Study Area

The Forest is located inside the mayor's office of Gustavo A. Madero in Mexico City at an average altitude of 2 240 m (^{Conabio, 2006}). It is surrounded by the following roads: to the north by 412 and 510 avenues; to the west, by José Loreto Fabela and 508 avenues; to the south-east by 608 avenue. It is 162 ha, of which 70 % are green areas and 30 % are used for various cultural and recreational activities, which includes: a lake, a children's living center, a track to skate, an outdoor theater, a projection room and a spa. Migratory birds, mainly ducks, go to the lake in their way to their final destination, Texcoco Lake, State of Mexico.

Around 3.5 million people visit BSJA each year. Every day, hundreds of athletes use the facilities to do their training routines in one of their three outdoor gyms or on a 1 200 m cyclic track.

Methods

MVC is part of the so-called declared preference methods, and consists in raising direct questions from DAP for an improvement, or Willingness to Accept (DAA), for worsening of a particular good or service; in both cases the current and final conditions are specified (^{Valdivia et al., 2011}). The MVC has great acceptance, given its ability to generate valuations that are disclosed, if there is a real market (^{Camargo et al., 2013}).

A survey was applied in which detailed what was being valued and under what market context the transaction would be carried out; this implies that the specific aspects of the good should be clearly described, as well as the means of payment (^{Vásquez, 2007}). Referendum-type questions (the most commonly used) involve letting the individual answer yes or not willing to pay a certain amount for accessing the benefits of the conservation program offered (^{Valdivia et al., 2011}).

^{Azqueta (2007)} points out that the main advantage of the MVC is that it is only with it that the non-use values can be quantified. In addition, it does not require any previous assumption, nor to estimate the demand function of the person, and is the only one that allows determining the compensatory variation.

Its disadvantages are based upon the distrust that arises from the answers obtained. The suspicion of a dishonest conduct by the interviewee leaves an essential doubt about the true meaning of his (her) response. The central problem is that there is no way to verify the validity of the results. Also, and given that he (she) works with hypothetical assumptions, in the presence of uncertainty, the utility that a person expects to receive from a particular good before knowing the state of nature that will accompany him (her), may differ greatly from what he (she) will obtain, once this unknown condition is resolved.

Method Application

A questionnaire for BSJA visitors was made, which consisted of five components: profile of the respondent, economic variables, description and precept of the site (place of origin, estimated travel time, number of visits to the BSJA, most valued attributes, current perception of the park), valuation of the park and conservation of the environment (DAP) and level of satisfaction with the visit.

The number of visitors is approximately 3.5 million per year; Therefore, the questionnaire was applied to a sample calculated by sampling for infinite populations (^{Morales, 2011}) with the following model:

Where:

n = Number of subjects that make up the sample

z = Value of Z corresponding to the confidence level

p = Proportion of people who would be willing to pay

q = Proportion of people who would not be willing to pay

e = Sample error or margin of error accepted

By using the following data and replacing in the formula, a 95 % confidence level was determined; Z = 1.96, p = 0.5, q = 0.5, e = 10 % and n = 96.

Regarding the DAP, the interviewer first asked if the visitor was willing to pay a small fee per entry to invest in the conservation of the forest's environment and infrastructure; if the answer was yes, another question was made: if the person would be willing to pay MEX$ 4, MEX$ 8, MEX$ 12 or MEX$ 16 per entry to the park, in order to improve the current attributes and / or add others. Low amounts were considered given the circumstances and socioeconomic characteristics of the visitors (previously observed).

In addition, even though the sample size resulted from 96 surveys, it was decided to apply 120 for greater accuracy. Thus, 30 referred to the amount of MEX$ 4; 30 to MEX$ 8; 30 to MEX$ 12 and another 30 to MEX$ 16. The people surveyed were the visitors of legal age because their ability to pay is more feasible. The interviews were applied three times, including weekends (due to the greater flow of visitors) and intermediate days of August.

