SciELO - Scientific Electronic Library Online

 
vol.43 número1Growth analysis of potted gerbera conducted with mineral fertilization and organic fertigationKinetic parameters of ruminai degradation of Marandu grass silage supplemented with brewer’s grain índice de autoresíndice de materiabúsqueda de artículos
Home Pagelista alfabética de revistas  

Servicios Personalizados

Revista

Articulo

Indicadores

Links relacionados

  • En proceso de indezaciónCitado por Google
  • No hay articulos similaresSimilares en SciELO
  • En proceso de indezaciónSimilares en Google

Compartir


Ciencia e investigación agraria

versión On-line ISSN 0718-1620

Cienc. Inv. Agr. vol.43 no.1 Santiago abr. 2016

http://dx.doi.org/10.4067/S0718-16202016000100011 

 

AGRICULTURAL ECONOMICS/RESEARCH PAPER

Attitudinal determinants of willingness-to-pay for river ecosystem improvements in central Chile: A choice experiment

 

Carlos A. Huenchuleo1, Jan Barkmann2, and Rainer Marggraf2

1Pontificia Universidad Católica de Valparaíso, Facultad de Ciencias Agronómicas y de los Alimentos, Escuela de Agronomía, Calle San Francisco s/n, La Palma, Quillota, Chile.

2Georg-August-Universitat Gottingen, Department of Agricultural Economics and Rural Development, Platz der Gottinger Sieben 5, D-37073 Gottingen, Germany.

Corresponding author: carlos.huenchuleo@pucv.cl


Abstract

A respondent’s preference for non-market goods depends not only on the attributes of the goods but also on the respondent’s attitude towards the goods being valued. Accounting for these characteristics may substantially improve the ability of stated choice models to represent preference heterogeneity. In this regard, we analyzed the influence of respondent attitudes on the valuation of river ecosystem quality attributes affected by pulp mill wastewater in two central Chilean watersheds. We applied the Choice Experiment (CE) method to assess preferences on river pollution risk, water quality effects, threatened species and the yield in local fisheries. The payment vehicle was an additional annual charge to the electricity bill. All three environmental attributes and the payment attribute were highly significant determinants of choice (P<0.001). A calculated mean WTP value for an optimistic policy scenario was 13 USD/year per household. As hypothesized, attitudes influenced stated respondent preferences with some differences between the two studied watersheds. Overall, our findings suggest that attitudinal variables deserve more attention in the analysis of preferences for water resource improvements in emerging economies.

Key words: Attitudes, choice modeling, economic valuation, water resource.


Resumen

Las preferencias por bienes sin mercados dependen no solo de los atributos del bien, sino también de las actitudes hacia el bien que está siendo valorado. La consideración de estas características puede mejorar sustancialmente la capacidad de los modelos de elección declarados para representar la heterogeneidad de preferencias. En este sentido, se analizó la influencia de las actitudes de los individuos sobre la valoración de atributos de la calidad ecosistemas rivereños, afectados por aguas residuales de la celulosa en Chile central. Se aplicó el método Experimentos de Elección para la evaluación de preferencias acerca del riesgo de contaminación del río, los efectos de la calidad del ecosistema rivereño sobre las especies amenazadas y la productividad de las pesquerías locales. El vehículo de pago fue un cargo anual adicional a la cuenta de electricidad. Los atributos ambientales y de pago fueron determinantes significativos de las elecciones (P<0,001). La disposición a pagar media calculada para un escenario optimista fue 13 USD/ año por hogar. Como era esperado, las actitudes influenciaron las preferencias declaradas por los entrevistados con algunas diferencias entre las dos cuencas estudiadas. Los resultados sugieren que las variables actitudinales merecen mayor atención en el análisis de preferencias por mejoras en los recursos hídricos de economías emergentes.

Palabras clave: Actitudes, modelación de elecciones, recurso hídrico, valoración económica.


