Research Desk: What if all income groups turned out to vote equally?
By Dylan Matthews
Following up from the desk's response on Friday: If turnout among different income levels was homogeneous, would it be expected for Democrats to really dominate nationally?
This is a tricky point, but cstar's right that we can come to an estimate. The Census Bureau keeps data breaking down the number of households by yearly income (under $2,500, $2,500-$7,500, etc.) and CNN has exit poll data from the 2008 presidential election broken down on the same lines. CNN uses fewer categories, so I merged the ones in the census data and put together this table:
"% Population" gives the percentage of total households falling into a category, while "% Voters" gives the percent of CNN exit poll respondents in that category. The Obama and McCain percentage show the support levels for each category for each candidate. We can use this data, then, to figure out what the vote percentages would be if "% Voters" were equal to "% Population". To be more precise, I used the Census Bureau's exact numbers of households in each category and multiplied each by vote percentages, then added up the totals for Obama, McCain and "other". Dividing any one of those totals by the sum of the three produces the percentage for that candidate.This method assumes 100 percent turnout, so it cannot provide an estimate of the total number of votes each candidate would receive, but it does accurately reflect what the percentage outcome would be if each income level voted in proportion to its size.
I found that, with even turnout across income levels, Obama would have received 55.2 percent of the vote and McCain 42.7 percent. The actual results (PDF) were 52.9 percent for Obama and 45.7 percent for McCain. Obama would gain 2.3 points, and McCain would lose 3, with other candidates picking up the rest. These numbers are not huge, and given that exit polls are not always reliable and this is some very back-of-the-envelope math, I would avoid concluding that even turnout would help Democrats.
However, the data do show that the change in influence from even voting numbers would not be from the rich to the middle class, but from the middle class to the poor. Voters making less than $30,000 would increase their say significantly, while those making between $50,000 and $100,000 would see theirs drop. While Obama split the middle class evenly with McCain, he pulled way ahead among poorer voters, which is reflected in the outcome estimate at which I arrived.