17 Mar 2011

Measuring subsidies to kill them?

Coincident to the post earlier this morning, I got this email:

The Global Subsidies Initiative (GSI) announces the release of a new study: Measuring Irrigation Subsidies in Andhra Pradesh and Southern India: An application of the GSI Method for quantifying subsidies [PDF]

The second report in GSI's series measuring irrigation subsidies provides the starting point for a debate on the use of irrigation subsidies in India. Subsidies to irrigated agriculture in major irrigation projects in just the four south Indian states (Tamil Nadu, Andhra Pradesh, Karnataka and Kerala) were conservatively estimated at US$575 million per year from 2004 to 2008. The full subsidy is certainly substantially more, as significant numbers of small- to medium-size schemes were not included in the study, nor were substantial electricity subsidies to private groundwater irrigators.

The majority of the subsidies were for interest on capital, and operation and maintenance (O&M) costs, over and above the revenue generated from providing irrigation services, hydro power generation and pollution abatement fees. Irrigation projects that were not completed to schedule were found to have significant cost overruns leading to government funding tied up in projects not delivering any benefits or revenue. When governments were unable to collect revenue from farmers, re-investment back into completed systems stopped and system maintenance declined.

Collecting basic data for the study was a challenge. Government-funded supply costs were not recorded and compiled with the notion of generating accurate cost estimates for irrigation activities. What data are available are often spread over a variety of government departments. The study recommended that the government encourage water authorities to maintain better records and publicly provide information on water costs, revenues and subsidies, in a more organized and accessible manner.

There may be clear public policy reasons for providing subsidies. However, without accurate information on subsidy volumes, public policy debate is hampered and the relative costs and benefits of alternative paths to policy goals impossible to assess.

This study was prepared on behalf of the GSI by International Water Management Institute.
I asked Tim Shah to comment on the report, and he said:

Irrigation subsidies in India are indeed a huge problem. Based on the literature though, and the stuff I read for my project, Andhra Pradesh has been a unique case which I summarized in my blog post to you. I don't know as much about the other southern states (e.g. Tamil Nadu and Kerala), but farmers in AP have generally become more effective in managing their crops in spite of the generous subsidies thrown at them. As AP is highly reliant on monsoonal rains to replenish its groundwater aquifers, the need to conserve, manage crops more efficiently and sell such crops to markets is even more critical than other places in the country (such as Punjab) which have more friendly hydrogeological conditions and higher storing capacity which has mistakenly given farmers this idea that water is abundant and limitless.

While I have not read much about this GSI method of quantifying subsidies, I do not think it is the "be all end all" solution by any means. This could be carried out in any state in India and would simply tell institutions that they have been idiots for years through providing OVERLY generous subsidies. Obviously there have been cost overruns which have led to government funding tied up in projects not delivering any benefits or revenue. But by using a more grounded and participatory approach, like the one we researched, it can act as a better model to manage and maintain records of state subsidies to farmers. This way, farmers are utilizing all of the hydrological data possible for groundwater conservation, they are engaged in their practices and learn about how to efficiently and strategically manage crops based on limited water resources. From here, the state government could provide subsidies accordingly based on how much water and electricity they need. If the subsidy providers become more strict about this, then the farmers will already have their hydrological data which the state can use to assess funding allocation. The GSI approach seems inaccessible, disconnected from rural reality and not participatory!

Better public policy in these Indian states is not going to only come from the quantification of subsidies. It will require a much more participatory and "two-way" approach where farmers and subsidy providers are actually communicating with one another. This leads to better policy because it is transparent and can ultimately lead to the increase in the welfare of rural farmers through the efficient allocation of resources.

DZ's Bottom Line: Subsidies can be troublesome because they distort behavior. Although it's possible to coordinate action to overcome this behavior, success is neither easy nor guaranteed. So GSI's initiative to highlight the explicit cost of subsidies is welcome, as is Tim's observation that farmers are capable of responsibly managing their water in a sustainable way.

I'd reconcile these views with the political desire to throw money at farmers by suggesting that governments just send money to farmers. Such a "solution" delivers politicians' gifts of taxpayer money to farmers while minimizing the distortionary effects of a subsidy on behavior.

Unfortunately, it also highlights the bribery, so perhaps it's not something we're likely to see :)