An important part of the economic valuation related to goods and services derived from biodiversity, has focused on non-market goods and services, because the vast majority of them belong to the category of public goods. The reasons why these types of goods and services are not reflected in the market and complicate their valuation are two: non-excludability and non-rivalry. The first refers to the fact that when goods and services are produced there are many difficulties to be excluded from the use of the beneficiaries, generating the inability of assigning charges (prices) and, consequently, hindering or limiting their commercialization. The second fact means that the aggregation of beneficiaries (consumers) occurs with a marginal cost zero, which occurs, in some cases, even before congestion occurs in the use of the good or service considered.
The estimation of value measures for goods and services that lack of market has been the concern of many economists over the past 70 years. Since the pioneering work of Clawson (1959), Davis (1963) and Hotelling (1949), researchers have developed techniques for measuring values of non-market goods and services, whose acceptance has gradually increased over time. The demand for these value measures arises from the growing trend in the deterioration and destruction of natural assets worldwide, as well as the inefficiency in the use of resources. This generates the need to recognize these values and use them in the definition of policies that influence the behavior of producers and consumers of resources to help to stop deterioration.
Ecosystem economic valuation in Mexico is relatively recent. The applications of valuation methodologies date from the last decade of the last century and mark an increasing tendency in number and variety of studies, with a growing interest in the academic and governmental sphere. The aim of this paper was to review the economic valuations made in Mexico, specifically those developed in the field of forest biodiversity, in order to characterize the methods used, their level of utilization and the main aspects of biodiversity.
Materials and methods
This study performed searched for economic valuations carried out in Mexico related to forest biodiversity. The review was carried out in the library of the Universidad Autónoma Chapingo and the electronic search engines such as Google, Consorcio Nacional de Recursos de Información Científica y Tecnológica (CONRICYT), CrossRef y Scientific Electronic Library Online (SciELO). The research included concepts related to the economic valuation of forest biodiversity, such as: contingent valuation, travel cost, biodiversity value and forest value. The research found was characterized and analyzed.
The process of economic valuation of forest biodiversity demands several topics, including the forest-biodiversity relationship, the classification of the services derived from it, the typification of values and valuation methods, which are described below.
Biodiversity and forest resources
Biodiversity is defined as the variability between living organisms from all sources, including terrestrial, marine and the ecological complexes organisms of which they form part (United Nations Environment and Development Program [UNEP], 1992). The concept of ecosystem services refers to the benefits that people derive from ecosystems (Millenium Ecosystem Assessment, 2005). Humans rely on natural systems to produce a wide variety of goods and services, ranging from the direct use of certain species as a source of food or medicine, to functions provided by ecosystems such as water purification, nutrient retention or regulation of the climate (Polasky, Costello, & Solow, 2005). With respect to services derived from biodiversity, four levels are distinguished: genes, species, ecosystem and functional services (Table 1). The level of genes is considered to be the most basic and is expressed physically on nucleotides, chromosomes and individuals. Important crops could not maintain their commercial status without the genetic support of their ancestral, primitive and wild relatives, which are essential in crop productivity, as well as in the change and improvement of certain properties (flavors, resistance to pests and diseases, and adaptation to environmental conditions), through research on biotechnology and genetic engineering. De Groot, Wilson, and Boumans (2002) offer a broad disaggregation of ecosystem services.
|Gen||Genetic resources and raw material|
|Species||Pollination, biological control, pharmaceutical services, raw material and food production|
|Ecosystem||Gas regulation, climate regulation, disturbance regulation, water regulation, water supply and quality, sediment retention, soil formation, nutrient-soil fertility recycling, waste treatment, species shelter, raw material-food production, recreation, culture, scenic beauty and biodiversity production|
|Funtional||Due to the added character of the level, the services are practically the same as the ecosystem level|
Classification of values and forest values
Economic values are based on the preferences of individuals (Brown, 1984), so they are oriented towards an instrumentalist view (Pearce, Moran, & Biller, 2002). Economic values depend on: (1) People’s perception of the object and all other related and relevant objects; (2) the values held by people and their associated preferences; and (3) the context of the valuation. Although the terminology used has not been fully agreed upon, several environmental economists have advanced in a typology of values related to ecosystem services.
