3) 27 May 2022: Wrestling with measuring attitudes toward cooperation

An update:

Still trying to refine the method of measuring baseline attitudes toward cooperation with opponents.

This is really difficult. We’ve made several series of iterations to our approach toward measurement and are actually making progress, thankfully. But we’re not there yet. And as we dig more deeply into this challenge, we’re uncovering new obstacles and influences we had not earlier considered. All of this is REALLY good, in one respect, because it means we are learning a lot, and quickly. And these are things that need to be understood if the problems we face are to be fixed. So that’s good. But it’s frustrating that we are not getting to the goal more quickly.

The biggest question we’re wrestling with at the moment is this:

• Our political climate is one in which we see deep, passionate division, unwillingness to work with each other, gratuitous damonization of opponents, all resulting in, essentially, gridlock. Yet …
• Our surveys repeatedly, across all geographies and ideologies, no matter how we word them, show a public overwhelmingly in support of working with opponents to reach some agreeable compromise.

We’ve been trying to reconcile these two conflicting observations.

There are a long list of potential explanations for why our surveys might pick up sentiment that is different from what one observes in action in actual politics. Having thought this through extensively and tested several methods of confirming or eliminating different explanations, we’ve zeroed in on the following. The current perspective is that the answer is relatively complex, because there are likely multiple different influences, all contributing to the variance we see. Although, I suppose that could be a good thing, in that rather than the problem being primarily one, seemingly invincible force, it might be overcome by incrementally tackling multiple weaker influences, one at a time.

… Clearly I’m an optimist.

Here are the five factors we currently think collectively influence the problem.

• Abstract beliefs vs specific challenges.
  • When we ask someone (to simplify) “what do you think about working with others?” they may answer that question differently than if you ask them, “how flexible are you on the issue of abortion rights?” We can certainly measure this difference; that is an iteration we are about to launch. But if this turns out to be a major factor — that people believe in compromise only in abstract — then the challenge of how you change that becomes tricky.
• Stated intent vs actual behavior.
  • In the first effect mentioned above, a survey should be able to get an individual to admit that they believe differently about something in abstract than they do when it becomes more specific and real. But there may be a subtly different effect, which is that people may truly believe they think something, but when they have to actually do it — to act upon it — they don’t actually behave as they claimed they would. Sort of, “would you jump in front of a bus to push your friend out of harm’s way?” This may be especially true when the repercussions of doing some negative thing one disavowed would not be immediately felt. For instance, on the topic of compromise, if one is not compromising with one’s wife in an active dispute, the negative consequences of that are immediately felt. So one may advocate for compromising behavior. But in an online chat, if one is not compromising with a political opponent that one does not know, interacts with only briefly in a chat room and never will see again, then there are no immediately felt implications of that. So one may not that cooperative behavior is important, but one may not actually act upon that in every moment.
• Disconnect between diverse public sentiment and binary electoral choices.
  • If we poll 1,000 people about their attitudes toward compromise, it is likely we will see a broad distribution of responses, across a very granular scale. For instance, if we scored each person on a scale of 1-inflexible to 5-compromising, we may find that Person A scores 2.35, while Person B scores 4.21. So, in this scenario, we have a public with a wide range of views on compromise. But then they have to vote for a candidate to represent them. And most often there are only two choices. So that range of views suddenly becomes a binary of only two positions on the compromise scale. Said differently I may be in favor of compromise, but may not have a candidate I can vote for who represents my view. Or said more differently, candidates may be more extreme than voters.
  • This is likely exacerbated by the facts that:
    • Many candidates feel they need to celebrate controversial, polarizing positions in order to energize their voters.
    • Extreme voters are of disproportionate value to candidates, because they tend to be more motivated and willing to be involved in campaigns, and a candidate needs to appear amenable to their extreme views in order to win their support.
• Extremists disproportionately drive political discourse
  • Further to the point immediately above, it may be that our surveys are correct and the broad electorate is indeed more compromising than aggregate political behavior seems to suggest, but that it is the extremes that disproportionately drive political activity.
• Skews in our survey audience
  • There are also almost certainly skews in the audience that (a) is online and (b) self-selects to respond to our surveys. This is ultimately measurable if we need to, but it doesn’t on initial thought appear to be an insurmountable problem.

OK, so we are now devising methods of:

1. Identifying which of the influences above are the most significant drivers (if any), and then
2. Brainstorming what that tells us about how we effectively measure and influence attitudes toward cooperation.

Enough for today. More info as we learn it!

• This topic was modified 1 year, 6 months ago by Hank.
• This topic was modified 1 year, 6 months ago by Hank.
• This topic was modified 1 year, 6 months ago by Hank.
• This topic was modified 1 year, 6 months ago by Hank.
• This topic was modified 1 year, 6 months ago by Hank.
• This topic was modified 1 year, 6 months ago by Hank.
• This topic was modified 1 year, 6 months ago by Hank.
• This topic was modified 1 year, 6 months ago by Hank.
• This topic was modified 1 year, 6 months ago by Hank.