 

Introduction

Although significant improvements have been made in past years, river ecosystem quality in Chile is still poor in some places due to untreated wastewater, mainly from households and industries (Banco Mundial, 2011). In particular, the negative effect of pulp mill wastewater on river ecosystem quality has received attention in Chile (Vasconi, 2006; Espinosa, 2001). Despite substantial efforts to reduce wastewater toxicity, wastewater qual-ity is still believed to impact aquatic organisms negatively (Orrego et al, 2006). Consequently, river ecosystems downstream of pulp mills are severely compromised even today (Vasconi, 2006). The protection of river ecosystem quality in Chile rests mainly on the wastewater emission standard D.S.90-2000 (Ministerio Secretaría General de la Presidencia, 2001). This standard is supervised by the Superintendency of the Sanitary Service and monitored by an industry self-controlled report-ing system. Legal norms for water and ecosystem quality themselves are still under development for most watersheds in Chile, including the area of the present study.

Chile is an emerging economy that has been following a general strategy towards social and economic development that gives substantial weight to economic considerations. From a welfare economic perspective, the design of economically focused policies needs to pay adequate attention not only to enterprise and industry concerns but also to the economic preferences of the population. "The population", however, is not a set of individu-als with necessarily uniform preferences. With respect to preferences for environmental goods, including goods referring to river ecosystems, studies conducted recently in Chile have shown substantial preference heterogeneity (Cerda et al., 2012). Against this background, a deepened insight into the socio-economic and attitudinal causes for preference heterogeneity can help po-litical analysts to devise policy options that better reflect citizen preferences. Our study responds to this assessment by analyzing the preferences of residents living downstream from two pulp mills in central Chile that affect the quality of the river ecosystem. In particular, we analyzed attitudinal determinants of preference heterogeneity (expressed as differences in willingness-to-pay; WTP) for potential river ecosystem improvements with a choice experiment (CE; a stated preference method) (Louviere et al, 2000).

There is evidence from several stated preference studies that attitudinal factors affect respondent preferences for river ecosystem quality improve -ments (Poppenborg and Koellner, 2013; Álvarez-Farizo et al., 2007; Morrison and Bennett, 2004). Moreover, the predictive capacity of discrete choice models can be enhanced when attitudinal variables are included as covariates (Greiner, 2015; Barkmann et al., 2008). Most importantly, respondents having a ‘pro-environmental’ orien-tation are more likely to pay for river ecosystem quality, while those who have a ‘pro-development’ orientation are less willing to choose options for river ecosystem quality improvements (Morrison and Bennett, 2004). The respondents who have no particular interest in the river but are aware of its poor ecological condition may still be more likely to pay for water quality improvements based on altruist motivations (Álvarez-Farizo et al, 2007). Those with business interests in the river ecosystem quality may also be more likely to take action to improve river ecosystem quality (Álvarez-Farizo et al., 2007).

CEs were applied to the valuation of different attributes of river ecosystem improvements. Attributes addressing river ecology such as the presence of certain water plants, fish, birds and other animal species were studied by Morrison and Bennett (2004), Hanley et al. (2006ab), and Álvarez-Farizo et al. (2007). Recreation attributes (boating, fishing, and swimming) were analyzed by Morrison and Bennett (2004). The effect of an invasive alga was analyzed by Beville et al. (2012). Finally, the importance of river flow rate and local jobs (Hanley et al., 2006a), aesthetics (Hanley et al., 2006b), and of water supply (Álvarez-Farizo et al, 2007) have been analyzed. In all cases, river ecosystem quality was a significant determinant of choice.

In spite of the relative abundance of stated preference studies on non-market aspects of water quality and river ecosystems, none of the studies were conducted in Chile. Consequently, the acute lack of data on the non-use and indirect use value of water quality attributes is frequently mentioned as a restriction in applied cost-benefit analyses prepared for legislative purposes in Chile (CONAMA, 2006; EULA, 2006). Against this background, the objective of this study is two-fold. First, we attempted to determine the WTP of central Chilean citizens for improvements in several policy-relevant attributes of river ecosystem quality at the Mataquito and Itata rivers. Specifi-cally, we included attributes on river pollution risk and on water quality effects impacting the river ecosystem and the river ecosystem services such as the habitat for potentially threatened species or yields in local fisheries. Second, the analysis of the influence of attitudinal factors on WTP allows for the Identification of variables affecting respondent preferences when water resources are impacted by pulp mill emissions.

