1

The Distributive Impact of the Water Market in Chile: A Case Study in Limarí Province, 1981 - 1997

Donato RomanoDepartment of Agricultural and Resource Economics

University of Florence, Italyand

Michel LeporatiInstituto Nacional de Desarollo Agropecuario

Santiago, Chile

Published in the Quarterly Journal of International Agriculture vol. 41 (1/2): 41-58

2

• analysis of water market distributive impacts• determinants of peasants’ participation in the water

market

• peasant agriculture • agricultural sector• non-agricultural sectors

Limarì Province• research area of previous studies on the efficiency of

water market• highly representative of the Chilean agriculture

transformations

Objectives:

Perspectives:

Where?:

3

• water can be privately exploited through transferable use rights

• water property rights are separable from land property rights- ‘real’ rights - divisible- mortgageable- no use priorities

• only demand and supply will determine the resource allocation

• the State only guarantees the rules of the game.

The water market in Chile (Código de Aguas, 1981)

4

The state of the art (I): efficiency

(Theoretical) pros of private water market vs. public provision (Hearne and Easter, 1997):

• concentrate resources in higher-value uses better allocation

• more rational resource use conservation• higher share of private investments public budget

savings

5

The state of the art (I): efficiency

Real effects of the water market:

• positive assessment:- Gazmuri (1994): more efficient water allocation- Holden and Thobani (1995): infrastructures and institutional competencies- Ríos e Quiroz (1995): water scarcity and infrastructure

flexibility- Bauer (1997): incentive to investment and flexibility

• negative assessment:- Peña and Retamal (1992): market power and asymmetric

information - Barrientos (1995): few transactions - Bauer (1997): “ “

6

• By and large: pros > cons• Hearne and Easter (1995):

The state of the art (I): efficiency

The water market in Limarì province :

«the market transfer of water-use rights does produce substantial economic gain-from-trade. [These economic gains] are about three times the value of each water share from the Cogotí dam.»

But, in the Limarí valley:• low transaction costs (on average, 0.02% for buyers and 0.05%

for sellers)• presence of the reservoir storage (Cogotí dam), modern

infrastructures (gated canals, flow meters, etc.) and well organized user associations.

7

The state of the art (II): distributive effects

Theoretically: neutrality, provided that there is• access to information• no market power• extension and technical assistance

Practically:• positive effects

- Gazmuri (1994) and Thobani (1995): poverty reduction, thanks to trickle down effects

• negative effects- Ríos and Quiroz (1995) and Bauer (1997): peasants disadvanta-

ged because of lack of knowledge about new procedures, lack of financial resources, and weak bargaining power

8

The state of the art (II): distributive effects

The water market in Limarì province (Hearne and Easter, 1995):

«the market transfer of water-use rights does produce substantial economic gain-from-trade. These economic gains occur in inter-sectoral trades and in trade between farmers, and they produce rents for both buyers and sellers. Yet buyers, especially large table-grape producers (…), receive higher rents than sellers.»

Consistent with the broader Chilean context of «exclusionary growth» (Carter and Barham, 1996).

9

Analysis of all registrations (both original registrations and transfers) of rights among agents which took place between 1981 and 1997 in Limarí Province.

The case study (I): market functioning

Water market is a very active institution in Limarí Province:• total registrations: 1,367 • yearly average: 80.4• corrected yearly average: 39.3• market transactions: 84% of total registrations

10

0

100

200

300

400

500

600

1981 1983 1985 1987 1989 1991 1993 1995 1997

Years

Tot

al r

egi

stra

tions

agriculture non-agricultural sectors

constitutional plebiscite (1988)

presidential elections (1993)

inception (1981)

Total registrations of water-use rights, 1981-97

11

Intra- and inter-sectoral registrations of water-use rights, 1981-97 (percentage on total registrations)

12

Characteristic Non-agricultural sectors Agriculture total Capitalist agriculture Peasant agriculture

Type of registration Original regs. = 13.95 Original regs.= 56.28 Original regs.= 21.46 Original regs.= 81.63 (% on sector total) Mkt. transactions = 86.05 Mkt. transactions = 43.72 Mkt. transactions = 78.54 Mkt. transactions = 18.37

