Features Commentaries on Reform by CVD's Rob Richie and John Anderson

January 24, 2001

The news web site featured articles by The Center for Voting and Democracy's leaders and other voting system experts in its "op-ad" that ran simultaneously in the New York Times and on its web site.

CVD executive director Rob Richie writes about reform of the Electoral College, CVD president John Anderson makes the case for instant runoff voting, Behind the Ballot Box author Douglas Amy writes about proportional representation and Australian scholar Ben Reilly provides an international perspective on instant runoff voting.

See the commentaries by Richie and Anderson below. For the full set of articles, go to

Meddling with Reform

By Rob Richie, The Center for Voting and Democracy
January 24, 2001

If you thought Election 2000 was distorted by partisanship, keep your eyes on the push for electoral reform.

Summary: The nationís outrage with the electoral system is catalyzing long-needed reform, but will partisans exploit this opportunity for their own advantage?

Letís call it ëFloridamokí the protracted partisan battle that yielded Electoral College victory for George W. Bush despite Al Goreís half-million ballot lead in the national popular vote. The resulting furor has spurred a national reexamination of the quirky Electoral College system by which we elect our presidents.

But reformers, beware! Partisans from both major camps can see opportunity in the outrage inspired by Bush vs. Gore: the chance to hijack the reform impulse and manipulate it to their own advantage.

Take Grover Norquist, for example. Heís the renowned conservative strategist whose group, Americans for Tax Reform, is to some observers synonymous with ëGrand Old Party.í In December Norquist revealed to the National Review how partisans could tweak electoral reform toward his favorite team. He suggested that Republican-controlled states like Michigan, New Jersey and Pennsylvania ñ all of which gave their electors to Al Gore in 2000 and to Bill Clinton in 1992 and 1996 ñ switch to allocating electoral votes by congressional district, thereby assuring that Republicans would win at least some electoral votes in those states in 2004.

Norquistís suggestion reminds us that individual states are vested by the Constitution with the sole power to decide how electoral votes are allocated. Electoral reform will be hammered out in the state houses, subject as they are to control by one party or the other - and to partisan shenannigans.

The question is whether statehouse leaders will take advantage of the current reform climate to seek to benefit one party over others or to benefit all voters and the nation as a whole.

The Fixes

There are several potential fixes to the presidential election system.

The most obvious one is scrapping the Electoral College in favor of direct election by popular vote, just as we elect nearly every other office in the nation. Direct election is a pre- condition to full political equality in presidential elections. Only direct election can ensure that all votes count equally no matter where people live. Only direct election can provide clear incentives for campaigns to give at least some degree of attention to every voter instead of only those in selected states.

Despite these advantages and consistent majority support in the polls, however, direct election is for the moment unlikely. Such a significant Constitutional change would take strong bi- partisan support -- currently not in evidence -- to address some small-population states' mistaken worry that they would be ignored in national elections under a direct vote.  But states right now can take three significant steps to mend the Electoral College without ending it: instant runoff voting (IRV) and allocation of electoral votes by proportional representation (PR) or by congressional district, as Norquist suggests. There are important differences, including the degree to which each reform creates partisan advantage.

Of the three proposals, instant runoff voting is least subject to partisanship and would do the most for a state's voters It would maintain winner-take-all allocation of electors, but at least ensure that the winner has a clear majority by simulating a traditional runoff election.

Hereís how IRV works: Rather than just vote for a single candidates, voters have the option to rank the candidates in order of preference: first choice, second choice, third choice. If a candidate receives a majority of first choices, he or she wins. But if not, the candidate with the fewest votes is eliminated -- thus failing to advance to the runoff -- and a second round of counting occurs. Ballots cast for the eliminated candidate now count for the votersí second choice on each ballot, just as if those voters had come back to the polls for a runoff election. Rounds of counting continue until candidate wins a majority of valid ballots -- which always will happen by the time the field is reduced to two, just as in a traditional runoff election.

Instant runoff voting would eliminate all talk of "spoiler" candidacies. It would create the situation where, as independent presidential candidate John Anderson has put it, a voter can vote for his favorite candidate without fear of electing his least- favorite candidate.

Well-tested from decades of use in Australia and Ireland, IRV would only require states to purchase modern voting equipment that can allow voters to rank-order the candidates -- purchases many states expect to do anyway in the wake of the controversy in Florida. It already is under serious consideration for presidential elections in such states as Vermont (where the proposal is supported by the League of Women Voters, Grange and Common Cause, among others), Alaska (where it will be on the statewide ballot in 2002) and New Mexico.

ëProportional representationí allocates electoral votes in proportion to the statewide popular vote rather than by the current "winner-take-all" method. Supported as a national change by both Franklin Roosevelt and Richard Nixon during their presidencies and already used in most presidential primaries and in legislative races in most well-established democracies around the world, proportional allocation would ensure that more voters can help the candidate of their choice win electoral votes. If a candidate were to win 40% of the popular vote in the state, that candidate would win 40% of the state's electoral votes -- a truer reflection of voter intentions than the funhouse mirror generated by winner-take-all. A minimum threshold of support necessary for candidates to win electoral votes could be set to allay concerns about splinter candidates gaining electors despite having no chance to win.

