31 March 2008

Decline in Southern Democrats: 1952-1994

(Click to enlarge.)

Here's a graph I just made, that shows the proportion of Democratic and Republican U.S. Senators, U.S. Representatives, and state Governors in the former confederacy states, over time.

The first data point is 1929-30 (prior to F. Roosevelt-to show that the high level of Democrats early on is not a function of the Great Depression), the second is 1953-55, and each successive one is the next session of Congress. Point 4 represents the years immediately following the Brown v. Board of Education decision that invalidated segregation in the South. Point 9 represents the years following the 1964 Civil Rights Act (which was bitterly resented in the South), and the last point represents the 1995-96 Congress, when Republicans took control of the House for the first time in over 40 years, and many Southern Democrats switched to the Republican Party.

Now I just need to figure out how to put the appropriate dates on the X-axis. Excel is a bit of a pain on that kind of thing.


Anonymous said...

Is your graph a line graph or a scatterplot? That affects how you insert the date.

In any event its an interesting graph. It appears the south is growing more contested. What is the date range?

James Hanley said...

It's a line grap. The black lines represent Democrats, the gray lines represent Republicans. The circular markets denote the lines for Senators, the diamond markers denote the lines for Representatives, and the asterisk markers denote governors.

The date range is 1953-1995, except that the first one is 1929-to show that 1952 is not a dishonest starting point, where maybe Democrats had an abnormally large advantage due to the great Democratic landslides of the Great Depression, but I might have to cut that one out to make the axis work.

Anonymous said...

Yeah, that's your problem. Line graphs requires a constant frequency for their date range. You will either need to drop this observation or switch to using a scatterplot, or add extra years containing no data between the 1929 date and the others.