pageok
pageok
pageok
[Ilya Somin (guest-blogging), March 29, 2006 at 2:39pm] Trackbacks
Political Ignorance and Israeli Coalition Politics:

One of the major themes of my academic work is that modern democracies suffer from a serious problem of political ignorance (see, e.g., here and here). Most voters are "rationally ignorant." Because there is so little chance that any one vote is going to be decisive in an election, individual voters have almost no incentive to learn about the competing parties and their policies, and as a result it is rational for them to devote very little effort to acquiring political knowledge (except for the few who have reasons for doing so unrelated to improving the "quality" of their votes).

In a proportional representation (PR) system such as that in Israel, the problem may be even worse than in the US. Voters in a PR system need to know not only what the policy differences between the parties are, but also what effect voting for a particular party will have on the resulting coalition government that emerges from an election. In some cases, voting for a right-wing party might actually increase the chance of creating a more left-wing coalition government or vice versa.

Yesterday's Israeli election is a good example of this. In order to form a government, Israeli politicians must put together a coalition with at least 61 seats in the 120 seat parliament. Yesterday, the centrist Kadima Party got 28 seats, while right-wing parties (Likud, NU-NPR, Yisrael Beteinu) got 32, and parties to the left of Kadima got 31 (Labor 20, Meretz 4, Pensioner's Party 7). Various special interest parties, got most of the remaining seats. Kadima is unlikely to form a coalition with the right-wing parties because these parties oppose Kadima's central policy agenda: unilateral withdrawal from large parts of the West Bank. But because Kadima got only 28 seats, they will almost certainly have to form a coalition with the Labor Party (20) and perhaps other leftist parties as well. Had more right-wing voters picked Kadima rather than the parties closer to their views, Kadima might have won enough seats (say 40) to be able to form a government without Labor (which many Kadima leaders would have preferred to do), and therefore a government that would be less leftist.

Ironically, by voting for right-wing parties instead of Kadima, Israeli rightists may well have ensured a more left-wing government than would have resulted from their voting for Kadima instead! They "achieved" the opposite result from the one they probably intended. I suspect that this occurred at least in part because Israeli right-wing voters (like most other voters in PR systems) simply had insufficient incentive to put in the time necessary to think systematically about the impact of picking a particular party on the resulting coalition.

The extra knowledge burden imposed by the need to calculate coalition possibilities is an important (and generally ignored) weakness of PR electoral systems.

Related Posts (on one page):

  1. Political Ignorance and Israeli Coalition Politics II:
  2. Political Ignorance and Israeli Coalition Politics:
Pritesh:
One of the major themes of my academic work is that modern democracies suffer from a serious problem of political ignorance

Can you name a time and place were this was not the case?
3.29.2006 3:45pm
Zaoem (mail):
Apply the same logic to left-wing voters and everyone should have voted for Kadima! I guess that something is wrong with the median voter theorem.
3.29.2006 3:55pm
Arthur (mail):
Fortunately, extremist third party candidates can't affect the outcome of U.S. elections, as President Gore will attest.
3.29.2006 3:57pm
Defending the Indefensible:
Replace first-past-the-post and proportional representation with a Condorcet method that isn't subject to strategic voting.
3.29.2006 3:57pm
Grant Gould (mail):
Every voting method is subject to strategic voting -- indeed, that was Condorcet's theorem. The Condorcet methods are merely subject to strategic voting in only a well-defined set of cases that may be rare.
3.29.2006 4:03pm
Oscar Volij (mail):
For the record, many right wing voters (like myself) would prefer almost anything to Kadima.
3.29.2006 4:07pm
Blue:
And this is why PR is a terrible voting system. Particularly the Israeli version, with low barriers to minor party representation.
3.29.2006 4:12pm
The Original TS (mail):
Ilya -- nice post, very thought provoking!

This ties into one of the biggest problems with the U.S. primary system. Many voters have little incentive even to vote, much less gather sufficient information. As a result, hard-core party idealogues have a disproportionate influence on the outcome. In the general election, middle-of-the-bell-curve voters are left with a classic lesser-of-two-evils choice.

It's interesting to contemplate swapping the primary system for a single-election Condorcet system. On the other hand, we might end up with James K. Polk president for life.
3.29.2006 4:16pm
jackson dyer (mail):
You left out Shas which got 13 seats.

Kadima could form a coalition without labor if it includes Shas but why would it? Why leave out Labor when at a time when major decisions about the countries security will be taken you will need imput from every political point of view?
3.29.2006 4:17pm
The Original TS (mail):
The Condorcet methods are merely subject to strategic voting in only a well-defined set of cases that may be rare.