Avg. registrations size 16.24 12.24 15.55 7.48(shares/registration)

Concentration index 1.59 1.44 1.60 1.15(registrations/individual)

Purchasing/selling ratio:   - Registrations 1.94 0.90 1.80 0.48   - Prices 0.91 1.07 1.51 1.04

Re-sales 60.84 32.35 60.17 13.16(% on sales)

Agents’ behaviour among sectors

Different behaviour:

Peasantagriculture

Capitalist agriculture

Non-agricultural

sectors

13

Conclusions:• quite far from a perfectly competitive market

- market power - access to information?

• peasant agriculture dependent

The case study (I): market functioning

14

Evolution of water-use rights among different population quintiles on the period 1981-97:

The case study (II): distributive effects

Gini coefficient Lorenz curves

Whole populationPeasant agriculture

15

Evolution of water-use rights distribution,all sectors, 1981-1997

0

10

20

30

40

50

60

70

80

90

100

0 20 40 60 80 100

% population

% w

ate

r share

s

1981 1987 1992 1997

p.c. shares = – 16.7%

years Gini

1981 0.54

1987 0.54

1992 0.56

1997 0.57

users = + 27.0%

16

0

10

20

30

40

50

60

70

80

90

100

0 20 40 60 80 100

% population

% w

ater

sha

res

1981 1987 1992 1997

Evolution of water-use rights distribution,peasant agriculture, 1981-1997

p.c. shares = – 83.4%

years Gini

1981 0.50

1987 0.51

1992 0.56

1997 0.58

users = – 53.7%

poor = + 146.7%

17

Economic, social and institutional determinants of peasants’ participation and behaviour in the water market: survey on 114 peasant households (2.38% of total HHs)

The case study (III): determinants of peasants’ participation

logistic regressiontransact/do not transact

buy/sell

18

Logit model estimates for peasants’ involvement in the water market transactions

Dependent variable y=1 do not transact y=0 transact Degrees of freedom 5

-2logl restricted 144.406-2logl non restricted 107.865

2 36.541 Percentage of correct forecasts 82.8%

Variables Coefficients Pr>2

Intercept 1.2146 0.0482

JFED: Age of household head 0.1989 0.0153

MEDNI: Household education attainment -0.3258 0.0013

OTRO: Being member of organisations -3.3168 0.0005 other than water-users

PAS: Farm grazing area 0.9293 0.0154

INF: Level of information about local -3.3251 0.0290 water market

19

Logit model estimates for peasants’ behaviour in the water market

Dependent variable y=1 sell y=0 buy Degrees of freedom 6

-2logl restricted 73.325 -2logl not restricted 58.682

2 28.683 Percentage of correct forecasts 87.6%

Variables Coefficients Pr>2

Intercept 0.1895 0.0918

JFED: Age of household head 0.0759 0.0321

MEDNI: Household education attainment -0.0857 0.0119

CRE: Access to credit -1.9503 0.0225

DIR: Being manager of water-users -1.3870 0.0293 associations

ACC: Total number of water shares -0.1358 0.0291

INF: Level of information about local -7.5430 0.0009 water market

20

• Private water market has become a very active institution in the Limarí valley

• But, there are evidences that the market is highly imperfect:- asymmetric information - barriers to entry due to social, cultural, economic, and

institutional conditions existing among different agents

Conclusions

• The distribution of water resources is heavily unbalanced and it tended to worsen :- general decrease in the amount of per capita rights - accumulation of resources by the most powerful social-economic

groups - peasants are the group that suffered the greatest loss: water is a

strategic determinant of agricultural development

21

• Determinants affecting peasants’ participation and behaviour in the water market: crucial role of the variables relative to information, education and participation in agricultural organisations and in water users associations potential role of government interventions in:- training of human resources

- promotion of adequate and accessible information for the whole population

Conclusions

22

Conclusions

• Distribution dynamics influenced by the Government’s abdication from its role as ‘referee’ of the game (consequence of the radically liberalistic environment where the reform was designed and implemented)

• The Concertaciòn Governments’ ‘growth with equity’ objective has not been achieved