texto en 
Inglés (pdf)
Artículo en XML
Referencias del artículo
Traducción automática
Enviar artículo por email
Citado por SciELO 
Similares en
SciELO 
Permalink
Introduction
One of the issues that receives the most attention from researchers in environmental economics is whether environmental goods and services provided by natural resources are luxury goods or not. Regarding this debate, there are also reasons related to income distribution since it is of interest to know whether low-income social groups have a higher willingness to pay (WTP) for an improvement in the quality of environmental goods than high-income groups (Kristrom & Riera, 1996). This is an important aspect that requires attention in order to make empirical quantifications on the magnitude of the income elasticity of demand for environmental services. From the point of view of those who make decisions on environmental policy, another issue of concern relates to how price changes affect the quantity of environmental services demanded. Technological innovations could involve reduced costs in environmental services supply, and knowledge of the price elasticity of demand would predict the response of consumers to such change. It could also be of interest to predict the response to introducing economic policy instruments such as taxes, fees or subsidies, in order to influence the behavior of businesses and consumers towards the environment (Haneman, 1984).
In the present study, indicators of WTP for environmental improvements in the Basaltic Prisms ecosystem of Huasca de Ocampo, state of Hidalgo, were obtained through a survey applied to visitors. This research was conducted with the aim of making a segmentation of consumers of recreational services, which serves as a starting point for calculating the WTP by income level, and estimating the income elasticity of demand to determine whether the environmental recreational services of the Basaltic Prisms ecosystem are an inferior, normal or luxury good. The research hypothesis was that the income elasticity of WTP of consumers of recreational services is less than unity but greater than zero, so that recreational services can be classified as a normal good for the typology of consumers that could be identified.
Materials and methods
Basaltic Prisms Park is located in the municipality of Huasca de Ocampo, Hidalgo State, 38 km northeast of the city of Pachuca on Federal Highway 105. The municipality of Huasca has a temperate, cool climate and a mean annual temperature of 15 °C.
The characterization of the consumers of recreational services was made with a sample of 285 observations drawn at random from a survey conducted in 2012 in Huasca de Ocampo, Hidalgo. The conceptual tools of multivariate analysis and the binary logit discrete-choice statistical model, as well as the concepts of environmental economics, were used. Consumer groups with similar characteristics were identified by means of multivariate analysis, using two-step cluster analysis, and for the empirical estimation of the theoretical microeconomic concept called compensating variation, the dichotomous logistic probability model was used. The empirical estimation of the compensating variation is the indicator of the measure of central tendency called willingness to pay.
The variables used to define the typology of consumers are described in Table 1. The discrete dependent variable was WTP, and the rest were the respective independent variables. The environmental improvement project posed to visitors to the Basaltic Prisms, for which they were asked if they were willing to pay, consisted of the conservation of the prisms and their scenic beauty, conservation of the water and the improvement of its quality, and conservation of the trees and green areas that benefit the ecosystem.
Table 1 Variables used to define the typology of visitors in the Basaltic Prism ecosystem of Huasca de Ocampo, Hidalgo.
| Variable | Description | Type | Units or Attribute |
|---|---|---|---|
| WTP | Willingness to pay | Nominal | 0= No, 1 = Yes |
| PRI | Price | Continuous | Amount of money to pay ($ per member) for access to the recreational site |
| HI | Household income | Continuous | $·month-1 |
| AG | Age | Continuous | Years |
| SCH | Schooling | Continuous | Years of schooling |
| FS | Family size | Continuous | Members per family |
| GEN | Gender | Nominal | 0= Male, 1 = Female |
| MS | Marital status | Nominal | 0= Married, 1 = Single |
Source: Made from the survey applied to visitors from Huasca de Ocampo (2012).
In the case of neoclassical environmental economics, welfare economics, which provides the theoretical basis for compensating variation as well as guidance on how to empirically calculate its observable monetary expression, that is, the WTP of the consumer of recreational environmental services, is used. The other microeconomic concept that is widely used is the income elasticity of willingness to pay, which measures the sensitivity in WTP of a consumer for recreational services when there is a change in his or her income (Azqueta, 2007; Bateman, Mace, Fezzi, Atkinson, & Turner, 2011).
