State polls by age
       |           

Welcome, Guest. Please login or register.
Did you miss your activation email?
March 28, 2024, 05:44:13 AM
News: Election Simulator 2.0 Released. Senate/Gubernatorial maps, proportional electoral votes, and more - Read more

  Talk Elections
  Presidential Elections - Analysis and Discussion
  Presidential Election Process
  Polling (Moderator: muon2)
  State polls by age
« previous next »
Pages: [1]
Author Topic: State polls by age  (Read 4496 times)
SingingAnalyst
mathstatman
YaBB God
*****
Posts: 3,639
United States


Show only this user's posts in this thread
« on: August 17, 2017, 03:52:32 PM »

Are state polls by age reasonably accurate? I've seen maps on this forum of, say, 18-29 voters only or 65+ voters only.

Are they accurate? I would think the subsample sizes would be on the small side. Are there really significant state-to-state variations in the age gradient of vote breakdowns?
Logged
15 Down, 35 To Go
ExtremeRepublican
Moderators
Atlas Icon
*****
Posts: 11,659


Political Matrix
E: 7.35, S: 5.57


Show only this user's posts in this thread
« Reply #1 on: August 17, 2017, 10:40:48 PM »

The exit polls are reasonably accurate, and there are a couple strong trends:

-The age gap is much smaller in states that trended towards Trump and much larger in states that trended towards Clinton.  This is particularly evident in Iowa, Wisconsin, Minnesota, Kentucky, and Maine, which all had some form of "reverse age gaps", with youngs being more Republican than olds.

-The age gap is also much smaller in very white states and larger in states with lots of minorities (particularly lots of Hispanics).  What is likely happening is that, in very white states, the youngest voters are a lot more demographically similar to the oldest ones (it may also be true in states like Alabama, Mississippi, and Arkansas, where the white-black ratio has pretty much remained constant and that have no real trendy areas, although none of those were exit polled).  In states like Arizona and Texas, even if the youngest whites voted exactly like the oldest whites, there would still be a pretty significant age gap due to demographics, so that is important to keep in mind.
Logged
Southern Senator North Carolina Yankee
North Carolina Yankee
Moderators
Atlas Institution
*****
Posts: 54,123
United States


Show only this user's posts in this thread
« Reply #2 on: October 15, 2017, 12:01:05 AM »

The exit polls are reasonably accurate, and there are a couple strong trends:

-The age gap is much smaller in states that trended towards Trump and much larger in states that trended towards Clinton.  This is particularly evident in Iowa, Wisconsin, Minnesota, Kentucky, and Maine, which all had some form of "reverse age gaps", with youngs being more Republican than olds.

-The age gap is also much smaller in very white states and larger in states with lots of minorities (particularly lots of Hispanics).  What is likely happening is that, in very white states, the youngest voters are a lot more demographically similar to the oldest ones (it may also be true in states like Alabama, Mississippi, and Arkansas, where the white-black ratio has pretty much remained constant and that have no real trendy areas, although none of those were exit polled).  In states like Arizona and Texas, even if the youngest whites voted exactly like the oldest whites, there would still be a pretty significant age gap due to demographics, so that is important to keep in mind.

I would love to see accurate data breaking down MS Whites down by age.

MS Whites 65 and up are probably voting close to 90% Republican, while those under 50 are probably in the mid 60's.

I ran some numbers the other day. Just a slight 5% increase in AA turnout, and Whites voting 75% Republican!, would put the Republicans below 50% in MS.
Logged
PoliticalShelter
Jr. Member
***
Posts: 407
Show only this user's posts in this thread
« Reply #3 on: October 16, 2017, 03:43:21 PM »

Here is a 2012 exit poll how on Mississippi age groups voted:



Obviously for 2016 we would have to account for lower AA turnout, but it does show how quickly Mississippi could end up moving with the massive gap between over 65s and under 65s.
Logged
SingingAnalyst
mathstatman
YaBB God
*****
Posts: 3,639
United States


Show only this user's posts in this thread
« Reply #4 on: August 15, 2018, 06:50:25 PM »

Very interesting point, about the age gradient being stronger in states that trended D.

