Reasons for Supporting Spoilers

Reasons for Supporting Spoilers:
Three Models for Minor US Political Parties

Peter A. Taylor
November, 2000 (updated 11-9-2004)

[A more recent essay, Duverger's Law and the Libertarian Party, covers much of the same material.]

There has been a lot of traffic on Green Party email lists recently about the possibility of Ralph Nader being a "spoiler" for Al Gore. Voters who express their strong environmentalist preferences by voting for Nader instead of Gore are punished under our election system by an increase in the likelihood of Bush being elected. In the short term, voters who would otherwise vote for Democrats are diverted to the Greens, allowing the Republicans to defeat Democrats in our "plurality" elections (no runoffs in most of our elections, so the largest vote getter wins, even though a majority opposed him). In the longer term, if the Democrats move to the left to try to absorb the Greens, they are moving away from the center (for an exception, see the "Wallace" addendum, below). Unless this has very dramatic effects on Democrat voter turnout, but not on Republican voter turnout, this is political suicide for the Democrats. If the Democrats move to the right, towards where the center would be if the Green voters didn't exist, they are moving away from where the Greens want them to go. Either way, Ralph Nader appears to be the Republican Party's best friend. How, then, can one ever justify support for a minor party?

I would like to provide several possible answers to this question, along with some historical context. Specifically, I would like to suggest three models for the role of minor parties in countries that use our "First Past The Post" (plurality) voting rules (which generally lead to a two-party system, with minor parties tending to act as spoilers):

1. The Socialist Party (USA) in the early 20th century: In this case, a minor party is seen as a publicity device for the ideas its activists want to promote. Activists want to get enough votes to attract the attention of the press, but otherwise, not enough to tip any important elections. The premise here is that the major parties do an okay job of competing for the middle-of-the-road voter, but the political system is screwed up because of the voters' ignorance, which minor parties can fix without needing to be able to compete on a level playing field. Activists may hope their party's minor party status is permanent so their party doesn't get corrupted by big money or power-seekers.

2. Abraham Lincoln's early Republican Party: In this case, because of inertia or oligarchy, the major parties (the Whigs and Democrats of Lincoln's day) are unresponsive to the middle-of-the-road voter. A more responsive minor party can hope to replace one of them in short order, hopefully without soon becoming like the party it replaces. The voters don't need to be educated (ie. everyone knows about slavery). The new minor party merely organizes a new majority of voters, and when the dust settles, there will again be a two-party system. The role of spoiler is tolerable because it is only transitional.

3. The Liberal Alliance in the mid-1980s United Kingdom: The Alliance tried to take advantage of the spoiler effect in order to blackmail the dominant major party (Labour) into supporting any of several electoral reforms which would eliminate that effect, allowing all voters to vote their consciences without throwing their votes away. (Plurality voting is actually unusual among voting systems in having the defect that one must make such a choice.) This would have allowed the Alliance (now the Liberal Democrats) to compete on a level playing field, which would have made their long term growth tremendously easier. This is good both for broadening public discourse and for undermining oligarchies in the current (mainly) two-party system. Here again, the spoiler effect is tolerable because it is only transitional, and there is hope that the threat will not have to be carried out very often.

In my opinion, the Liberal Alliance model is the most appropriate model for minor parties in the modern US. I think it is not only the most realistic model, but also the one that holds the most promise for long-lasting positive change. Unfortunately, this model does entail running candidates who will sometimes act as spoilers, and it seems likely that this situation will continue for some time. This strategy makes sense, as long as the "electoral reform or else" threat is made clear. My fear is that some Green activists seem to be trying to downplay the seriousness of the spoiler problem instead of educating people about the alternatives to it (such as Instant Runoff Voting).

Addendum (4-9-2001): There is an important variation on the 1850s Republican model which deserves separate consideration. This variation is to combine the Republican model with the modern Canadian model, where there are two-party systems within any one province, but not the same two parties throughout the whole country. This model may be hard to follow in the US because of the importance of nationwide presidential "coattails," but still it is clearly easier for Jesse Ventura to be elected Governor of Minnesota under the Reform Party than for Ross Perot to be elected President. Like the 1850s Republican model, the major parties are seen as unresponsive to the (local) median voter (even if they compete fiercely elsewhere in the country). Again, a minor party wins in this situation more by forming a new coalition of previously disaffected voters than by educating the voters. The role of spoiler is again tolerable here because it is seen as a brief transition to a new (local) two-party system.

In short, there are a number of ways minor parties can try to cope with the spoiler effect that they have under First Past The Post election rules. My preference would be for them to work vigorously for electoral reform. The one thing that they cannot do and still be effective is to deny the problem.