In general, the values of ecosystem services are classified into use and non-use values. The former include all direct and indirect forms in which an agent expects to make physical use of a natural resource (Blomquist & Whitehead, 1995). In turn, these values are divided into values of direct and indirect use. The former are goods or services that can be consumed directly. Indirect uses are, essentially, the ecological functions provided by resources (Munasinghe & Lutz, 1993). Non-use values do not mean unused, it refers to the fact that people assign values, even though they will never make use of the valued resource (McCollum, Peterson, & Sorg-Swanson, 1992). The concept of non-use value includes the option values and existence values. The option value is the value represented for an individual when knowing that some natural resource will available in the future; that is, it maintains the option of being able to benefit. The concept of value of existence includes vicarious, inherent and inheritance values. The vicarious r value refers to the value of knowing that other individuals may use a particular resource. The inherent value derives from knowing that the resource exists, and the inherent value of knowing that the resources will be available for future generations.
Some economists frequently use the term “value of existence” or “off-site” to refer to non-use benefits (Bishop & Heberlein, 1990; Mitchell & Carson, 1989). Some examples of these values in the field of forest resources are (Secretariat of the Convention on Biological Diversity [CBD], 2001):
a) Direct use values. Wood, carbon, firewood, some non-timber products, genetic information in agriculture and pharmaceuticals. Recreation/tourism, research/ education and culture/religion.
b) Indirect use values. Basin functions: soil conservation, water supply, quality of water, protection against floods and storms, protection for fishing activity. Global climate: carbon storage, carbon sequestration, biodiversity and local amenities.
c) Values of option and existence. General conservation of biodiversity and recreational ecosystems.
Economic valuation methods of biodiversity
|Approaches||Economic valuation methods|
|Market information||Market prices Productivity Costs|
|Stated preference||Contingent valuation Selection experiments|
|Revealed preference||Travel cost methods|
|Transfer of benefits||Use of estimates (WTP) from other studies|
Because almost all valuation studies conducted in Mexico are restricted to travel cost and contingent valuation methods, the review of this section focuses on these two methods.
Travel cost method (TCM)
The travel cost method is a revealed preference technique which considers the fact that people from different origins travel to a particular site and therefore can be expected to visit the site at different rates (Mendelsohn & Brown, 1983). This technique uses the variations that occur in the travel cost to estimate the demand for the site of interest. Once the demand function is known, it is possible to estimate the consumer surplus of the visitors and, consequently, the value of the site. The method can generally be applied per zone or individual. The first applications have used the zonal method, including the following steps (Freeman III, Herriges, & Kling, 1993):
a) For a given recreation site, the surrounding area is divided into circular areas for the purpose of measuring the round travel cost from each zone to the site.
b) Visitors to the site are sampled to determine their zones of origin.
c) Visiting rates for each zone of origin per period are defined.
d) The measure of round travel cost from the zone of origin to the area of recreation is obtained.
e) A regression is made considering visiting rates and a set of socioeconomic variables such as average income, level of education, family size, etc. The regression proves the hypothesis that visiting rates depend, in part, on travel costs.
The recreational demand function is estimated under several assumptions: 1) individuals can be grouped in residential areas or municipalities; 2) individuals within each zone have similar preferences; 3) people react to increased travel costs in the same way as they would react to the increase in site entry fees (Dixon, Scura, Carpenter, & Sherman, 1996). The demand function obtained expresses the number of visits (rate) that people from different origins are willing to make at different prices (travel costs). The complete demand curve is plotted by determining the distance (and consequently the price) at which zero tourists occur, as well as the number of visits occurring at zero price (Figure 1), in the case of a linear demand function. In this way, consumer surplus, defined as the area under the demand curve and above the price, represents the benefit that each visitor derives from the recreational experience, which is also interpreted as the value of what is lost If the site were destroyed. The annual value of the benefits generated by the site is the sum of the total estimated consumer surplus for each origin of visitors.
Some of the most frequent problems in the application of this method are the bias of time, substitution and trips with multiple destinations. Time bias occurs when it is not considered as an element of the behavior of recreation practitioners (Cesario & Knetsch, 1970). The extent of bias depends on the significance of the variable omitted, in this case travel time, and the correlation between the variable omitted and the variable retained. If the monetary cost of time and travel are positively related; that is, high travel costs also have high travel times, the slope of the demand curve would be expected to be undetermined (i.e. the true slope of the curve is more negative slope) (Dwyer, Kelly, & Boews, 1977 ). One of the most accepted solutions to this type of bias is the conversion of the observed time cost to a monetary value using an appropriate shadow price (Cesario, 1976).
