Estimating Urban Residential Water Demand Determinants and Forecasting Water Demand For Athens Metropolitan Area, 2000-2010
Estimating Urban Residential Water Demand Determinants and Forecasting Water Demand For Athens Metropolitan Area, 2000-2010
Estimating Urban Residential Water Demand Determinants and Forecasting Water Demand For Athens Metropolitan Area, 2000-2010
Abstract
Aiming at sustainable water resources management and use, the current water
policy requires an essential analysis of water demand formulation and evolution.
In this context, the present paper considers the city of Athens in Greece, where
domestic water use is analyzed, estimated and forecasted. The policy-relevant
variables, mainly income and water prices, are systematically considered and
their effects on water demand are appraised. The study concludes that a drastic
increase in water demand induced by increasing income will occur, while the
economic instruments have little potential to influence water use.
*
Corresponding author: Department of Economics and Regional Development, 136, Leoforos
Sigrou, 176 71 Athens, Greece, e-mail: kbithas@panteion.gr
48 K. BITHAS, C. STOFOROS, South-Eastern Europe Journal of Economics 1 (2006) 47-59
1. Introduction
The continuously intensifying scarcity of water resources is a crucial problem in
almost all contemporary societies. Even in areas where there are adequate quantities
of water, the problem of scarcity is usually confronted through the deterioration of
water quality resulting in increasing costs for certain - mainly indoor - water uses.
The problem of water scarcity has manifested itself in different ways in recent years.
The most common ways are: a) increased cost of water usage, b) intensified competi-
tion over access to water resources and c) social insecurity (breakout of diseases) due
to the lack of water.
The traditional policy, which dominated the 20th century and aimed at confront-
ing the scarcity of water, depended on utilizing new water resources. Technological
advancement and economic growth accentuated water transferability from remote
sources, as a practical and relatively inexpensive practice. The socio-economic ben-
efits of transferring water were greater than the costs of exploiting new resources
and therefore the so-called supply policy prevailed. However, the efficacy of such a
supply-side measure is nowadays questionable, since the extensive use of water re-
sources has increased globally. Indeed, the hunt for new water resources would have
continued indefinitely if it had not reached an impasse: water resources are finite and
therefore eventually a continuously increasing number of potential users would be
competing for a given number of resources, as all these potential users are acting on
the basis of water supply policy.
This impasse becomes more intense as the demand for water supplies continu-
ously increases, pushed upward by the following main factors: a) increases in popula-
tion –those already experienced and those anticipated–, b) economic growth and c)
western life-styles.
The problem is more severe as far as domestic use is concerned. Domestic use re-
quires certain qualitative characteristics related to human health. Therefore, although
domestic use accounts for 10% of total water use, the availability of appropriate wa-
ter resources is limited and the costs for domestic use are rapidly increasing.
In this context, there has appeared a new model of policy that aims at maximizing
the benefits of the utilization of water resources already in use in order to minimize or
even to eliminate the need for new water supplies (Baumann et al, 1998). This policy,
defined as “water demand management”, is gradually becoming popular, especially
in developed societies. “Water demand management” mainly consists of the follow-
ing actions: a) the minimization of losses in transport and storage systems, b) the re-
use of water, c) the containment of water use by avoiding waste and d) the efficient
use of water resources (Renwick and Archibald, 1998).
In the framework of demand management, it is vital to analyze and to under-
stand the characteristics of water demand. How demand is formulated, which factors
determine it, how demand responds to changes in income and relative prices and
eventually how future demand will be shaped (Griffin and Sickles 2001, Arbues et
K. BITHAS, C. STOFOROS, South-Eastern Europe Journal of Economics 1 (2006) 47-59 49
2. Theoretical basis
Water Demand Analysis and Water Pricing
Essentially, the analysis of water demand aims at offering all the necessary informa-
tion and knowledge for designing an effective water demand policy, and specifically
a policy that pursues the efficient use of water resources. Accordingly, efficient use is
defined as a pattern of use that maximizes the benefits arising from the exploitation
of water resources (Tietenberg 1991, Pearce 1999). A pure competitive water market
would ensure efficient use by defining the optimum use of water and its optimum al-
location among competitive users.
