House Elections

So, I ran the numbers. With the help of Wikipedia, the Clerk of the House and the FEC, I built a chart going back to 1938 (see yesterday’s post). It shows the popular vote accumulated by each party during the Congressional general election, as well as the seats actually received. Then, I did the math to see how many seats the popular vote should have translated into.

Vote Return on Investment (ROI)

Lastly, I calculated the return that every voter received who voted for that party. You could argue the utility or approach, but I chose to divide the percentage of seats obtained by the percentage of the popular vote actually received.

So, in 1938, the Democrats received 17,715,450 votes in House elections across the country out of a total of 36,359,419 votes cast. That comes out to 48.72% of the popular vote. They actually captured 262 out of 435 seats in the House whereas they earned only 212 (less than a majority). So, if you voted for a Democratic congressman (win or lose) in 1938, the final results meant that you received representation worth 1.236 votes, which is, in fact, the highest rate of return that I recorded since 1938. Well played! Reprehensible, too!

Skewed Results

Over the course of 40 elections, the Democrats have been able to obtain more seats than they “earned” an astonishing 78% of the time. The Republicans have only managed this 32.5% of the time.

If you’ve been voting Libertarian any time since 1980, then you have received no representation despite earning as many as 7 representatives in Congress. Essentially, voting Libertarian counted as a vote for whichever party obtained the majority in Congress, regardless of their policies. This is also true of any of the other futile third parties.

For that matter, the results make it appear that the better the winning party performs, the more value that their voters receive, i.e. the value of votes improves as the party’s performance improves, essentially creating an exponential increase in representation.

The counter-reality is that four times in the years reviewed, the wrong party ended up with control of the House:

• 1942, when the Democrats managed to capture 222 seats on a million less votes
• 1952, when the Republicans took the majority of the seats on a couple hundred thousand less votes
• 1996, when the Republicans too the majority of the seats on thirty thousand less votes
• 2012, when the Republicans managed to capture 234 seats on four hundred thousand less votes

Even Recently

Moreover, seven of the last eleven elections should have ended with neither party controlling a majority. The other four ended with each party controlling the lower legislature twice. During those twenty-two years, Americans cast over a billion votes for Congressional representatives. Less than a million and a half votes separate the ballots cast for the two parties (511,948,991 or 50.08% for Democratic candidates and 510,386,62 or 49.92% for Republican candidate).

Matters get particularly interesting if you factor in the popular vote for third parties, which reduces those percentages to 48.16% for Democrats and 48.01% for Republicans. This raises the rather interesting idea that the major parties should have to cooperate with one or more third parties to pass any legislation if they are unable to obtain support from their supposed opposite number.

Redressing the Lack of Representation

On a primordial level, the solution appears to be at-large elections. Essentially, seats in the House would still be distributed among the states based on population, but the states would vote for slates of candidates from each party. Then the seats would be apportioned by the results. Surprisingly, at-large seats still existed in some states until Congress outlawed them in the 1960’s. The outlawing is not the surprising part. (Arguably, it was meant to improve the viability of minority candidates. That sort of worked as long as they ran under the auspices of one of the major parties.)

For example, the imaginary state of New Wyoming gets three seats. The Republicans, Democrats, and Greens all run slates of three candidates, with each candidate prioritized 1, 2 and 3. Voters pull tabs for one party only for Congress. The results are Republican- 48,345; Democrats – 12,944; and Green – 20,231. New Wyoming would place Republican #1, Republican #2, and Green #1 in the House.

Of course, myriad variations are possible: voting options for each seat; list the candidate, not the party; etc. The challenge is ensuring representation of the widest swathe of voters. Assumptions like random drawing of districts will do little to prevent unequal representation.

Detailed Suggestion

So, here’s a thought, however impractical. The only assumption behind the below is that representation by the party outweighs representation by someone local. (Discuss elsewhere how this operates as a hold on the opposing party from legislating too aggressively.)

• Every ten years, say on August 1 of any year that ends in a 1, re-allocate the number of Representatives by state.
• Use an average of the voters in the previous five Congressional elections by state for the allocation.
• Every state receives a minimum of one Representative no matter how few people voted. For every 220,000 actual voters (averaged from the previous five elections and rounded down at 109,999) each state receives one representative, no longer restricting the number of seats to 435.
  • The 220,000 figure was arrived at by using the current voting percentage of 31% times the entire population of the represented states and dividing by 435, rounding the result to the nearest 10,000. For 2010:
    (308,143,815 * .31)/435 = 219,597
    Every decade, we raise the percentage by 2% until it reaches 50%. That will take almost a century. This should strongly encourage states to incentivize voting.
• Each party nominates a full slate of Representatives for each state.
• Seating in the house by party is determined by national voting. Parties are awarded their seats in reverse order of performance. For this purpose, independent candidates are treated as one party. This is to ensure that the most popular candidates from poor performing parties receive seats.
• Actual representatives are allocated within the party by state results.
  That feels like the closest approximation of actual representation of voting choices while living in a world dominated by political parties.

Problems with the suggestion

This leaves one glaring problem- unaffiliated candidates don’t have an avenue to election. My solution is to bundle independent candidates together as a single party when the counting happens, but then allocate the seats to those candidates that receive the most votes when nationally ranked. This also reduces the seats within any state containing independent winners that can be allocated to party-affiliated candidates.

The viability of minority candidates could remain problematic, though aspects of this proposal do increase the likely success of any candidate supported by a nationwide party organized around any unifying principal. On its surface, that sounds potentially divisive unless you consider the unifying value of greater investment in the democratic process.

Let’s look at an example:

We will work with West Virginia, Wisconsin and Wyoming because they come last in the alphabet. On August 1, 2011, it is time to allocate Congressional seats. In our world, West Virginia receives 3 seats, Wisconsin has 8 seats, and Wyoming has 1 seat, all based on the 2010 census. Voting turnout for Congressional elections in 2002, 2004, 2006, 2008 and 2010 averages as follows: West Virginia – 547,270; Wisconsin – 2,287,646; and Wyoming – 210,184. The proposed system would result in 2 seats for West Virginia, 10 for Wisconsin, and 1 for Wyoming. This is a close run matter for West Virginia and Wisconsin. In both cases, encouraging turnout likely would result in an additional seat for West Virginia and a better buffer for Wisconsin.

Then the election happens. National results require the apportionment of candidates based on poorest performing parties first. West Virginia and Wyoming only manage to field major party candidates. Wyoming goes mostly Republican and fills their seat that way. West Virginia splits close enough that the top Democrat and the top Republican take seats in Congress. The most successful candidate in the nation for the resurgent Framer Labor Party resides in Wisconsin and takes their lone seat. Otherwise, Wisconsin splits pretty close with the Democrats winning slightly, so it goes 5 to 4 in their favor. Bear in mind that we would have to play the example out entirely to confirm these results.

Part 1 of the series is where all this begins