Why I left the Libertarian Party and joined the Republicans
By Paul Hager, ©2002,2003



For an extremely abbreviated statement of the above, I direct the reader to the press release elsewhere on this site. The brevity of the press release was dictated by necessity: long, involved releases aren’t going to get disseminated by the media. Thus, I could only devote one paragraph (the fourth) to the topic of "democratic voting reform" and the Libertarian Party’s rejection of same, despite its being the principal reason I left. Now, however, my only constraints in addressing this issue are your interest and patience. In what follows, I will endeavor to abuse neither.

Why I joined

Nearly everyone who joins a "third party" in the U.S. will do so because of profound disillusionment with the dominant two party system. That was certainly my reason when I first joined the Libertarian Party in 1976. At that time I also thought there was a decent chance the Libertarian Party could grow and either become a major party or have such a significant impact that one or both of the major parties would co-opt most of its platform, as had happened with the Socialists in the early part of the 20th century. After the poor showing of the 1980 Libertarian Presidential candidate, Ed Clark (still the best showing to date), it became apparent that the party wasn’t attracting enough support to realize either goal. Consequently, I switched to the Democrats as the party with the best chance of winning on the particular set of freedom issues I considered most important at the time. It turns out that I wasn’t the only Libertarian to desert after the 1980 debacle. A number of prominent Libertarians (including David Koch, the 1980 VP candidate) left to join the Republican Party and/or become active in the Cato Institute, a philosophically libertarian think tank.

By the early 1990’s, I was getting fed up with the Democratic Party, which seemed to have abandoned its commitment to civil liberties. I had also become even more philosophically libertarian over the intervening years. Two events finally tipped the scales in favor of rejoining the Libertarian Party. The first was Ross Perot’s showing in the 1992 Presidential election. Not only did Perot garner the second highest third party vote percentage in the 20th century (behind Teddy Roosevelt’s showing in 1912), he did it while avoiding the usual erosion of support that afflicts third party candidates as the election approaches. Perot’s support actually increased.

The second event didn’t involve the U.S. at all. In late 1993, the dominant Progressive Conservative Party of Canada went from 178 seats in the 295-seat Parliament to 2 seats. Its position on the center-right was taken over by the Reform Party, a regional party dominant in British Columbia and Alberta. The possibility that a major political party could disintegrate and be replaced had been one I hadn’t seriously entertained. The last time it had happened in the United States was when the Republican Party supplanted the Whigs in 1854.

Was the U.S. ready for a major political realignment? There was some data suggesting such a thing was not impossible. Voter turnout had been declining since the 1960’s and polling data showed widespread frustration with the two party system. Most significantly the polling data also showed substantial public interest in the emergence of a third party.

With the above facts in hand, in early 1994 I decided to give the Libertarian Party another go.

My experiences as a candidate

For my first act as a newly minted Libertarian, I chose to run for U.S. Representative in Indiana’s 8th district. At that time the Libertarian Party did not have automatic ballot status, which meant that candidates had to petition to get on the ballot. For a Congressional candidate the requirement was around 3,000 or so valid signatures. Because a large proportion of signatures will be disallowed for various technical infractions, in practice it is best to secure close to twice the required number. For this and other reasons, experience has shown that volunteer petition drives often come up short, which was what happened to me – I didn’t make the ballot. That same year, Libertarian Steve Dillon managed to get just over 2% of the vote in the Indiana Secretary of State race. This gave the Libertarian Party automatic ballot status for the next four years.

In 1996, with ballot status assured, I ran again. Perhaps because of the novelty of it, I was invited to participate in all of the candidate debates. Another factor may have been that the Republicans and Democrats were unsure as to whom I might take votes from and didn’t bother to raise objections. I got a lot of exposure and generated some favorable press. One example: the 8 October 1996 Bloomington Herald-Times, editorializing about the three-way candidate debate said, “Libertarian Hager was the only likeable, civil fellow among the three”, and “[W]e bet he won some votes.” I ended up getting less than 2% of the vote, though it was more than the margin separating the winner, Republican John Hostettler, from the Democrat loser, Jonathan Weinzapfel.

I ran again in 1998 with every expectation of improving on my 1996 result. My 1996 campaign had generated interest in the Libertarian Party, as evidenced by some new county organizations forming in the 8th district along with growth of the established ones. The growth of the Evansville Libertarian organization was nothing short of spectacular, with the first meeting of 1997 consisting of over 40 people (in 1996, there were typically 5). I was the featured speaker and was asked, more than once, if I was going to run for Congress again in 1998. When I did choose to run, I was therefore able to attract enough people from all around the district to field a real campaign staff with a dedicated campaign manager. Most importantly, I had developed some important media contacts, which improved the chances I could get my message out. Despite, or perhaps because of these apparent advantages, I found myself shut out of nearly all of the candidate debates. Considering the amount of time and effort that my many supporters put into the campaign, the fractional percentage improvement over 1996 was a bitter disappointment.