I can see how, for example, WV in 2016, would have little or no difference in age groups-- and other states, like AZ, CA, and FL, would have a huge difference.

I am willing to bet that was not always the case. Georgia, for example, went from being Kennedy's 2nd best state in 1960 to Nixon's 2nd best state in 1972. Clearly, it trended R. But I have a hard time believing that Georgians under 30 were less likely to vote for the candidate of "acid, amnesty, and abortion" than older Georgians in 1972 (a year in which voters under 30 nationwide were far more likely to vote Dem than those 30 or older). Thus, then R trend in Georgia between those two years, if one only includes those of voting age in 1960, was probably even stronger than the raw numbers indicate.
Logged
SingingAnalyst
mathstatman
YaBB God
*****
Posts: 3,639
United States


Show only this user's posts in this thread
« Reply #5 on: August 16, 2018, 12:14:37 PM »

Do we know if the age brackets in exit polls are properly weighted by race? (so each age bracket has the correct racial composition). If they aren't then that can produce some weird results, just because an age cohort is more or less White than it is supposed to be.

Florida's 2016 exit poll (particularly when looking at voters under 50) seems like a good example to suggest that age brackets aren't properly weighted by race, do these numbers make any sense at all otherwise?

18-24: Clinton leads 63-27
25-29: Trump leads 49-43
30-39: Clinton leads 56-38
40-49: Trump leads 50-46
50-64: Trump leads 55-43
65+: Trump leads 57-40
I am almost certain they are not race-weighted.

Fortunately, CNN, in its national sample, broke the white vote down by age. They found relatively little difference in support for Clinton by age, to wit:
18-29: 42D/48R
30-49: 37D/56R
50-64: 34D/63R
65+: 39D/59R

That might be hard to do accurately at the state level, though, except in large states like FL.
Logged
15 Down, 35 To Go
ExtremeRepublican
Moderators
Atlas Icon
*****
Posts: 11,659


Political Matrix
E: 7.35, S: 5.57


Show only this user's posts in this thread
« Reply #6 on: August 16, 2018, 10:30:40 PM »

Do we know if the age brackets in exit polls are properly weighted by race? (so each age bracket has the correct racial composition). If they aren't then that can produce some weird results, just because an age cohort is more or less White than it is supposed to be.

Florida's 2016 exit poll (particularly when looking at voters under 50) seems like a good example to suggest that age brackets aren't properly weighted by race, do these numbers make any sense at all otherwise?

18-24: Clinton leads 63-27
25-29: Trump leads 49-43
30-39: Clinton leads 56-38
40-49: Trump leads 50-46
50-64: Trump leads 55-43
65+: Trump leads 57-40
I am almost certain they are not race-weighted.

Fortunately, CNN, in its national sample, broke the white vote down by age. They found relatively little difference in support for Clinton by age, to wit:
18-29: 42D/48R
30-49: 37D/56R
50-64: 34D/63R
65+: 39D/59R

That might be hard to do accurately at the state level, though, except in large states like FL.

Unfortunately, we don't have this on presidential exit polling, but most polling data I have seen shows a very large gender gap with young white voters too, with young white men being overwhelmingly Republican, but young white women leaning Democrat.  It's true overall that men are more Republican, but it seems to be even more prominent with young voters.  It's also prominent with black voters (regardless of age), as congressional Republicans actually got nearly 20% of black men in 2016 but barely got any black women.
Logged
morgankingsley
Junior Chimp
*****
Posts: 5,018
United States


Show only this user's posts in this thread
« Reply #7 on: August 17, 2018, 04:55:17 AM »

I wonder how Oregon would be
Logged
SingingAnalyst
mathstatman
YaBB God
*****
Posts: 3,639
United States


Show only this user's posts in this thread
« Reply #8 on: August 17, 2018, 07:08:03 AM »

Not sure; I didn't find any age polls for OR.

