Failed Experiment

In describing recent Israeli elections here, I glossed over the nature of some of those elections.

Israel experimented with a way of choosing a head of government that, as far as I know, was unique among parliamentary democracies. In 1996, 1999 and 2001, they held direct elections for prime minister.

One step in a leadership transition in any parliamentary system is to designate a potential prime minister who will take office if he or she is able to form a government. That means creating a cabinet that has either majority support in parliament (usually the lower house of parliament), or at least is not actively opposed by a parliamentary majority. There are three variations on that theme:

1. One party has a parliamentary majority. That is the simplest outcome; that party's leader more or less automatically becomes prime minister.
2. Two or more parties form a coalition that, in total, has the support of a parliamentary majority. In a coalition cabinet, the leader of the largest party usually becomes prime minister. Cabinet jobs are doled out to the smaller parties, roughly in proportion to how many seats they have in the parliament.
3. Neither a single party nor a coalition has majority support. Unless/until other parties gang up against it and vote it out, the largest party can lead a minority government, and its leader becomes prime minister.

The head of state (such as a monarch or a president) is usually the one whose job it is to designate a potential prime minister who will attempt to form a government.

In scenario #1, that is a mere formality. For example, in 1997, when the Labor Party won a huge majority in the British House of Commons, Queen Elizabeth asked the leader of that party, Tony Blair, to form a government. But she had no other choice.

Scenarios 2 and 3 can be more tricky. Usually the nod goes to the leader of the largest party, but other questions, such as who might best be able to form a coalition, or whether a new election needs to be held, might enter in to it.

The British general election of February 28, 1974, left no one party with an overall majority in the House of Commons. Labor was the largest party. Prime Minister Ted Heath of the Conservative Party balked at resigning, hoping to remain in office with help from the Liberals and/or the Ulster Unionists. Four days after the election, it was clear that that was not going to happen, and Heath resigned. The Queen was then able to summon Harold Wilson, the Labor leader, to form a minority government. The Queen had held back in the meantime, trying, as she always does, to stay out of party politics.

In Israel, the president designates a party leader to form a government. That has been the case through most of that country's history. However, in 1992, a law was enacted that took that task away from the president and gave it to the electorate as a whole.

That change was intended to strengthen the office of prime minister by giving him or her a direct popular mandate, and de-emphasizing the need to build a coalition within the Knesset (parliament), a process that had magnified the power of the small parties. As I described here, Israel's system of pure proportional representation has resulted in many parties being represented in the Knesset.

Any change instituted by a government is subject to the law of unintended effects. In other words, an action that is intended to produce a certain effect, often produces other, unexpected, effects, and those other effects are often directly the opposite of what was intended.

In this case, the unintended effect was increased scattering of votes among small parties in the Knesset, and an accompanying decrease in the share of Knesset seats going to the major parties.

Why did that happen? Take the example of a right-wing voter. Such a voter could either vote for what was, at the time, the main right-wing party, Likud, or could vote for a smaller party with whose philosophy the voter was more in synch. The problem with the latter approach was that a decrease in Likud's vote total could contribute toward making the left-wing Labor Party the largest party, and thus the party empowered to build a coalition. Therefore, our right-wing voter had an incentive to stay with Likud, in order to improve the Likud leader's probability of becoming prime minister.

A similar effect was at work in the American presidential election of 2000, when Ralph Nader ran as a third-party candidate. If those who voted for Nader had instead voted for the major-party nominee to whom they were presumably closer ideologically, i.e., the Democratic Vice President Al Gore, he would have defeated George W. Bush. The reluctance to cause such a scenario holds down third parties' share of the U.S. electorate.

But direct election of the Israeli prime minister to some degree decoupled the issues of the parties' parliamentary strength, and the choice of prime minister. Likud's leader, Benjamin Netanyahu, barely won the first such election, in 1996. After having voted directly for Netanyahu, right-wing voters were free to vote for smaller parties for the Knesset, confident that that vote would not affect Netanyahu's prospects. The same effect worked on the left, in relation to the losing candidate in the prime ministerial election, the Labor leader and incumbent Prime Minister Shimon Peres.

This effect is reflected in the numbers. At the 1992 general election, the two largest parties at that time, Labor and Likud, between them won a total of 76 of the 120 Knesset seats. In 1996, the year of the first election for prime minister, those parties' total had been reduced to 66 seats. Then, in 1999, it fell to 45.

Thus, the change that was supposed to reduce the power of the small parties, instead caused them to grow.

The experiment was scrapped in 2001, when Israel amended its law on the formation of governments, to return to the president the task of designating a prime minister.

As I see it, the lesson here for the U.S., whether or not one is inherently interested in the politics of Israel, is that a country needs to be careful when considering changes in electoral law.

Proposals such as the introduction of proportional representation or the abolition of the electoral college should be carefully scrutinized, in order to minimize the possibility of creating unintended effects.