Electoral College weighting of votes by race

Factual Questions
1

I’m putting the question here because I’m looking for a mathematical answer. If you think it’s a silly question, I apologize in advance :).

So I’ve seen some analyses of how the electoral college gives disproportionate influence to voters from low-population states compared to voters from high-population states: a voter in Wyoming, for example, influences the electoral college something like 3.5 times as much as a voter in California. (This is an opinion piece, but the bar graph at its heart appears factually correct).

My question is, how does this intersect with race? Different racial groups in the United states have different geographic patterns: for example, I believe that 78% of rural Americans, compared to 64% of all Americans, are white. Does this disproportionate electoral college advantage to small-population states give disproportionate influence to members of particular racial groups?

For example, if we consider the median African American voter to have one vote’s worth of influence, how much influence does the median white voter have?

A proper analysis would, I think, look at the demographic breakdown of every state, and compare that to the electoral college votes given to that state as well as to that state’s total population. I have no idea how to conduct that analysis, but wonder if it’s already been conducted, or if someone knows how to set up a fancy spreadsheet to do it.

2

I think you have entirely misunderstood how the electoral system works. The number of electors for each state, matches the number of Representatives, plus the number of Senators (always two).

The reason why smaller states are said to be disproportionately represented, is because of the two Senator-matched electors. Several states have so few residents, that they only have ONE representative in Congress, for the entire state. But every state gets at least THREE electors anyway.

The number of Representatives, and therefore electors, your state gets, has been declining steadily over time. This is because it was decided not to expand the membership of the House of Representatives, and instead, they increase how many people each Rep represents.

Anyway. That means that more populated states, have fewer Electors per voter than the states with the smallest population. Only by a very few, however.

But electors are NOT awarded to small states “because they are small and we felt sorry for them,” or anything like that. They were awarded in this way as a part of sticking to the basics of this being a REPUBLIC, and not a DEMOCRACY. Therefore there is no place for asking if electors be awarded by race, or other such designations.

3

Nobody suggested that electoral votes be allocated by race, or any other change to the current system. The OP is asking what the relative weighting of races is under our current system. It’s probably not 1:1.

4

…no, I haven’t misunderstood at all. I’ve read the founding documents, the relevant sections of the constitution, and many other documents beside. I’m talking about effect, not about anything else. If you think the question is silly, though, that’s fine.

5

It’s a good thing that isn’t what the OP was asking or suggesting, then.

Electoral votes don’t represent an equal number of people across the board because of the 3-electoral-vote minimum; one of WY’s electoral votes represents fewer people than one of CA’s and therefore it appears that WY voters’ votes are “worth more”.

The OP’s question, I think, is if such disparities exist across racial, rather than state, lines. (Obviously, the racial answer depends on the racial backgrounds of various states.)

6

Example setup, exaggerated for effect to show what I mean.

The Triple States of America have a system somewhat like ours: each state gets two electoral college votes, plus one vote for every ten citizens. There are three states.

CityState has 180 residents and 20 EV. 60% of the residents, or 108 residents, are black, and the remaining 72 are white. Each resident influences the election by 1/180 of 20 votes, or 1/9 of an EV each.
TownState has 80 residents and 10 EV. 20% of the residents, or 16 residents, are black, and the remaining 80%, or 64 residents, are white. Each resident influences the election by 1/80 of 10 votes, or 1/8 of a vote.
FarmState has 10 residents and 3 EV. 1 of the residents is black and the remaining 9 are white. Each resident influences the election by 1/10 of 3 votes, or 3/10 of a vote.

There are 145 white residents. The median influence wielded by them is 1/8 of a vote.
There are 125 black residents. The median influence wielded by them is 1/9 of a vote.

This is drastically simplified. How does it work in the United States?

7

According to the linked article,

Is there any language in the Constitution that is contradicted by the electoral college system? If not, there is no basis for a challenge.