Materials and methods

The study area is located at the downstream sec-tions of the Mataquito and Itata river watersheds in the VII Maule region and the VIII Bío-Bío region of central Chile, respectively. The Mataquito river basin occupies an area of 619,000 ha and includes land uses such as agriculture (80,051 ha), forestry plantations (32,074 ha), and native forests (62,365 ha). The main agricultural products are fruits, such as apples and cherries, vineyards and cereals. The Itata river basin covers a surface of 1,129,400 ha comprised mainly of agriculture (156,108 ha), forestry plantations (60,020 ha), and native forests (146,236 ha) (INE, 2009; Figure 1). The main crops are cereals, such as wheat and oats, forage, industrial crops, and vineyards. Plantations of exotic Pinus radiata (Monterrey Pine) dominate both river basins and supply the wood to the pulp mills that have harmed the local fisheries at the estuary.

The design of the choice experiment

The policy option outcomes (‘scenarios’) were described through selected attributes of river ecosystem quality and a payment to be made. Based on local information collected, three attributes on river ecosystem improvements were constructed: the river pollution risk, the number of threatened species, and the local fishery yields. Each attri-bute and its respective levels were explained to the respondents using examples with visual aids. For each attribute, one level corresponds to the current situation (status quo).


Figure 1. The location of the study area (adapted from SINIA, 2008); the actual study area hatched.

 

The pulp milis are the main industrial installations in the research area (at the Mataquito research site, the pulp mill is the only industrial installa-tion, while there are two industrial installations at the Itata study area, one of which is the pulp mill). Pollution spills leading to violations of national water quality standards have been a re-current feature of the operation of the pulp mills. Therefore, we decided to construct an attribute on river pollution risk that directly addresses such violations. In our CE design, a violation 25% of the time corresponds to the current situation or status quo. Additionally, we offered respondents reductions of river pollution risk to 20%, 15% or 10%. The threatened species attributes addresses the number of fish and waterfowl species living in the river ecosystem with a poor ("threatened") conservation status. Currently, there are on average 25 such threatened species in the Mataquito and Itata rivers (Cade-Idepe, 2004ab). The proposed attribute levels were a reduction to 20, 15, and 10 threatened species. The yield of local fisher-ies attribute refers to the yield of small fisheries operating mainly in the estuaries of the two rivers. We chose the estuarine species Robalo (Eleginops maclovinus), which is a well-known species in the study area. Recently, the current average yields in local fisheries for Robalo were 1 ton/ year (Mataquito) and 10 ton/year (Itata) (Araya et al., 2006). Given the high yield fluctuation during the last five years, we proposed to respondents an increase in fishery yields of 10%, 25%, and 50%. For the payment attribute, an additional mandatory annual payment per household to the electricity bill was utilized. We explained to the respondents that the money raised would be used to enforce pulp mill compliance with the potentially more stringent legislation on improving river ecosystem quality. Based on pre-study results, we proposed four payment levels (500, 1,000, 2,000 and 3,000 Chilean pesos (CLP) per household per year in addition to a zero status quo payment (1 USD ~ 606 CLP, February 2009).

From a full-factorial design of four attributes with four levels each, 256 (44) possible attribute level combinations could be obtained. We included the zero payment level only in the status quo card of the choice sets. Therefore, only the four other attribute levels are included in the experimental design of the non-status quo choice cards. From an orthogonal main effects design, we generated a reduced orthogonal experimental design with 16 different scenarios (=choice cards) (Hensher et al, 2005). The choice cards were organized into 16 choice sets using the shifting procedure (Chrzan and Orme, 2000). The choice sets were randomly assigned to four blocks of four choice sets each. Each choice set consisted of two river ecosystem improvement scenarios (A and B) and an opt-out scenario that represented the status quo with zero payment (scenario C). Each respondent was asked to choose one scenario from each of the four choice sets.

Attitudinal variables

For the analysis of attitudinal impacts on prefer-ences and WTP, the results from theoretically elaborated approaches are available. More elaborate approaches tend to use formal social psychology theories for analyzing attitudes, beliefs and motivations (Greiner, 2015; Pop-penborg and Koellner, 2013; Barkmann et al, 2008; Lam, 2006; Forsyth et al, 2004). Less elaborate approaches lack explicit underpinning in theory. Instead, they offer more freedom to directly investigateprima facie policy-relevant attitudinal impacts on preferences. This study follows the second, theoretically less elaborate approach because of its more applied research goals. However, there is a need in Chile for studies applying more formal social psychology theories for stated preferences and in other fields of the applied environmental sciences (with respect to soil conservation, see Huenchuleo et al., 2012).