On the whole, 1998 wasn’t particularly good for the Libertarian Party nationally either. The party’s best hope for a good showing was in the Indiana 6th Congressional District where a mentally disturbed, convicted felon had managed to win the Democratic primary and was on the ballot. The Libertarian candidate still finished in last place with around 11% of the vote. Around the country, in three-way races Libertarian candidates got buried, typically receiving around 2% of the vote. The few Libertarian victories occurred in two-way races, where one of the majors didn’t participate, or in at-large races for positions like city council or school board. The performance of Libertarians around the country indicated that the two party system was as entrenched as ever.

The lack of any observable progress was somewhat perplexing since polling data showed a significant part of the electorate was philosophically libertarian. Furthermore, politicians espousing libertarian ideas like drug legalization and the right to keep and bear arms (in the same campaign) had won elections, though not as Libertarians. Most notable were Ron Paul of Texas, who had left the Libertarian Party (he was its 1988 Presidential candidate) to run for Congress as a Republican and won, and Republican Governor Gary Johnson of New Mexico, who had made national news calling for an end to the drug war.

Although I was expecting to see some evidence of progress, I also knew that the ultimate solution for the Libertarian Party lay in what I term "democratic voting reform". By this, I mean changing the voting system so that the method we use to elect candidates no longer favors a two party system. I had mentioned changing the voting system in both of my Congressional campaigns (note the next-to-last paragraph in this press release from my 1998 campaign), suggesting that Indiana adopt a voting method known as approval voting.

Since the next part of my narrative deals with the subject of voting reform, this is a good place to interrupt my account and discuss some important theoretical concepts relating to voting and what is known as social choice theory.

The narrative interrupted: why we have a two party system in the U.S.

For essentially all of its history, the U.S. has had a two party system. Why? Why did the two party system arise, and why are the two parties seemingly impervious to serious challenge? It is certainly not mandated by law – there is nothing that says, "thou shalt have only two parties."

One clue to the answer is to consider how a voter deals with the situation where three candidates are seeking the same office. If the voter only likes one candidate, then the choice is easy: vote for that candidate. Things become complex when the voter finds two of the three candidates acceptable and is forced to choose between them. While it will often be the case that the voter will have a clear preference between the two, choosing the preferred candidate is often the wrong strategy. Actually, if the voter knows that one of the two candidates is more likely to win, the best strategy is to vote for the more popular of the two candidates irrespective of preference. This is because sticking with an unpopular first choice risks helping to elect the unacceptable candidate. This strategy is simple and intuitive. It is so common that we have labeled it, "voting for the lesser of two evils."

The mathematics of game theory, developed in the mid-20th century, allows us to rigorously evaluate an intuitive strategy like "lesser evil". In game theoretical terms, an election is a non-cooperative game. In other words, there is no practical way for all voters to get together and coordinate their strategies. In the U.S. electoral system, the rules of the game are that every voter gets one vote and whichever candidate gets the most votes wins. This system is known as plurality voting. An optimal strategy in a non-cooperative game is known as a Nash Equilibrium. In plurality voting the equilibrium strategy is, in fact, "lesser evil." This equilibrium strategy, in turn, serves as the cornerstone of what is known as Duverger’s Law. This law states that plurality voting in single-member districts will tend to produce a two party system.

To explain how Duverger’s Law works, consider the following. An election is the aggregation of individual voter choices, which are themselves representative of individual voter strategies. Over a series of elections involving three (or more) parties, voters who choose the "lesser evil" will operate both to select two dominant parties and iteratively reinforce their dominance. In practice it works something like this. After a first election, voters will see from the result which two parties are ranked first and second. In the next election, these two parties will be known to the voters as being popular choices and will therefore be beneficiaries of the "lesser evil" or equilibrium strategy. Any difference between second and third place will be magnified in successive elections, siphoning support away from the third party to the more successful party that is most like it. The effect is even stronger now than in earlier decades, because the process I just described is characteristic of the period before scientific polling, when the only sure way to gauge popularity was by the result of an election. Polling reliably informs voters which candidates/parties are popular ahead of election day. This makes any sort of breakthrough by a third party much more difficult today than was true in the 19th or early 20th centuries.

Duverger’s Law and U.S. history show that third parties almost never achieve major party status because the only avenue for success is replacing one of the major parties. The most dramatic and least likely possibility is for one of the two major parties to disintegrate outright, allowing a well-situated third party to take its place. A somewhat more likely possibility is for a third party to become the regional beneficiary of Duverger’s Law. This entails supplanting one of the majors in a single state or group of states and then building from there.

Practically speaking, the regional approach is nearly as difficult as a nation-wide coup. Adding to the problems faced by a third party in supplanting one of the majors is the suffocating effect of new campaign finance regulations limiting party-building contributions (for a discussion, see this press release from my 2002 Secretary of State campaign).

