Economist polls.

Quote:

---

Originally Posted by GGT

Right, I'm just saying that expectations of all these polls, models, data, and predictive powers should be put in perspective. As you say, some 'conventional wisdom' can be false because it's based on outdated premise. iirc, you were wrong about the youth vote in 2008, claiming that students aren't likely voters, and dismissed Ron Paul's growing influence in the GOP. ;)

I'm 99.99% certain that pollsters and demographers will continue to do evermore polls/surveys, analysts will continue to make their predictions, media will treat them as fact or spin them as political, and campaigns will spend millions of dollars for the data-mining and adverts. I'm 100% skeptical that means better candidates, more choices for voters, or a well-informed electorate.

---

I think you'll find that the people who do these things don't claim to know more than they do. If pundits take their research out of context, it's the pundits' fault.

Agreed. But it's nearly impossible to predict which events will happen, and when. Even extended-time weather forecasts have limitations, using state-of-the art science and predictive computer models. That said, in-person voter turn-out is always lower during inclement weather, even simple rainy days. That's one good reason for early voting and absentee/mail-in voting -- to extend the voting time.

States that don't have early voting are more vulnerable to weather events, disasters, and/or power outages that happen right before, or ON, election days.

Quote:

---

Originally Posted by GGT

Agreed. But it's nearly impossible to predict which events will happen, and when. Even extended-time weather forecasts have limitations, using state-of-the art science and predictive computer models. That said, in-person voter turn-out is always lower during inclement weather, even simple rainy days. That's one good reason for early voting and absentee/mail-in voting -- to extend the voting time.

States that don't have early voting are more vulnerable to weather events, disasters, and/or power outages that happen right before, or ON, election days.

---

Virtually no one in poli sci tries to predict rare events; they try to figure out which factors make those events more likely. That knowledge could then be used to make necessary precautions. You're never going to be able to predict specific events because they happen for systematic (i.e. predictable) and idiosyncratic factors (particularly ones caused by free will).

We're saying basically the same thing, in different ways. Those "necessary precautions" (about voting) include voter registration efforts, expanding early voting, logistics/transportation (of groups, and during bad weather) for in-person voting, using social media, all sorts of things.

But it doesn't take a "political scientist" or polls/surveys to do those things. Just motivated and organized team work.