Annual expenditures on management and maintenance of irrigation systems in the Punjab province amounts to over Rs5 billion. Revenue collected through water charges remain much below the expenditures , resulting in provision of large subsidies to the sector.

Substantial amount of funds is spent annually on the collection of water charges. Appropriate charging policy is important not only for making water charge collection more efficient, but also for recovering the costs more effectively. It has strong implications for income distribution in the rural areas.

Water charging policy option as suggested in this article would result in significant increase in revenue and lead to re-distribution of Rs976 million every year, in favour of small and poor farmers.

On 10 June 2003, the Punjab government has taken a historic decision to change the canal water charging policy from crop-area-cum-crop type based charging to crop-area-based flat-rate charging. Under crop area-cum-crop based system, water charges were levied based on the area cropped and differentiated by the type, condition and season of the crop.

Charges were generally higher for high water consuming crops such as rice, and low for less water consuming crops such as wheat (for example, per hectare crop-based-water-charges prior to June 10, were Rs37 for fodder, Rs148 for wheat, Rs222 for cotton, Rs297 for rice, Rs432 for sugarcane).

Under the new flat rate system, per hectare water charges are to be fixed for Rabi and Kharif seasons, regardless of the type of crops grown in each season (new rate per hectare are Rs124 for Rabi and Rs210 for Kharif crops, regardless of the type of crops grown). Several factors have led to this change. These include: (a) the crop-based charging system was considered obsolete and not in line with the changing water and irrigated agriculture situation; (b) it was thought to be often manipulated by the influential farmers and revenue officials (such as the mis-reporting and mis-recording crop types and crop areas, e.g., charging for fodder rates when high water rate crops such as rice or sugarcane were cultivated); (c) as the water charge assessment was based on the discretion of the revenue officials, it is generally perceived that it has lead to creating an environment for rent seeking behaviour; (d) crop based charges were considered advantageous to large farmers and disadvantageous to small farmers that constitute majority in the farming community; and (e) old system has lead to increased pressure on public funds resulting from widening gap between the irrigation expenditures and revenue collection (e.g., in recent years total revenue collected through water charges in the Punjab province accounted for 31.4 per cent (or Rs1.6 billion) of the total expenditures (of Rs5.1 billion).

Also, the estimates suggest that the Punjab government has been spending three rupees for every one rupee of water charge collected).

Reaction to this policy change has been mixed. While, there are many supporters of this change who see this as a welcome development, some continue to resist and criticize this policy change.

There are three key issues in relation to this policy change. First, in the newly introduced flat rate system, the charge is based on the farm area cropped during Rabi and Kharif seasons (i.e., flat rate per hectare of area cropped in each season) and not based on farm area owned or farm area having water entitlement or area receiving water.

Under warabandi type of systems, water allocation is made on the size of farm landholdings. In aggregate terms, large farmers receive and use more canal water than small farmers. On per hectare basis, if we assume that small and large farmers receive similar amount of water, small farmers who generally have higher cropping intensity will end up paying more per hectare water charge than large farmers.

As in the past, under the new policy, area irrigated partially from surface water and partially from groundwater would be fully liable for canal water charges. Those who make more use of groundwater and other inputs to increase their cropping intensity would have to pay more in per hectare water charges. It is important to note that it is the farm size that forms the basis of water allocation, regardless of the proportion of farm area cropped. Therefore, it makes sense to levy flat rate charge based on farm size or farm area having water entitlement.

The new crop area based flat rate policy, though does not account for intra-seasonal crop differences, it accounts for inter-season crop differences and, like old system, water charges are levied based on area cropped and cropping intensity during a season. So while, the new system would help in addressing the crop type-misreporting issue, but it would not resolve the problem of crop area mis-reporting. The flat rate per unit of land based on land size, independent of crop type and cropping intensities is better option to address both of these issues.

Second, there are two critical questions: (a) will the flat rate policy result in increased revenue? and (b) what are the implications of such policy changes for small and poor farmers?.

The International Water Management Institute (IWMI) has recently undertaken an in-depth study on the implications of alternate water charging policies in Pakistan. The study is based on data sets collected through households level surveys from 891 farm households from 10 canal distributaries in four scale large surface irrigation systems in the upper Indus basin (namely, Upper Jehlum Canal system, Lower Jehlum Canal system, Lower Chenab Canal system and Hakra canal system). The survey covered the period 2001-2002 agricultural year. Following policy options are analyzed in this study:

Option 1. Present policy: No change in the structure and level of water charges, charges are based on cropped areas and cropping intensities (old system); Newly introduced so-called 'flat rate policy', which is based on cropped area, crop season and cropping intensity, independent of crop type in a season, is similar to the old system

Option 2: Flat rate per unit of land based on land size, independent of crop type and cropping intensities, applied uniformly across all farm size categories; (flat rate)

Option 3: Differential rate per unit of land based on land size, applied differentially across various farm size categories - progressive rate structure (similar to increasing block rate charging). Lower water charge for the first two hectares, applied uniformly to all land size categories, and charge increases progressively with increase in size of holdings above 2 hectares, by Rs50 per ha for each successive category of land size.