The Canadian experience in 1993 shows that it is not impossible for a third party to replace a major. There are, however, critical differences between the U.S. and Canada (which I didn’t fully appreciate in 1993) that militate against the same sort of thing happening here. One is that regionalism is a major factor in Canadian politics, unlike the U.S. where regional differences play, at most, a minor role. Exacerbating Canada’s regional differences is Quebec, a large proportion of whose dominant francophone population desires independence. A regional third party can exploit the situation where two major parties are seen to advocate national over regional interests, forcing the two majors to split the vote. Alternatively, a regional third party that is ideologically similar to a major party may exploit a temporary weakness to supplant it locally. A second difference is that Canada has a parliamentary system, in which the executive is chosen by the parliament. Voting for a member of Parliament is also an indirect vote for the Prime Minister as well. In contrast, the executive is a distinct branch of government in the U.S., and it is common for voters to "split" their tickets by voting for one party’s candidate for president but for the other party’s representative in Congress. Hence, the "coattail" effect (and its converse) is more pronounced in Canada than in the U.S. In 1993, widespread disgust with the campaign tactics of the Progressive Conservative Prime Minister Kim Campbell redounded disastrously to Progressive Conservative parliamentary candidates as a whole. Voters in the Western provinces defected to the regional center-right Reform Party as an acceptable alternative. In Quebec, voters had an additional reason to defect from the Progressive Conservatives: the Liberal Party leader, Jean Chrétien, who had been the object of a thoroughly repellent personal attack by the Campbell campaign, was himself French-Canadian.

As the foregoing shows, a U.S. third party whose strategy is to wait for a major party to do a Canada-style meltdown is in for a very long wait. Fortunately, there is a much better approach that is almost certain to allow multiple parties to flourish: change the rules of the game so that Duverger’s Law no longer applies.

How to break Duverger’s Law

Two alternatives present themselves. Both would involve replacing plurality voting with some other method. The first is to adopt proportional representation for legislative bodies. With proportional representation, a party is guaranteed to elect representatives if it achieves some threshold of support (for example, in Germany the threshold is currently 5%). Many (but not all, Canada being a case in point) parliamentary systems around the world have adopted some form of proportional representation. Experience and game theory demonstrate that a multiparty system will evolve when proportional representation is used. Unfortunately, proportional representation yields a number of unpleasant consequences that make it a poor solution, as I will indicate in due course.

The second alternative is to only change the voting method, while maintaining the current U.S. system of single-member legislative districts. The two best choices for a replacement voting method are approval voting and Condorcet voting. With approval voting, people can vote for more than one candidate for a given office. All votes are tallied and the winner is the candidate with the most votes. Approval voting eliminates the split-vote and spoiler problems that force voters into choosing the "lesser evil" strategy when three or more candidates are on the ballot. (See The Approval Voting Home Page for additional discussion and technical references.) For 500 years, the Venetian Republic used approval voting to select the Doge (leader) from among up to ten contenders. Today approval voting is used by eight professional engineering and scientific organizations world-wide with a collective membership of 600,000 (see Why the Academy Awardsâ may fail to pick the "best picture" – again for a suggested ninth organization).

Though approval voting is a very good voting method, the Condorcet method is generally regarded as the best. Condorcet voting takes its name from the 18th century French mathematician, the Marquis de Condorcet, who developed it. The Cordorcet method allows voters to rank order their preferences. The ranking information is then used to generate a series of head-to-head match-ups between every candidate. Thus if A, B, and C are running for office, Condorcet voting will produce the following match-ups: A-B, B-C, and C-A. The winner is the candidate who beats all others head-to-head. Condorcet’s is the only system that will always find the majority winner when three or more candidates are running. This is the reason that social choice theorists often call the majority winner the "Condorcet winner".

While Condorcet’s is the best voting method, and approval is a good runner up, it should be noted that there is no perfect voting system. Condorcet was the first to show that cyclical majorities (A beats B, B beats C, C beats A) were possible (Condorcet voting uses a tie-breaking method to find the winner when no majority winner exists1). Condorcet’s work was later extended by Kenneth Arrow in his famous impossibility theorem. The fact that perfection in social choice doesn’t exist shouldn’t trouble us. Our focus should be on choosing systems that have the best probability of producing a democratic result under the broadest range of circumstances.

I refrained from saying much about proportional representation above because, even if it were adopted, proportional representation would not obviate the need to use either approval voting or Condorcet voting for U.S. Senate, President, and state and local executive candidates, which are all single-winner. At the federal level, proportional representation could only apply to the lower house of Congress. Beyond that, trying to achieve a multi-party system in legislative bodies by using proportional representation (instead of applying approval or Condorcet to our system of single-member districts) has a number of practical and theoretical problems, which I will address now.

Weaknesses of proportional representation

Pure proportional representation, as used in Europe and elsewhere, employs a voting method (there are several) that will produce what is, in effect, a "representative sample" of the political preferences of the electorate. For example, if the voting population self-identified as 25% Republican, 28% Democrat, 17% Libertarian, 8% Green, and the remaining 22% independent, then proportional representation should produce an electoral outcome that would approximate that breakdown in the party affiliations of the representatives (the percentages, incidentally, are reasonable based upon opinion polling in recent years). In order to accomplish this, it is necessary for candidates to be selected at-large, either regionally or nationally. In the U.S., the Constitution would require that at-large elections for U.S. Representative be in individual states. Obviously, proportional representation schemes would achieve their best results in larger states.

