Alex Hilke writes:
I’m fascinated by the research that FiveThirtyEight did to construct a model of future voting behavior: http://www.nytimes.com/interactive/2013/04/30/us/politics/presidential-math-demographics-and-immigration-reform.html. I have two questions though that I hope you could answer for me. Following the Census’ recent report of stagnant minor decline in White population growth have the projections been updated to reflect the new assumption? And does the model take into account internal generational replacement and change in political attitudes? For instance the more conservative White seniors will die off in large numbers and be replaced by Progressive young Whites as well as the already progressive young Blacks, Latinos, Asians, Natives and Multiethnics. To substantiate this I cite CNN’s recent exit poll cross tabs from the last Presidential Election when “Other” is factored out.
18-29: 46% D, 54% R
39-44: 39%D, 61% R
45-64: 38% D, 62% R
65-100: 39%D, 61% R
Recent studies I’ve seen cited by the California Young Democrats claim political voter behavior to turnout every time for one party after three straight elections. Using this statistical assumption and chronological death statistics the model can further be improved for accuracy.
My response: I don’t know if the model underlying that New York Times visualization accounts for changes in the relative sizes of age cohorts. Speaking more generally, our research has found that a popular president helps his party’s long-term support among the cohort of people who are young during that presidency. (Sorry for that convoluted sentence; I’m sure it could be rephrased more cleanly somehow.) That’s all about support for the Democrats or Republicans; another question is what policies they will follow.