“Willingness to pay, borrow, and work for rural water service improvements in developing countries”
A. Abramson, N. Becker, Y. Garb and N. Lazarovitch
This recent publication in the leading water journal, Water Resources Research, highlights one aspect of my studies. The main research question we address is how different financing approaches compare in enabling cost recovery for rural water service improvements in developing countries, with Simango as a case study.
You can Download It Here
A summary, posted on the website, Global Water Forum, is below. This is the theoretical basis for the 2011 fieldwork conducted in Simango.
“Changing the paradigm for financing of rural water improvements”
Many rural communities of developing countries engage only partially in the modern cash economy, since they are characterized by traditional subsistence agriculture. A high proportion of their needs are met directly by non-cash transactions such as trading time in farm labor in return for a harvest, or bartering the surplus of that harvest for other household goods or services. Thus, only a fraction of their daily activities and decisions are negotiated through cash. Since basic infrastructure lies outside of this traditional economy, these communities often fail to finance improvements in their water and electricity, relying, instead,, on donations or subsidized government programs to meet these needs for them.
In a market economy, prices and market transactions often provide all of the information necessary for policy-making. But where an economy or market sector is undeveloped, the conventional approach is to use what’s called stated preference methods to find out what a functioning market would otherwise tell us. These methods often consist of asking people, either directly or indirectly, what they would be willing to pay for some hypothetical change to the status quo. These kind of stated preference methods have been promoted in rural areas of developing countries to inform water development policy, especially since the UN convention in Rio di Janeiro in 1992, which forwarded an approach to valuing water services as economic goods, instead of just as human rights that everyone should somehow or other be given,.
The idea behind this approach is that sustainable policies for water development can only be formulated when projects can recover their costs. And they can only recover their costs when we have a good idea about what the market demand is for improved water services. Since 1990, over 20 stated preference water demand studies in developing countries have been published in the literature looking at precisely this question. Our review of these studies showed that, as a whole, they suggest that people in rural areas are significantly less willing to pay for water improvements than their urban counterparts. The policy upshot of this finding has been taken to imply that rural areas are seemingly unable or unwilling to fully finance their own water improvements in a way deemed necessary by a cost recovery approach. (In economic terms, there really isn’t a difference between being unable and unwilling to pay.) Thus, these areas are considered basket cases where only government subsidies or donor funding can be used to successfully achieve water development goals. In practice, they are asked to pay for only the operation and maintenance of water improvements–an approach that obviously falls well short of full cost recovery.
But what happens if we broaden the modalities of payment to better reflect these kind of non-market contexts? For example, what kind of preferences would be stated or expressed if these communities were allowed to pay for water improvements in a different currency, such as time, or after their cash income is supplemented? If they really don’t want improvements to their water services, we might expect there to be little difference in demand. But if they do want improvements, but simply lack the cash to pay for them we would expect their demand to increase with alternative payment arrangements. This is exactly what we found when we conducted a discrete choice experiment that asked a rural community in Zambia how much they would pay while receiving a no-interest loan, or, alternatively, how much they would work toward water improvements. We compared these alternative indices, which we termed Willingness to Borrow and Willingness to Work, with conventional Willingness to Pay (WTP). The results show that while all improvement scenarios we investigated would echo the conventional findings and fail to recover costs under the conventional WTP approach, some would succeed if users were given loans (with which they could supplement their incomes), and all would succeed under a scenario of being able to convert their willingness to work into project revenue according to various feasible activities.
These are certainly encouraging results since they indicate, at least for the typical rural area we studied, that:
1) There really is demand for water improvements
2) This demand isn’t being fully captured by the current approach
3) Other approaches may do a better job at capturing this demand
But it isn’t enough to show that people actually want water improvements. If cost recovery is the goal, the real challenge is how to convert this demand into a monetary form that will reach or exceed the costs of improvement.
We provide two models that may be replicable, as they are based on existing, successful platforms. The first is one in which loan programs are combined with rural water improvements. Loans would allow communities to earn supplemental income, which could then be used toward water fees. The second is a water-for-work program that combines water improvements with irrigation of high value crops such as vegetables, which could be feasible as an outgrower or contractual irrigation scheme. Our calculations based on experience in Zambia show that only a fraction of work hours expressed as demand for water improvements would be necessary to actually recover costs in the short term (2-year period).
This case study and the results of this discrete choice experiment suggest that the tradeoffs that are conventionally imputed by choice models in a willingness-to-pay paradigm may not be well suited to partially-monetized economies. In such contexts, expressed willingness to work a large amount of time is most certainly related to the fact that their value of time is not very high, because the labor markets are greatly undeveloped, and highly seasonal. Similarly, the WTB results are most likely a function of the absence of credit markets in the village, and this demand for a loan may represent a large portion of the willingness to borrow results.
In practice, it appears that water-for-work programs may hold the greatest potential for financially sustainable water development. Preliminary results from a 2011 non-hypothetical field experiment in Zambia demonstrate the kind of caution needed in stated preference studies, especially when these invoke cash payments in an economy where these are the exception . When over 100 households were actually offered water improvements at cost, a high percentage (about 75%) indicated that they would come to pay in cash. So far, not one payment has been made. But, when another group of about 50 households was asked to work for these improvements, most began paying through work commitments. Thus, indices of willingness to work appear to capture desires for water improvement, reduce the problematic infidelity between stated and revealed preferences, and provide mechanisms for practically translating these desires into important development goals.