Interaction terms of CE attributes and attitudinal variables can improve the predictive capacity of CE models and can be used to test hypotheses about socio-economic and attitudinal influ-ences on preferences (Barkmann et al., 2008). Based on previous studies (Morrison and Ben-nett, 2004; Birol et al., 2006; Álvarez-Farizo et al., 2007) and results from our pre-study, we analyzed the influence of two types of attitudinal variables on respondent WTP (Table 1). These included the respondent’s attitudes and beliefs about river resources and pollutant effects and the respondent’s attitudes towards hypothetical payment scenarios (cf. Greiner, 2015). Moreover, we incorporated items on the respondent’s relationship with the river and demographic variables.

For the survey, we considered households fulfilling two conditions: (1) they were located downstream from the pulp mills on either the Mataquito or Itata rivers, and (2) they were located on the main roads adjacent to the rivers. The total number of households at the Mataquito and Itata river site was 883 and 4,001, respectively (INE, 2005). We sampled a minimum of 150 households from each site using a systematic sampling procedure (Babbie, 2008). First, the sampling ratio was calculated (Mataquito: 0.17; Itata: 0.04) to achieve a total of 300 observations. Based in the inverse of the sampling ratio, every 6th household at the Mataquito site and every 25th household at the Itata site were chosen for an interview.

Statistical analysis

From the CE data, we estimated two-level nested logit (NL) models for the Mataquito and for the Itata samples. A suitable NL model tree structure was identified (cf. Louviere et al., 2000), and the corresponding models were estimated with NLOGIT 4.0. The vector of utility coefficients was estimated with máximum likelihood estimation procedures (Hensher et al, 2005). As model statis-tics, we report the conservative pseudo-R2 values in relationship to a ‘constants only’ model. Values between 0.24 and 0.3 correspond to R2 values of 0.5 and 0.6 in their ordinary least squares (OLS) equivalents (Hensher et al, 2005).


Table1.The socio-demographic and attitudinal variables studied.

15 pointsLikert scales: (a) completely disagree = 1, disagree = 2, neither/nor = 3, agree = 4, completely agree = 5; (b) very bad = 1, bad= 2, neither/nor = 3, good = 4, very good = 5; (c) very uncertain = 1, uncertain = 2, neither/nor = 3, certain = 4, very certain = 5.

25 points ordinal scale: 1 (no formal education); 2 (primary education); 3 (secondary education); 4 (technical- higher education); 5 (higher education).

37 points ordinal scale: 1 (0-160,000 CLP); 2 (160,000-320,000 CLP); 7 (more than 960,000 CLP). The national monthly minimum income is approximately 160,000 CLP (January, 2009).

***significant at P<0.001; **significant at P<0.01; *significant at P<0.05.

 

All interaction terms were introduced into an initial model (only containing attribute terms and the alternative specific constant) one at a time. Then, all the statistically significant interactions terms (P>0.05) of these single interaction models were introduced to the initial model without interaction terms. The final ‘improved’, parsimonious model was generated by the stepwise exclusion of non-significant interaction terms one-by-one (Barkmann et al, 2008). For the NL models, we only utilized data from respondents who selected at least once a scenario a choice other than the zero payment status quo alternative (‘payers’, Mata-quito 79%, n=125; Itata 90%, n=138). Preliminary analyses showed that models often displayed a significant coefficient for the alternative specific constant (ASC) with a negative sign giving rise to an unwarranted ‘status quo bias’ (Meyerhoff and Liebe, 2009). In the displayed ‘payers’ model, the numerical value of the ASC coefficient was small and not significant. However, in calculating the aggregated scenario WTP values, all non-payers are included and represented as stating zero WTP.

The CE results can be used to calculate welfare measures, particularly for an estimation of sample WTP (Bateman et al., 2002). Based on the quantification of the utility coefficients of the NL models, a maximum marginal WTP (mWTP) for a 1-unit attribute level change can be estimated as:

where:

P: the utility coefficient of any of the environ-mental attributes and,

Pc: the marginal utility of income given by the coefficient of the payment attribute.