What proportional representation does very well is to elect candidates with minority political views to a legislative body. This is both a strength and a weakness. Minority ideas may be innovative and of societal value but they may also be completely crazy. A political system must be open enough to allow a fair discussion of all sorts of ideas, whether crazy or not. However, proportional representation effectively takes the discussion of ideas out of the electorate and places it in the governing body. It does this by guaranteeing that any minority party meeting the selection threshold will be able to place some of its candidates in the legislature. For example, if a political party that advocates everyone with blue eyes should be shot gets 8% of the vote, and the threshold is 5%, then some of this party’s candidates will be elected to the legislature. In contrast, in a system of single-member districts using approval or Condorcet, the anti-blue eye party would never find majority support and its 8% showing would elect no one to a position of responsibility in the legislature. The fact that proportional representation indiscriminately rewards minority ideas no matter how extreme or crazy, explains why parliamentary systems that use pure proportional representation often exhibit instability, with numerous changes of government. Even though any particular extreme views are held by a small minority of representatives, extremist parties that are necessary to the formation of a ruling coalition can join votes of no confidence and bring a government down. For this reason extremist parties can and do tilt a government toward more extreme polices.

The world has seen the worst case scenario of proportional representation in action with the Nazi takeover of the Weimar Republic. Members of the Reichstag were elected by strict proportional representation, which allowed the Nazis to gain around one third of the seats. This, in turn, positioned the Nazis to exploit the system and have Hitler named Chancellor. Ironically, the Nazi power grab happened even though the Nazis had lost seats in the last election. Had the Weimar Republic used single member districts, along with either Condorcet or approval voting, Nazi candidates would have seldom been majority winners since we know that around 70% of the voters were really anti-Nazi and voted for one or the other of the two major alternatives. With many fewer Nazis in the Reichstag, the Nazis would have never been in a position to take over the government.

With respect to the Nazi takeover, it is worth noting that after WW II the Germans abandoned their Weimar Republic system of proportional representation in favor of a mixed system in which half of the members of the Bundestag are selected in single-member districts by plurality voting and half are selected by proportional representation. Perhaps the aforementioned problems with proportional representation explain the fact that the German system has, in recent years, been adopted by a number of countries around the world to replace their own pure proportional representation systems. (France actually went so far as to abandon proportional representation altogether and switch to a runoff method for electing representatives to its parliament.) What is surprising is that none of these countries considered fixing their systems by adopting either approval voting or Condorcet voting for their single-member districts.

Aside from the instabilities that inhere to rewarding political extremism, proportional representation also produces other undesirable effects. One is that it completely eliminates representation of local interests. This further promotes government centralization. Such a result would have been anathema to the framers of the Constitution. (References by advocates of proportional representation to James Madison as a way of suggesting that he would support the system misrepresent history. Madison’s use of the term "proportional representation" referred to a state’s representation in Congress being proportional to its population and had no other meaning.)

To better understand why approval or Condorcet voting with single-member, local representation is a much better solution that any sort of proportional representation scheme, let’s consider a central question of democracy: how direct should citizen control of the governmental process be? By way of an answer, consider a group of people who are choosing between two alternatives, one of which is "correct". It can be shown mathematically that if the average member of the group has a better than 50-50 chance of making the correct choice, the probability that the group as a whole will make the correct choice approaches certainty as the number of people in the group increases. This is known as the Condorcet Jury Theorem, in recognition of the fact that Condorcet was the first to prove it. While the Jury Theorem might seem to argue for increased use of the referendum process or even pure democracy, the flip side is that a large group is certain to make the wrong choice if the average member has worse than a 50-50 chance of being correct. What it amounts to is that if a large group is well informed about a set of alternatives, it will make right choices; if it is not well informed it will make wrong ones. In the real world, direct democracy can be expected to produce large numbers of wrong choices whenever the average voter is unable to become sufficiently informed on complex policy matters to render a correct judgement. Hence, the referendum process should be used sparingly, if at all.

To increase the odds of producing correct choices requires a dedicated group of individuals who spend all of their time becoming informed on complex policy matters. When this dedicated group is selected by the general public the system is called representative democracy. In order for the average member to decide better than a flipped coin, this group should be composed of individuals who are as rational, well-informed, and independent-minded as possible. In the real world, the dedicated group – the legislature – achieve their positions via political parties, to which they owe fealty to a greater or lesser degree. The more individual representatives feel obligated to a party for the positions they hold, the more likely they are to abdicate independent judgement in favor of party discipline and party ideology – in other words, the more partisan they become. To the extent that discipline replaces deliberation as the dominant value among partisan representatives, decisions will be made by a handful of ideologically driven parties. According to the Jury Theorem, a few parties that act as single entities rendering ideological rather than rational judgements will produce bad choices much of the time, which undercuts most of the expected benefits of a dedicated group of individuals.