Substitution bias occurs when variables are omitted in the estimation of the demand curve of a site, which reflect the presence of substitute sites. The result of this omission is the underestimation of the true demand curve. Such underestimation is expressed when the cost of use of a site is increased to visitors and do not react as those located at a greater distance with the same level of monetary costs. Their participation will be higher because they have fewer substitutes available compared to the individuals located at a greater distance. To reduce this bias, the multi-site approach was proposed, which includes travel costs as The bias of multiple destinations refers to the fact that on some occasions, especially when visitors come from distant places, visitors make the decision to visit a site considering the inclusion of at least one other destination or other reasons in the trip. In this way, travel costs cannot be fully accounted for the recreational experience in estimation. One of the solutions formulated is the multi-destination estimation method proposed by Mendelsohn, Hof, Peterson, and Johnson (1992). This method estimates a system of inverse demand functions for the possible trips and the profit value of a site by estimating the value of the demand system without the site.
The application of the zonal method has evolved towards the individual method in the last two decades; one of the main reasons is that the zonal method assumes that the behavior of individuals is identical within an area. This assumption is difficult to maintain because of the inhomogeneous character of the individuals in the areas considered as the origin of the recreationists, so the aggregation and averaging of each zone is often inaccurate (Zhang, Wang, Nunes, & Ma, 2015). The individual method seeks to develop the relationship between the cost and the number of visits made by an individual to the site of interest during a period, including other explanatory parameters (Tourkolias, Skiada, Mirasgedis, & Diakoulaki, 2015). The fundamental difference between the zonal and the individual travel cost method is that the latter defines the dependent variable, number of visits made per period by individual i to site j (Pak & Türker, 2006), that is:
Contingent valuation method (CVM)
The use of contingent valuation surveys was proposed by Ciriacy-Wantrup (1947) as part of the measurement of the benefits generated by soil conservation programs. Subsequently, Davis (1963) was the initiator of the empirical application of contingent valuations in his doctoral thesis. At the beginning of this century, the number of applications of the method already showed a high tendency, “contingent valuations have been applied in more than 50 countries and now there are thousands of articles on studies applying the CVM” (Carson, 2001).
The CVM, as a stated preference method, uses interview techniques to estimate the economic benefit of non-market goods. Surveys are carefully developed to simulate a marketplace where people are questioned about the values they would assign to non-market commodities. Contingent valuation studies generally include the following steps: (a) a hypothetical market is defined for the good under study, for example, an environmental good; B) respondents are asked their maximum willingness to pay (WTP) for environmental improvement (or their WTP to prevent deterioration), and also to declare their minimum willingness to accept (WTA) if such improvement is not carried out; C) average WTA and WTP; d) WTP or WTA obtained are used in a regression against socioeconomic variables such as income, education and age; and e) data are aggregated converting the stated mean values into values of the relevant population (Hanley, Spash, & Walker, 1995). During the application of the survey, the interviewee is expected to reveal the amount of money he would be willing to give up (or accept) to restore the same level of original utility, given an increase (or reduction) in the quantity of the non-market good. It should be clarified that the estimation of the WTA presents a clear downward trend in the CVM studies. The rejection of the use of the WTA is part of the disparity observed with the WTP estimations. Such disparity is explained by psychological aspects, income effect, transaction costs, value involved and reasons for benefits. “Generally, analysts’ reluctance to use CVM measures for an environmental loss means that activities with negative environmental impacts should not encourage this measure because their real value associated with the loss will be underestimated” (Brown & Gregory, 1999).
Although CMV acceptance has grown over the past three decades, the accuracy of its results is still being discussed. The main problems observed in this method are related to the design of the questionnaires and their administration. The most common biases are: starting point, payment instrument, strategic bias, hypothetical bias and question format.
Starting point bias is generated when the form or means of payment introduces, directly or indirectly, potential values of WTP that influence the amounts given by the respondent. Most analysts agree that when starting points are used in contingent valuation studies, the evidence suggesting starting point bias is compelling. Mitchell and Carson (1986) argue that this type of problem should be controllable by designing card payments or using open-ended questions when a WTP statement is requested.