True. It breaks down when there are cyclical preferences.
3.29.2006 4:21pm
Ilya Somin:
In response to the question about Shas:

Even with Shas's 13 seats in the coalition, Kadima will probably have to include Labor's 20 seats to get to the magic number of 61. Shas and Kadima between them have only 41 seats. In fact, a Kadima-Labor-Shas coalition seems like an especially likely outcome. By contrast, had more rightists voted for Kadima, the likely result would be a Kadima-Shas-some other interest group party coalition. The latter is presumably preferable to the former from the standpoint of most Israeli rightists.
3.29.2006 4:28pm
Matty G:
As a political scientist, I agree with the underlying principle of this post: PR systems tend to often make strategic voting necessary, requiring greater citizen knowledge, and thus punishing uninformed voters more severely than other systems.

That said, it's not clear to me that the example given fits the case. A few points here.

1)We would need to outline the preference ordering of each type of voter. Let's play a simplified version of the election, in which there are only three types of voters (Kadima, rightist, and leftist) and only three parties (Kadima/centrist, right wing Likud, and left wing Labor). One assumption that many people (including political scientists) make about voters is that their preferences are single-peaked, meaing that sincere voters whose 1st choice was right wing would NECESSARILY prefer centrist over left wing, and vice versa. Unfortunately, this is demonstrably not true, especially when there are multiple issues in an election. Even a two issue universe - economic policy and Palestinian policy - could produce voters whose preference was: left, then right, then Kadima, or vice versa. So we have to be careful about that.

2)Information is missing even WHEN voters are fully informed. This problem will almost certainly never go away. Even if every voter understand the strategic consequences of their vote, they PROBABLY DON'T KNOW the outcome of the election prior to it happening. It's VERY tough to tell a rightist to voter for Kadima if THEY THINK THERE IS ANY CHANCE of Likud winning, since a Likud victory would pre-empt the strategic need to vote for Kadima. This is not a problem of PR systems, this is a problem of missing information, generally. And one that can't be pinned on "ignorant voter."

Anyhow, just two things to think about. Great post.
3.29.2006 4:40pm
Sam:
To echo Oscar, to true (right-wing) believers there isn't much difference between Labor (and the Pensioners) and Kadima, so voting for Kadima outright would not lead to a preferable outcome.

A more practical application of this theory to the Israeli election would be to analyze the voters for parties which did not make it into the Knesset. The Knesset keeps raising the barrier to representation, in an attempt to snuff out the more extreme or quirky small parties, currently requiring 2% of the vote (or more than 2 seats) before a party may be represented. Votes for parties which do not clear this threshold are generally discarded (vote sharing agreements aside).

What of the effects of this system on voters? Some right wing parties (ie-NU-NPR) campaigned against some others, by warning that votes for smaller right wing parties would likely not count, again helping Kadima. The left has this issue as well, with parties such as the Greens (sounds familiar?) and the "Green Leaf". Are voters convinced to vote for the mainstream parties by this logic? Does such logic perpetuate disenfrachisement of the smaller parties?
3.29.2006 4:43pm
Grant Gould (mail):
Left-Right "political spectrum" descriptions of politics exaggerate the strengths of Condorcet methods (by definition there are no cyclical preferences when politics is a linear spectrum). The real world, though, contains a lot of cyclical preferences, as described in the other comments here -- it is not a one-dimensional political spectrum.

It is not ignorance or unsophistication that has led most democracies to either two-party systems which lock out minority views or parliamentary systems which encourage strategic voting. These both seem to work in the presence of real voters much of the time. Of course, this is in part because the real purpose of an election system is to create public confidence in and obedience to its result -- something that Condorcet methods don't quite do.
3.29.2006 5:11pm
DJF:
I don't think the election is properly characterized as a left-middle-right division between the populace. I think this election was actually a referendum on a single issue dominating Israeli politics and policy right now, which is what to do about the Palestinians and the contested territory.

I think the left is more appropriately described as a concessions-for-peace camp, Kadima is the security barrier/ unilateral solution camp, and the right is the pry-it-out-of-our-cold-dead-hands camp.

If you disagree with Kadima's central policy platform, which is to unilaterally withdraw from most of the West Bank, and to unilaterally declare a permanent border, probably along the security barrier, you certainly don't want to vote for them for any reason.

As it is, there's a lot of belief that a center-left coalition is going to have a harder time accomplishing Kadima's goal than a centrist or center-right government would have, so rational ignorance voting may end up paying dividends for the Israeli right, at least until the Palestinians start asking for annexation and one-man, one vote.
3.29.2006 5:37pm
MDJD2B (mail):
I don't think the election is properly characterized as a left-middle-right division between the populace. I think this election was actually a referendum on a single issue dominating Israeli politics and policy right now, which is what to do about the Palestinians and the contested territory.