Remember that proportional representation selects candidates at-large as a way of generating a "representative sample" of the electorate’s party preferences. Since selection is primarily on the basis of party, the end result is strong parties and extremely partisan candidates. Aside from being theoretically non-optimal in a decision making group, these are traits that, according to opinion polls, American voters find most objectionable in political candidates.

The fundamental problem is that proportional representation, under the rubric of "representative democracy," has the perverse effect of removing true democratic choice from the electorate in the matter of whom it selects to represent it. The only justification for such a result – based up the Jury Theorem – would be if the average voter is worse than a coin flip in judging the quality of political candidates. So, whether stated or not, proportional representation assumes that voters really are incompetent judges when it comes to choosing their representatives.

As noted above, we know that the American electorate generally values traits that are desirable in a deliberative body. The problem is not with the judgement of the voters; the problem is that the ability of the voters to render that judgement is derailed when it faces three or more choices. This is where approval or Condorcet voting comes into the picture. Both will find the majority choice (when there is one) in each legislative district that most times, according to the Jury Theorem, will select the best quality representatives. At the same time, either of these voting methods will remove the "lesser evil" problem that marginalizes third party candidates and effectively closes the political marketplace of ideas to them. Approval and Condorcet will not provide an instant multiparty system, the way proportional representation would. Innovative ideas and the candidates who espouse them may have to run several times before they earn majority support. However, this is in keeping with principles of representative democracy. Proportional representation is not.

History is replete with disasters caused by failure to find the judgement of the majority because of an undemocratic voting system (either plurality voting or proportional representation). That is why democratic voting reform is such an important issue irrespective of its potential benefits for third parties.

The narrative resumed: Secretary of State

By knowledge and interest, I’m most comfortable running for Congress. However, getting hammered in the 1998 election left me disinclined to expend a huge amount of effort to run again in order to only get 2%. About the only plus to come out of the campaign was that I began making friends among the Republicans, some of whom had suggested that I abandon the Libertarians and join them. The invitation was flattering but I felt that success for the Libertarian Party was still possible if it adopted democratic voting reform as a major strategy. Unfortunately, I was unaware of anyone other than me talking about the issue. One possibility that crossed my mind was to run for Indiana Secretary of State in 2002. Secretary of State is the office responsible for administering the state election laws, and would be an ideal vehicle for bringing the issue before the public.

In 2000, I expected to sit out the election as I was not interested in struggling through another losing race. Shortly before the state Libertarian Convention, I was approached by Mark Rutherford, the state chair, who said there were a number of people in the party who wanted me to run for U.S. Senate against Republican Richard Lugar. Another Libertarian was seeking the party’s nomination, but it was felt that I would be a better representative of the party, particularly in the TV debates which I was assured were almost certain to include the eventual Libertarian candidate.

Debating Lugar and whatever sacrificial lamb the Democrats threw out was very attractive. It meant a contested nomination for U.S. Senator at the state convention, but I had every confidence that I’d prevail. I said I would run, so long as the party supplied me with a campaign chair and a treasurer, which was agreed.

After securing the nomination, I depended on the state party organization to set up speaking engagements, interviews, and so on. I have no complaints with what they were able to accomplish. Nor do I blame them for Lugar’s people categorically refusing to include me in the debates. The stated reason was that the Libertarian would lower the quality of the debates! I was insulted, though I knew the statement was probably based upon the performance of the 1998 Senate candidate and had nothing to do with me. It was the exclusion from the debate that led me to write and post Why Richard Lugar Shouldn't Be Awarded the Nobel Peace Prize as a way of demonstrating that I was competent to debate foreign policy with Lugar. The Indianapolis Star in a 21 September 2000 editorial agreed that I should have been included in the debates:

“Hager is as consistent a Libertarian as they get. In a recent editorial board meeting with The Star, he demonstrated command of the issues and the Constitution, which he carries in his jacket.” … “The guy’s got good ideas. Let him in.”

The principal reason I wanted to debate Lugar was to build up name recognition around the state in order to run for Secretary of State in 2002. Since that wasn’t going to happen, I came up with another idea that would accomplish the same purpose. This was the genesis of my open carry protest, which I document on my Why I Carry site. The primary goal was to mobilize people in Indiana concerned about infringements of the right of self-defense. The secondary goal was getting name recognition for the Secretary of State run.

After the November election, I met with Mark Rutherford and Brad Klopfenstein, the party Executive Director, to let them in on the carry campaign and my intent to run for Secretary of State. I wanted to give them an opportunity to officially disavow what I was doing if they felt the carry campaign was too radical and would hurt the party. The carry campaign and the right of self-defense took precedence for me over reforming the electoral system, and I was going to move forward irrespective of whether or not the Libertarian Party approved. No disavowal was forthcoming – both seemed to think my carry campaign was a good idea and expressed enthusiasm about my desire to make a vigorous run for Secretary of State.