Strategic bias is another potential source of imprecision in CVM estimates, which arises when respondents intentionally mislead researchers by over-stating and underreporting the true value of the non-market good. That is, if a respondent, correctly or incorrectly, believes that according to the WTP that he declares will be benefited or harmed in the provision of the good that is valued, then a motivation is generated for the respondent to manipulate the WTP stated, which will differ from the true value assigned. Freeman III (1986) and Mitchell and Carson (1981) argue that the strategic bias should not be a significant problem when CVM instruments are carefully designed. This assertion is supported by three considerations: a) the absence of significant evidence for the free rider hypothesis; b) the fact that most CVM instruments do not offer obvious opportunities or incentives to try to manipulate the outcome; and c) visual inspections of the distribution of offers do not suggest the existence of strongly biased responses, although it must be acknowledged that it is weak evidence.
The hypothetical bias refers to the bias related to the hypothetical nature of the contingent valuation method. Cummings, Brookshire, and Shulze (1986) argue that the cause-effect of WTP related to this type of bias has been poorly defined in the literature. In fact, Mitchell and Carson (1986) consider that the most serious methodological problems of CVM derive from their hypothetical character; however, they point out that there are reasonable arguments in the literature that support the idea that carefully designed hypothetical payment scenarios can approach to real payment situations with sufficient accuracy to be a useful component in the cost/ benefit analysis. An important part of the biases observed in the method, as well as the treatment of these, can be consulted in Mitchell and Carson (1986) and in Romo-Lozano (1998).
One of the most significant moments in the evolution of the method occurred in the context of the disaster caused by the oil spill in 1989 by a cargo vessel of the company Exxon Valdez. The court recommended to include the economic valuation of the damages, for compensation purposes, in which passive values (also called non-use or existence values) were considered. This led the United States Government’s National and Atmospheric Administration (NOAA) to form an expert commission to analyze and determine whether contingent valuation was valid for the measurement of non-use values. The commission included the participation of two Nobel Prizes in economics: Kenneth Arrow and Robert Solow. The results of this commission were expressed in a report that was in favor of the use of the CVM to estimate the non-use value in environmental disasters. The report included a rigorous set of measures with the purpose of reducing the biases which are traditionally observed in the method (Arrow et al., 1993).
At the end of the last decade of the last century, applications of the CVM marked a major trend in the question format for obtaining the WTP. The open question format of “How much would you be willing to pay for ...?” began to be replaced increasingly by the closed question format, “If it costs $x to get ..., would you be willing to pay that amount? This change increased the attention and need for statistical aspects.
Hanemann and Kanninen (1999), after a detailed review of the applied statistical models, point out that most of them violate some restrictions of the economic theory, since probability-response models based on the Box-Cox utility model are used, which include linear and logarithmic versions or nonlinear models of utility. Depending on the stochastic specification, the former give rise to probit and logit models of probability of response. The latter generate log-normal, log-logistic and Weibull models. Finally, the authors point out that such problems can be avoided by using a set of probability models of modified responses.
An important advantage of the CVM is that it can estimate both use and non-use values and can be applied to the valuation of environmental changes regardless of where they come from. The method has the disadvantages that it is relatively expensive and that when the scenarios are multidimensional, known as embedding effect, they can be complex for the interviewees; for example, the concept of biodiversity may be difficult for the respondents (Nijkamp, Vindigni, & Nunes, 2008).
Results and discussion
Table 3 shows the 11 studies found with applications of economic valuation in Mexico related to forest biodiversity. Additionally, it is known that there are at least four bioprospecting agreements (Barreda, 2001), these are: agreement between Diversa and the Universidad Nacional Autónoma de México (UNAM); agreement between the Organización de Médicos Indígenas Tradicionales and the Colegio de la Frontera Sur; agreement between Sandoz and the Unión de Comunidades Forestales Zapotecas y Chinantecas de la Sierra Juárez de Oaxaca; and agreement between the transnational companies American Cyanamid and American Home Products with the University of Arizona and the Botanical garden of the Biology Institute and the Faculty of Chemistry of UNAM. To know the economic conditions of such agreements would allow us to know the market values agreed for bioprospecting (gen level); unfortunately, it was not possible to know such conditions.