In the case of the probability model, where the dependent variable assumes discrete values, coded as “1” when the consumer gives an affirmative answer (yes) to the question about his or her willingness to pay, and “0” when the answer is negative (no). The expression for calculating the income elasticity of WTP is:
where:
ηI = |
Income elasticity of WTP |
I = |
Consumer income |
θ = |
Marginal effect on the dependent variable to a change in the independent variable in the logistic probability model, and is given by |
where:
βIng = |
Parameter associated with the variable average income (I) |
e = |
Euler’s constant (Hanemann & Kanninen, 2001) |
β' = |
Transposed vector of the parameters of the regression model |
x = |
Vector of the regression model’s independent variables |
The empirical estimation of θ is provided directly by the results output when the respective logistic regression has been run in NLOGIT Version 5 software (Econometric-Software, 2012). In the case of the denominator 0.5 of the expression for calculating income elasticity, two additional comments must be made. First, in the present study, the calculations are presented for the probability point of 0.5. Elasticity varies drastically depending on the section of the curve that is being analyzed because it is in the range of zero and unity (Horby & Soderqvist, 2003).
According to microeconomic theory, the environmental recreational service is an inferior good when its income elasticity is negative, since consumption decreases as its price increases. If the elasticity is between zero and unity, the recreational service is a normal service, and if it is greater than unity, it is a superior or luxury good (Vázquez, Cerda, & Orrego, 2007).
Once the relevant variables were defined, the segmentation of consumers of recreational services was made with the Statistical Package for Social Sciences version 16 (SPSS Inc., 2007) software and its Two-Step Cluster Analysis option. Categorical variables used for the segmentation were marital status (MS) and gender (GEN), while continuous variables were the price that the consumer was willing to pay for improvements in the project (PRI), household income (HI), age (AG), schooling (SC) and family size (FS).
The two-step cluster analysis method is based on the methodology called balanced iterative reducing and clustering using hierarchies (BIRCH). According to Bacher, Wenzi and Vogler (2004), in the first stage the observations are pre-clustered through distances quantified by the log-likelihood logarithm, from which a feature tree is generated. The resulting subclusters are subsequently added in the second step, comparing their distances with a specific threshold. Thus, if the distance is greater than the threshold, the two clusters merge (Bacher et al. 2004; Chiu, Fang, Chen, Wang, & Jeris, 2001; Haab & Kenneth, 2002). Since the number of existing clusters was unknown a priori, the computational algorithm determined them automatically based on the statistical criteria referred to above.
Results and discussion
Table 2 shows the results of the auto-clustering of the data of the consumers of recreational services. The clustering criterion, in this case the Bayesian Information Criterion (BIC), is computed for each potential number of clusters. The smaller the BIC value, the better the model will be, and will therefore indicate the best solution for determining the number of clusters. That is, the optimal number of clusters is obtained when there is the lowest change in the BIC and the highest ratio of distance measures. Table 2 shows that this criterion is met when the change in the BIC takes the value of -213,959, and the ratio of distance measures has a maximum value (2.134). Therefore, the optimal number of clusters is three.
Table 2 Auto-clustering based on information given by consumers of recreational services in the ecosystem of the Basaltic Prisms of Huasca de Ocampo, Hidalgo.
| Cluster number | Schwarz bayesian information criterion (BIC) | Change in BIC | Change ratio in BIC | Ratio of distance measures |
|---|---|---|---|---|
| 1 | 1,830.758 | - | - | - |
| 2 | 1,564.861 | -265.897 | 1.000 | 1.184 |
| 3 | 1,350.902 | -213.959 | 0.805 | 2.134 |
| 4 | 1,286.664 | -64.238 | 0.242 | 1.591 |
| 5 | 1,271.489 | -15.175 | 0.057 | 1.249 |
Source: Made from the analysis generated by SPSS 15 and its Two-Step Cluster Analysis option.
Table 3 shows the inter- and intra-cluster distribution of willingness to pay and of two categorical socioeconomic variables of consumers of recreational services.
Table 3 Percentage frequencies by type of consumer of recreational services of the Basaltic Prisms ecosystem of Huasca de Ocampo, Hidalgo.
| Types | Willingness to pay | Gender | Marital status | |||
|---|---|---|---|---|---|---|
| Male | Female | Married | Single | |||
| Yes | No | |||||
| I | 40.5 | 31.1 | 35.8 | 39.6 | 93.9 | 0.0 |
| II | 26.2 | 33.3 | 0.0 | 60.4 | 6.1 | 43.3 |
| II | 33.3 | 35.6 | 64.2 | 0.0 | 0.0 | 56.7 |
| Total | 100.0 | 100.0 | 100.0 | 100.0 | 100.0 | 100.0 |
Source: made from the analysis generated by SPSS 15 and its Two-Step Cluster Analysis option.
