Decentralise. Easy to say. Difficult to do.

Trevor Buising

Few states have had as long an experience of decentralisation as has Indonesia. The Dutch, primarily to increase administrative effectiveness, enacted the first law for decentralisation in 1903. It was the first of several designs. Yet Indonesia today is more centralised than it was a century ago. Many states, in particular developing ones, have attempted to decentralise for a bewildering variety of administrative, political and economic reasons. It is a technically complex policy, especially for developing countries. Yet decentralisation is a political as well as an administrative necessity for Indonesia. However daunting the task, Indonesia is so diverse that it has to decentralise, and sooner rather than later.

A recent World Bank study noted that the 'problems associated with decentralisation in developing countries reflect flaws in design and implementation more than any inherent outcome of decentralisation'. Policymakers may not sufficiently understand the specific problems they want to overcome through decentralisation, or they may adopt an ineffective strategy to solve them. Implementation is inherently even more difficult. Policymakers may give the field implementers unclear guidelines. Implementers may lack the required skills and commitment. The policy may lack sufficiently powerful political mentors and organised support to carry it through. Changing circumstances may make the policy redundant, or it may be insufficiently resourced.

Much of this has been the case in Indonesia. Flaws in the original design forced the colonial Dutch to revise the 1903 law in 1921. None of the three 1940's decentralisation laws was satisfactory - they did not apply to all of Indonesia, and they were framed during the anti-colonial struggle for independence, when expediency rather than longer-term considerations was the priority. The Dutch were still working towards implementing the amended design when the Japanese invaded in 1942.

The independent Republic of Indonesia enacted a replacement for these Dutch attempts in 1957. Law 1/ 1957 came out of lengthy negotiations, only to be rendered inoperative in many of its provisions by Presidential Decree No 6 of 1959. The PRRI-Permesta regional rebellions gave President Sukarno the opportunity to replace constitutional democracy under the 1950 provisional constitution with presidential rule under the 1945 constitution.

New Order

The New Order tried to decentralise as well. Law 5 of 1974 was potentially an effective general design, negated by a lack of detailed design and implementation. Like the Dutch, the New Order leadership accepted the need for decentralisation if only as a means of enhancing administrative effectiveness, particularly with respect to development and thus its claims to legitimacy through performance. However, Law 5/ 1974 left many details to be finalised in subordinate legislation. This applied in particular to the problem of sectoral decentralisation - that is, the allocation of specific functions in the various fields of government activity to particular levels of government.

If the break-up of functions between the various levels of government had been included in Law 5 then many (but not all) of the subsequent problems would have been avoided. Sectoral decentralisation is technically complex. Moreover, many bureaucrats in the affected departments perceived decentralisation as detrimental to their institutional interests. This made determining the details a protracted process.

Indeed, if the details had been included in the draft, Law 5 might never have been enacted at all. Thus there may have been good reasons for deferring sectoral decentralisation to supplementary regulations. Still, the longer it took to enact the regulations the more difficult it became to maintain the political will to decentralise in accordance with the original objectives. French decentralisation was on a lesser scale than is being attempted in Indonesia, yet it still took decades, and that by a state with a much greater capacity than that of the often ill-coordinated personal fiefdoms of the Indonesian state.

Law 5/ 1974 had an additional problem. One of its aims was to shift the focus of regional autonomy from the provinces to the regencies (kabupaten) and municipalities. This level was closest to the people and thus the most appropriate for administering services. Before 1974, legislation dealt only with transfers of functions from the central government to the provinces. It regarded sectoral decentralisation to the regencies and municipalities as an internal provincial matter. Furthermore, between 1950 and 1974 the number of provinces had grown from 9 to 26, 17 sectors needed to be decentralised, and the legislation was confusing. On top of that, the oil boom allowed New Order sectoral departments to subvert the objectives of decentralisation by coopting the regions with development money.

In the early 1990's the New Order, especially under dynamic Interior Minister Rudini, sought to revive the impetus for decentralisation. Regulation 45/ 1992 was designed to push through decentralisation to the regencies and municipalities. All functions except those specified as central or provincial functions could go to the regencies and municipalities. Regulation 8/ 1995 implemented these changes and launched the 'Autonomous Regions Pilot Project'. Activities in 19 sectors were to be transferred to the regencies and municipalities (so-called level 2 regions). Inaugurated with great fanfare, this initiative failed because it was under-resourced. The central government gave selected level 2 regions greater responsibilities but no greater funding to go with them.

'Justice'

Last year, the Habibie government brought down Law 22/ 1999 to replace Law 5/ 1974. The new law, it said, would enhance 'democracy, community participation, equitable distribution and justice as well as to take into account the regions' potential and diversity'. Actually it was hardly needed. Law 5/ 1974 could just as well have been implemented to do all this. What was really needed was the supplementary legislation.

The changes are not as great as often imagined. Although consideration was given to abolishing them, the provinces have been retained. (There are compelling reasons for retaining them - they bridge the centre and local communities). However, Law 22 is more specific about the role of the regencies and municipalities than was Law 5. They are no longer part of the hierarchy of 'administrative territories' which made them subordinate to the provinces and hence the centre. As with Regulation 45/ 1992, Law 22 states that the regencies and municipalities can assume responsibility for all aspects of government except those reserved for the central and provincial governments. These regions must in any case assume responsibility for a minimum of eleven fields or sectors, a provision similar to that of Regulation 8 of 1995.

Law 22 also clearly stipulates that the decentralisation of functions to the regions must include the transfer of the relevant resources - facilities and infrastructure, personnel and funding. Obviously the framers of Law 22 have learned something from the failure of the 'Autonomous Regions Pilot Project'.

Yet like Law 5/ 1974, Law 22/ 1999 requires considerable supplementary legislation. With one notable exception little of this legislation has yet been passed. Law 5 and Law 22 both required a replacement for Law 32/ 1956, the inoperable law determining fiscal relations between the centre and the regions. This was finally accomplished with the enactment of Law 25 of 1999. This law should increase revenue adequacy and certainty for the regions, improve regional equity, contribute to macroeconomic stability and enhance transparency, accountability and participation in the budgetary process. However, Law 25 itself also requires considerable supplementary legislation.

Regional development planning also still needs reform. In principle, bottom-up planning has been an important feature of Indonesian development planning processes (known as P5D) since 1982. But in practice the emphasis has been on implementing central government policies, programs and projects, and hence on increasing the effectiveness of regional sectoral agencies to implement rather than design policy. Nobody would argue that effective service delivery is not an important responsibility of the state, but this is not what decentralisation is all about.

At the heart of any decentralisation policy must be the realisation that effective policy requires a comprehensive understanding of local circumstances - so comprehensive that central planners simply cannot do it themselves. Diversity requires diverse policy inputs. If decentralisation is to be effective in Indonesia, regional development planning has to be reoriented towards the needs and potentials of the region itself.

Trevor Buising (tbuising@hotmail.com) is a consultant from Brisbane, Australia. He is a former colonial administrator in Papua New Guinea who recently completed a PhD on Indonesian decentralisation at Griffith University.