Had California, with its 39 million people, existed in 1790, and was given the opportunity to join the young United States, it would naturally ask “okay, how many votes do we get in your Senate?”. And after finding out that it would only be two senators for every state, would just as naturally reply “you mean we get the same number of senators as tiny Delaware? Is that democracy for the stupid? Hmmmm. We'll let you know.”

And at the Constitutional Convention, this was very similar to the position taken by the colonies with large populations, such as Virginia and Pennsylvania. While originally balking at the idea that the small states would have equal representation in the Senate, they were caught between a rock and a hard place, and for two very good reasons. First, with regards to ratifying the new Constitution, they started out with only one vote to begin with. Oh shit. Wait. What?

And second, small colonies like Delaware were threatening to join forces with foreign governments. “The English weren’t so bad after all, maybe we made a mistake. Maybe they would like to help us in our disputes with Virginia.” So they caved and the much ballyhooed American semi-democracy was born.

Thus begun what would ultimately result in a central government that would, on average, be more conservative than its public, shift political direction more slowly, and be continually at risk of electing presidents that lose the popular vote. And because of the very asymmetric demographic trends in the growing US population, Republicans are now more likely than not to be elected President while losing the popular vote. The last two Republican presidents, Bush (2000) and Trump (2016) were both elected that way, and this trend will become an ongoing feature in American politics.

But the inequalities in the Electoral College begin with the inequalities of the Senate, which was designed to provide the economic elite with checks to the truly democratic one-person one-vote philosophy of the House of Representatives. Further, it was also designed to reduce the influence of rapid swings in public opinion, so Senators were given six-year terms, with only one-third subjected to re-election every two years.

The Electoral College was then constructed to reflect the congressional delegation, with one elector for each House and Senate member, leaving the states to decide on how to select the electors. As such, it would become a combination of the democratic one-person one-vote rule of the House, with the “I don’t give a fuck how many people your state has” rule of the Senate. Add the strong tendency for large urban populations to be more liberal, and you have a voting system that not only biases both the Senate and the Presidency towards the right, but also the Supreme Court (since it is selected by the President and approved by the Senate).

Estimating the Republican Bias of Two Senators per State

But how much bias? Quantifying the “I don’t give a fuck” rule for the Senate has several options. On the Senatoral Voting Power.pdf table below, we can see some statistics that show this phenomenon.

This table is sorted in descending population order, and as can be seen, the number of people per Senator in each state varies from a maximum of 19.4 million in California to 292 thousand in Wyoming. One person in Wyoming has 66 times more weight than a person in California in Senatorial representation. If Senators were assigned proportionately to population, California would have 12.2, and Wyoming would have 0.18.

First, using 3.181 million people (the total 50-state population divided by 100 Senators) as the national average population per Senator, we then derive the “Senate Voting Power per Person”. A person in California is worth 0.16 in the calculation of how much a resident is worth in terms of Senators relative to the national average. And correspondingly, a person in Wyoming is worth 10.9 times the national average. The democratic one-person one-vote rule is a wild misrepresentation of the Senate, and this disparity has grown substantially since 1790.

To illustrate the degree that the one-person one-vote rule is violated in the Senate, we can see that the sum total of “Senate Voting Power per Person” across all states is 132.2. This number would be close to 50 if Senators were assigned in proportion to each state’s population. This is a substantial difference and a tremendous advantage to residents of states with small populations.

But which party gets the advantage? Based on the positive relationship between population density and Democratic voting trends, it is certainly reasonable to assume the Democrats are punished by the absence of the one-person one-vote rule in the Senate. And two variables on the table illustrate this point. “Democratic Voting Power” is computed by multiplying the “Senate Voting Power per Person” times the percentage of population that is either identified as Democrat or leans Democrat based on Gallup 2015 estimates. “Republican Voting Power” is derived the same way.