By way of example, in the case of WTP, 195 of the 285 respondents were willing to pay for improvements in recreational services that the ecosystem provide, and 90 expressed no willingness to pay. Of the 195 consumers who expressed willingness to pay, 40.5 % are in consumer type I, 26.2 % in consumer type II, and 33.3 % in type III.
Table 4 shows a summary of the microeconomic indicators of the consumers of recreational services for the general model and for the three types of consumers that allows jointly analyzing the socioeconomic characteristics, the indicators of the willingness to pay and the income elasticity estimated from the probability model.
Table 4 Microeconomic indicators of the consumers of recreational services of the Basaltic Prisms ecosystemof Huasca de Ocampo, Hidalgo.
| Model / Type | Willingness to pay ($) | Household income($) | Age (years) | Schooling (years) | Family size (members) | Income elasticity |
|---|---|---|---|---|---|---|
| I | 43.20 | 9,295.00 | 29.3 | 4.1 | 4.3 | 0.56 |
| II | 36.90 | 8,751.00 | 41.1 | 3.5 | 3.7 | 0.18 |
| III | 47.80 | 8,689.00 | 40.6 | 4.1 | 4.8 | 0.17 |
| Complete | 42.00 | 8,934.00 | 36.5 | 3.9 | 4.2 | 0.27 |
Source: made from the analysis generated by SPSS 15 and its Two-Step Cluster Analysis option.
Considering the macroeconomic indicators of the WTP for improvements in the Basaltic Prisms ecosystem that increase the welfare of the consumers of recreational services and income elasticity as the common thread of the analysis, the following types of consumers were defined:
Consumer type I. The highest income elasticity was observed in consumer type I (0.56), which is characterized by having the highest average household income ($ 9,295.00 per month), 4.1 years of schooling and a family size of 4.3 members, and being the youngest (29.3 years old on average). The WTP to access increased satisfaction derived from potential improvements in the Basaltic Prisms ecosystem is the second highest ($ 43.20 per person). The income elasticity value of consumer type I indicates that if there is a 10 % increase in his or her income, there would be a stimulus for the WTP to increase by 5.6 %. This indicator means that for consumer type I, the recreational services provided by environmental goods are a normal good.
Consumer type II. The income elasticity for consumer type II is 0.18. This consumer is characterized by having an average income of $ 8,751.00 per month, less schooling (3.5 years) and a smaller family size (3.7 members), and being older (41.1 years old). This type of consumer has the lowest WTP ($ 36.90 per family member) to access improvements in the Basaltic Prisms ecosystem, which would improve the quality of the environmental recreational services. The elasticity of consumer type II indicates that if there is a 10 % increase in his or her income, there would be a stimulus for his or her willingness to pay for an improvement project in the Basaltic Prisms ecosystem to increase by 1.8 %. This indicator means that consumer type II considers the recreational services provided by environmental goods as a normal good.
Consumer type III. The lowest elasticity (0.17) is observed in consumer type III, which is characterized by having the lowest monthly income ($ 8,689.00), an average age of 40.6 years (11 years older than consumer type I), 4.1 years of schooling and 4.8 members per family. In contrast to the other two types of consumers, consumer type III has the highest WTP ($ 47.80) with the aim of obtaining greater satisfaction derived from potential improvements in the Basaltic Prisms ecosystem. The elasticity of consumer type III indicates that if there is a 10 % increase in income, there would be a stimulus for his or her WTP to increase by 1.7 %. This means that the recreational services provided by environmental goods are a normal good for this consumer.
The estimated results are consistent with the research of Horby and Soderqvist (2003) in Greece. These authors calculated that the estimated income elasticity was greater than zero and less than unity, so that environmental recreational services behave as a normal good and not as a luxury good. Likewise, Kristrom and Riera (1996) found the same in a sample for Spain, except that lower-income consumers of environmental recreational services showed a higher WTP than higher-income consumers.
Conclusions
The results allow concluding that it is possible to identify three types of consumers of recreational services provided by the Basaltic Prisms of Huasca de Ocampo, Hidalgo. For the three types of consumers, the recreational services provided by the ecosystem are a normal good. It was found that the consumer willingness to pay for recreational services provided by an ecosystem is greater in those consumers with the lowest household income than in those in higher-income brackets.