8

I’m not sure that question can be answered as posed. Although there are certainly patterns to voting along lines of ‘race’, the reality is that racial groupings are not a homogenous collection of voters; witness the predominately black voters in Michigan who were outraged about the ineffectual response to the contaminated water in Flint and widely voted against traditional Democratic trends, or the Latino voters who voted for Trump because of the conservative values he espoused despite his general disdain and undertone of racism. In a normal race most of the Electoral College vote is already pre-accounted for by the cultural dominance of states compared to alignment of the candidates’ stated values and positions, leave a few key swing blocks which really control the election. Here is an article from FiveThirtyEight.com from late last year prior to the primary season which discusses some of the demographic breakdowns, and for those hating on Nate Silver and his team for miscalling the election, the article makes the point nearly a year ahead of the election that even at that point, before the impact of Trump agitating the Republican base and drawing out voters disaffected with the system, that it would take relatively small shifts in key demographics to swing critical states in the supposed “Blue Wall”.

Since minority voters are typically concentrated in urban areas, their votes would seem to count for a less influence than the predominately ‘white’ middle states, but they also may represent a potential for greater sway specifically because of the lack of representation of their interests by both parties, whereas even states with proportionally higher nominal strength of normalized Electoral College votes such as Kansas or Idaho can be essentially called before the candidates are even selected. So to do this kind of analysis you can’t just count the distribution of minorities normalized by electoral votes but also how much potential for swing they have in a particular state, and how much potential that state has to shift the Electoral College in a given election. And then you have to account for how the vote may be distributed within each socioeconomic group; college educated blacks and Latinos vote in a manner decidedly different than those without college educations.

The phrase 'All politics is local," has been bandied about a lot recently to try to explain how the polls and pundits were so mistaken about the election (even though an objective look at polling data through the campaign should have given those predicting an easy win for Clinton cause for concern). Several people who ‘predicted’ the result based upon their personal experience with specific groups of voters that dramatically shifted from traditional voting patterns (e.g. Michael Moore) are now basking in their supposed prescience, but in large measure they happened to be associated with a swing group that matched the mood in other areas that they were not as informed about, which tells us something about how seemingly different demographics may respond in similar ways to specific types of issues. Interpreting the probability of those swings, however, requires some insight into the local socioeconomic culture beyond what can be discerned from a supposedly random poll or simple demographic breakdown. A better method of making predictions using local pundits who are ranked and weighted by the post-hoc accuracy of their predictions is probably necessary to get better predictions in a close race like this one in which many voters were clearly voting in opposition to a candidate rather than for the candidate or specific issues.

Stranger

9

Emphasis added. Easy-peasy. Just go to wikipedia to get the total population of each state and demographic breakdown and do the math. Am I missing some subtlety here?

Or go here to start.

10

Overall, African Americans comprise 13.3% of the total U.S. population, so in any state with a lower percentage, African American will necessarily be underepresented, no matter how many electoral votes the state has.

On the flip side, the District of Columbia has three electoral votes. The population is 672,000, with 48.3% African American.

11

Kunilou and Stranger on a Train, look at my example: I think your posts are missing the point of the question. It’s not about who gets spoken to the most, or about who’s underrepresented in which state. It’s a matter of population and electoral college math.

So here’s what I’m thinking. I’m imagining a table, which I’m not sure how to code here. It’d have columns like this:
-State name
-Total population
-EV votes
-People per EV vote (calculated) *
-Several columns with percentages of population per racial group
-Several columns with total population per racial group (calculated) *

The asterisked columns would be the interesting ones. I’m still not sure how derive a median value from them, though. Thoughts on this?

12

John, I’m playing with that site, and it’s annoying as hell. It won’t let me select the table, and when I try to download the raw data, I get a .csv full of “[object]”,“[object]” etc.

Does anyone else have a source for this data?

13

View the source code for the page, and it’s pretty easy to copy out the relevant hunk and reformat to spreadsheet use. (I can help if you want.)

I don’t know that you need to derive a median, though. Perhaps just calculate for each state what proportion of the electoral votes derive from each race, add them up, compare that to the total number of electoral votes?