|Authors||Site||Biodeiversity level||Type of ecosystem||Valued service||Number of respondents||Methods||Bias covered*||Results|
|Romo-Lozano (1998)||Ocampo, Michoacán||Ecosystem||Temperate rainforest||Recreational||200 y 300||CV and TC||All biases||17.5 USD of recreational benefit per visitor; WTP = 16.5 USD|
|Larqué-Saavedra, Valdivia-Alcalá, Islas-Gutiérrez, y Romo-Lozano (2004)||Ixtapaluca, México||Ecosystem||Temperate rainforest||Services of a forest||385||CV||None||WTP = $180.00 in La Paz, Mexico / DAP = $180.00 en La Paz, México|
|Del Ángel-Pérez, Mendoza-Briseño, y Rebolledo- Martínez (2006)||Coatepec, Veracruz||Ecosystem||Vegetation cover with varied structure||Vegetation cover services||161||CV||None||Average WTP = $2,886.00|
|Sanjurjo-Rivera e Islas-Cortés (2007)||Río Colorado, Sonora||Ecosystem||Watershed||Recreational||85||CV||Open format bias was avoided||93 % of the sample has a WTP between $15.00 and $25.00|
|López-Paniagua, González-Guillén, Valdez-Lazalde, y de los Santos-Posadas (2007)||Cuenca Tapalpa, Jalisco||Ecosystem||Watershed||Hydrological services||243||CV and TC||None||The economic value of the environmental hydrological service was estimated at $7,256.50|
|Del Ángel-Pérez, Rebolledo-Martínez, Villagómez-Cortés, y Zetina-Lezama (2009)||San Andrés Tuxtla, Veracruz||Ecosystem||Watershed||Hydrological services||241||CV||None||Average monthly WTP value of $6.02|
|Martínez-Cruz et al. (2010)||Región Izta-Popo, México||Ecosystem||Watershed||Hydrological services||95||CV||None||The mean of the WTA was $7,992.97∙ha1∙year-1|
|Flores-Xolocotzi, González-Guillén, y de los Santos-Posadas (2010)||Mexico city||Ecosystem||Urban forest||Recreational||187||CV||Zero-value responses and protest responses were identified||The average annual WTP per visitor was $543.60 La DAP promedio anual por visitante fue de $543.60|
|Silva-Flores, Pérez- Verdín, y Návar-Cháidez (2010)||Pueblo Nuevo, Durango||Ecosystem||Watershed||Hydrological services||242||CV||None||The WTP per person is equal to $0.003∙L-1 ∙day-1 The WTA is equal to $0.054 L-1 ∙day-1|
|De Yta-Castillo (2013)||Pluma Hidalgo, Oaxaca||Ecosystem||Cloud forest||Cloud forest services||140||CV||None||Families provide an economic value of$6,388.00 per month|
|Jaramillo-Villanueva, Galindo-de-Jesús, Bustamante-González, y Cervantes-Vargas (2013)||Montaña de Guerrero||Ecosystem||Watershed||Hydrological services||115||CV||None||The average WTP was $132.90|
The search inquired the existence of documents that report the use of value measures resulting from the valuations found, but no documented evidence was found. In agreement with the analysis, of the wide set of methodologies available, in Mexico only the CVM and the TCM have been used.
The frequency of evaluation in services is: three studies of valuation of recreational services; five hydrological services valuation studies and three valuation studies of aggregate environmental services. The only level of biodiversity explored is ecosystem. There are no known studies developed in terms of non-market goods and services at gene and species level, although some of the studies characterized at the ecosystem level could well be included in the functional level category.
The CVM is the most used method (11 studies) while the TCM was used in a single study carried out in conjunction with the CVM. It is remarkable that, from the set of applications of the contingent valuation method, eight studies did not consider the biases that the literature reports as commonly present. The application of the TCM considered some of the frequent biases, such as time bias.
In Mexico, economic valuation applied to forest biodiversity, although this topic has progress, is limited both in the number of studies carried out and in the variety of methods. Few studies have dealt with the problems that the methods have; that is, biases, which reduce the level of reliability of the results. Likewise, the expansion of studies towards other levels of biodiversity has important challenges and opportunities. The trend of recent years is the increasing use of statistical tools in the application of the methodologies aforementioned. Such tools have proved to be a support not only for the treatment of inaccuracies of estimation, but also in the conciliation of the restrictions that the models of economic theory impose. Within this framework, the possibilities for moving towards better estimates are promising. Contingent valuation, with the advances that it currently reports, continues to be an increasingly robust method and a great help to the creativity of researchers interested in entering into economic valuations of the different levels of forest biodiversity.