February 2005

More than a century ago, John Stuart Mill explained how it was misleading to describe a winner-take-all election system as "majoritarian" because "a majority of a majority is not the same as a majority of the whole."

Here's what he meant. Suppose each representative has been elected by a majority in a single-member district. Suppose a very controversial bill passes by a margin of just one vote in the legislature (for example, 218-217 in the U.S. House). This legislation theoretically could have the support only of those voters who had elected the representatives supporting the legislation. If that were true, all the voters who did not help elect one of these representatives  perhaps 75% or more  would be denied their policy preference. That does not even reflect the fact of non-voters who did not participate at all.

Such a polarized breakdown of voters is unlikely, but the broader point is a powerful one:  that winner-take-all elections in no way ensure passage of policy supported by the majority. This potential is exacerbated in the United States by three particular trends: 1) voter turnout is extremely low, with far less than half of American adults voting in congressional races in off-presidential years; 2) most general elections are not competitive due to partisan imbalance and to how single-member districts are shaped in redistricting, which in turn makes the even-lower-turnout party primary the most important election in most districts; 3) more races are being won without a majority of the vote, particularly in primaries. Minority rule is even more likely to occur in the U.S. Senate, given its great divergence from the principle of one person, one vote because each state has two Senators, regardless of population.

We received the analysis below from a long-time Internet democracy activist who has compiled statistics on potential minority rule in the U.S. House of Representatives and the U.S. Senate based on the November 2004 elections and, in the case of the U.S. Senate, elections in 2000 and 2002 as well (one-third of the Senate and all House Members are elected every two years). Note that even though his analysis does not factor in the problems of our low turnout and non-competitive general elections, it still shows that even with all Members voting, a particular bill could pass the U.S. House with the support of only 26% of voters in 2004 and the U.S. Senate with the support of only 10% of voters in 2000-2004.

His charts also have interesting information about popular vote totals. For example, the total votes cast for the 96 Democratic candidates for the U.S. Senate in 2000-2004 in fact is greater than the total votes cast for the 97 Republican candidates for Senate in that period – yet Republicans hold 55 of 100 Senate seats.

See Also:
Minority rule potential, November 2000 Elections
Dubious Democracy 2000

U.S. House Elections, November 2004:
Minority Rule Percentage (MRP)

The U.S. House of Representatives has 435 House Members, each elected from a one-seat district for a two-year term. The chart below indicates that a bill could receive a majority of 218 votes in the House from legislators who received less than 27% of the votes from those at the polls in November 2004. A bill could pass with the votes of 218 Republicans whose share of the vote was barely 32%.

Total votes is based on the total valid votes cast in the presidential race in November 2004 (see chart below).

Chart on Votes Cast in November 2004 House Races

"Non-votes" in U.S. House races is determined by subtracting the number of votes cast in House races from the number of votes cast in the Presidential race. It does not include the people at the polls who did not cast a valid vote in the presidential race (which was more than 1% of all votes).

U. S. Senate Elections, 2000-2004:
Minority Rule Percentage (MRP)

The U.S. Senate has 100 Members. Senators serve six-year terms, with one-third of the Senate elected every two years. Two Senators are elected from each state, regardless of that state's population. The chart below indicates that a bill could receive a majority of 51 votes in the Senate from legislators who received only 10% of all votes cast in U.S. Senate races in 2000, 2002 and 2004. A bill could pass with the votes of 51 Republicans whose share of that total vote was less than 20%.

Total U.S. Senate votes is based on adding all votes cast in the 100 U.S. Senate races from 2000 to 2004. It does not include the votes of anyone at the polls who did not cast a valid vote in the U.S. Senate race. See chart below.

** Vermont Senator Jim Jeffords was elected as a Republican in November 2000 and is counted as one in this chart. He has since become an independent.

Chart on Votes Cast in U.S. Senate Races, 2000-2004

"Non-votes" in U.S. Senate races is determined by subtracting the number of votes cast in Senate races from the sum of all votes cast in the presidential races in 2000 and 2004 and all votes cast in Senates races in 2002. It does not include the people at the polls who did not cast a valid vote in the presidential race (which was more than 2% of all votes in 2000 and more than 1% of all votes in 2004).