For example, Alabama has 9 electoral votes, and 27% of the population is black, so 9 X .27 = 2.43 electoral votes “belong” to blacks. If you figure that for each state, then add up how many total for each race/ethnicity, that should tell you the answer to your question. Blacks comprise about 12% of the U.S. population, so if everything else is equal, the sum of your state-by-state figures ought to be 64.56 (12% of 538).

14

It’s not that many numbers-- you could enter them manually into a spreadsheet.

15

Because of the winner-take-all way of allocating electoral votes, Black people in Alabama do not actually have the equivalent of ‘2.43 electoral votes’, it’s fairly certain for the foreseeable future that in presidential elections Alabama is always going to go in the opposite direction that most of its Black population would prefer. Black people in Alabama have essentially zero influence over selecting the president.

16

Yes. Which was exactly Stranger’s point.

There’s a lot more to the EC’s impact on the election than just the low-population states having fewer populace voters per EC member.

The OP can certainly calculate the number he wants. The math is trivial. But that won’t answer the deeper question of who’s powerful and who’s not and by how much. No more than the oft-quoted figure of 3.5 to 1 for WY vs CA really tells the whole story on WY’s vs CA’s influence.

Imagine a country where 50 of the 51 states + DC were either guaranteed safe red or guaranteed safe blue. And each side adds up to 260-some EC votes.

In that country there’s exactly one swing state. It decides. Everybody else is along for the ride. That state has all the voting power and the other 50 have zero voting power. That state could be purebred white or purebred black or purebred Latino. If so, what would that say about the voting power of that ethnic group on a national level? Would it even make sense to talk about the relative abundance of the ethnic groups in the other 50 states/DC?

Interesting questions all.

17

OK. I copied the table into Excel and added the number of EC votes for each state. I multiplied the percentage black for each state by the number of EC votes, and totaled these.

As **slash2k **posted, based on the national percentage black of 12%, you would expect this total to equal 64.56. It actually totaled 64.83.

So, the effect questioned in the OP is negligible. It is also somewhat pointless since the Constitution calls for the President to be selected by the States, not the People.

18

I hacked a counter before noticing Excavating had already done so. Since my result differs slightly from his, I may as well post it. I used this Wikipedia page for the demographic breakouts, but adjusted numbers to add to 100%.

I’ll include the date table I extracted from the Wiki in case someone else wants to use it. I may as well include the source code as well.

[SPOILER]char *Ethname = {
“White”, “Latin”, “Black”, “Nativ”, “Asian”, “Pacif”, “Mixed”,
};

struct stademog {
char *sd_name;
int sd_dev, sd_rev;
double sd_ethn[7];
int sd_pop;
} Statd[51] = {

