Is it true that TWO electoral colleges are better than one? Improving the American political system by adding a second “mega-region”-based electoral college.

Background:

Every four years, interest is re-kindled in the odd “Electoral College”-based method of tallying up American presidential votes.

Since the electoral votes are actually assigned on a per-state[*] basis (Figure 1), this has the side effect of making the presidential elections hinge on a few “battleground” districts—tiny subsets of swing states whose voting outcome isn’t already predictable.

Thus, only voters in these regions actually need to be granted presidential boons in order to persuade them—voters in the vast majority of states can be safely ignored.

[*] Some states split their electoral votes between candidates, but this is not common.

Fig. 1: Since electoral votes are (generally) assigned on a per-state basis, you don’t actually need to know the percentage of voters in each state who voted a certain way. A map like this one would be sufficient!

The issue:

People frequently discuss the idea of changing the electoral college system to a one-vote-per-person system.

However, very little consideration has gone into the other direction—having a SECOND electoral college, essentially an “electoral college for the electoral college.”

If one electoral college is good, maybe two would be better?

Proposal:

Here, we propose that the United States be grouped—for the purposes of elections only—into 10 five-state “electoral mega-regions.” Washington D.C. will retain its 3 electoral votes, and will be counted as a secret “Megaregion Zero” (not shown on map).

(To increase the level of mystery, its votes will only be used in cases of ties, and will not be included in tallies otherwise.)

Fig. 2A: Ten megaregions, each consisting of five states, have been gerrymandered together into this proposed grouping. See figure 2B for an examination of their electoral votes

Then, each of these megaregions would cast its aggregated “mega-electoral-votes” just as the normal electoral votes are determined.

Fig. 2B: One interesting failure mode of this method of grouping—and something that may make it a difficult sell—is that some groupings actually just “delete” the votes of certain states. For example, in Megaregion 6, Texas has a majority of votes for the entire region, as does California in Megaregion 7. Thus, this system serves to remove the following eight states from the electoral pool entirely: Alaska, Hawaii, Louisiana, Mississippi, Arizona, New Mexico, Colorado, and Oklahoma.

In order to fix the disenfranchisement problem described in Fig 2B, we could assign the votes of each megaregion based on a a simple majority of its states: instead of allowing Texas to entirely determine the outcome of Megaregion 6 (as it would if we weighted states by population or electoral votes), we would count each state as a single vote: so it would be a “best 3 out of 5” for Texas, Louisiana, Mississippi, Hawaii, and Alaska.

With this “one state / one vote” method, every state would be important in assigning the votes of the megaregion.

Conclusion:

Thanks to this reformed system, politics will be saved forever. Also, this may showcase the electoral college system, leading other countries to adopt it.

PROS: Improves the electoral college system by adding a second layer, thus multiplying its benefits.

CONS: It’s somehow theoretically possible that this might lead to gerrymandering???

See below for an additional example of this system in action.

Fig. 3A: If the states voted as above, for the RED, BLUE, YELLOW, and PURPLE parties, the updated megaregion-voting-assignment process would allocate the votes as shown in Figure 3B (not described: a tiebreaking process for dealing with situations in which the split is 2:2:1 or 1:1:1:1:1—presumably ties would be broken by a coin flip or something).
Fig. 3B: Now that each state has cast its (single) megaregion vote (as shown in figure 3A), we see how the state’s votes are distributed on a per-megaregion basis.