Election Data Geekery

So, if there is any question about whether being a financial analyst is a good job for me, I think I just answered it. On my lunchbreak, I was tooling around with this. After some fun playing with the map, I delved into the data. Then it occurred to me, I could cut and paste the data into Excel and sort to get a better analysis. Yup, that's what I did on my LUNCH BREAK and I had fun. Here are my initial findings for the ballot questions:


My initial process was just purely sorting by straight votes. But as you can see, of course the biggest cities make the top 4 and bottom 4. So, analytically it wasn't too interesting. BUT what is interesting to look at the pattern of cities that crop up. Boston, Worcester, Springfield, Quincy, and Cambridge always pop up. But look at Question 1 and 3, no Newton. And on Question 3 Brockton pops up out of nowhere on the No side. Maybe that's because the Raynham-Taunton Greyhound Park is nearby?

Since I spotted the error in my sorting and analysis, I realized that I should weight the votes somehow. So, I weighted votes as a percentge against total votes. Remember, this is FUN for me:



Look at the variety of towns! Also look at Question 3. It appears that I was correct that opinion was heavily influenced by location (and probably employment). Sad to say, a lot of these towns I've never heard of. Apparently Wendell is a stoner friendly town! New Marlborough loves their doggies. Leverett loves taxes so much, we should give them more! It is also interesting to see which questions were the most divisive. On the no side, Question 1 had the highest percentage of either question followed by Question 3, and then 2. On the yes side, it went 3,2,1. Stay tuned for my presidential analysis and I might try another method of weighting voting results which would include town population.