I think we can all agree that the election was a lot closer than we had expected according to the polls. This isn't the first time that the polls have been off, we all vividly remember their inaccuracy in swing states in 2016, resulting in an unexpected Donald Trump victory. Despite efforts to correct for their polling methodology, it appears the pollsters were just as an accurate, if not more so, in many of the states which they polled extensively.
Joe Biden was predicted to win by large margins in many states which ended up being won by extremely narrow margins. Luckily the overall predicted margin of victory was so overwhelming for Biden that he still managed to win. Yet we can’t ignore the inaccuracy of this polling when so many are expressing skepticism. Some people are going so far as to say that we cannot trust polling anymore at all and that we should simply ignore the polls. The is certainly much to be skeptical about, but this total rejection of polling would be throwing out the proverbial baby with the bathwater.
From both a tactical and strategic standpoint, it would be mistake, to reject polling, and thankfully, this view is actually a misrepresentation of the larger context in which polls are taking place. At DSOT we are very interested in the importance and accuracy of polls, because a large part of our thesis is that polling on issues can give us insight into what the American people want, and that insight shows us that Americans are in fact overwhelmingly left wing in their policies, and want many of the same things that traditional progressives advocate for. Throwing out polls especially if they can still provide us with useful information will only aid those who want to perpetuate minority rule and unpopular policy, because without the ability to objectively analyze opinions through polling and data analysis, there's no ability to determine whether or not any actions by a government are unpopular or unwarranted. After all, who can say what the people think, if polling means nothing.
So today we are going to analyze whether or not the polling was accurate and if we can still trust it. In this case, we should draw a distinction between polling about candidates, and issue polling. It is hard to argue that polling around candidates has problems. There seems to be for one reason or another something that pollsters are not understanding about the voting population when it comes to candidate preferences. In their modelling when they try to extrapolate the data that they gather from conversations with people clearly support for Republican candidates is somehow being underrepresented. However, the story is completely different when we get to issue polling. In fact, in 2020, issue polling was dead on accurate.
How can we save this with confidence? The key is that we can look at ballot initiatives across the country and see whether or not the polling around those initiatives in the lead up to the election was accurate or not to the actual voting. For example:
Polling Support: 64%
Voting: 61%
Polling Support: 64-67%
Voting: 66%
We can see that across the country these polls were extremely accurate in New Jersey, in Florida, in Oregon, in California and elsewhere. Time after time, across the country, when we remove partisanship, when we remove the confounding issue of political party and culture, we find that people are much more honest and accurate, and we have a much better understanding of how they actually want to vote. Thus, issue polling still seems to give us a valuable insight into the minds of the American people. While we should be skeptical about polling predictions for candidates until these polling institutions demonstrate a track record of having accurate data, we at DSOT think it’s appropriate to still be confident of support for our super majority policy issues.
This is very important because it shapes what we can expect the Democratic party to do looking forward over the next few years issue-wise. To face the major challenges currently underway and heading down the pipeline, we are going to need confidence to pursue aggressive policies and if we cannot trust polling we will have neither the confidence nor the authority to push the Democrats forward. So remember, that policy polling can be accurate, is accurate, and is something that we need to demand and push our democratic representatives into using that information to pass popular policy and make sure that we get a government that represents all of us.
Join us every Tuesday, in growing solidarity and power until we take back our democracy from amoral special interests, and greedy politicians: Don't Shop on Tuesdays!
Comments