Observers have traditionally agreed that extreme multi-partyism is incompatible with the political framework of presidential systems, and Indonesia has emerged as a prominent example in this respect.

Indonesia's party system could in fact be classified as "atomized" multi-partyism, with two large, five medium-sized and 10 smaller parties represented in the national legislature. This has made it difficult for sitting presidents to maintain stable coalitions for the duration of their terms.

The political fragmentation in parliament has now triggered calls for the "simplification" of Indonesia's party system. In this context, several of the larger and medium-sized parties have proposed the raising of the existing electoral threshold.

At its recent leadership meeting, Golkar was the first party to present a comprehensive reform package for the revision of the political and electoral laws. Not surprisingly, the smaller parties have opposed these attempts to reduce their influence, and have tried to mobilise civil society groups to defend their cause.

Most of the larger parties have focused their reform proposals on raising the electoral threshold from its current level of 3 per cent to 5 per cent, with National Mandate Party (PAN) leader Soetrisno Bachir even suggesting 10 per cent.

Yet, what has been overlooked by most politicians is the fact that the successful application of an electoral threshold is not determined by its percentage level, but by the functionality of its institutional mechanism.

In this regard, Indonesia has so far applied one of the most ineffective and unpractical systems in the world. Based on existing laws, parties that fail to reach 3 per cent of the votes are still allowed to place their representatives in the national and local legislatures. However, they are legally required to disband themselves prior to the next elections, forcing them to either join those parties that have passed the threshold or create new parties altogether.

This unique set of regulations has produced comical results. Parties that did not meet the cut-off in the 1999 and 2004 elections have almost invariably changed their names instead of following the intention of the law, which was to encourage them to merge with larger parties.

Some parties simply added random words to their existing names, turning the Justice Party (PK) into the Prosperous Justice Party (PKS) or the Justice and the Unity Party (PKP) into the Indonesian Justice and Unity Party (PKPI).

The Crescent Star Party (PBB) has been particularly creative, planning to change its name into Star Crescent Party to qualify for the next elections. However, despite its obvious failure to achieve the intended results, the threshold mechanism as such has only been questioned by a handful of politicians and observers.

The current system should be reformed for several reasons. First, an electoral threshold remains ineffective as long as even the smallest of parties are still allowed to take up their seats in parliament. These seats are often the only sources of income for the "bonsai" parties and their officials, and thus form important incentives for them to cling to their privileges and perpetuate them beyond the next elections.

Parties like Sukmawati Soekarnoputri's PNI Marhaenisme, which gained only 0.8 per cent in 2004, nevertheless obtained one seat in the national legislature and 82 seats in local parliaments across the country. The availability of these resources discourages the party from merging with larger counterparts, knowing that even a similarly dismal performance at the next elections would still deliver the necessary parliamentary representation to keep the party organisation alive.

Second, the fact that even tiny parties can gain parliamentary representation has worsened intra-party factionalism in the larger parties. With so few obstacles to reaping legislative benefits, party functionaries dissatisfied with Golkar, PDIP, PKB, PPP, PAN or the Democrat Party tend to split instantly from their respective parties and create their own political vehicles to contest elections.

By contrast, in democracies with higher thresholds for parliamentary representation, party officials are much more likely to seek compromises within the framework of their existing parties. In Germany, for example, both the Christian Democrats and the Social Democrats contain highly antagonistic factions, but these groups negotiate their rivalries internally because the cost of leaving the party would be unbearable.

In Indonesia, on the other hand, an expanding number of mini-parties have resulted from internal party splits, most of which could have been avoided with a higher barrier for new parties to gain entry into parliament. The recent creation of PKNU as a splinter party of PKB is only the latest example.

Third, from a more theoretical perspective, while parliamentary thresholds are necessary to create effective parties as institutions of political representation and aggregation, there is no reasonable justification for requiring a party to disband itself based on its electoral performance.

Parties should be allowed to compete in the next election even if they performed poorly in the last ballot. In many Western European countries, numerous obscure parties have participated in elections without ever winning seats in parliament, and their continued existence has had no negative effect on the stability of the party system.

Thus the only workable mechanism to reduce the number of political parties is to introduce a strictly implemented parliamentary threshold, not to raise the levels of the current ineffective model. Parliamentary thresholds are widely accepted tools of preventing the fragmentation of party systems, and they have worked well in most developed democracies.

Of course the appropriate percentage is subject to discussion in every democratic system, and Indonesia may well begin with a relatively low threshold. The 10 per cent threshold introduced by Turkey, for example, has certainly been too high, and would be inapplicable to Indonesia. However, a barrier of 4 per cent as practiced in Austria may be an adequate instrument for Indonesia to avoid the kind of extreme multi-partyism that could potentially obstruct its further democratisation.

Marcus Mietzner

Jakarta

The author received his doctorate from the Australian National University, and is currently senior visiting fellow at the Indonesian Institute in Jakarta.