“Alabama”, 0, 9, 67.0, 3.9, 26.2, 0.6, 1.1, 0.1, 1.5, 4822023,
“Alaska”, 0, 3, 64.1, 5.5, 3.3, 14.8, 5.4, 1.0, 7.3, 731449,
“Arizona”, 0, 11, 57.8, 29.6, 4.1, 4.6, 2.8, 0.2, 3.4, 6553255,
“Arkansas”, 0, 6, 74.5, 6.4, 15.4, 0.8, 1.2, 0.2, 2.0, 2949131,
“California”, 55, 0, 40.1, 37.6, 6.2, 1.0, 13.0, 0.4, 4.9, 38041430,
“Colorado”, 9, 0, 70.0, 20.7, 4.0, 1.1, 2.8, 0.1, 3.4, 5187582,
“Connecticut”, 7, 0, 71.2, 13.4, 10.1, 0.3, 3.8, 0.0, 2.6, 3590347,
“Delaware”, 3, 0, 65.3, 8.2, 21.4, 0.5, 3.2, 0.0, 2.7, 917092,
“DistCol”, 3, 0, 34.8, 9.1, 50.7, 0.3, 3.5, 0.1, 2.9, 632323,
“Florida”, 29, 0, 57.9, 22.5, 16.0, 0.4, 2.4, 0.1, 2.5, 19317568,
“Georgia”, 0, 16, 55.9, 8.8, 30.5, 0.3, 3.2, 0.1, 2.1, 9919945,
“Hawaii”, 4, 0, 22.7, 8.9, 1.6, 0.3, 38.6, 10.0, 23.6, 1392313,
“Idaho”, 0, 4, 84.0, 11.2, 0.6, 1.4, 1.2, 0.1, 2.5, 1595728,
“Illinois”, 20, 0, 63.7, 15.8, 14.5, 0.3, 4.6, 0.0, 2.3, 12875255,
“Indiana”, 0, 11, 81.5, 6.0, 9.1, 0.3, 1.6, 0.0, 2.0, 6537334,
“Iowa”, 6, 0, 88.7, 5.0, 2.9, 0.4, 1.7, 0.1, 1.8, 3074186,
“Kansas”, 0, 6, 78.2, 10.5, 5.9, 1.0, 2.4, 0.1, 3.0, 2885905,
“Kentucky”, 0, 8, 86.3, 3.1, 7.8, 0.2, 1.1, 0.1, 1.7, 4380415,
“Louisiana”, 0, 8, 60.3, 4.2, 32.0, 0.7, 1.5, 0.0, 1.6, 4601893,
“Maine”, 4, 0, 94.4, 1.3, 1.2, 0.6, 1.0, 0.0, 1.6, 1329192,
“Maryland”, 10, 0, 54.7, 8.2, 29.4, 0.4, 5.5, 0.1, 2.9, 5884563,
“Massachusetts”, 11, 0, 76.1, 9.6, 6.6, 0.3, 5.3, 0.0, 2.6, 6646144,
“Michigan”, 16, 0, 76.6, 4.4, 14.2, 0.6, 2.4, 0.0, 2.3, 9883360,
“Minnesota”, 10, 0, 83.1, 4.7, 5.2, 1.1, 4.0, 0.0, 2.4, 5379139,
“Mississippi”, 0, 6, 58.0, 2.7, 37.0, 0.5, 0.9, 0.0, 1.1, 2984926,
“Missouri”, 0, 10, 81.0, 3.5, 11.6, 0.5, 1.6, 0.1, 2.1, 6021988,
“Montana”, 0, 3, 87.8, 2.9, 0.4, 6.3, 0.6, 0.1, 2.5, 1005141,
“Nebraska”, 0, 5, 82.1, 9.2, 4.5, 1.0, 1.8, 0.1, 2.2, 1855525,
“Nevada”, 6, 0, 54.1, 26.5, 8.1, 1.2, 7.2, 0.6, 4.7, 2758931,
“New_Hampshire”, 4, 0, 92.3, 2.8, 1.1, 0.2, 2.2, 0.0, 1.6, 1320718,
“New_Jersey”, 14, 0, 59.3, 17.7, 13.7, 0.3, 8.3, 0.0, 2.7, 8864590,
“New_Mexico”, 5, 0, 40.5, 46.3, 2.1, 9.4, 1.4, 0.1, 3.7, 2085538,
“New_York”, 29, 0, 58.3, 17.6, 15.9, 0.6, 7.3, 0.0, 3.0, 19570261,
“North_Carolina”, 0, 15, 65.3, 8.4, 21.5, 1.3, 2.2, 0.1, 2.2, 9656401,
“North_Dakota”, 0, 3, 88.9, 2.0, 1.2, 5.4, 1.0, 0.0, 1.8, 699628,
“Ohio”, 18, 0, 81.1, 3.1, 12.2, 0.2, 1.7, 0.0, 2.1, 11544225,
“Oklahoma”, 0, 7, 68.7, 8.9, 7.4, 8.6, 1.7, 0.1, 5.9, 3814820,
“Oregon”, 7, 0, 78.5, 11.7, 1.8, 1.4, 3.7, 0.3, 3.8, 3899353,
“Pennsylvania”, 20, 0, 79.5, 5.7, 10.8, 0.2, 2.7, 0.0, 1.9, 12763536,
“Rhode_Island”, 4, 0, 76.4, 12.4, 5.7, 0.6, 2.9, 0.1, 3.3, 1050292,
“South_Carolina”, 0, 9, 64.1, 5.1, 27.9, 0.4, 1.3, 0.1, 1.7, 4723723,
“South_Dakota”, 0, 3, 84.7, 2.7, 1.3, 8.8, 0.9, 0.0, 2.1, 833354,
“Tennessee”, 0, 11, 75.6, 4.6, 16.7, 0.3, 1.4, 0.1, 1.7, 6456243,
“Texas”, 0, 38, 45.3, 37.6, 11.8, 0.7, 3.8, 0.1, 2.7, 26059203,
“Utah”, 0, 6, 80.4, 13.0, 1.1, 1.2, 2.0, 0.9, 2.7, 2855287,
“Vermont”, 3, 0, 94.3, 1.5, 1.0, 0.4, 1.3, 0.0, 1.7, 626011,
“Virginia”, 13, 0, 64.8, 7.9, 19.4, 0.4, 5.5, 0.1, 2.9, 8185867,
“Washington”, 12, 0, 72.5, 11.2, 3.6, 1.5, 7.2, 0.6, 4.7, 6897012,
“West_Virginia”, 0, 5, 93.2, 1.2, 3.4, 0.2, 0.7, 0.0, 1.5, 1855413,
“Wisconsin”, 10, 0, 83.3, 5.9, 6.3, 1.0, 2.3, 0.0, 1.8, 5726386,
“Wyoming”, 0, 3, 85.9, 8.9, 0.8, 2.4, 0.8, 0.1, 2.2, 576412,
};

