How people elect parliaments
This page examines the specific results for the US House of Representatives elected in November 2012 (with a term running from January 2013 to January 2015). This includes an examination of the many ways in which votes are ‘missing’ from the official tallies due to the absence of major party candidates in dozens of districts.
In November 2012 around 121.6 million Americans turned out to vote for the election of their national House of Representatives. This represents around 56% of the US electorate, which was estimated to consist of around 215 million potential voters.
Around 27.6% of America’s eligible voters supported Democrat Party candidates. Around 27.0% supported Republican candidates. About 1.5% supported candidates of other parties or independents.
The remaining 44% of potential voters, for one reason or another, did not participate in the elections.
In absolute numbers, the participating voters gave around 59.7 million votes to Democratic candidates and 58.6 million votes to Republican candidates. Around 0.9 million votes were cast for Libertarian candidates, and around 2.4 million for all the remaining candidacies.
Despite the Democrat lead in the nominal vote total, the actual results in the 435 individual seats saw 201 only Democrats elected to the House, against 234 Republicans. The Republicans thus won a solid political majority without winning a majority of the votes.
This sort of outcome, sometimes called an ‘inversion’, is unusual. The elections of 2012 were only the second instance of such a situation since World War II (the other occasion being the 1996 elections). Inversions also occurred in 1912 and 1942.
How do inversions election results happen? As with any electoral system based on single-member divisions, the political make-up of the House of Representatives is subject to distortions caused by geographical concentration of partisan support across the electorate. This is examined in more detail in part one and part four of this series of pages.
How is the House currently elected?
As stipulated by federal statute for nearly a century, the House consist of 435 seats.
(Under present law there are a further 6 non-voting seats for representatives of DC, Puerto Rico and the territories, which we will not deal with in this analysis).
Numbers of members for each State
The number of seats elected by the voting population of each State is determined by the US Constitution, which requires a general apportionment of the seats among the States in proportion to population.
This apportionment approach means that the number of votes required to elect each Congressman varies significantly across the nation, in particular between the smaller States.
On average, each US Congressman is elected from a district with a potential voting population (generally known as VEP – vote-eligible population; not the same as the number of registered voters) of around 495,000 voters, of which around 280,000 turn out to vote. Typically the successful candidate is supported by around 180,000 voters, which is therefore around 37% of the average district electorate.
However, the actual numbers of voters in question vary quite considerably from district to district.
Firstly, the guarantee of a minimum of one House seat for each State means that voters of those States with small populations which round up to a number of representatives at the next integer higher than their quota – including Wyoming (1 seat for 241,000 electors) – are receiving representation at a significantly higher rate than the national average. Conversely, States with small populations which round down to a number of representatives at the integer below their quota – including Montana (1 seat for 479,000 electors) – are receiving representation at a significantly lower rate than the national average.
The effect diminished with the larger State population, where the number of voters in each division increasingly approaches the national average.
Even after that structural variation, the actual numbers of voters who register to vote, and then the number of registered voters who turn out to vote, also vary very significantly.
This means that overall, very substantial differences in voter influence occur across the country in electing House members.
In all States, nominations for election are controlled by state legislation.
Nominations are also filtered through party primary voting exercises in a variety of ways. The most common method is the party primary.
Subject to federal law, the 50 States each determine (by State legislation) their own voting methods for electing their representatives to Congress.
A federal statute from the 1960s currently requires that States elect their members in single-member divisions. This law was adopted to prevent States from using multi-member districts together with plurality voting (that is, the ‘block vote’ system) to generate unfair election results which supressed the representation of minorities.
As for the specific vote counting method, for many decades, all States have chosen to maintain the plurality voting method.
366 of the 435 House seats are elected by the standard single-member-district simple plurality voting method.
63 seats – those in the states of California and Washington – are elected by a two-round runoff system. Unlike the well-established 2-round-runoff system used in France, in California the first round of voting happens many weeks earlier than the national election day, and has the appearance of being a primary election. Turnout at the first round is therefore significantly lower than at the final round held on national election day.
Finally, the 6 seats for Louisiana are also elected by a two-round runoff system, but the poll held on national election day is the first round of that system. This means that the party vote in the district electorates is often shared between 2 or more candidates. (Where second round ballots are needed for these districts, they are held a few weeks later, often in circumstances of somewhat lower voter turnouts.)
The effect of use of single member districts on the 2012 election result
The numbers of votes cast in each US congressional electoral division vary quite significantly.
As mentioned earlier, voter registration is voluntary in the United States. Registration rates vary significantly between States and between divisions within each State.
Voting is also voluntary. Turnout rates also vary significantly in every division.
To some extent turnout is influenced by voter mobilisation efforts and advertising impacts related to the ‘swing’ status of the division and of the State in which is located. This influence may be driven independently by the swing status of separate House, Senate or Presidential races, and the impact of all such races may combine. (Hence, divisions which are themselves swinging, which are located in presidential swing states and which are further boosted by a competitive Senate race will have generally high relative turnout rates.)
As a result of these factors, the difference in total votes cast – and in votes needed to win a seat – is substantial across the 435 divisions.
In 2012 the number of votes cast in House divisions ranged from 95,900 (in Texas’ 29th district) to 479,700 (in the Montana district), around a mean of 280,100. (All values given here are rounded to the nearest 100).
The number of votes recorded by the seat winners ranged from 46,600 (New York 12th district) to 318,100 (Pennsylvania 2nd district) around a mean of 180,200.