Indeed, in a market that operates under competitive conditions, the price of water
would be determined by the interaction of market demand and supply to reflect the
actual costs of water usage. This price would induce users to purchase the optimum
quantity of water. In this context, no exogenous administrative intervention would be
necessary, as the “invisible hand” would, by itself, ensure the efficient level of use
induced by an equilibrium price that reflects water costs. Furthermore, the “invisible
hand” would lead to defining the appropriate investments in order to attain the effi-
cient use of water in the future. However, perfect competitive market conditions do
not and probably cannot exist in the majority of cases (Briscoe 1997, Pearce 1999). In
most cases, the supply of water is a monopoly whose characteristics closely resemble
those of a “natural” monopoly. Specifically, the extremely high infrastructure costs
for transporting, treating and delivering water make difficult the operation of multiple
water suppliers.
50 K. BITHAS, C. STOFOROS, South-Eastern Europe Journal of Economics 1 (2006) 47-59
The economic characteristics of the water sector, in combination with the fun-
damental social perception that water is a socially sensitive good related to human
existence and health, led to a strict administrative framework for the operation of the
water supply sector and hence of the water market (Kaika 2003, OECD 1989). In
effect, the fundamental decisions, like the determination of investments and prices,
have been strongly influenced by administrative rationale. In such a framework of
direct or indirect government interventions, the estimation of demand parameters
and characteristics acquires a special significance, since the decision-makers require
sufficient knowledge and information. Furthermore, if the objective of water policy
is to ensure socially efficient use, demand analysis is a precondition of designing such
a policy, since it defines the optimum socioeconomic water use and the respective
water price (Martiner-Espineira et al. 2004, Espey et al. 1997, Arbues et al. 2003).
The conditions for determining the efficient use of water and the relevance of demand
analysis, which offers the necessary information, are traced through Figure 1.
In Figure 1 curve LRMC represents the long-term marginal cost of water sup-
ply. The data and the information for the estimation of the LRMC curve are, to a
significant level, available to the administrative agencies responsible for formulat-
ing water policy. However, the authorities are hardly aware of the functionality and
operationability of the long run (marginal) costs of water supply. Curve D represents
K. BITHAS, C. STOFOROS, South-Eastern Europe Journal of Economics 1 (2006) 47-59 51
the marginal benefits of water use. However, the information for defining curve D
could be revealed in a free water market as well as through market surveys in the
administratively supervised markets. If a competitive market exists, then the curve D
defines the respective demand curve. The efficient use of water is determined by the
intersection of the LRMC and D curves. ΟΑ represents the efficient quantity of water
use (Tietenberg 1991, Pearce 1999). In this light, the efficient optimum use of water
would have prevailed in a competitive market. In the administratively regulated mar-
kets, the achievement of effective use, OA in Figure 1, can be ensured if the decision
makers define the appropriate price, OP, being equal to the marginal cost of OA water
use. To define the appropriate price, decision-makers should have sufficient informa-
tion on water demand. This information is not directly available through the function
of the existing water markets functioning under administrative supervision.
Furthermore, the analysis of demand can be an important tool in formulating the
relevant investment policy. The key factor for making appropriate investment deci-
sions is the expected (future) demand. Forecasting demand leads to the determination
of the optimal future level of water use and hence to the design of the future capacity
of the water supply system (delivery networks and processing plants). Water supply
infrastructure is considered to be expensive, especially as far as urban uses are con-
cerned. Therefore, the forecasting of water demand, and the definition of the level
and the appropriate magnitude of investments, is of crucial importance for the deci-
sion-makers.
ens Water Company together with the national government are interested in analys-
ing the current consumption patterns as well as predicting the future level of demand
for designing optimal investments as well as pricing policies.
There are several approaches to forecasting water demand (Granger, 1980, Gar-
diner and Herrington 1986, Luay-Froukh 2001,). Empirical models can take into ac-
count the effects of earlier events and as a result they can even explain the past fairly
well (Arbues et al. 2003, Dalhuisen et al. 2003). However, new unforeseen events
will occur and the future will always appear more uncertain than the past (Clements
and Hendry 2003). Furthermore, as Clements and Hendry (2003) note, the economic
forecasts end as a mixture of science - based on econometric systems that embody
consolidated economic knowledge - and art, namely judgements about perturbations
from recent unexpected events.
Single equation water demand estimation may not need to be an ad hoc assump-
tion. Following a standard dual approach, we assume an underlying structure of con-
sumer preferences based on a Stone-Geary utility function. The model is based on the
static theory of optimising household behaviour assuming similarity of preferences,
homogeneity of goods and perfect information (Mertinez-Espineira et al., 2001).
