Effect of Participatory Irrigation Management on Agricultural Yields and System Performance

This paper assesses the effects of PIM on agricultural yields in Mexico, Colombia, Sri Lanka and Indonesia where the International Water Management Institute (IWMI) has conducted extensive and in-depth studies of transferred irrigation systems. The paper also attempts to provide more general information about the impact of IMT and PIM interventions in those same countries and others. Towards the end of the paper, the author examines the economic productivity of land and water irrigation systems / sub-systems in 11 countries.

Selected Countries

PIM was introduced in Mexico, Colombia, Sri Lanka and Indonesia. The national IMT program in Colombia can be characterized as a significant but only partial devolution of management to water users. By the end of 1996, 17 of the 23 public irrigation districts in the country had been transferred to WUAs. This constitutes 115,695 ha of a total of 241,077 ha of irrigated area under public management.

In Mexico, the WUAs had taken over the financial and managerial responsibilities for operating the system below the main canals in the first phase. In the second phase of transfer process, the responsibility of O&M of the main systems has been handed over to a Limited Responsibility Society (LRS), which is a federation of the WUAs at the district level. As of today, 75 of the 81 irrigation districts have been transferred to 406 WUAs representing an area of 2.9 million ha.

Following a decade of field experiments, in 1988 Sri Lanka adopted a policy of transferring full responsibility of O&M below the distributary canal head of medium and major schemes to farmer organizations. The program was implemented in 170 schemes.

In Indonesia, the government in 1987 decided that all small scale irrigation systems of less than 500 ha would be transferred to WUAs by the year 2003, which would affect 2.1 million ha of public irrigation service area. Systems with service areas of less than 150 ha were to be given first priority. Implementation began in West Java and West Sumatra.

A key difference between these approaches exists in Latin American and Asian countries. In the former, the programs aim not to maximize direct user farmer participation in the O & M but to involve farmers in representative governance. In the latter, however, farmers are usually direct participants in maintenance activities.

PIM and Agricultural Productivity

The author reviews IWMI studies of the impact on crop yields as well as in crop intensities using a sample drawn from a variety of system types, sizes, agro-hydrological conditions and socio-economic conditions. In the case of Colombia transferred and non-transferred entities were compared; in Mexico, only the transferred units were selected; in Sri Lanka, transferred and non-transferred units along with rehabilitated and non rehabilitated units were taken; and in Indonesia, small scale irrigation systems were selected that had the longest experience with IMT.

In the case of Colombia, crop yields remain conspicuously low in both groups of transferred and non transferred systems. Crop intensities in only one system out of four increase from 110% before IMT to 165% after IMT. However, even that increase is attributable to a shift to sugarcane production (semi-perennial) and to perennial grape and crops.

Evidence drawn from Mexico, over a 15-year period, shows an upward trend on bean yields and a more pronounced upward trend on wheat and barley yields over the period. However, increases started several years before IMT took place. All that can be said is that in Mexico, the IMT program has not resulted in any yield declines.

In Sri Lanka there has been a significant improvement in crop yields in schemes that have undergone both IMT and rehabilitation. The analysis also indicates that agricultural production has stagnated in groups with only one type of intervention (rehabilitation or IMT) and may have declined slightly in schemes where there has not been any intervention. In terms of cropping intensities, results show that in all types of schemes, the trend in pre-IMT period has been relatively stable. In the post IMT period, the schemes which have been rehabilitated and transferred show a statistically significant increase in cropping intensities. In the same period, there may have been a slight decline in cropping intensities in non-IMT schemes.

In Indonesia, comparisons were made between systems with and without IMT, and between before and after IMT in the case of transferred systems. The results do not show significant yield differences between the two groups nor in the trend lines between the two groups. However, before and after IMT comparisons show contrasting results in terms of farmer reporting.

In conclusion, for the cases reviewed, it has not been proven satisfactorily that participatory irrigation management has led to significant positive changes in agricultural yields as a result of implementation of IMT programs. A few encouraging cases exist and it is possible that it is still too early to be able to observe the full benefits of PIM.

IMT and System Performance

The basic idea is to treat the irrigation system as a 'black box' where inputs like water, land, labor and financial resources give origin to output, in the form of crop produce and financial gains. These indicators not only allow comparisons to be made among different systems, but also within a system at various levels such as main, secondary, and on farm. An added advantage is that such indicators are suitable for tracing performance over time.

IIMI has developed a minimum set of indicators that can be used to undertake across system performance evaluation within countries and on a global basis with an end objective of establishing determinants for the successful performance of irrigation systems under various settings and conditions.

