Immigration voting patterns

There is a lot of speculation about how Senators in Congress will vote on the Immigration bill currently being debated in the Senate.  So, I thought it might be useful to reflect on what happened the last time such a measure was debated in the Senate, i.e. 2006.

In 2006, the comprehensive immigration bill then proposed was strongly supported by Senator Kennedy (D) and President Bush (R) and passed the Senate 62-48.  It failed in the House despite strong support from President Bush.

What is more interesting is an analysis of the Senate vote itself.  I'll caveat this analysis by pointing out that there are just 100 Senators and a lot of their voting must have been based on personal convictions, so its rather a small sample.  Also, with support from President Bush, the bill presented very different risks for both parties, so this may not be representative of attitudes in general.

Also, let me acknowledge the data sourced.  To analyze the vote, I used the data posted here.  Further, the Hispanic percentage in each state was obtained from Wikipedia.

So, what does the data tell us?

• Not surprisingly, the single most important predictor of whether someone would vote for the bill seemed to have been political party, with the Democrats being significantly more likely to vote for the bill than Republicans.

• What's more, as expected, when the percentage of Hispanic exceeded 5% of the state's population, the probability that the Senator voted for the bill increased in both parties.

• In looking at the data, I noticed a pattern  To the naked eye, it seemed that most of the States where people seemed to be voting against immigration were slave states at the time of the Civil War.  So, I checked the data.  According to the data, it appears that whether a state is a former Slave state or not makes little difference to Democratic voting patterns, but slave state GOP voters seem substantially less likely to vote for immigration. 

• However, this isn't really conclusive.  After all the proportion of Hispanics in former slave states may be substantially different than in non slave states.  So, I looked at both together. 

As the last chart shows, in general Hispanic proportion does cause the propensity to vote for the bill increase in both parties and in both slave and non slave states.  However, while for Democrats, the Hispanic proportion crossing 5% completely overwhelms any impact slave states have, among the GOP Senators, the increase in probability to vote for immigration as Hispanic proportion increased failed to offset the impact of being a slave state.  In fact, the probability of voting for immigration among GOP Senators hailing from high Hispanic population in slave states is still less than corresponding probability among GOP Senators hailing from low Hispanic non slave states.

In the May 3, 2013 episode of Backstory they discussed how wars end and one of the wars they discussed was the US Civil War.  They point out that the Confederates and the Union had very different conceptions of what actually happened.  The Union believed they won the war and the South capitulated.  The Confederates however saw it as a ceasefire in which the Union agreed to not interfere in exchange for the Confederates stopping the violence and returning to the fold.  

This data suggests that atleast when it comes to immigration, some in the GOP are still suffering from a hangover from that war.