Document Library

Cambodia’s Security-Sector Reforms: Limits of a Downsizing Strategy

Author: D Hendrickson
Date: 2001
Size: 16 pages (56 KB)

Access full text: via document delivery

Summary

Just how broad should International Monetary Fund (IMF) and World Bank security sector reforms be, given that they are anxious not to be accused of political interference? Are long-term improvements possible by reforming only parts of the sector, rather than total restructuring?

The IMF’s loan assistance to Cambodia is conditional on economic liberalisation and stabilisation programmes, including a reduction in public spending. Military expenditure was judged excessive and thus a key security sector reform was downsizing.

Demobilisation reforms were through the Cambodia Veteran’s Assistance Programme (CVAP), re-established by the World Bank in 1999 after political obstacles hindered its implementation in 1994. However, a paper from the Conflict, Security and Development Group argues lessons were not learnt from previous, failed, reform efforts. Furthermore, a downsizing strategy alone is insufficient to bring public expenditure reductions. Security sector reforms are likely to fail unless donors and governments agree an appropriate reform strategy and loan conditions.

Security sector reforms have not been achieved as donors had hoped. Critically, current and future IMF loans are conditional on the government reducing military spending quickly. However, the government needs time and resources to enable security sector restructuring before the World Bank’s CVAP can be implemented successfully and military expenditure reduced. Other findings include:

• IMF loan conditions focused on downsizing targets, without appreciating the need to also, and indeed first, resolve security sector governance and accountability
• Military expenditure has not fallen, despite demobilisation, as the army argued the need to use funds for retraining. Poor financial management and auditing may allow military expenditure to be ‘hidden’ or diverted via other budgets (for example, public works)
• Poorly managed, CVAP designers and implementers failed to consider, (1) Cambodia’s inadequate capacity to implement restructuring, and (2) the Cambodian government’s weak commitment to reforms
• Demobilisation did not proceed on time and the World Bank began CVAP implementation without, (1) overcoming the donors’ and government's differing views of CVAP methods and aims, and (2) agreeing funding commitments with donors
• Needs were inaccurately assessed as outdated data concerning military size was used in CVAP financial and structural decision-making.

Before designing and implementing reforms, donors should have full knowledge of security sector problems and needs, in addition to a country’s institutional, technical and managerial capacity. This would prevent unobtainable targets being set. Also critical is the degree of commitment to reform on the part of donors, security forces and government. Other policy needs include:

• Allowing time to seek all stakeholders’ views – government, donors and security forces – regarding CVAP design and implementation
• Establishing a framework for reforms: Considering current and future defence needs; incorporating sector-wide restructuring (strengthening governance and accountability); addressing ex-combatants’ needs regarding reintegration into civilian society
• Providing sufficient resources to build the technical and administrative capacity needed to implement security sector reforms, before setting a timetable for reforms
• Incorporating demobilisation programmes into wider security reforms that also address structure and governance of the sector. This includes mechanisms to encourage cost effective spending
• Reform programmes must address and improve, (1) transparency, (2) civilian control and monitoring, and (3) accountability.

Access full text: via document delivery

Source: Hendrickson, D. 2001, 'Cambodia’s Security-sector Reforms: Limits of a Downsizing Strategy', Journal Of Conflict, Security and Development, vol. 1, no. 1, pp. 67-82