Excerpt from Arrogant Capital
Disbelief in the results produced by the present party system is beginning to stir interest in wider-ranging reforms. One such debate is whether it's time for Anglo-America to give up the old first-past-the-post, single-member-constituency approach to electing national legislatures. Choosing legislators this way favors the two-party system, but with the old major parties losing appeal, the dilemma is stark: whether to further entrench the existing parties despite their failings or to accept the transformation and establish a system of proportional representation designed to sort out (but also facilitate) the emergence of new groupings.
Beyond Anglo-America, most nations already employ a form of proportional representation in their national legislative elections and accept the multiparty results. For the English-speaking countries to change in this direction would break with long-standing tradition, but debate is rising.
New Zealand shifted to the German system of proportional representation in 1993, when parliamentary elections were already increasing minor party strength. Gains for Britain's third-party centrists have spotlighted the issue there, and commentators have been predicting that introduction of proportional representation will come about in the 1990s to seal a coalition bargain between the Liberal Democrats and Labour. In Australia, a form of proportional representation is already used in elections to the Senate. And in Canada, the nation's leading newspaper, the Toronto Globe and Mail, has endorsed proportional representation as a way to sort out the increasing chaos of the party system. Interest in proportional representation has grown in the United States, too, but it is small by comparison. . . .
In order not to remain the only member of the Group of Seven relying solely on the restrictive first-past-the-post system of choosing legislators, Americans should at least begin thinking about how to modify our system in a proportional direction. This could be done boldly at the federal level by congressional action -- or piecemeal at the state level. Many possible variations exist, most of which sound complicated on first reading. Discussion will not mushroom overnight. Even so, one plausible approach would be to keep electing members of the Senate, two from each state, on today's single-winner basis because proportionalism would be unworkable, but to reorient the House of Represen-tatives by enlarging its membership to, say, 650 members from the present 435 and then establishing some variation of proportional representation.
Small parties would get little or no benefit in small states electing only two, three, or four House members. But in large states, the effect of proportional selection would be considerable. For example, in California, where minor parties are already making some impact in congressional races, a changeover would make the state's potentially enlarged House membership of seventy-five House members elected in 2002 come in a wider variety of flavors. As one hypothetical approach, the state could be divided into five large regional districts, each of which would elect fifteen House members. Any party that crossed (say) a 7% threshold would get one House member. So the San Francisco Bay region, for example, might elect eight Democrats, three Republicans, one Nationalist, one Rainbow/Peace and Freedom, one Green, and one Gay Rights. Diversity would flourish. The nature of interest-group access to the House would change: voters and voter blocs would at least partially replace lobbies and hired guns.
There are other approaches to proportional representation. One is the "additional member" system, by which a nation's voters elect a fixed number of legislators from single member districts but then also vote for a national candidate list, from which winners are selected in proportion to overall support for their party.
Still other related variations are not quite proportional representation, like "alternative voting" [majority preference voting] and the "single transferable vote" [preference voting] system. Alternative preference, used to select Australia's lower house in ordinary single-member districts, simply requires voters to rank candidates as their first, second, or third choices. If one candidate secures 50% of the first preferences, he or she wins. However, if nobody does that well, then the weakest candidate is dropped and his or her supporters' second-choice votes are distributed among the remaining contenders. Sooner or later someone will get 50%. But an opening-round plurality is not enough. Even for the mathematically inclined, these systems can seem cumbersome and thus undesirable. Yet all are used efficiently in some country or other, and in the computer age, all are now easy to manage, which allows national attention to focus on their other merits -- promoting a broader representation of parties and minority viewpoints, as well as the much greater turnout of voters that this breadth almost always encourages.
Were the United States to move toward partial proportional representation for these reasons, the major change, as we have seen, would concentrate in the House of Representatives, not in the Senate or in the presidency. But although the presidency isn't an obvious candidate for proportional representation, there are several possible indirect approaches. John Anderson, the 1980 independent presidential candidate (and subsequent proportional representation supporter), has suggested that he and other third-party contenders would have been helped by the use of an Australian-style alternative preference system of popular voting (without the electoral college).
In this system, as noted, voters rank the top group of candidates by preference. In 1980 an independent Republican voter choosing Anderson would presumably have marked his ballot Anderson (1), Reagan (2), and Carter (3). Had this option existed, Anderson contends, his support would not have sunk from 20% in the September polls to 7% on election day. People could have voted for him and still have cast a backstopping second-preference vote for Reagan or Carter, precluding any possible constitutional malfunction.
That might indeed have allowed the third-party vote to remain high instead of being forced back into a two-party preference. This option, too, is worth considering.
Interestingly, even the unpopular electoral college unfolds avenues to proportional representation. Either legislation by Congress or enactment of a constitutional amendment could require states to split their electoral votes. But the states do not need to approach Washington to allocate their electoral votes any way they want. Each can cast them en bloc for its winner, award one for each congres-sional district carried or allocate its electoral vote proportionally.
And the itch is spreading. Maine now splits electoral votes by congressional district, Florida has considered the same approach, and in 1993 the state of Washington held hearings on a bill to allocate its eleven electoral votes proportionally.
Had this last system been in place in 1992, Bill Clinton would have won five electoral votes in Washington, George Bush three, and Ross Perot three. Indeed, had the electoral votes of all fifty states been awarded on this basis, no 1992 presidential candidate would have won a majority in the electoral college, sending the choice of the president to the House of Representatives. To take that one more step: were the House forced to choose the president with any frequency, our system would start to turn quasi-parliamentary.
The reader may sense that the previous five or six paragraphs have not produced firm recommendations. Just so. I am inclined to think that some form of proportional representation makes sense for the House of Representatives, not just because of its disruptive effect on existing interest-group relationships but because it would also reduce or eliminate the flagrant racial gerrymandering of the early 1990s. It is no coincidence that after the U.S. Supreme Court cast doubt on the legitimacy of using grossly distorted boundaries to create black-majority districts in 1993, interested parties and federal judges began discussing the alternative usefulness of proportional and related approaches like cumulative voting; if a county board, for example, had five seats, voters would be allowed to make five choices, with the option of picking five different candidates or cumulating all five votes for one contender.
As for abolishing the electoral college and choosing the president through popular vote, that option would gain merit if the process also involved selecting first, second and third-place choices so that the second-place votes of the weakest candidate could be redistributed if nobody got over 50%. Declaring a winner based on a 38% popular-vote plurality is unwise and unnecessary. But then again, some of the possible changes in the electoral college could also make that ancient mechanism more friendly to twenty-first-century circumstances...
Public support for reform will depend on how much parties and independent candidates proliferate in the presidential elections of 1996 and 2000. Continuing electoral fragmentation in the United States, while the other English-speaking countries deal with their own party changes by adopting some degree of proportional representation, would turn this country's preliminary discussion into a full-fledged national debate.
This article is from Arrogant Capital (Little, Brown and Company, 1994) by Kevin Phillips. Phillips is editor of American Political Report and author of several books on American politics.