Furthermore, the welfare change generated by an improvement in river ecosystem quality was calculated. The change from river quality level (Q0; status quo) to an improved level (Q1) was calculated as a compensating variation (CV) (Louviere et al., 2000). CV is the change in respondent income that is needed to adjust the status quo utility level (U0) of the respondents with respect to the utility level of the improved level (U1). In other words, the CV can be expressed as a respondent’s maximum willingness to pay to achieve the river ecosystem improvement:

 

To calculate the CV, we used an optimistic scenario with River Pollution Risk that was reduced to 15%, the number of Threatened Species was decreased to 15, and Fisheries Yield was increased by 50%. Adjusting for the respondents that never deviated from the status quo, 3,585 (90%) and 694 (79%) households at the Itata and Mataquito river sites, respectively, were included for the estimation of total aggregated WTP on the two project regions.

Results and discussion

On average, the respondents were 52 years old at both river sites. For the Itata respondents, 69% were women compared with only 52% at the Mataquito river site (P<0.01). A total of 87% of the Mataquito respondents were rural compared with 36% of the Itata respondents (P<0.001). A large percentage of the respondents did not have any formal education (Mataquito: 38%; Itata: 31%).

Of the respondent households, 62% received less than the monthly national mínimum income at both river sites (~264 USD). While 60% of the Mataquito respondents indicated that their oc-cupation (business) was related to the river, only 19% of the Itata respondents indicated the same (P<0.001). A minority of respondents (Mataquito: 21%; Itata: 29%) had conducted recreational ac-tivities related to the river in the last two years.

The estimated models were overall significant (Table 2; P(x2)<0.0001). All attributes were significant predictors of choice with high pseudo -R2 values of 0.24-0.30 corresponding to R2 values between ~0.5 and ~0.6 in their OLS model equivalents (Hensher et al, 2005). The attribute coefficients displayed the expected signs. The results indicated a disutility for higher river pol-lution risk, for higher numbers of species that are threatened, and for higher payments for river ecosystem quality improvements. Positive utility was derived for higher fishery yields. The ASC was not significant in the models. The attributes interacting with attitudinal and socio-demographic variables yielded several significant predictors of choice (see below). Judged by nested logit model results, the Threatened Species attribute had a significant relationship with most of the attitudinal and socio-demographic variables. Although this attribute represents a non-use value of river water quality, the respondents considered it important to improve the ecological status of the local rivers.

In the Mataquito model (a), the respondents that strongly believed the Mataquito river was already well protected by law were likely to be willing to pay less for a reduction in the number of threatened species (Threatened species*Protection). Our payment scenario directly proposed a morer estrictive regulation to increase supervisión of the pulp milis in the study area. Therefore, respondents agreeing with current regulations had no motivation to pay for more restrictive policies for river ecosystem quality improve-ments. Although, only 13% of the sample held those respective beliefs, the significance of the interaction term suggests that the stated prefer-ences meaningfully conform to the stated beliefs (internal validity). This result also highlights that most of people believe that the current river water quality regulation is weak. Therefore, the implementation of secondary water quality norms for protecting river ecosystem quality needs to be considered a priority for adequate watershed management.


Table 2. The valuation of river water quality improvements in central Chile.

***significant at P<0.001; **significant at P<0.01; *significant at P<0.05.

1Payment attribute in hundred Chilean pesos; n.s.: no significant coefficient.

2Null hypothesis on the positive effect of Business on WTP was rejected at 95% level of significance (two-tailed t-test).

 

On average, the Mataquito respondents from explicitly rural areas were willing to pay more for reductions in river pollution risk compared with the urban respondents (River pollution risk*Residence). In contrast, Alvarez-Farizo et al. (2007) indicated that respondents working in the agricultural sector were less likely to take action to improve river water quality. However, even from our rural Mataquito sample, only a minority (16%) of the respondents were farmers. Moreover, the respondents with a higher level of formal educational were more likely to choose higher payments on the choice set scenarios (Payment*Education).

In the Itata model (b), the respondents that be-lieved fertilizers negatively affect river quality were, on average, willing to pay less for higher fishery yields (Fisheries yield*Fertilizers). The majority of the Itata respondents believed that fertilizers negatively affect river quality (88%). This counter-intuitive result can be explained by the very low percentage of respondents involved in the local fishery business (0.02%, n=3) because they do not see any direct positive impact of improving fishery yield on their welfare. Under the emerging economy situation in rural Chile, environmental knowledge on fertilizer effects does not appear to translate into a behavioral intention (willingness-to-pay) if a respective environmental improvement does not affect the material welfare of the respondents. In contrast, the respondents without strong beliefs on fertilizer effects were willing to pay more. In applied terms, this finding shows that most of the local population is prone to supporting river quality regulations if it directly benefits them. The current development of secondary water quality norms is mainly focused on measurements of physical-chemical parameters, however, more attention should be considered to river ecosystem services recognized by people.

