Water allocation, crop choice, and priority services

• Published date: 01 February 2022
• Author: François Salanié,Vera Zaporozhets
• Date: 01 February 2022
• DOI: http://doi.org/10.1111/jpet.12538

J Public Econ Theory. 2022;24:140–158.wileyonlinelibrary.com/journal/jpet140

|

© 2021 Wiley Periodicals LLC

Received: 30 March 2021

|

Revised: 25 June 2021

|

Accepted: 5 July 2021

DOI: 10.1111/jpet.12538

ORIGINAL ARTICLE

Water allocation, crop choice, and priority

services

François Salanié |Vera Zaporozhets

Toulouse School of Economics, INRAE,

University of Toulouse Capitole,

Toulouse, France

Correspondence

Vera Zaporozhets, Toulouse School of

Economics, INRAE, University of

Toulouse Capitole, 1, Esplanade de

l'Université, 31080 Toulouse Cedex 06,

France.

Email: vera.zaporozhets@tse-fr.eu

Funding information

Agence Nationale de la Recherche,

Grant/Award Number: ANR‐17‐

EURE‐0010

Abstract

We analyze the problem of allocating irrigation water

among heterogeneous farmers when water supply is

stochastic. If farmers are risk‐neutral, a spot market for

water is efficient; while the oft‐used uniform rationing

system is inefficient, both ex ante and ex post. Indeed,

we show that it leads farmers to overexpose to risk,

thus making shortages more severe and more frequent

in case of drought. We propose instead a regulation by

priority classes extending Wilson, and we derive an

efficiency result. We characterize the set of farmers

that would win or lose from such a reform. We also

argue that a system of priority classes may be preferred

to a spot market system, because scarcity is easier to

manage ex ante than ex post, and because this system

facilitates the supply of insurance to risk‐averse agents.

1|INTRODUCTION

Some resources, such as water for irrigation, are subject to severe supply shocks, and this

feature makes it difficult to allocate them across competing uses. While economists often

advocate markets as a solution, it is fair to acknowledge that they also raise practical and

theoretical difficulties, and that other systems may sometimes perform as well. This article

proposes to consider the institution of a priority class system, when it comes to solving the

difficult and important problem of allocating irrigation water across users.

Let us recall that agriculture relies on irrigation to produce 42% of crop output, and is

responsible for 70% of total water withdrawals (FAO, 2012). An unfortunate consequence is

that agricultural water demand contributes significantly to water scarcity during droughts. In

addition, climate change is expected to aggravate the severity and duration of droughts in many

regions (IPCC, 2002; Swain et al., 2018; UN, 2012,2015). These considerations have led various

countries to examine reforms of water allocation systems to make them more robust to un-

certain supplies.

1

Indeed, irrigation is often managed thanks to coarse regulatory instruments, such as ra-

tioning. These instruments are economically inefficient, as they do not allocate water supply to

the most productive users. Consider for example the French system. In this country, farmers

have to apply for yearly irrigation licenses, which specify a reference volume that can be

extracted at each location. In the river basins that have adopted volumetric water management,

water is allocated to users proportionally to these licenses (Lefebvre et al., 2014). Short‐term

policies for managing water scarcity include temporal restrictions of water use (such as night

irrigation) which are implemented by the prefects. Between 2011 and 2014, there were 8950

prefectural decisions to introduce temporary restrictions in 1320 zones. The concentration of

these decisions on certain zones is striking. A 50% of zones got less than five decisions each,

25% saw more than 10 decisions, and 3% got more than 30. Among the 100 French depart-

ments, four departments got more than 500 decisions.

2

It is apparent that the system does not

work well, as some zones face almost permanent restrictions on water use, even in rainy years

(Erdlenbruch et al., 2013). Similar remarks apply to many countries. The absence of a price

system is seen by many observers as the main cause of the actual crisis in global water

management (Barbier, 2019).

An alternative would be to create water markets, which should allocate water efficiently,

even when supply is scarce. It turns out that water markets are not widely implemented, with a

few exceptions, such as in Australia (Connor & Kaczan, 2013; Grafton et al., 2015), in the

western states of the USA (Hagerty, 2019), and in Chile (Bjornlund & McKay, 2002; Howitt,

1998). In reality, multiple market failures, difficulties associated with implementation, and

sometimes political obstacles, have led to the adoption of various other systems, depending on

the objectives of water authorities (Dinar & Mody, 2004).

In this paper, we propose an alternative system aiming at a better balance between efficiency,

complexity, and social acceptability. A few priority classes are created. Ex ante, typically in fall,

farmers register in a class by paying the corresponding price. Ex post, in spring or summer, once

the water supply is observed the members of the highest priority class are first served, and then the

members of the second highest priority class, and so on until all available water is allocated.

Priority classes ensure that high‐value users have a higher probability of receiving water,

which is good for efficiency. In fact, in a framework with 0/1 demand, Chao and Wilson (1987)

and Wilson (1989) show that such a priority system achieves all the efficiency gains attributed

to spot markets. This system was implemented in the electric power industry to manage peak

demand and the consumers' reliability needs. In practice, the difficulty is often to determine the

prices that apply to different levels of reliability, in different places, and for different periods

(Beenstock & Ephraim, 1997; Strauss & Oren, 1993). Industrial users sometimes also want to

choose different reliability levels for different equipments. These requirements raise technical

difficulties that are specific to the management of a power distribution network.

3

Our

1

The adaptation of agriculture to water scarcity and climate change is listed among the key objectives of European

water and agriculture regulation (e.g., WFD, 2000/60/EC and EC, 2013).

2

On the basis of own calculations using data from http://propluvia.developpement-durable.gouv.fr/propluvia/faces/

index.jsp.

3

See Davidson et al. (2019) for the analysis of a recent incident that leads to an interruption of service for some

consumers. Astier (2021) discusses the trade‐offs involved when using price schedules that may be more or less

complex. For a recent reference on priority classes, see Gershkov and Winter (2020).

SALANIÉ AND ZAPOROZHETS

|

141