Boston Rambles

Boston Rambles

A Rambler Walks and Talks About the Hub of the Universe

The Red White and Blue Divide

In my previous entry I described an interesting observation about the demographics and voting patterns in Pennsylvania. I noticed that the sum of two numbers, the Percentage of residents with a Bachelor’s Degree age 25+ and the Percentage of Non-White residents, seemed to correlate with the likelihood of voting for Hillary Clinton or Donald Trump. Specifically, the higher this number, the higher the likelihood of voting for Clinton. Below is a graph showing all 67 counties. On the X-axis is the sum of the % Bachelor’s and % Non-White population, and on the y-Axis is the percentage vote Hillary Clinton received in the county.

Percentage of Bachelor’s Degree among residents 25 years or older added to the percentage of non-White residents graphed against the percentage voters voting for Clinton in 2016 Election by county. Sources: American Community Survey, Pennsylvania Department of State.

I do not even need to draw a trend line on the graph to see that it is obvious that this number correlates very well with the Clinton vote.

Some observations.

1. Every county with a College Degree + Non-White number (CNW#) that exceeds the figure for Pennsylvania (College Degree=28.1; Non-White=22.6: CNW=50.7) was a county in which Clinton won. (10 counties). The only county Clinton won that had a CNW below 50 was Lackawanna County (CNW#=38.7, home of Scranton) which, as I discussed in the previous article, is Joe Biden’s hometown and was a place in which he campaigned frequently). In fact, Lackawanna was the only county with a CNW below 40 to give the Democratic candidate more than 40% of the vote!

2. Six counties had a CNW of 40-50; three (Erie, Northampton, Berks) gave Clinton more than 40% of the vote and three (Cumberland, Lancaster, Pike) gave her between 35-40% of the vote. All were won by Trump, but two of the first three were won by Obama in 2012 and he lost Berks County by less than 1,000 votes. Every other county has a CNW below 40 and (except for Scranton), every county gave Clinton less than 40% of the vote.

3. The population of the ten counties with CNW>50  is 6,285,010 (49.1%) of the population of Pennsylvania. The population of the six counties with CNW between 40-50 is 1,833,040 (14.3%) of the state population. These counties will be the battleground in future elections, particularly Berks and Lancaster County, which are changing rapidly as the influence of Philadelphia is increasingly apparent.

4. Pennsylvania has a CNW of 50.7 yet Trump won the popular vote by about 67000 votes. The ten highest CNW counties cast 3,061,326 votes, 51.37% of the total, with Clinton taking 1,857,037 votes to Trump’s 1,102,278. This means Trump won the states with CNW<50 by 1,809,353 to 987,444. Thus Trump support among his base was stronger than Clinton support among her core constituencies.

5. Lackawanna and Erie, two traditionally Democratic counties saw large drops in votes for Clinton, but their CNW# indicates they are still outliers, which implies that further losses will occur in future elections, much the same as Democratic support in Washington, Beaver, and other counties in Western Pennsylvania have declined precipitously in recent years.

6. Clinton received less than 20% of the vote in six counties in Pennsylvania. All six have CNW<20. Four additional counties have CNW<20 and Clinton received less than 27% of the vote in all four counties.

Conclusion: White voters without college degrees broke for Donald Trump at an astoundingly high rate. Essentially the race was between these three sets of voters: Whites without or with college degrees, and non-whites, with the latter two breaking for Clinton. Red is white, Blue is less white and has more education as a rule. I will test this formula out on other states and see if it holds up for more than Pennsylvania.


Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes:
<a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>