People from outside any country oten make the error of believing that the nation's voters are more interested in foreign/security affairs than they really are. Indeed, objective national interests often compel national leaders to adopt security policies different than they originally planned to. (Didn't George W. Bush, for example, run against nation-building in 2000?) The difference in polls between the close to 50 seats Kadima was projected to win when Sharon was in charge, versus the 28 seats it actually won might suggest that there was a radical change in Israeli views toward security over the last few months. But it is more likely to reflect either the relative importance of domestic concerns or the relative importance of the percieved competence of individual candidates (as opposed to any issues)in the minds of the voters.

Certainly, the many voters for religious parties were voting primarily for the status of religion. And the Labor Party candidate, Peretz, deemphasized security issues and ran on a populist platform.

Security issues are very important to Israel. And Israel's security decisions are more important to Americans than are the friendliness of the government to private firms, or subsidies to religious schools. But Israelis are concerned with what kind of country they live in, and have a right to be concerned with these issues. This is just as true as the proposition that many American voters will vote on the basis of abortion, the status of homosexuals, the deficit, support for various special interest subsidies, etc., rather than on the basis of issues that are of most concern to the French, Israelis, Arabs, or Indians.
3.29.2006 6:29pm
Mark F. (mail):
Isn't the bottom line that most people would like the system that gets them the government closest to their own political views?
3.29.2006 6:59pm
Sam:
The difference in polls between the close to 50 seats Kadima was projected to win when Sharon was in charge, versus the 28 seats it actually won might suggest that there was a radical change in Israeli views toward security over the last few months. But it is more likely to reflect either the relative importance of domestic concerns or the relative importance of the percieved competence of individual candidates (as opposed to any issues)in the minds of the voters.

Most likely, it reflects the utter incompetence (or virulent bias, if you prefer) of most Israeli pollsters.
3.29.2006 8:31pm
Justin (mail):
I suspect a majority of Israeli citizens are perfectly fine with Kadima/Labor economic policies, but does anyone have any actual evidence for or against this hypothesis? IS seems to believe this isn't the case, but I think he ignores that strategic voting isn't something that only liberterian tax folk are incapable of :)
3.29.2006 10:02pm
m (mail):
Let me register a little dissent about how great this post is. Although I commented on an earlier post about this, there is still no evidence above that Israelis suffer from the same levels of voter ignorance that afflict other constitutional democracies. Despite Ilya's theoretical assertions about "rational ignorance," it'd be nice to have some numbers to back this up. I'm not saying those numbers don't exist, but it would be helpful to have them on the table. If a large number of Israelis are well informed about party politics, then the "rational ignorance" thesis looks that much less persuasive as an explanatory account of voter knowledge (or lack thereof).

The problem of voter ignorance is separate from the problem of strategic voting under conditions of uncertainty. As Matty G pointed out, even extremely well informed voters may not be able to predict electoral outcomes, or even assign accurate probabilities to them. Ilya seems to have run these problems together. Not that this detracts from the strategic voting analysis in a PR system. But it would help to distinguish the two issues more clearly.
3.29.2006 11:30pm
Blue:
"It is not ignorance or unsophistication that has led most democracies to either two-party systems which lock out minority views or parliamentary systems which encourage strategic voting."

There are many of us that argue that locking out minority views, and forcing the creation of two centrist parties, is an exceedingly positive feature of FPTP systems. Only single-issue crazies benefit from PR.
3.30.2006 12:32pm
Matthew Shugart (mail) (www):
The claims made in the post about PR would be hard to sustain from a reading of the extensive literature on this topic.

Political information is higher in multiparty systems than in two-party systems, and governments are more likely to be located near the median voter than in two-party systems.

If you followed the campaign, you know that every Israeli papaer daily speculated on the likely coalition arrangements, who would be minister of 'X' under various scenarios, which parties would refuse to join a coalition without certain concessions, etc. Even the most "rational" voter, who had too many other more important concerns to spend lots of time acquiring information, would not have a difficult time at all knowing the consequences of his or her vote.

There are, of course, cases under proportional representation (PR) in which increased votes for the right might increase the odds of a left-wing government (or vice versa), but this election was not one of those. More importantly, the odds of that happening are far, far lower than in a plurality/majority system.

Of course, Israel is just about the most extreme example of PR and multipartism that you can find, and any arguments against PR are going to look a lot better in the context of a discussion of Israel (especially this election!). But those arguments should not be mistaken for general arguments, nor should Israel (especially this election!) be mistaken as typical of PR.

Notwithstanding the Israeli case of "extreme" PR, can anyone seriously doubt that the Israeli electorate knew what the biggest issue was in this election and knew how to vote to achieve it? There is a clear majority in favor of disengagement, and the electoral system revealed it. That could not have been assured with a non-PR system.

The best analysis I have seen of the election yet was a Ha'aretz piece on Shas, the Pensioners, and the reduced size of the major parties. Highly recommended.
3.31.2006 6:42pm