In particular, the approach adopted in this paper is based on the hypothesis that
all available sources for consumption are distributed evenly. According to microeco-
nomic theory, the individual choice is conceived as an interrelationship among the
quantity of goods that the consumer wishes and is able to buy in terms of price, p,
income, y, his preferences as well as social and demographic characteristics. In other
words, the consumer divides his income between quantities of goods and services,
q, so that an increase in the utility level, u, derived by the individual consumption, is
ascribed as:
ln D = βi ΣCij
ln Pj + Σγ ik
ln Zk (2.1)
where, lnD is the logarithm of the corresponding demand, lnP is the logarithm of the
price of residential water and lnZ other shift variables (income, trend, etc.).
A two-stage procedure is applied in order to forecast future water demand. In
particular, the first stage consists of the econometric model previously described and
in the second stage a synthetic model is constructed for conducting policy-relevant
simulations. The current simulation is based on a model where the pattern of water
consumption is the endogenous variable and price, income, and a weather index are
the main determinants of the system. The functional form of the equations and the
K. BITHAS, C. STOFOROS, South-Eastern Europe Journal of Economics 1 (2006) 47-59 53
parameters needed for forecasting purposes are derived from equation 2.1. The model
is then calibrated for the base year (1999, in this case) and the simulations are run
using the calibrated term together with the parameter estimates.
Dw = α + βi ln Pw + γi ln ln + δi ln X + εi
where Dw stands for the quantity of residential water demanded, Pw for the water
price, in real disposable income, X for the vector of other variables (in this particular
case a trend variable as a proxy to weather variations) and ε the error term. Moreover,
54 K. BITHAS, C. STOFOROS, South-Eastern Europe Journal of Economics 1 (2006) 47-59
price and income are expressed in real terms using CPI as the deflator. It is important
to stress that per capita real GDP was used as a proxy for the available income and
the size of the population was used for deriving per capita measures.
The results of the econometric analysis are presented in Table 1. In particular, all
variables have the expected, from the literature, signs and are statistically signifi-
cantly different from zero (t-statistic), the explanatory power of the model is high
(2.2)
(R2 and R2 adjusted) and finally, no problem of first order autocorrelation is apparent
(Durbin-Watson).
Source: Estimated
*Note: real GDP refers to per capita real GDP
The elasticities obtained from the present model are also presented in Table 1. Spe-
cifically, the residential water price elasticity is very low, which is also the outcome
of previous research for other countries, (Martinez, 2002) a finding that could be
partly explained by the fact that household expenditure on water is a very small frac-
tion of its total expenditure. So, if the government, for scarcity reasons, decides to
reduce actual water consumption, tariffs are not a very effective policy measure. The
income elasticity is less than unity but much higher than the own price elasticity.
Thus, as income increases, accordingly, the demand for residential water will in-
crease by a smaller proportion. Finally, the trend variable, which has been used as a
K. BITHAS, C. STOFOROS, South-Eastern Europe Journal of Economics 1 (2006) 47-59 55
Figure 2. Actual and Fitted Values - Per Capita Consumption of Residential Water
Source: Estimated
In order to assess the credibility of results, the model was run over the sample
period (1981-1999). In Figure 2, comparisons between actual and fitted values are
presented. It is clear that conducted simulations capture the actual time path of the
variables reasonably closely. Additionally, two statistics for examining the forecast-
ing accuracy, namely the root mean square percent error (RMSPE) and the mean
percent error (MPE) are reported. The statistics suggest (the RMSPE is equal to 0,01
and the MPE is equal to –0,01) that the model tracks historical developments in water
demand patterns fairly well. Overall, the results for both stages can be considered
promising, a fact that permits us to continue with the post sample prediction through
policy scenarios.
56 K. BITHAS, C. STOFOROS, South-Eastern Europe Journal of Economics 1 (2006) 47-59
Source: Estimated
58 K. BITHAS, C. STOFOROS, South-Eastern Europe Journal of Economics 1 (2006) 47-59
5. Conclusions
Growing water scarcity requires sustainable water use - especially in urban areas
where domestic use requires high quality characteristics. To design effective water
policy, urban water demand should be sufficiently analysed and broken down. The
case of the Athens metropolitan area presents certain interesting aspects in this con-
text. In the Athens area, the population density is relatively high, the climatologi-
cal conditions lead to intensified water use, while water resources are substantially
scarce and the various economic growth scenarios alter water use patterns and habits.
Indeed, The Athenian water system will face severe problems in the future, if current
trends continue. In this context, to design a sustainable water policy is vital for the
future of the city.
The present paper identifies the factors that determine water demand in Athens.