The congressional district plan breaks up statewide winner- take-all majorities differently. Already used by Maine and Nebraska, it assigns electoral votes according to the results in each congressional district in a state, with the two Senate electors going to the overall statewide winner (states get one electoral vote for each senator and representative). Although one candidate has won all electoral votes in each election in Maine and Nebraska since they adopted the system, bigger states typically will have a mix of districts that lean strongly toward one party or another. Candidates without a chance to win the statewide vote might still be able to win certain districts. Third-party candidates would be unlikely to win electoral votes -- Ross Perot did not come close to winning any congressional district despite winning nearly a fifth of the popular vote in 1992.

Enter Grover

The problem with the last two reforms is that, unlike IRV, they have very different partisan impacts based on whether states enact them alone or in coordination with other states. Whatever genuine civic support there might be for them, state leaders are unlikely to pursue them on mere philosophical merits lest they put their state at a disadvantage in national elections. Thatís because any reform likely to guarantee a more-equal division of electoral votes in one state could lead presidential candidates to divert their resources to competitive winner-take-all states where campaign energy could swing a state's entire bloc of electoral votes. This phenomenon is common in Republican presidential primaries, which now use a mix of approaches but are generally moving toward statewide winner-take-all allocation in order to compete with other winner-take-all states.

More importantly, reforms that allocate electors proportionally or according to congressional districts are vulnerable to political manipulation at the state level. Although there are well-intentioned reformers who support these changes, they also draw the attention of deeply partisan strategists - enter Grover Norquist.

Suppose one party controls both the legislature and governor's mansion in a winner-take-all state. But suppose also that this party's presidential candidate has a good chance of losing the state. State leaders suddenly can become interested in "fair" allocation of electoral votes because half a loaf - some proportion of the state's electors - is better than none at all.

Given the partisan nature of how congressional districts are created in redistricting, allocation by congressional district is particularly problematic. Most districts strongly favor one party or the other, often due to political gerrymandering to protect congressional incumbents -- incumbents' 99% re-election rate in 1998 and 2000 is no accident.

Problems with the congressional district approach are exemplified by Pennsylvania, one of Norquist's proposed targets. When the presidential vote is counted statewide, Democrats do well -- over the last three elections, Democratic candidates won the stateís 23 electoral votes by comfortable margins. But Democratic voters are relatively concentrated in a few areas, and when it comes to the stateís U.S. House districts, there is a decided Republican tilt. In 1996 Bill Clinton ran behind his statewide vote average in 13 of 21 districts, and despite Goreís victory in the state in 2000, George W. Bush probably won a majority of the districts in 2000 (presidential results by district will not be available until this spring).

This Republican edge exists now even before Pennsylvaniaís Republicans have a whack at redistricting later this year. By tweaking a few districts here and there, they could (and probably will) enhance their control of more districts. This makes it quite possible that under a congressional district allocation method of choosing presidential electors, a Republican candidate could win most of the state's electoral votes even when losing the statewide popular vote. Michigan Republicans might have a similar opportunity, given that they control redistricting and Democratic votes are heavily concentrated in Detroit.

Allocating electoral votes by congressional district would give a distinct advantage to Republicans if adopted nationwide, although the approach would not benefit Republicans in every state -- which is why some states controlled by Democrats, particularly in the South, may seek to change statewide winner- take-all rules. . The Democratic vote typically is concentrated in urban areas -- a concentration that led Democrats to win 24 of the 26 two-major-party congressional races won in 2000 with more than 80 percent of the vote. George W. Bush almost certainly won more House districts than Al Gore in 2000. Combined with his victory in ten more states (giving him 20 more electoral votes than Gore under the district system), he would have won a comfortable electoral college win with the system despite his losing popular vote. If the method had been used in 1960, Richard Nixon would have beaten John Kennedy.

Besides opening the door to gerrymandering, the congressional district method has other problems. Just like the current system, it leaves most voters in a position where their vote won't matter much in a nationally competitive election, as most people will live in a state and a congressional district that tilts strongly to one major party candidate. Thus, all the campaign resources still will be focused on a relatively small portion of the electorate - the "swing vote" in the swing districts and states. In addition, unless combined with instant runoff voting, congressional district allocation maintains the problem of candidates winning with less than a majority of the vote due to "spoiler" minor party candidacies.

The simplest, most powerful change for electing our president would be direct election combined with instant runoff voting to ensure that the winner represents as many Americans as possible. Instant runoff voting for now can be pursued state by state. Proportional allocation of electoral votes is worth serious consideration if done all at once across the nation, but that would require Constitutional change. Allocating electoral votes by congressional district is the most problematic reform. It shares the downsides of proportional allocation, but has the significant additional problem of perverting fair results due to partisan gerrymandering.