Addendum (8-27-2001): Another role for minor parties is implied by McKenzie and Tullock in their discussion of the 1968 Presidential campaign of George Wallace in chapter 10 of The New World of Economics, 3rd edition. In this model, instead of a single left-right political spectrum, there is a political "issue space" with two or more axes on it. The major parties are responsive to the median voter on one axis, but not on the other, because they have misjudged where the median voter is on the second axis. A minor party that picks up significant support can "help" the major parties find the median voter again, even if it never wins a major election. Note that if the Democrats of California in 2000 moved closer to the Greens, they would have moved away from the center, whereas, according to this model, if the Democrats of Alabama in 1968 moved closer to Wallace, they would have moved closer to the center, along a different axis.

Figure 1
1968 Presidential Election

Sociological "Issues"

Wallace
x


Left	median voter	Right
	x


x	x
Democrats	Republicans

This "Wallace" model seems to reflect views that are common among Libertarians: the left-right political spectrum is an inadequate way of looking at politics, and the Democratic and Republican parties are equidistant from where the Libertarian party stands. However, the Green party seems to me to lie pretty much on the same left-right axis as the major parties, just further to the left. While Libertarians can plausibly claim to be pretty much indifferent about their candidates acting as spoilers (or claim that a slight libertarian flavoring moves them closer to the center), in my opinion, Greens cannot. Furthermore, the Greens cannot plausibly claim that the major parties need their help to locate the median voter on a single axis: if the Republicans were that far to the right of the median voter, the Democrats would consistently win. I disagree with Libertarians who embrace this "Wallace" model: I think (1) the major parties know where the middle is, (2) the major parties have internal restraints that restrict their competition, and (3) the Libertarian party is too radical to be very helpful in gauging moderate libertarian sentiment. I think Libertarians would be much better off following the Liberal Alliance model and focussing on electoral reform. But still, the role of spoiler isn't as embarassing for Libertarians as it is for Greens.

Figure 2
Libertarian View of "Issue Space"

Neither
x

Libertarians

Uncle Sam In		Uncle Sam
My Wallet, Not	median voter	In My Bedroom,
My Bedroom	x	Not My Wallet

x Greens
x	x
Democrats	Republicans

Both

Figure 3
Libertarian Nightmare Scenario

Neither
x

Libertarians

Uncle Sam In		Uncle Sam
My Wallet, Not		In My Bedroom,
My Bedroom		Not My Wallet

x Greens
x	x
Democrats	Republicans

x
median voter
Both

Addendum (3-26-2002): It is probably more realistic to view the Libertarian Party not in terms of the Wallace model, but as a variation of the Socialist Party "publicity device" model, judging by the party's distance from the political center. However, there are different explanations for this distance. Maurice Duverger, in Political Parties, distinguishes between political activists who are sincerely attempting to influence government policy ("fight to win") and those whose politics are primarily a form of self-expression ("fight to feel good about losing"). Bernard Crick has some less conciliatory things to say in In Defense of Politics about this sort of self-expression, or as he calls it, "student politics." In some cases (ie. the "ban the bomb" movement within British socialism), the positions taken seem to Crick to be calculated to ensure that everyone understands that these activists are not really interested in participating in politics in any serious way. While some Libertarians seem to be using the US Socialist Party as a model, to me, the apparent dominance of activists who are oriented towards self-expression over those interested in influencing policy represents a significant departure from this model.

As Duverger put it, the relationship between plurality voting and the two-party system is so close that it comes as close as anything does in the field of sociology to being a natural law. (This is "Duverger's Law."). It's hard for me to understand how anyone who is serious about influencing government policy through a minor political party could not be interested in alternatives to plurality voting. I find it hard to escape the conclusion that electoral reforms such as Instant Runoff Voting would make minor parties much more effective in promoting ideas that have been downplayed by the major parties.

Addendum (11-9-2004): One way to try to avoid the spoiler effect is vote trading. For example, Green voters in a swing state like Florida will agree over the internet to vote Democrat in exchange for Democratic voters in safe states like Texas or Massachusetts agreeing to vote Green. These agreements are unenforcable, but since the motive for voting is largely a sense of moral duty in the first place, the honor system might work well enough.

One problem with this idea is that the Democratic party's supporters have to cooperate. If the Democratic party starts to view the Greens as a serious competitor rather than a nuissance, that willingness to cooperate may disappear. Another problem is that this idea only gets you votes in races that are not competitive. This may get the Green Party a few extra votes in Texas, but what it does not do is to make it possible for three parties to coexist for long as rough equals in the same electoral district under plurality voting rules.

Consequently, I view vote trading as a tactic to be used by a minor party that is following one of the other strategies, rather that a strategy unto itself. Vote trading is compatible with both the "Socialist" voter education model and the "Canadian" regional model. (It could also help reassure voters that their "student politics" are harmless.) Vote trading might also make sense as an interim measure to a party that seeks publicity or a regional foothold before switching to another model. But it does not get around the spoiler effect. Plurality voting still leads (at least locally, in the long term) to a two-party system.

Previous: Electoral College

Next: Instant Runoff Voting

Comments?