There is a very strong correlation, then, between a state voting for Republicans and receiving more in federal spending than its residents pay to the federal government in taxes (the rust belt and Texas being notable exceptions). In essence, those in blue states are subsidizing those in red states. Both red and blue states appear to be acting politically in opposition to their economic interests. Blue states are voting for candidates who are likely to continue the policies of red state subsidization while red states are voting for candidates who profess a desire to reduce federal spending (and presumably red state subsidization).
One cautionary note on this is that the last time I tried to look into this subject, I was assured that there’s no really methodologically rigorous way of assessing the issue and was cautioned away from overreliance on the very same 2005 Tax Foundation study that’s the basis for this map.
My larger concern, however, is with the “ecological fallacy” of attributing the quality of whole states to specific individuals. There are millions of people in states like Alabama, Mississippi, Arkansas, etc. who voted for Barack Obama. Those people are primarily lower income people. If we slashed federal spending on Medicaid, Title I education assistance, food stamps, etc. and used the savings to cut taxes on high income individuals then “Alabama” and “Arkansas” would suffer while “New Jersey” and “Connecticut” would gain. But at an individual level, low-income people would lose and high-income people would win. So there’s nothing really irrational about a well-to-do businessman in Mobile voting for tax cuts, or a poor person in Newark voting for anti-poverty programs.