Was mono-lingual polling an issue?

[Kurt_and_Hunter]

I've read a few things lately about how the most notable Democratic under-polling effects were in states with large latino populations.

It seems that pollsters conduct the surveys in English and then try adjust for Spanish speaking households by multiplying the latino-english numbers, assuming that all hispanic sub-groups (Like English speaking or not) have the same attitudes.

It's interesting stuff.

http://www.themonkeycage.org/2010/11/are_polls_more_likely_to_be_of.html

http://www.themonkeycage.org/2010/11/more_on_bilingual_polling.html

[BzaDem]

1. That's an interesting question. On the one hand, English proficiency is required to get citizenship

so one would suspect most naturalized citizens could speak English.

On the other hand, it might be that people who speak another language as their first language hang up the phone if someone (particularly a robo-pollster) calls in English.

Nate Silver looked at this issue recently. This could very well be a factor, and we might have more evidence for it in future elections. The fact that pollsters keep totally missing Nevada (in 2010 and 2008) certainly indicates that something is going on.

http://fivethirtyeight.blogs.nytimes.com/2010/11/03/did-polls-underestimate-democrats-latino-vote/