My first foray into the area of voting reform came in January 2001 when I participated in a Green Party sponsored panel on electoral reform. In my opening remarks (Approval Voting: rational, constitutional electoral reform), I defended the Electoral College and introduced the idea of approval voting. I was not yet a declared candidate, but I began the process of discussing and becoming identified with the issue. The forum was also an early indicator that the 2000 election and the Florida election debacle had fortuitously elevated electoral reform to a major issue. I began to feel confident that I could generate interest in the idea, raise some serious money, and most important, employ an electoral strategy that would turn the perversities of plurality voting to my advantage.

My reason for promoting approval voting (rather than Condorcet) was that it is extremely simple to understand and would cost nothing to implement – points that I always made when I gave speeches on the subject. Any new idea is going to meet resistance from established interests – to have the best chance a reform (like a piece of engineering) needs to adhere to the KISS principle: Keep It Simple, Stupid.

Early on I made contact with Stephen Brams, professor of politics at New York University. Brams, along with Peter Fishburn, wrote the 1983 book Approval Voting. I’d recommend it to anyone who is interested in learning about the empirical and mathematical arguments in favor of approval voting. Brams was kind enough to act as an unofficial advisor on matters pertaining to how I presented approval voting in the course of my campaign.

I began to discuss the issue and my interest in running for Secretary of State with Libertarians in the Spring of 2001. (There was a brief period early on when two other people expressed interest in running for the nomination but because my campaign was already picking up steam they dropped out.) Around this time, I also became aware of a group of Libertarians who were concerned about voting reform. A fellow named Bill Redpath was running for Governor of Virginia in 2001 on the single issue of voting reform; however, he was promoting proportional representation and, for single-seat races, a voting system called Instant Runoff Voting or IRV (see the Center for Voting and Democracy, a group that advocates both). Although better than ordinary plurality voting, IRV often fails to find the majority choice and, empirically at least, fails to put a dent in the two party system. In Indiana, few Libertarians seemed to be aware of the need for voting reform, which was bad, but that also meant that few knew anything about IRV, which was good. My goal, therefore was to educate my fellow Libertarians about voting reform in general, voting systems in particular, and demonstrate why the party should embrace approval voting. At the same time, by discussing the problems with IRV, I could inoculate Libertarians against it.

I also strove to educate the national Libertarian Party about voting reform because I wanted to get national support for my Secretary of State run and get a national effort on behalf of voting reform going at the same time. Early attempts to discuss the issue with some of the national party officials were frustrating. No one seemed to be very interested in voting reform except, perhaps, for selecting party officers or candidates at the Libertarian National Convention. The voting system that most seemed to know about was IRV. When an opportunity arose to be selected for the national platform and bylaws committees, I applied for both. I got my second choice: the bylaws committee.

Educating Libertarians was only half of my task. I also had to educate the public. My campaign website indicates the approach I was using. My campaign logo says "I Approve" and has four boxes with two checkmarks (note: although the logo design was mine, the rendering was done by my then 9-year-old daughter). First, the logo can be interpreted as saying, "I approve of Paul Hager." Second, it shows graphically how approval voting would work: four boxes equals four candidates, two of whom are approved of and receive votes. The choice of four boxes rather than three was deliberate. It was a way of showing Greens and others that my campaign was seeking to benefit independents and other parties besides Libertarians. The slogan, "Vote your heart and your mind," deals with the dilemma voters face under plurality voting where their true preference – their heart – tells them to vote for one candidate but the necessity not to waste their vote – their mind – tells them to vote for the "lesser evil." The message is that approval voting eliminates the dilemma. Reaching the public also involved setting up speaking engagements, appearances, and having articles about voting reform and approval voting appear in newspapers and periodicals, all of which the campaign did starting from the beginning of 2001.

Throughout the summer, I pursued both my carry campaign and my candidacy. An early goal was to mobilize gun owners by explaining why we really needed a third party to protect our rights. Here is a flyer I handed out at those speaking engagements. As I noted in the flyer, the two party system has proven to be a poor protector of our right of self-defense. Opening up the system so that three (or more) parties could fairly compete for majority support was essential in reversing the institutional drift toward gun prohibition. Although the flyer identifies the Libertarian Party as the best party for gun owners, the mention of a fourth party (the fictional "aegis party") was a way of saying that my campaign was really non-partisan.

My efforts to educate Libertarians began to pay off. Andy Horning, who was the party’s candidate for Governor in 2000, had secured a weekly op-ed column in the Indianapolis Star. He became a convert to approval voting and wrote a column explaining it that appeared 13 August 2001 (link not available), which greatly assisted in the process of getting the information out to the public.

In September, I was planning to begin serious fund raising. Indiana places no limits on individual contributions to candidates for state office. If I could interest just a handful of wealthy backers that voting reform was a good idea, I could immediately become a credible candidate. This is the only way a third party candidate with no personal wealth has a chance of putting together a successful campaign. Mark Rutherford had arranged for me to meet a person who had contributed five figures to the National Libertarian Party in the past. This person had become disenchanted with the party when it failed to make any progress and was, therefore, the perfect person for me to pitch my reform proposal to. The occasion was to be a Libertarian conference featuring Don Gorman, a candidate for the 2000 Libertarian Party nomination for President who lost out to Harry Browne. A few days before the conference, terrorists attacked the WTC and Pentagon. Commercial aircraft were grounded and the conference was called off. All of a sudden, there was zero interest in the topic of voting reform.