The Itata respondents that agreed with the adequacy of our payment vehicle (electricity bill) tended to be willing to pay more for reductions in the number of threatened species, i.e., fewer species with con-servation problems (Threatened species*Vehicle). A total of 55% of the respondents believed that the electricity bill was an adequate payment vehicle. This result highlights the importance of a credible payment scenario that properly elicits respondent preferences. Moreover, the respondents that were more certain about their choices were likely to express a higher WTP for reductions in the number of threatened species (Threatened species*Certainty). These two findings are in line with the finding from Greiner (2015) who found that a positive attitude towards a payment for environmental service (PES) scheme was a significant predictor of choice. In this study, we found that 85% of the respondents were certain about their choices. This result indicates that the respondents substantially understood our choice experiment. It also showed that scenario plausibility play a significant role in respondent preferences for river based ecosystem service at-tributes. In applied terms, this finding highlights the importance of a good understanding about water quality regulations, which facilitates the acceptance and support of a new policy.

The Itata respondents with a higher level of formal education were more likely to be willing to pay for reductions in the number of threatened species (Threatened species*Education). Similar results were shown by Álvarez-Farizo et al. (2007) who indicated that respondents with a higher educa-tion were more likely to pay for an improvement in river ecosystem quality attributes in Navarra, Spain. In this regard, lack of access to formal education may represent a significant barrier to respondents taking action to improve river ecosystem quality in the study area. This issue is substantially important considering that 31% of the Itata respondents at the study site had no formal education.

The Itata respondents who related their business (occupation) with the river were more likely to be less willing to pay for reductions in the number of threatened species (Threatened species*River business). This result was contrary to our expec -tation. The respondents may see more restrictive river ecosystem quality polices as a restriction to carry out their business. For example, local farmers may be forced to treat river water after using it or reduce the application of fertilizers and plant protection chemicals. Nevertheless, only 19% of Itata respondents actually related their business with the river. Moreover, the Itata respondents with a higher income level were more likely to choose the higher payments on the alternative choice set scenarios (Payment*Income). A previous study showed that farmers with a higher income level express positive attitudes toward ecosystem services in a South Korean watershed (Poppenborg and Koellner, 2013). A similar influence of income interacting with the payment attribute on improved river health was found by Morrison and Bennett (2004) in New South Wales, Australia, which is expected by basic economic theory.

A higher percentage of the Mataquito respondents lived in explicitly rural areas (87%) and related their business with the river (60%) compared with only 36% and 19% of Itata respondents, respectively. In fact, nested logit models displayed a significant influence of Residence (Mataquito) and Business (Itata) on respondent WTP for attributes of river ecosystem quality. Morrison and Bennett (2004) also found that respondent preferences for river ecosystem quality are site-specific. However, when residents outside the river sites were sampled, they found that value estimates were similar.

The results of the simplified policy scenario analysis are shown in Table 3. The marginal mean WTP values per household ranged between 162236 CLP/year by 1% reduction in River Pollution Risk, 199-210 CLP/year by 1 fewer Threatened Species, and 41-45 CLP/year by 1% of increased Fisheries Yield. We obtain a total mean WTP per household of 7,828 CLP/year and 8,571 CLP/year for the Mataquito and the Itata sites, respectively. Given the difference in the number of residents at the two sites, the estimated total aggregated WTP (compensating surplus) for the scenario differed substantially (Mataquito: 5.43 million CLP/year; Itata: 30.73 million CLP/year). Considering that 62% of the households earn less than the national minimum income (~264 USD per month), the valuation scenario probably captures a considerable share of total WTP for river ecosystem quality improvement. It may represent a significant amount, at least, for improving gov-ernment supervisión downstream from the pulp mills. The WTP values were also substantial if we consider that a relatively low percentage of respondents conduct recreational activities in the river ecosystem and its surroundings.


Table 3. The marginal stated preferences, WTP and exemplary policy scenario calculations.

1USD ~ 606 CLP (February, 2009).

