Correlations, Causes and the Logic of Obscuration: Donor Shaping of Dominant Narratives in Indonesia's Irrigation Development

Abstract

This article analyses policy trends in Indonesian irrigation, particularly during the last five decades, from the perspective of dominant narratives, as authored, suggested and pushed by international donors. It argues that international donors' adherence to ‘deferred maintenance’ as the core element of irrigation policy problem framing does not match with farmers' and the irrigation agency staff perceptions and practices. The logic of obscuration and the discursive manoeuvers that maintain it are analysed. The article concludes that there is space for more profound conceptual contestation and for alternative actions pathways even within the ‘dominant paradigm’ to address management problems more effectively.

Notes

1. These policy elites include politicians, national and international policy-makers, policy advisors, and international donor agencies. On policy elites as epistemic communities, see Haas (1992) and Yee (1996). We do not suggest that donors carry sole responsibility for the priorities of Indonesian irrigation policy. We do highlight donors' prominent role in providing broader international justification for it.

2. Unusually defensive as compared to other natural resources sectors like forestry. In the urban water supply and sanitation sector there has been more institutional and organisational dynamics, particularly through privatisation and public–private partnership approaches.

3. The resonance with Scott's (1998) perspective on large (infrastructural) projects as instruments of establishing state rule is obvious, but not pursued in this article.

4. Prior to the introduction of high yielding varieties, rice was already an irrigated crop. However, from a ‘productive discourse’ perspective existing levels of hydraulic water control were lower than required for ‘green revolution’ agriculture (Ishikawa, 1967).

5. ‘Poor performance’ was expressed as ineffective irrigation water use resulting in low irrigation efficiency (see Bottrall, 1981; O'Mara, 1990).

6. Apart from the World Bank, Indonesia's donors in irrigation development in the 1980s have included the Asian Development Bank and the governments of the Netherlands, Japan, Australia, and the United States. Though diverse in their approach and mode of operation, these donors share common characteristics in the way they based their development agenda on the international policy trend of a shift from construction towards O&M to catch up with ‘deferred maintenance’.

7. Though WUAs were formed at tertiary level, farmers had always been responsible for irrigation management at this level prior to IMT. The tertiary level represents the network of irrigation canals directly connected to farmers' fields. In larger scale irrigation it is the local irrigation unit that receives water from the upstream primary and secondary levels, the conveyance system conventionally managed by the irrigation agency. The secondary level represents part of the conveyance system initially managed by the irrigation agency.

8. The principles of management transfer as stated in the Letter of Sector Policy of 1999 were rendered highly ambiguous by the subsequent Water Law of 2004.

9. Consensus existed among the policy elites on the problem definition – a consensus possible through exclusion of irrigators and line staff (cf. McRae, 1993).

10. See Araral (2005) on the vicious cycle problem of public irrigation in the Philippines.

11. See the five core elements of dominant narratives (agent, act, scene, agency, and purpose) in Kaplan (1993).

12. And neither were the sometimes idealised understandings of FMIS questioned in policy discourse (cf. Mosse, 2003).

13. See Mollinga (2003) for how, in the Indian case, under the Command Area Development (CAD) programme, the irrigation agency actively kept WUA formation and activity at the level ‘below the outlet’, not allowing it to affect ‘above the outlet’ governance and management, being its own domain. In Indonesia institutional reform initiatives for ‘proper’ operation of the main system was initiated under Irrigation Sub Sector Project I, but with little sustainable impact.

14. This belief originated at least discursively from the experience with the Gal Oya irrigation project in Sri Lanka, where the combination of physical rehabilitation and farmers' involvement resulted in the successful establishment of a farmer organisation and a significant increase in water productivity (Uphoff, 1986). However, later findings show that the positive lessons from Gal Oya have not been repeated even within the country (Aluwihare and Kikuchi, 1991).

15. The terminology ‘irrigation management transfer’ is itself evasive, if not misleading. What is actually implied in strong formulations of IMT is governance transfer, that is, transfer of the power to allocate water and funds. While irrigation agencies may be willing, even happy, to transfer management tasks, they are generally highly reluctant to transfer governance power (see Mollinga and Bolding, 2004).

16. Deferred maintenance is a symptom, we suggest, of institutionalised corruption within the irrigation agency, and corruption a significant part of the X factor in X → DM + PP. The ‘system of political and administrative corruption’ (cf. Wade, 1982 for India) that operates in the Indonesian irrigation bureaucracy will be discussed in detail in a separate article (Suhardiman and Mollinga, in preparation; see also Repetto's analysis (1986) on rent-seeking in irrigation). The key element for the present analysis is that the flow of international irrigation funds significantly supports the so called upeti system, which constitutes a strong logic for continuation of donor projects quite unrelated to irrigation management as such. As an integral part of upeti system donor funds become an important financial means for the national elite to gain and sustain political power (MacIntyre, 1994). We argue that this political value of donor funds necessitates the irrigation agency to shape its development orientation as to fit into donors' investment strategy and comply with the dominant narratives. See also Anderson (1991) on the continuity between the colonial bureaucracy and the present day bureaucratic mechanisms and procedures.

17. Rationales for reform vary with actors' and agencies' perceptions. Many irrigation officials may have believed that performance improvement was important, but for the Ministry of Finance, the budget savings that IMT promised were clearly an important part of the reform rationale. Similarly, different actors within one agency could position themselves differently as regards to certain policy problems (see Suhardiman, 2008 for detailed discussion).

18. In addition, this stratagem allows the reproduction of a self-perception of donors' developmental role as being paramount, or, in reverse, the reproduction of a relationship with national government officials that is one of dependence with mutual benefits (cf. Mosse, 2004).

19. In both cases the analytical black box is ‘incentives’. In addition, there may be additional structural determinants.

20. The way this service provision based relation between farmers and the irrigation agency resulted in increased yields highlights the paradoxical notion of poor systems performance/deferred maintenance arguments in donors' dominant narratives. In Indonesia, ‘poorly performing’ systems produce average yields of 3–4 ton/hectare with 2–3 rice crops/year, while Indonesia became a rice exporting country in the Suharto period. Whether achieved yields and other system parameters constitute poor performance depends on the standards adopted, what is incorporated in these, and how they are expressed. A discussion of the ‘politics of standards’ that could be imagined in this connection, and which would involve mobilising critical debates on indicators and cost/benefit calculations, is outside the scope of this article. We are of the opinion that there is a performance issue in Indonesian irrigation, otherwise we would not have written the article, but the argument of the article also suggests that actual levels of performance are perhaps not the key factor in poor performance policy discourse.

21. These doctrines include, in Indonesia, the pasten or k-factor system as well as operation guideline for specific irrigation infrastructure such as the Romijn gates.