Dependable Erection

Wednesday, May 07, 2008

Odd numbers

CNN's got some exit polls up from NC. In some cases they're a bit confusing, i think. Multiple variations of the same question (see the first page with the Church attendance question, for example) or the "Vote for President in November" question, which i can almost wrap my head around before it slips away from my comprehension.

But these numbers on page 4 jumped out at me:
Vote by Income

Clinton Obama No Preference

Less Than $100,000 (80%)

41% 57% 1%

$100,000 or More (20%)

42% 56% 1%

Vote by Income

Clinton Obama No Preference

Less Than $50,000 (48%)

37% 60% 2%

$50,000 or More (52%)

45% 54% 1%

Are they really saying that 52% of North Carolinians have an income above $50k? Even if that's household income, which i suspect it is, that's a pretty high number, no? And 20% above $100k? The 2000 census put the median household income in NC at just under $41k. I haven't read anywhere that household incomes have gone up nearly 25% in the past 8 years, but maybe i'm just completely out of touch. Wouldn't be the first time that's happened.

Labels: ,

posted by Barry at 9:47 AM Digg this

12 Comments:

  • Folks with higher incomes have more time to stop and participate in exit polls?

    By Blogger toastie, at 11:07 AM  

  • Take a look at the next set of numbers, you'll see it breaks out the incomes even better. I suspect that in the 50K column, it may actually be a breakdown of the less than 100K; meaning that of the 80% of North Carolinians who make less than 100K a year, 52% are between 50K and 100K, while 48% make less than 50K. A bit confusing I agree, which is why I like the full breakdown better.

    By Blogger capt. shitknife, at 11:10 AM  

  • Maybe the median income of NC as a whole is not the same as the median income of NC voters?

    By Blogger Unknown, at 11:19 AM  

  • I think bjv and toastie have it - lower SES groups are generally underrepresented in elections. Transportation and scheduling difficulties are big issues that keep these folks from making it to the polls.

    Plus, I'd imagine that a lot of people with lower incomes feel like the system is already failing them, and that their vote is meaningless anyway.

    By Blogger JeremyT, at 11:47 AM  

  • just looking at the numbers and the turnout around the state, i have a difficult time believing that yesterday's turnout skewed toward the higher end of the income spectrum in North Carolina. Certainly in my precinct it didn't.

    Probably says more about where CNN did their exit polling than anything else.

    By Blogger Barry, at 12:33 PM  

  • Capt. shitknife - i'm looking at those numbers, and that's not the way i read it.

    $50 - 75k - 18%
    $75 - 100k - 14%
    $100 - 150k - 11%
    $150 - 200k - 5%
    $200k + - 4%

    add those up, and you get 52% of the voting pool from yesterday. Which would put the median income of yesterdays voters (and again, i'm assuming they mean household income, even though they didn't state so directly) somewhere around $53k.

    as noted above, i have a hard time believing that yesterday's turnout skewed that heavily in favor of higher incomes, or that median income in NC has risen about 25% over the past 8 years.

    more likely, CNN's numbers on this are bad.

    By Blogger Barry, at 12:38 PM  

  • The exit polls must have been held in areas with high incomes. The non-Triangle, non-Charlotte areas of NC are not near this income level.

    By Blogger Valerie at We Love Durham, at 1:52 PM  

  • I missed this thread and posted something similar over here.

    In any case, I don't think it's just an SES skew. I think it's also an age skew. Median household incomes are necessarily lower for the 18-35 set, not just because they're in entry-level positions, but also because they're much more likely to be single person households, so lack the dual income effect. Even with young turnout high this year, check out the vote by age list at CNN, and compare to census numbers. The median age in NC was 36.6 in 2006. The median age in their exit poll was over 40.

    By Blogger Unknown, at 2:47 PM  

  • Let me try again to do math. The median age for CNN's exit poll was over 50, not just over 40. 55% of voters were over the age of 50.

    That's your explanation right there.

    By Blogger Unknown, at 2:49 PM  

  • So where the hell was CNN exit polling that they came up with a median age of 50? i simply don't believe that that 50 represents the median age of voters in yesterday's Democratic party primary.

    and what does that do for the rest of their analysis? i don't think you can pull much meaning out of their data, to be honest.

    By Blogger Barry, at 2:55 PM  

  • Time of day of the exit polling may have something to do with it too: I'd suspect higher ages during the midmorning/midafternoon time frames, and lower after say 3 or 4 pm.

    By Blogger Ted in Durham, at 3:18 PM  

  • "i simply don't believe that that 50 represents the median age of voters in yesterday's Democratic party primary."

    I do. Why is that a hard number to believe? Turnout was up in the young adult vote Tuesday, but it was also up among all other groups, and the young vote is typically far below all other groups, with senior citizens typically far above the average. Add to that, the echo boom still isn't fully of voting age yet, which means that the population is bulked up in the baby boom, with the Gen X trough sitting in the 35-50 set. And even assuming an even age distribution and even turnout across a range of ages from 18 to 92, the midpoint would be 60. This didn't occur to me yesterday, but the median population age includes everyone from newborns to centigenarians, but the voting population starts at 18.

    With all of this, I still say $52k isn't that odd a number as a median household income for voters, considering it's almost exactly the same as the median family income in 2006.

    By Blogger Unknown, at 12:20 PM  

Post a Comment

<< Home