So, when partisans like Grover Norquist suddenly take interest in this "fair" method or electoral reform, watch out. The partisan spirit of Floridamok has not subsided, and it may distort the reforms demanded by the voting public.

A Clear Majority Winner in 2000
By John B. Anderson

As the nation grows accustomed to George W. Bush having captured Florida's electoral votes and, with them, the presidency, we must not overlook an important fact. In 2000, a majority of valid, unquestionably legal votes in Florida -- and indeed in the nation -- were cast for either Democrat Al Gore or Ralph Nader. Given that most Nader voters preferred Gore to Republican George W. Bush, there is little doubt that in a one-on-one race on November 7, Al Gore would have defeated Bush in Florida. But because of Nader's candidacy and our antiquated plurality voting system, that majority vote for Gore was fractured. As a result, Bush is our president.

Some argue that Nader "spoiled" the election for Gore. The spoiler charge that plagued Nader's campaign this year is a familiar one. In 1992, Ross Perot undercut the re-election bid of Republican George Bush. In 1980, when I ran for president as an independent after abandoning the Republican primaries, I was labeled a spoiler even when polling near 25 percent. My candidacy was said to deprive voters of the clear choice between Republican Ronald Reagan and Democratic incumbent Jimmy Carter. Never mind that my platform clearly attracted many people uncomfortable with this choice, just as Ross Perot, Ralph Nader and other third party candidates have attracted genuine support.

I sympathize both with the major party candidates who don't want to have their candidacies spoiled and the minor party candidates who want to be considered as more than spoilers. On the one hand, it is understandable that major party candidates don't want to lose only because some of their constituency support a third party candidate. We should not be surprised when they seek to keep third party candidates off the ballot and exclude them from debates.

But such exclusionary practices must not be the answer to the spoiler dilemma, particularly given our shrinking voter participation. Having more candidates run in general elections and take part in debates strengthens democracy. As indicated by Perot's impact on the 1992 presidential campaign and Jesse Ventura's impact on Minnesota's gubernatorial election in 1998, third party candidates can increase turnout and interest in elections. More people watch debates, more ideas are discussed, and, ultimately, more people vote.

Fortunately, there is a way to resolve this conflict between the self-interest of major party candidates and democratic principles. There is no need to accept electoral rules in which voting for your favorite candidate can contribute directly to the election of your least favorite.

The root of the problem is that, unlike in most democracies, here the candidate with the most votes wins all, even if opposed by a majority of voters. By allowing candidates to win without a majority of the vote, third party candidates and independents can be "spoilers" if they split a major party candidate's vote.

One way to resolve the problem is allow two rounds of voting. In the first round, all candidates run. If no candidate wins a majority, then the top two candidates face off in a "runoff." In a two-person race, a candidate cannot win without obtaining a majority.

There is another, better solution that is practiced in London, Ireland and Australia: instant runoff voting. Any state could implement instant runoff voting immediately for its federal elections, including the presidential race. There are significant efforts underway to enact instant runoff voting for federal elections in such states as Alaska, New Mexico and Vermont. Instant runoff voting also would be the best way to conduct a direct election for president.

Here's how it works. In instant runoff voting, people are given the freedom to do more than vote for just one candidate. Instead, they can rank the candidates in order of preference: 1, 2, 3. If a candidate receives a majority of first choices, the election is over. But if not, the ballot-counters simulate a runoff by eliminating the candidate with the fewest votes In this round, all the ballots cast for a remaining candidate stay right with that candidate, while the second choice of voters who cast ballots for the eliminated candidate are now tallied. If this new count results in a majority winner, the election is over. If not, the next weakest candidate is eliminated, and everyone's remaining top choice is counted. The rounds continue until there is a majority winner -- there always will be a majority winner once the field is reduced to two, just as in a traditional runoff election.

With instant runoff voting, we would determine a true majority winner in one election - avoiding all the extra expense, campaign financing and hassle of a second round of votingí. And we would forever banish the spoiler concept. Voters would not have to calculate possible perverse consequences of voting for their favorite candidate. They could vote their hopes, not their fears.

With instant runoff voting, liberals who liked Nader but worried about Bush could have ranked Nader first and Gore second. Similarly, some conservatives could have ranked Reform Party candidate Pat Buchanan first and Bush second. Rather than contributing to Gore's or Bush's defeat, Nader and Buchanan instead could have better stimulated debate and mobilized new voters that would have helped the major party candidate closest to them in political views.

Our primitive voting system was this year's real spoiler. Instant runoff voting would give us a more participatory, vital democracy where candidates could be judged on their merits and the will of the majority could prevail. And this year, we would have settled the election at the ballot box, not in the courts.

[John B. Anderson is the president of the Center for Voting and Democracy ( in Takoma Park, Maryland. He served in Congress from 1961 to 1981 and was an independent presidential candidate in 1980.]