The study indicates that average annual water charge per ha (area weighted) was Rs420 per hectare in 2001-2002. Average per hectare water charge was found to be inversely related to land size categories; landholders between 1 to 5 hectare paying significantly more than the overall average, and those with greater than 10 hectare paying less than the overall average.

This is due to differences in cropping intensities. Average cropping intensity varies significantly across farm size categories, with highest cropping intensity of 181 per cent on smallest farm size category of less than one hectare and lowest cropping intensity of 115 per cent on largest farm size category of over 10 hectare. Smaller size farms, constrained by the size of land tend to use their plots more intensively, making greater use of available family labour (and other inputs), and more importantly they make greater use of groundwater for irrigation.

The study indicates that average groundwater cost per hectare is inversely related to farm size categories, smaller size farms using more groundwater and incurring higher costs and vice versa. For example, average annual per hectare groundwater cost for farms of sizes between less than one hectare to less than 10 hectares varies from Rs3209 to Rs4555. On the other hand, average ground water cost for farms over 10 hectares is Rs2779.

Since under the present charging system, crop area that is partially irrigated with canal water and partially with groundwater is fully liable for canal water charges, the system tends to penalize small and poor farmers who use more groundwater for irrigation. On average, poor farmers incur Rs56 more in total per hectare cost of irrigation than the non-poor, due to greater use of groundwater and resulting higher overall cost.

Overall, cropping intensity based water charging is pro-large farmers, and anti-small and poor farmers who make greater use of groundwater to increase their cropping intensities. The new crop area based flat rate system (independent of crop type) is not much different from the old crop type-cum-crop area based system.

The study indicates that option one (current policy) is pro-large landholders and penalizes poor small farmers whose cropping intensity is generally higher due to relatively greater use of groundwater and other factors. Option 2 is better than option 1, but it is not pro-poor. Option 3 is better than both option 1 and option 2 from both revenue/cost recovery and equity perspectives. Option 3 is pro-poor, as per hectare charge for the poor would be less than for the non-poor, and overall revenue would be more than that in the first two options.

In Punjab province, adopting policy option 2 (flat rate based on land size independent of crop type and cropping intensity, based on average rate) would result in annual gains for smaller size farmers (up to 5 hectare) through reduced costs by Rs74.45 million, and cost to larger farmers would increase by Rs326.77 million, and total revenues will increase by 5.3 per cent.

Policy option 2 is better option than policy option 1 in terms of equity and revenues. Under policy option 3, smaller farmers, as a result of reduced costs, would gain annually by Rs346.88 million, and larger farmers would contribute more towards costs by Rs529.76 million, and overall revenue would increase significantly by 21.8 per cent. With policy option 3, Rs876 million would be re-distributed, with significant amount in favour of small and poor landholders in the province.

For country as a whole, adopting policy option 2 would result in annual gains for small farmers through reduced costs by Rs130.06 million, and cost to larger farmers would increase by Rs605.97 million, and annual total revenues will increase by 5.6 per cent.

Under policy option 3, smaller farmers, as a result of reduced costs, would gain by Rs519.65 million, and larger farmers would contribute more towards costs by Rs842.45 million, and overall revenue would increase significantly by 22.7 per cent. With policy option 3, over Rs1362 million would be re-distributed each year, with significant amount in favour of small and poor landholders in Pakistan.

It is concluded that water charging policy that is based on cropped areas and cropping intensities favours larger landholders, and disfavours small and poor landholders. Policy changes towards flat rate per unit of land based on land size, independent of crop area, crop type and cropping intensities, applied uniformly across all farm size categories or differential rate based on land size applied differentially to various land size categories would result in redistribution of significant amount of rural incomes, largely in favour of small and the poor landholders.

These policy changes could be implemented with existing institutional arrangements in place, and as such do not involve any major implementation costs. Major benefits from such a policy change would include: (a) more funds available for operation and maintenance of irrigation systems, with resulting improvements in performance of the systems leading to increased efficiency in water supply and productivity; (b) significant benefits in terms of reduced costs to small and poor landholders; and more importantly, it would be a step forward to reversing existing inequities in water charges. What is needed is a clear understanding of the problem and strong political will to a policy change that improves irrigation management and is pro-poor.