If there were no bias to either party, we would expect that the sum of the “Democratic Voting Power” and “Republican Voting Power” across all states to favor the more numerous Democrats. But it favors the Republicans: 59 to the Democrats 51.6. Compare the estimated 137 million Democratic leaners to the 127 million Republican leaners, and the Republican advantage in the Senate is even more remarkable.

From this, the argument naturally follows that whites are also favored, given the much higher percentage of whites identifying with Republicans. This is illustrated In the Electoral College White Votes Matter More by the Center for Economic and Policy Research.

Further, the District of Columbia, which is both heavily Democratic and with a population greater than Wyoming, has no Senatorial representation. This further skews the Senate in favor of the white Republicans. This also does not take into account US Territories such as Puerto Rico, which would also favor Democrats given its large Hispanic population.

Estimating the Republican Bias of the Electoral College

Since there is one Electoral College vote for each member of the state’s Congressional delegation, this Senatorial bias should be carried over to the Electoral College, but to a lesser extent than in the Senate.

In the Electoral Voting Power.pdf above, we can see this carry over. Using the number of electoral votes allocated to each state, we calculate the “Electoral Voting Power per Person” which is relative to the national average of 1. Note that California’s average person is worth 0.84 of the US average, and that Wyoming is worth 3.04. This is not nearly as skewed as the Senate, but certainly a factor in electing a president.

Summing up the “Electoral Voting Power per Person” across all states and the District of Columbia, we get 68.2. If population was evenly distributed across all the states, this number would be 51.

Does the Electoral College favor the Republicans, like it does in the Senate? Again, calculating the “Democratic Voting Power” and “Republican Voting Power” using the percentage of the population assigned to the respective political parties, we can see the overall impact by summing up across all states. That number is 28.6 for the Republicans and 28.4 for the Democrats. This is very close, except when one considers that there are 137 million allocated Democrats, and 127 million Republicans, yet the “Electoral Voting Power” still slightly favors the Republicans. And this disparity makes it difficult for Republicans to win the Presidency without the Electoral College.

The New Democratic Political Underclass

Okay Democratic scum, you are on notice. At this point in US history, you have never had such a low value of your vote, and Republicans have never been higher. You are functionally second-level citizens, and the delusion of American democracy has never been so obvious for you. The Constitution was written to reduce the impact of the one-person one-vote rule of the House of Representatives and the Democrats have just been steamrolled by it.

But even the House, which is highly susceptible to the legal process of gerrymandering, has been seized by the Republicans. The sophisticated 2010 Republican REDMAP operation to gerrymander congressional districts, as described in David Daley’s book, Ratf**ked: The True Story Behind the Secret Plan to Steal America’s Democracy, was a fantastic success for the Republicans to seize control of the House while having fewer popular votes. This effectively made Barack Obama a 6-year lame duck president.

Shortly after Donald Trump takes office on January 20, 2017, the Republicans will control all four branches of the Federal government, despite having cast the minority of votes. This remarkable situation should enrage the Democrats, that have now been reduced to a political underclass.

And this raises the question, can a Republican president ever win the popular vote again? Demographic trends suggest not, and we are now in the era of minority Republican presidents that started with George W. Bush in 2000.

James Madison, the main architect of our four branches of the federal government, very much favored limiting the influence of the popular vote and giving disproportionate power to an educated and economic elite. It worked well then. It works well today. Good job, Jim.

Madison was from Virginia, and the new Electoral College managed to elect Virginians for 32 of the first 36 years of its existence. Including him. In his old age, many years after serving as President, Madison pat himself on his back for this great achievement:

“The happy Union of these States is a wonder; their Constitution a miracle; their example the hope of Liberty throughout the world. Woe to the ambition that would meditate the destruction of either!”

That was a bad miss, James Madison.

** Author's note: This data was assembled from 2014 population estimates and 2015 political party affiliation estimates. It also presumes that age-sex distributions are constant across all states, which they are not. However, adjustments for age-sex distributions might change the magnititude of the Republican bias, but not the direction.