Interestingly, the influence of attitudinal and socio-economic variables was site-specific. This finding highlights the difficulties of benefit transfer for economic valuation studies (Morrison and Bennett, 2004; Hanley et al, 2006a). We found interesting influences of respondent attitudes and beliefs on WTP for improving the regulative and monitoring capacity of government agencies with respect to future, more stringent water quality regulations and the resulting improvements in river water quality. In particular, the CE models displayed only two significant attitudinal variables, which were also site-specific: attitudes towards the ef-ficiency of water quality laws (Mataquito river site) and attitudes towards the negative effects of fertilizers on water quality (Itata river site). These results either indicate a genuine lack of influence, influences masked by correlated socio-demographic variables, or deficits on the part of the construction of our attitudinal and belief items. In this regard, Kumar and Kumar (2008) argued for interactions of economic sciences with disciplines such as psychology, sociology and other social sciences to make a more comprehensive valuation of ecosystem services. With respect to variable construction, there is certainly room for improvement as formal social psychology theo-ries could be used more stringently for variable construction. The Protection Motivation Theory (Rogers and Prentice-Dunn, 1997), the Theory of Planned Behavior (Ajzen, 1991), and the Value Belief Norm Theory of environmentalism (Stern et al, 1999) have all been shown to influence pro-environmental action although they have not yet been used systematically in the context of river ecosystem quality in emerging economies.

The success of environmental policies for protect-ing river ecosystem services not only depends on economic considerations but also on attitudes and beliefs of local communities who are the primary beneficiaries. We demonstrated with a basic ap-proach that beliefs on the efficacy of water policy and the awareness of the impact of fertilizers on river quality influence people’s support of water quality regulations. Therefore, a clear description of a policy is crucial for motivating participation. It is critical to fit an environmental protection scheme to the needs of the target biodiversity, the local situation and the people’s preferences (Greiner, 2015). The current development of secondary water quality regulations at the study area needs the consideration of these social psychology issues in adapting its application to local attitudes and beliefs.

Acknowledgements

Project Sponsors: Deutscher Akademischer Aus-tausch Dienst - DAAD Germany, and Comisión Nacional de Investigación Científica y Tecnológica - CONICYT Chile.

References

Ajzen, I. 1991. The Theory of Planned Behavior. Or-ganizational behavior and human decision processes 50:179-211.         [ Links ]

Álvarez-Farizo, B., N. Hanley, R. Barberán, and A. Lázaro. 2007. Choice modeling at the "market stall": Individual versus collective interest in environmental valuation. Ecological Economics 60:743-751.         [ Links ]

Araya, I., R. Quiñones, J. Dresdner, O. Barriga, P. Carrasco, N. Campos, M. Leiva, A. Hernandez, R. Ponce, and R. Rivas. 2006. Diagnóstico social, económico y productivo de las comunidades de pescadores artesanales de la desembocadura del Itata y zonas adyacentes, entre Cobquecura y Dichato. Programa de Estudios Económicos y Sociales del Sector Pesquero. Universidad de Concepción, Chile.         [ Links ]

Babbie, E.A. 2008. The basics of Social Research. Fourth Edition. Thomson Learning, Inc. USA. 549 pp.         [ Links ]

Banco Mundial. 2011. CHILE: Diagnóstico de la gestión de los recursos hídricos. Departamento de Medio Ambiente y Desarrollo Sostenible, Región para América Latina y el Caribe. Available online at: http://documentos.dga.cl/ADM5263.pdf (Website accessed March 01, 2011).         [ Links ]

Barkmann, J., K. Glenk, A. Keil, C. Leemhuis, N. Dietrich, G. Gerold, and R. Marggraf. 2008. Confronting unfamiliarity with ecosystem func-tions: The case for an ecosystem service ap-proach to environmental valuation with stated preference methods. Ecological Economics 5:48-62.         [ Links ]

Bateman, I., T. Carson, B. Day, M. Hanemann, N. Hanley, T. Hett, M. Jones-Lee, G. Loomes, S. Mourato, E. Ozdemiroglu, D. Pearce, R. Sug-den, and J. Swanson. 2002. Economic Valuation with Stated Preferences Techniques: A Manual. Edward Elgar Publishing, Cheltenham, U.K. and Northampton, M.A. USA.         [ Links ]