int main(int argc, char **argv)
{
int eix, stix, evtot, tpop;
double dpp, eprop, pprop;

for (tpop = evtot = stix = 0; stix < 51; stix++) {
	tpop += Statd[stix].sd_pop;
	Statd[stix].sd_dev += Statd[stix].sd_rev;
	for (dpp = eix = 0; eix < 7; eix++)
		dpp += Statd[stix].sd_ethn[eix];
	for (eix = 0; eix < 7; eix++)
		Statd[stix].sd_ethn[eix] /= dpp;
	evtot += Statd[stix].sd_dev;
}
for (eix = 0; eix < 7; eix++) {
	for (eprop = pprop = stix = 0; stix < 51; stix++) {
		eprop += Statd[stix].sd_ethn[eix] * Statd[stix].sd_dev;
		pprop += Statd[stix].sd_ethn[eix] * Statd[stix].sd_pop;
	}
	printf("%s have %4.2f%% of pop and %4.2f%% of ev

",
Ethname[eix], 100 * pprop / tpop, 100 * eprop / 538);
}
exit(0);
}[/SPOILER]Here’s what that program printed:

White have 62.84% of pop and 64.23% of ev
Latin have 16.08% of pop and 15.00% of ev
Black have 12.45% of pop and 12.16% of ev
Nativ have 0.94% of pop and 1.07% of ev
Asian have 4.65% of pop and 4.45% of ev
Pacif have 0.17% of pop and 0.20% of ev
Mixed have 2.87% of pop and 2.90% of ev

19

Only for a very peculiar definition of “power”. Consider that, under that same definition, if the “safe blue” states totaled 270 electoral votes, they would be considered to have no power whatsoever. It makes more sense to measure power by results, and to say that power consists in getting what you voted for (in which case, in that one-swing-state country, the swing state and whichever half of the other states it votes with in any given cycle share all of the power).

20

Targeting the OP’s premise, there is clearly not an intent, in the electoral college, to exercise racial discrimination. The founders didn’t intend for anybody to have a vote, except white male landowners, and any effect on any hypothetical scenario of universal suffrage could not have been further from their minds. Surely, Jefferson et.al. did not say “Gee whiz, fellas, some day black women are going to be voting, and we should sneak in this electoral college thing as our last bastion of defense for our threatened privileged class.”

In fact, absent the trajectory of industrilization, minority and low-income classes might have remained predominantly within the rural agricultural and natural resources sector, and the electoral college would actually favor blacks.