Future demand is estimated under certain plausible assumptions concerning its cru-
cial determinants. We conclude that water use will increase regardless of the evolu-
tion of water prices in the future, since income elasticity is more sensitive than price
elasticity.
If correct, these conclusions could lead to policy-relevant recommendations. The
most cogent is that water saving plans should be based on quantitative restrictions
and on voluntary actions prompted by information campaigns focusing on increas-
ing the environmental awareness of inhabitants. On the other hand, the paper creates
prospects for further demand analysis and forecasting so that policy makers will be
able to define focus groups and other factors that could shape an effective water de-
mand policy.
K. BITHAS, C. STOFOROS, South-Eastern Europe Journal of Economics 1 (2006) 47-59 59
References
Arbues, F., Garcia-Valinas M.A. and Martinez-Espineira, R. (2003). Estimation of residential water
demand: a state-of-the-art review. The Journal of Socio-Economics, 32, 81-102.
Baumann, D.D., Boland, J.J. and Hanemann, W.M., (1998). Urban Water Demand Management
and Planning. New York: McGraw-Hill.
Briscoe, J. (1997). Managing water as an economic good in M. Kay, T. Franks and L. Smith (eds.),
Water: Economics of Management and Demand. E and FN Spon, London, pp: 339-361.
Dalhuisen, J.M., Florax, R.J.G.M., de Groot, H.L.F. and Nijkamp, P. (2003). Price and income elas-
ticities of residential water demand: a meta-analysis. Land Economics 79(2), 292-308.
Chambers, M.J. (1990). Forecasting with Demand Systems: A Comparative Study, Journal of
Econometrics, 44, 363-376.
Clements, M.P. and D.F. Hendry (2003). Economic Forecasting: Some Lessons from Recent Re-
search, Economic Modeling. 20, 301-329.
Espey, M., Espey, J. and Shaw, W.D. (1997). Price elasticity of residential demand for water: a
meta-analysis. Water Resource Research 33(6), 1369-1374.
Gardiner, V. and Herrington, P. (1986). The Basis and Practices of Water Demand Forecasting.
Norwich: GeoBooks.
Granger, C.W. (1980). Forecasting Methods New York: Academic Press.
Griffin, R. and Sickles, R. (2001). Demand Specification for Municipal Water Management: Evalu-
ation of the Stone-Geary Form. Land Economics, 77(3), 399-422.
Hansen, L. (1996). Water and Energy Price Impacts on Residential Water Demand in Copenaghen.
Land Economics 72(1), 66–79.
Kaïka, M (2003). The Water Framework Directive: a new directive for a changing social, political
and economic European framework. European Planning Studies 11 (3), 303-320.
Luaky-Froukh, M. (2001). Decision Support System for Domestic Water Demand Forecasting and
Management. Water Resources Management, 15, 363-382.
Martinez-Espineira R. Griffin R. and Sickles R. (2001), ‘Demand Specification for Municipal Wa-
ter Management: Evaluation of the Stone-Geary Form’, Land Economics, 77 (3), 399-422.
Martinez-Espineira R. (2002), ‘Residential Water Demand in the Northwest of Spain’, Environmen-
tal and Resource Economics, 21, 161-187.
Martinez-Espineira R. and Nauges C. (2004), ‘Is All Domestic Water Consumption Sensitive to
Price Control?’, Applied Economics, 36, 1697-1703.
METRON 2000 (2000). Metropolitan Areas and Sustainable Use of Water: The Case of Athens.
Metron Project Report, European Commission, Laboratory of Environmental Planning, My-
tilini: University of the Aegean. ISBN - 960- 86789- 3-5”.
Nauges, C. and Thomas, A. (2000). Privately Operated Water Utilities Municipal Price Nego-
tiation, and Estimation of Residential Water Demand: The Case of France. Land Economics
76(1), 68–85.
Nieswiadomy, M. L. and Molina, D. J. (1989). Comparing Residential Water Estimates Under
Decreasing and Increasing Block Rates Using Household Data. Land Economics 65(3), 280–
289.
OECD, (1998). Pricing of Water Services in OECD Countries: Synthesis Report. Paris: OECD.
Pearce, D. (1999). Pricing Water: Conceptual and Theoretical Issues, Paper for European Com-
mission for the Conference on Pricing Water: Economics, Environment and Society. Portugal:
Sintra.
Renwick, M. and Archibald, S. (1998). Demand Side Management Policies for Residential Water
Use: Who Bears the Conservation Burden? Land Economics 74(3), 343–359.
Tietenberg T. (1996). Environmental and Natural Resources. Harper Collins. N Y.