- Purchase/rental options available:
Journal of Democracy 12.4 (2001) 109-122
[Access article in PDF]
Why Direct Election Failed inIsrael
Less than a decade ago, Israel became the world's first parliamentary democracy to adopt direct popular election as the method of choosing its prime minister. The passage of this controversial reform was spurred by a three-month-long government crisis in the spring of 1990, during which Israelis looked on in shock and dismay as members of their parliament (the 120-seat Knesset) indulged in an unprecedented public orgy of floor-crossing and unseemly bargaining, with parties and individual legislators scrambling for place, preferment, and political advantage. This episode--known ever after as "the stinking maneuver"--led Israelis to change their Basic Law in 1992 to require that a separate popular election of the prime minister (PM) be held concurrently with balloting for the Knesset, although the PM's mandate would continue to depend upon a parliamentary vote of confidence.
But the reform did not work as planned. After using it three times--in the 1996 and 1999 general elections and in the February 2001 by-election in which Ariel Sharon of Likud defeated Ehud Barak of Labor--Israel ended the experiment. In March 2001, the country reverted to a slightly modified version of the pre-1992 system for choosing its premier. It will use this system in its next general elections, which are currently scheduled for November 2003.
Israel turned to direct election to counteract the fragmentation, instability, and immobilism that came to a head in the crisis of 1990. These maladies are often associated with proportional representation (PR), particularly the low-threshold Israeli variety, which allows parties winning as little as 1.5 percent of the vote to take Knesset seats. 1 Yet the [End Page 109] very effort to secure the passage of reform introduced distortions that caused it to fall short and even backfire. Much to the chagrin of reform's supporters, the 1996 and 1999 elections each led to a troubled time of coalition building. Small parties, far from being sidelined, vied to play kingmaker. And each time the result was an awkwardly patched-together coalition government with little coherence or staying power. 2
Perhaps the most disappointing thing about the reform was its failure to give either the Labor or the Likud leader "coattails" that his colleagues could ride to victory in the at-large, list-based contest for the Knesset's 120 seats. On the contrary, the system's use of two ballots per voter (one for prime minister and one for the Knesset) encouraged ticket split-ting. Many voters rejected Labor and Likud Knesset candidates, opting instead for smaller parties with sharper issue profiles, leaving the two big parties with less bargaining power than ever and increasing the frag-mentation that ultimately caused the system's undoing.
What lessons may we draw from Israel's attempt to introduce elements of presidentialism or quasi-presidentialism into its parliamentary form of government?
Direct election was premised on a number of conditions. Voting for both the PM and the Knesset was to take place on the same day, so as to maximize the prospects for prime-ministerial "coattails" and to minimize the risk of French-style "cohabitation," in which the PM would be from a party not included in the Knesset majority. Stability would be ensured by the fact that the Knesset and the PM each had the institutional means to oust the other, but in doing so would have to face new elections as well.
To be chosen PM, a candidate would need at least 50 percent of the popular vote, with a runoff between the top two vote-getters to be held two weeks later in the event of an inconclusive first round. In the three elections under this arrangement, only two candidates for the premiership ran each time, and there was never any need for a runoff. Otherwise--and crucially--Israel's PR system for electing Knesset members remained unchanged. It continued to be highly proportional, with a single nationwide constituency, a low threshold of 1.5 percent, and...