In late October, the company I worked for closed its Indianapolis office and laid off nearly everyone, including me. For the next two months, my primary concern became finding a job. The campaign was functionally dead in the water.

Near the end of this unplanned hiatus in the campaign, I received a call from fellow Libertarian Rebecca Sink-Burris telling me she was going to contest the nomination with me. The first thought that ran through my head as I was hearing this was that the IRV people must have recruited her. But when I asked her what she was going to be running on, she said, "trust". "What about voting reform?" I asked. Her answer amounted to, it’s a good idea but it’s not what I’m going to be talking about.

After I got off the phone I mulled over this new development. The idea of running a political campaign solely on the basis that you can trust her or Libertarians in general more than the other guys was ludicrous. "Trust" is what I call a campaign grace note – at most an embellishment to a general campaign message. Saying "you can trust me" can carry a faintly negative connotation as well – implying that the other candidate(s) can’t be trusted – and should therefore be employed carefully, if at all. In a Libertarian Secretary of State candidate, making "trust" the campaign issue was particularly dangerous. In 2000, the Libertarian Party engaged in a maneuver to keep Green (and other) Party candidates off the Indiana ballot. For a party putatively in favor of opening up the political process and being dedicated to principle (as in "the party of principle") this was an astonishing political gaffe. In order to do damage control, I put together a press release deploring the maneuver and got Andy Horning and Steve Dillon to sign on. Talking about "trust" in a 2002 campaign risked dredging this episode up again: why trust a Libertarian when the party behaved just like the Republican and Democrat parties in 2000?

The more I reflected on my challenger’s campaign, the less sense it made. Something else had to be going on. I had my suspicions about who was behind her candidacy. There was a small clique (composed mostly of Marion County - that is, Indianapolis - people) that felt it should dictate policy for the state party. I’d had a disagreement early on with one of these people concerning a promotional tabloid-style publication. I said it should focus on the Secretary of State candidate and the issue of voting reform, since Secretary of State was the most important race, and the issue was the core of the campaign. I was told that there’d be something about the Secretary of State candidate - say in September 2002. But the first tabloid that was to be published would deal with general libertarian issues - voting reform wouldn’t be mentioned. But, I countered, the purpose of the campaign was to educate the public about voting reform. That process had to begin immediately, I said. The discussion had been in the late Spring of 2001.

The Secretary of State candidate in 2002 would be the top of the Libertarian ticket and therefore would be the titular head of the party. I and my supporters intended that the bulk of the resources of the state party headquarters and leadership should focus on the Secretary of State campaign. Clearly, this posed a threat to the primacy of the clique, which wanted to control the party’s agenda and message. Rebecca, on the other hand, was willing to defer to the clique, which made her its idea candidate.

When I was talking to fellow Libertarians, I talked exclusively about my issue and why it was so important to the party. I was also making progress getting my message out to the public. The article about the Academy Awards I co-wrote with Stephen Brams (cited above) was one example. It served as the basis for an article that appeared in the 17 March 2002 issue of the New York Times. I also wrote opinion pieces and appeared at public forums, the last being a "meet the candidates" event held by the Monroe County Farm Bureau shortly before the Libertarian Convention. Obviously, what I said to the public differed in some important particulars from what I was saying to Libertarians.

The message to Libertarians was essentially the following:

I explained that the problem wasn’t our message. Republicans who were philosophically libertarian won in two-way races. The problem was the voting system. If we focused on fixing the voting system so it no longer favored a two party system, we would have a fighting chance to become a major party. If we didn’t, the party was going to die. I had five, ten, and fifteen minute variants of a speech that hit the above points.

By the time of the convention, I knew that the odds were running against my campaign. Based upon the response I got in the debates I was able to have with my opponent, I expected to win the majority of delegates who had actually seen us standing side-by-side. However, I was weak in the Indianapolis metropolitan area. The Marion County delegation was, by far, the largest in the state, and the leaders of that County’s organization were supporters of my opponent. Thus, they were in a position to influence the selection of a slate of delegates that would be committed to my opponent before the convention began.

The convention was a weekend affair. The day before the selection of party nominees (including Secretary of State), a dinner was held at which the three candidates for Libertarian National Chair spoke. I got a chance to talk to all three about voting reform. The first was an academic who knew a little bit about different voting methods and made much of the fact that none of them is perfect. He was totally fixated on a faculty election that used alternative voting and produced what he considered an unsatisfactory result. The second professed to know something about voting reform but considered it unimportant. He also said that he was opposed to any voting method that fostered a multi-party system since it would help the Greens. The third fellow said that he considered voting reform important, though he didn’t seem to be well informed about the detailed arguments. He knew about approval voting but kept confusing it with IRV. Meeting the first two candidates reinforced my view that a sizeable percentage of people at the national level were not just indifferent to voting reform – they were actively hostile. From my perspective, the best hope continued to be to get reform moving in Indiana – national couldn’t be counted on assist in any meaningful way.

