Comment by pmyteh | Hacker News Reader

Comment by pmyteh | Hacker News Reader

Parent Full thread

These panel surveys are reasonably good. The better pollsters (like Ipsos MORI) spend a decent amount of effort recruiting a broad sample of people for their panels, and results are statistically adjusted for non-response bias etc. That said, the usual problem with political polling (that you can only poll people willing to answer pollsters' questions) applies extra here, as all respondents have agreed to be part of a panel and be repeatedly surveyed. As most panels are done through the Internet, it's at least possible that the participants are more terminally online than the average American.

That said, I don't think the results are crazy, on their face (though I don't myself expect a civil war). There are a number of groups openly proposing political violence, increased partisan polarisation, and a breakdown in the normal political processes (Congress has been essentially non-functional for a while, the Supreme Court have taken a decidedly activist turn, state legislatures are actively discussing subverting the next Presidential election, and trust in politicians of all parties is at historical lows). It's not clear how these will be resolved, and a further increase in political violence is certainly a possibility. Will that amount to civil war? Not close, in my view, but we live in unusual political times.

I don’t personally expect a US civil war in the next decade, but I think we are much closer to that point than is safe. What’s really concerning is IMO the odds have been getting worse over time and nobody seems to be trying to calm things down.

Gerrymandering for example is extremely beneficial for elites in both political parties, but increasingly corrodes peoples faith in the system. It’s the perfect example of people actively harming the United States, and never spoken about in those terms.

It's also an unfortunate consequence of game theory. If one party Gerrymanders and the other does not, that party will solidify their hold over the House tighter and tighter. They'll also solidify their grip on the states (who run the elections, and who set the rules on who can vote and when they can vote). Trying to play nice and believe in the system is a recipe for being eradicated - it's exactly what's happened to the Democrats for the last 20 years.

At the same time there are ways in game theory too that can give tools to almost avoid all gerrymandering even where players are all against each other.

In order to estimate how close a country is to a civil war, the people they should poll is those involved in the military. An attempt to take power through violence without military backing will look something like the 6 January 2021, which is to say a bunch of crazy people dressed up as Indians and cowboys who thought that storming a building means controlling a country.

> which is to say a bunch of crazy people dressed up as Indians and cowboys who thought that storming a building means controlling a country.

I don't know why people are still pushing this narrative when the hearings have made it perfectly clear at this point that there was orchestration and support among the government, Capitol police and likely even the Secret Service. The "crazy people dressed up as Indians and cowboys" were useful idiots meant to make the whole thing seem like a clumsy farce.

>I don't know why people are still pushing this narrative when the hearings have made it perfectly clear at this point that there was orchestration and support among the government, Capitol police and likely even the Secret Service.

Nobody takes the hearings seriously since they're political theatre. But if what you say is true, I definitely look forward to those people in government and the secret service to be charged with treason or sedition. It's been 18 months, so when should we expect that to happen?

If we want to see what a insurrection look like, let take a look at the 2016 Turkish coup d'état attempt. There were tanks taking control of airports, people abducted by military splinter groups, buildings bombed by military helicopters, town squared occupied by military troops, TV stations taken, and attacks on government positions occurred in multiple cities. Multiple generals was involved.

We can also observe the reaction (overreaction) that turkey had to the coup, and how the world perceived those actions.

Now compare that to what ever secret conspiracy that led to a bunch of useful idiots making a clumsy farce. How realistic is it that those events will be the start of an US civil war?

>and results are statistically adjusted for non-response bias etc.

...Can you explain this?

My brain translates that to "We fudge the numbers because we know dome portion of them are being deceptive."

I'm pretty sure that isn't what that's supposed to mean is it?

That sounds like they have a model of population demographics and try to “correct” their sample for various biases that effect likelihood of reaching and getting a response from people by reweighting the actual sample to match the expected demographics.

If actual views on the target issues are predicted well by demographics and independent of the response biases beyond their correlation with the demographics, this should work well.

If they aren't, though, this potentially magnifies the impact on the overall result of subsamples that are potentially most unreliable because of small size.

Reply on news.ycombinator.com