Financial

The primary interest of the respective governments in different countries was to reduce the government expenditure on O&M costs. In Colombia, from 1985-1995, government expenditure for selected surface systems fell from a range of $20 to $80 per ha before transfer to near zero after transfer. In the case of a lift irrigation scheme transferred in 1995 during the development stage government expenditure rose to $200-250 per ha up to IMT but has remained at the same level after the IMT.

In the case of Mexico, data are given on one irrigation system that increased its self-sufficiency from 50 percent before transfer to over 120 percent in post-transfer years. Data for the 4 years that follow transfer show an average collection rate of 120 percent.

The studies in Sri Lanka indicate that between 1985 and 1990, there was a statistically significant decline in the trend in government expenditure for O&M, which has continued in the post-IMT period.

Quality of Operation

The quality of irrigation operations can be assessed through two variables: Relative Water Supply (RWS) and Relative Irrigation Supply (RIS). Both relate water supplies to crop water requirements. Measurements taken after IMT in relation to these variables generally indicate poor management because after IMT water continues to be in excess supply in all four countries. The conclusion emerging from the cases from the four countries is that IMT has not led to improvement in systems management.

Sustainability of Infrastructure

The question of who bears the cost of rehabilitation closely related to recurring maintenance cost. If WUAs think that the system rehabilitation cost will eventually be borne by the government, WUAs may have a tendency to defer maintenance in order to lower the recurring costs.

A comparison of network conditions before and after IMT would provide a clear-cut assessment of the sustainability of the infrastructure after the IMT. However, in the absence of data concerning pre-IMT conditions, the author has relied on data relating to the post IMT position. In the case of Colombia, with one exception that has peculiar maintenance problems, schemes show that 11 percent of the infrastructure is dysfunctional. In Sri Lanka, only five percent of all structures were dysfunctional, though 15 percent of the main canal length that was inspected proved dysfunctional, which was relatively high. However, at the distributary level (under WUA control) the defects observed were three to 15 percent. The same exercise carried out in the Indonesian systems showed excellent canal conditions.

Economic Productivity

The indicator called Standardized Gross Value of Production (SGVP) was developed for conducting cross-system comparison mainly because there are differences in local prices at different locations throughout the world. To obtain SGVP, equivalent yield is calculated based on local prices of the crops grown and, compared to the local price of the predominant, locally grown, internationally-traded base crop. The second step is to value this equivalent production at world prices. The four basic comparative performance indicators are output per cropped area, output per unit command, output per unit irrigation supply and output per unit water consumed. These indicators were applied in 18 systems or parts of irrigation systems, located in 11 countries. The systems represented a wide range of agro-climatic regions having different characteristics, crops and cropping patterns, water distribution patterns, water resource availability and management style. The results of these four indicators are described below.

The measurements with respect to SGVP per unit command show that systems which mostly grow rice with low cropping intensity are found in the low range value. Middle range value is represented by rice with high cropping intensity of 200 percent. High range values are found in orchards, industrial crops and some cereals. Measurements of SGVP per unit cropped land indicate that non-rice producing irrigation systems can be more productive than the rice producing irrigation systems by 100 to 200 percent. The findings in respect of SGVP per unit irrigation supply indicate the lowest value for purely rice based systems. Irrigation systems which grow rice during rainy seasons and other crops during the dry season obtain middle range values. The highest value is obtained by systems which grow orchards, industrial crops and vegetables. SGVP per unit water consumed shows that purely rice-based systems with abundant water supply with cropping intensity at less than 100 percent give low returns, whereas water-scarce systems with orchards and industrial crops watered by private well pumping give values between 100 and 600 percent.

Based on the examination of the studies, the author infers the following:

1. At best, the effects of PIM on system performance in the context of IMT are modest.

2. It is too early to expect benefits emanating from the IMT.

3. Evidence for one or two issues like decreasing costs to governments or WUAs attaining self sufficiency is significant to dramatic.

4. In terms of other improvements such as increases of crop yields, cropping intensities, the quality of operation of the system or the sustainability of irrigation infrastructure, results are less clear. In fact, in a good number of cases, they suggest negative impact. Farmers perceive IMT as an improvement that motivates them to try harder.

By Carlos Garces Restrepo
IWMI c/o CIMMYT, Lisbon 27, Col Juarez, AP 6-641
CP 06600 Mexico D.F., Mexico
cgarces@cimmyt.mx