On the day of the vote for Secretary of State, my worst fears were realized. Turnout was terrible. Only 67 delegates were present. There ended up being two full debates, one in the morning and one in the afternoon before the final vote. I felt that I still had a fighting chance if I could peel off delegates I knew were committed to Rebecca. Most of the Marion County delegates had never heard us debate and really didn’t know anything about my issue. The only regret I have is that neither debate was taped. I feel that I performed at my best, and I know that my concluding statement in debate number one was the best piece of public speaking I’ve ever done in my life. It proved to not be enough.

The final tally was 44 for Rebecca to 23 for me - slightly worse than our projection of 25. In the final analysis, I couldn’t win over the people who were already committed.

My decision to leave the party if I lost had been made sometime in February. There is no question in my mind that the party will die soon without voting reform. Although I always struck a positive note in my presentations, I did say that death was imminent if the party didn’t get the voting method changed. If I believed what I’d been telling Libertarians for over a year, it was pretty hard for me to stick around. And personally, I didn’t want to continue fighting for a lost cause. A political party exists only to advance issues. Since the Libertarian Party couldn’t even save itself, if I was going to have an impact, I’d have to do so in another political party.

The choice of the Republican Party was quite easy. The fact is that the bulk of philosophical libertarians are in the GOP. The Democratic Party is burdened with the legacy of FDR and the Great Society, and is committed to a highly centralized, activist federal government. Libertarians can influence Republican policy. They’d be ignored by Democrats. (One of my first acts after leaving the Libertarian Party was to join the Republican Liberty Caucus, a libertarian PAC in the GOP.) The press release announcing my departure was intended to speed my transition into the GOP and, quite frankly, to make things harder for the Libertarians. Until we have voting reform in Indiana and elsewhere in the U.S., having Libertarians on the ballot will only hurt libertarian-oriented Republicans. Helping the Libertarian Party expire more rapidly is, therefore, a sad necessity.

Let me emphasize that the necessity is a sad one. I think there should be a viable political party that espouses libertarian ideas and works to render them into concrete policy. The Libertarian Party could be that party. The people who worked against my candidacy were, if not close friends, certainly comrades-in-arms.


I sent a copy of the press release announcing my departure from the Libertarian Party to Mike Delph, one of the Republican candidates for Secretary of State. We met shortly thereafter and he invited me to join his campaign, which I accepted. The Delph campaign issued a press release welcoming me aboard, which generated some more press coverage. In June, I had the privilege of helping the Delph campaign during the Republican State Convention. Mike lost, unfortunately, but he’s a young guy and I’m certain that he’ll have a successful career in public service.

I stopped following the activities of the Libertarian Party after leaving. In July, the National Convention was held in Indianapolis. To my surprise, I was contacted by the Associated Press to comment on the convention. It was somewhat ironic that the national press coverage quoted my comments about the party’s irrelevance while Rebecca’s Secretary of State candidacy produced no coverage at all.

It was only much later that I found out about two important developments that came out of the national convention. The first was that Geoff Neale was elected as the national party Chair. Geoff was the only one of the three aspirants who considered voting reform important. The second was that the convention adopted a platform plank calling for proportional representation for legislative elections and IRV for single-seat elections. There was no indication while I was still in the party that there was much interest in voting reform at the national level so this came as a shock.

Proportional representation and IRV are terrible solutions to the problem, but at least the problem is now recognized as existing. Bad as these solutions are, I would have stayed if Rebecca’s campaign had merely been about the wrong reform. There would have at least been agreement about the problem.

Rebecca got 4% in the general election, which assured the Libertarian Party ballot access for the next four years. (Between the majors, Republican Todd Rokita buried Democrat John Fernandez.) With ballot status assured, the Indiana Libertarian Party has three choices facing it:

Time will tell which of the above strategies will prevail.

After the election, I began to devote my efforts to creating a chapter of the Republican Liberty Caucus (RLC) in Indiana. Shortly thereafter Andy Horning contacted me to say that he was issuing a press release announcing that he was leaving the Libertarian Party to become a Republican. He wanted to join the RLC and “help-out”. Other ex-Libertarians (including my 1998 campaign manager) as well as a group of young Republicans joined up and the Indiana RLC became an official chapter of the national RLC in early 2003.

Andy Horning has since announced that he is seeking the Republican nomination for Congress.

As I write this, I am seeking the Republican nomination for the Indiana House.


  1. Aside from the tie-break suggested by Condorcet himself, a number of viable approaches work. Empirically, cycles are fairly rare. It is important to understand that cyclical majorities reflect real ambivalence or uncertainty in the electorate, possibly due to voters having insufficient information about the choices. If a true majority choice is present, Condorcet will always find it. If not, a cycle will exist.