First off, I am not from Alabama. But I did live there briefly in 1981, and have had a long-distance interest in it’s affairs ever since.
Second, I have no idea what will happen there tomorrow, and anyone who says they do is either a liar or onto something I’d like to know about.
That said, I saw something a few days ago that piqued my curiosity and I want to share some ideas about what it could mean for tomorrow’s election. I don’t recall where I saw this information (so there is no link here), but I was intrigued by the notion that Blacks in Alabama have recently voted over 97% (as I recall) for Democrats. Add to that, the notion that recent polls indicate up to 40% of GOP women in Alabama will not vote for Roy Moore, and there is a reasonable shot at Doug Jones winning a Senate seat tomorrow.
But before that, what does the 2010 Census have to say about Alabama’s racial and ethnic makeup. Traditional breakdowns of it’s population leave ambiguities about the exact numbers in each category, but there is data in the census that lets us get a more precise breakdown.
Based on block-level data in the Census Bureau’s Summary File 1 (SF1) data that was made public after the census, the population of Alabama can be categorized as follows:
Race / Ethnicity in Alabama
Categorization |
Population |
Total Population |
3,087,299 |
White Non Hispanic |
1,862,622 |
Hispanic White |
49,731 |
Hispanic Non White |
81,611 |
Black Non Hispanic |
994,254 |
Asian Non Hispanic |
44,157 |
AIAN Non Hispanic |
11,274 |
NHPI Non Hispanic |
1,423 |
Other Non Hispanic |
2,976 |
Multi Racial Non Hispanic |
39,251 |
Now, these numbers do not account for demographic changes since the Census was taken in 2010, and they represent total population rather than Voting Age Population (VAP). They also do not account for variations in actual registration percentages due in part to voter suppression tactics like Voter ID requirements, nor do they account for actual voter turnout (which nobody seems inclined to guesstimate for tomorrow).
Caveats not withstanding, these numbers give some reasonable basis for thinking about what scenarios might be required for Jones to win the election. Let’s start with the Black vote: 97% of 994,254 is 964,426, which is 579,224 short of the 1,543,650 needed to reach half the population. Where might those other needed votes come from? The following table gives some possibilities. (For presentation purposes, and due to the dominance of the White and Black segments, all other segments are grouped together.)
Targets for a Doug Jones Win
white pct |
black pct |
other pct |
white tot |
black tot |
other tot |
grand tot |
21.3 |
97 |
80 |
396,738 |
964,426 |
184,338 |
1,545,502 |
21.9 |
97 |
75 |
407,914 |
964,426 |
172,817 |
1,545,157 |
22.5 |
97 |
70 |
419,090 |
964,426 |
161,296 |
1,544,812 |
22.3 |
95 |
80 |
415,365 |
844,541 |
184,338 |
1,544,244 |
22.9 |
95 |
75 |
426,540 |
844,541 |
172,817 |
1,543,898 |
23.6 |
95 |
70 |
439,579 |
844,541 |
161,296 |
1,545,416 |
Think about these numbers. A 2% decline in Black Pct requires a corresponding 1% increase in White Pct for Jones to hit his numbers. Whether the decline is due to registration, turnout or voter choice makes no difference — the required numbers are reachable.
If White women hit 40% for Jones (which may be doable if the recent GOP polling is accurate and higher Dem rates are included), then the worst scenario above needs at best 8% of White men to join them (assuming an even 50% gender split — and women significantly outnumber men in Alabama).
I know there are problems with this simple model, but I fully believe that a Jones win is doable.
We will find out for sure tomorrow night!