I do find OR quite interesting politically, how Trump finished in the teens in Multnomah County while Clinton finished in the teens in tiny Lake County; how Trump actually carried Oregon excluding Multnomah by a few thousand votes.

OR was ahead of the curve, perhaps the most environmentally conscious state in the 1970s; throwaway bottles have been banned in OR since 1972. In each election from 1972 through 1988, while Dems were "wandering in the wilderness" at the Presidential level everywhere else, OR was one of the most Dem (and least Republican) states in the West at this level, always giving strong support to minor parties (as is the case today).

For this reason, I'm going to guess that Oregonians aged 50-64 are much more Democratic, relatively speaking, than residents of other states in this age group.

I'm sure you know a lot more about Oregon than I do....
Logged
15 Down, 35 To Go
ExtremeRepublican
Moderators
Atlas Icon
*****
Posts: 11,659


Political Matrix
E: 7.35, S: 5.57


Show only this user's posts in this thread
« Reply #9 on: August 17, 2018, 05:23:08 PM »


2016:

PRES-
18-29: Sample size too small
30-44: Clinton 51-40
45-64: Clinton 50-42
65+: Clinton 46-45

For 18-29's, we can see that the 18-44 vote was Clinton +21 and 18-29's were only 3/8 of that group.  With 30-44's being only Clinton +11, you are probably looking at a massive Clinton win in 18-29's.  But, they might not have gotten enough to be confident in that in their sample, so take that with a grain of salt.

For Senate, all age groups were strongly Democratic, but 45-64's were actually the least Democratic.  That age cohort (particularly the older half of it) may have been liberal at one point, but it switched sides.  Whether today's young voters do so remains to be seen.

For reference, nationally, Clinton won the 18-29 vote by 19, but House Republicans were more competitive with it, only losing by 10.  Trump won white 18-29's by 4, but House Republicans won them by 10.  Possibly underreported was that House Republicans did better with young minorities than with old minorities, winning 15% of black 18-29's (vs. 10% of the overall black vote) and 31% of Hispanic 18-29's (in line with Hispanics under 65, but 65+ Hispanics are far more Democrat).
Logged
morgankingsley
Junior Chimp
*****
Posts: 5,018
United States


Show only this user's posts in this thread
« Reply #10 on: August 18, 2018, 01:57:01 AM »


2016:

PRES-
18-29: Sample size too small
30-44: Clinton 51-40
45-64: Clinton 50-42
65+: Clinton 46-45

For 18-29's, we can see that the 18-44 vote was Clinton +21 and 18-29's were only 3/8 of that group.  With 30-44's being only Clinton +11, you are probably looking at a massive Clinton win in 18-29's.  But, they might not have gotten enough to be confident in that in their sample, so take that with a grain of salt.

For Senate, all age groups were strongly Democratic, but 45-64's were actually the least Democratic.  That age cohort (particularly the older half of it) may have been liberal at one point, but it switched sides.  Whether today's young voters do so remains to be seen.

For reference, nationally, Clinton won the 18-29 vote by 19, but House Republicans were more competitive with it, only losing by 10.  Trump won white 18-29's by 4, but House Republicans won them by 10.  Possibly underreported was that House Republicans did better with young minorities than with old minorities, winning 15% of black 18-29's (vs. 10% of the overall black vote) and 31% of Hispanic 18-29's (in line with Hispanics under 65, but 65+ Hispanics are far more Democrat).

I still can't believe more people did write ins on Oregon than votes for stein
Logged
Pages: [1]  
« previous next »
Jump to:  


Login with username, password and session length

Terms of Service - DMCA Agent and Policy - Privacy Policy and Cookies

Powered by SMF 1.1.21 | SMF © 2015, Simple Machines

Page created in 0.042 seconds with 14 queries.