The linear model proposed as well as the variables of interest are described below.

Econometric Model

Based on the NLogit program, the following linear model was designed:

Where:

P(SI) = Probability of saying “yes” to DAP

PP = Proposed price to be payed

INGRE = Monthly family income

SEXO = Sex

ECIVIL = Civil status

EDAD = Age

INTEGFAM = Number of family members

SITUACI = Present situation

SATISFAC = Satisfaction level

E = Error

Once the model was specified, each of the variables is described in the Table 1.

Marginal effects and elasticities

Since the values of the coefficients of the variables cannot be interpreted directly, as it would be in the case of a linear regression, the marginal effect is sought. The Table 4 shows the results for the marginal effects of each variable used in the final model.

In order to know the real effect of each variable, the respective antilogarithms of each one were obtained. To that result is subtracted 1 and multiplied by 100, to interpret the result in percentage terms (^{Gujarati and Porter, 2010}).

PP = -1.78 % indicates that for each peso the proposed entrance fee is increased, the probability of responding yes to the DAP decreases by 1.78 %. EDAD = 1.4506 %, it is interpreted that, for each year that age increases, the probability of answering “yes” to the DAP, (keeping everything else constant) would increase by 1.45 %. SEXO = following the same procedure as in the previous cases, -25.0695 % would indicate that if the person is male, the probability of responding “yes” to DAP decreases by 25.07 %. ECIVIL = 28.6355 % indicates that when the person is married increases the probability of answering “yes” to the DAP by 28.63 %. INGRE = 14.7669 % means that for each level of additional income, keeping everything else constant, people they would probably be 14.76 % more willing to pay for the entrance to the Forest. INTEGFAM = -17.2011 % means that for each additional member of their family, people decrease their probability of being willing to pay by 17.21 %. SITUACI = for this variable a result of -16.0883 % was obtained, which is interpreted as, for each level of additional degradation perceived by visitors, 16.08 % would be less willing to pay for the entry to BSJA. SATISFAC = -8.28 % means that if the level of user satisfaction increases by one unit, the probability of responding “yes” to the DAP would decrease by 8.28 %.

It can be observed that the variables that have the greatest effect (whether positive or negative) on the probability of responding “yes” to the availability to pay are: sex, civil status, family income, family members and current situation. For other authors, the level of education was important (^{Tudela and Leos, 2017}); however, this variable was not significant in this study and it was discarded.

Estimation of the Willingness to Pay (DAP)

^{Valdivia et al. (2011)} suggested the following formula for the calculation of DAP:

Where:

DAP = Willingness to Pay

( = Coefficient of PP

When clearing the DAP (or PP) of the P (SI) model:

When replacing the parameters that resulted and calculating the average, the result was MEX$ 7.32. That is, on average, visitors to the Forest of San Juan de Aragón would be willing to pay MEX$ 7.32 per ticket, under the condition that the money collected is used for environmental conservation and infrastructure maintenance. The Table 5

shows the results that were obtained when running the NLOGIT program.

If this value is multiplied by the number of visitors (3.5 million), the Economic Value of environmental services of the Forest of San Juan de Aragón amounts to MEX$ 25 620 000.00 per year, considering only the use value.

Surveys reveal that many visitors are not aware of the environmental services that BSJA provides, which could explain the low DAP. However, when asked about infrastructure and additional recreational services, walkers were more willing to cover a fee on the condition that the money be invested in proposing more activities or restoring some others such as the train, the spa and additional ones, which suggests that visitors value BSJA more for recreational activities than for environmental services.

Another possible explanation is that variables such as “the level of knowledge of environmental services” or “place of residence” were not included, which ^{González (2009)}, for example, did take into account. These variables are important since there could be a positive relationship with the DAP. ^{Valdivia et al. (2011)} and ^{González (2009)} consider the number of dependents or children instead of the number of family members, but did not take into account the level of satisfaction of the visit to the place of study or the services they receive.