United States

House of Representatives

Specific elections: 2000 – 2002 – 2004 – 2006 – 2008 – 2010 – 2012 – 2014 – 2016

Background

The House of Representatives is an assembly of 435 members. Seats are divided among the 50 States of the Union using the Huntington-Hill modified divisor method, with a minimum seat rule requiring at least one seat for each State.

(A further 5 non-voting members (termed ‘delegates’) are also elected to represent five non-state territories, including the federal capital territory of the District of Columbia, containing the city of Washington. These delegates are not dealt with in the discussion given below.)

Members are directly elected in the 435 single member divisions (‘districts’) by three different voting methods, as follows:

• In the states of California (51 seats) and Washington (10 seats) members are elected in contests limited to just two candidates, who are earlier selected in open primary elections by the SNTV method. This system sometimes results in a final choice being between two candidates of the same party, or of one party candidate and an independent.
• In the state of Louisiana (6 seats) members are elected by the two-round runoff method, with nomination open to multiple candidates from parties, with a runoff election (if needed) held a few weeks after the initial election day.
• Elections in each of the other 47 states (368 seats) members are elected by the plurality method.

Redistribution of seats among the States and reviews of district boundaries are conducted after the national census held every 10 years, the most recent being the Census of 2011.

Reviews of district boundaries (‘redistricting’) are administered by State governments and/or legislatures. Concentration distortion has historically been extensive in many States, and is made worse by redistricting often being gerrymandered by partisan legislators and officials.

Redistricting that took place during 2011-12 was significantly marred by gerrymandering. Across the 50 states control of the process was in the hands of different classes of decision-makers:

• Control of boundary reviews by independent commissions and/or courts – 13 states (140 seats)
• Control of boundary reviews by legislators of the Republican party – 17 states (209 seats)
• Control of boundary reviews by legislators of the Democratic party – 6 states (44 seats)
• Control of boundary reviews by politically divided legislatures or bi-partisan party-controlled commissions – 7 states (35 seats)
• (the final 7 states have a single district and thus have no boundary reviews)

The dominant position of Republican partisans in the 2011-12 redistricting round appears to have led to a clear Republican advantage in gerrymander-driven concentration distortion of election results. The impact of these boundaries can be estimated to have assisted in generating around 22 additional Republican district wins in the 2012 elections.

Terms of office for all House members are two years.

Commentary and analysis

Post: Gerrymandering still rife in US Congressional districts (6 November 2016)

Half-way review of the presidential primaries (March 2016)

The crucial voters don’t choose the presidential nominees (January 2016)

A brief history of presidential primaries (January 2016)

Preview of the 2016 presidential primaries (January 2016)

Preview of the United States elections of November 2016 (January 2016)

(unpublished) –

Inequality of the 2012 US House of Representatives district boundary changes (January 2016)

The US House of Representative could be more representative (January 2016)

Critical analysis of the US electoral system (January 2016)

Elections

Overview 2000-2016

Inequality in the effective influence of voters caused by variations in House of Representatives district enrolments has not been analysed, as district registration figures are not generally available.

Freedom of nomination is restricted in the United States. The regulations are controlled separately by state legislation in each of the 50 states. Nominees of the Democratic and Republican parties are generally unconstrained, although party primaries will be held to select one official nominee for each party (other than in Louisiana, where multiple party nominees may run). However in many states the nomination of other candidates is hindered by ‘ballot access’ laws which require extensive pre-nomination signature collections and/or high payments.

In addition, in California and Washington states only two candidates may nominate for the general election, these being the two plurality leaders of an official primary election held several weeks earlier, at which turnouts are often low.

[summary of results]

Inequality in the effective influence of voters caused by variations in House district turnouts (formal votes) has been significant in recent elections. Inequality is generally higher at mid-term elections (those not featuring the election of the President, where voter turnout is around 40% of registrations), that in Presidential election years (where turnout is around 60%).

The standard deviation of variations compared to the mean enrolment was 23.2% in 2000, 21.8% in 2004, 20.8% in 2008, and 20.8% in 2012.

For the mid-term elections in this period, the standard deviation was 26.3% in 2002, 28.4% in 2006 and 24.4% in 2010,

[inequality by margins]

Specific elections

2000 – 2002 – 2004 – 2006 – 2008 – 2010 – 2012 – 2014 – 2016

Data

Sources

All US elections are actually administered at the state and territory level, mainly by executive officials titled the Secretary of State in each jurisdiction.

However for all US-level elections (President, House and Senate) data on final election results is collated early in the year following each election by the Clerk of the House of Representatives and published online. These statistical publications therefore form a compact and authoritative source. However more localised data, and data on primary elections, must be sourced from each state’s electoral authorities. In the United states a substantial industry exists for the collation of electoral data, much of which is not available free of change.

Data on populations of eligible voters and registrations is not publicly available in a consistent manner. The most comprehensive non-commercial online source is probably that of the United States Elections Project, maintained by Dr Michael McDonald, currently at the University of Florida.

[data completeness – anomalous contests – augmentation]

Dataset

[Datasets are not yet published]

• United States election results – 2000-2010
• United States election results – 2012-2016

This dataset covers all general elections from 2000 to 2016 for:

• the United States House of Representatives
• the United States Senate
• delegates to the Electoral College to elect the President of United States

The House district boundaries for these elections fall into three phases based on decennial reviews:

• elections of 2000
• elections from 2002 to 2010 – except in regard to the state of Texas, which underwent a boundary revision effective from the 2004 election
• elections of 2012 to 2016

The dataset contains the following data for each electoral division:

•  (in most cases) the number of persons who voted, even if their vote was not formal
• the number of formal votes cast
• the number of votes for each significant political party (defined as one with a vote share consistently above 5% in a significant number of divisions at the election or at elections closely preceding or following), and for all independents wining 5% or more of the first preference vote in their division
• the party of the winner in each division

The information which can be derived from the above data includes the following:

• the average enrolment for all divisions*
• the variance in the enrolment for each division from the mean enrolment
• the formal (and in most cases, the total) voter participation rate in each division*
• (in most cases) the informal vote total and rate in each division*
• the percentage share of the total formal vote of each party and significant  independent in each division*
• the percentage share of total enrolment represented by the vote for each party and significant  independent in each division*
• the proportion of the total enrolment that is represented by the winner of the seat
• (* also allowing state and national averages to be calculated)