Beville, S.T., G.N. Kerr, and K.F.D. Hughey. 2012. Valuing impacts of the invasive alga Didymos-phenia geminata on recreational angling. Eco-logical Economics 82:1-10.         [ Links ]

Birol, E., K. Karousakis, and P. Koundouri. 2006. Using economic valuation techniques to inform water resources management: A survey and critical appraisal of available techniques and an application. Science of the Total Environment 365:105-122.         [ Links ]

Cade-Idepe Consultores en Ingeniería. 2004a. Diagnóstico y clasificación de los cursos y cuerpos de agua según objetivos de calidad - cuenca río Mataquito. Dirección General de Aguas, Ministerio de Obras Públicas, Chile. Available online at: http://www. sinia.cl/1292/articles-31018_Mataquito.pdf (Website accessed June 26, 2008).         [ Links ]

Cade-Idepe Consultores en Ingeniería. 2004b. Diagnóstico y clasificación de los cursos y cuerpos de agua según objetivos de calidad - cuenca río Itata. Dirección General de Aguas, Ministerio de Obras Públicas, Chile. Available online at: http://www.sinia.cl/1292/articles-31018_Itata. pdf (Website accessed June 26, 2008).         [ Links ]

Cerda, C., J. Barkmann, and R. Marggraf. 2012. Application of choice experiments to quantify the existence value of an endemic moss: a case study in Chile. Environment and Development Eco-nomics 18:207-224.         [ Links ]

Chrzan, K., and B. Orme. 2000. An overview and comparison of design strategies for choice-based conjoint analysis. Sawtooth Software 20002002, Research paper series.         [ Links ]

Espinosa, C. 2001. Evaluación de los impactos de la producción de celulosa. Análisis de Políticas Públicas N° 4, Publicaciones Terram. Santiago, Chile. 12 pp.         [ Links ]

EULA. 2006. Análisis general del impacto económico de la norma secundaria de calidad de aguas en la cuenca del río Bío-Bío, en el sector sil-voagropecuario. Universidad de Concepción. Available online at: http://www.sinia.cl/1292/ar-ticles-50925_AGIES2006.pdf (Website accessed April 01, 2010).         [ Links ]

Forsyth, D.R., M. Garcia, L.E. Zyzniewski, P.A. Story, and N.A. Kerr. 2004. Watershed pollu-tion and preservation: the awareness - appraisal model of environmentally positive intentions and behaviours. Analyses of Social Issues and Public Policy 4: 115-128.         [ Links ]

Greiner, R. 2015. Motivations and attitudes influence farmers’ willingness to participate in biodiver-sity conservation contracts. Agricultural Systems 137:154-165.

Hanley, N., S. Colombo, D. Tinch, A. Black, and A. Aftab. 2006a. Estimating the benefits of water quality improvements under the Water Framework Directive: are benefits transferable?. European Re-view of Agricultural Economics 33:391-413.

Hanley N., R.E. Wright, and B. Alvarez-Farizo. 2006b. Estimating the economic value of im-provements in river ecology using choice experiments: an application to the water framework directive. Journal of Environmental Management 78:183-193.         [ Links ]

Hensher, D.A., J.M. Rose, and W.H. Greene. 2005. Applied choice Analysis: A Primer. Cambridge University Press, Cambridge, UK. 744 pp.         [ Links ]

Huenchuleo, C., J. Barkmann, and P. Villalobos. 2012. Social psychology predictors for the adoption of soil conservation measures in Central Chile. Land Degradation and Development 23:483-495.         [ Links ]

INE - Instituto Nacional de Estadísticas. 2005. Chile: Ciudades, Pueblos, Aldeas y Caseríos. Censo 2002. Instituto Nacional de Estadísticas. Avail-able online at: http://www.ine.cl/canales/chile_ estadistico/familias/demograficas_vitales.php (Website accessed April 20, 2008).

INE - Instituto Nacional de Estadísticas. 2009. Censo Agropecuario 2007. Instituto Nacional de Estadísticas, Chile. Available online at: http://www. censoagropecuario.cl/ (Website Accessed May 5, 2010).

Kumar, M., and P. Kumar. 2008. Valuation of the ecosystem services: A psycho-cultural perspective. Ecological Economics 64:808-819.         [ Links ]

Lam, S.P. 2006. Predicting intention to save water: theory of planned behaviour, response efficacy, vulnerability, and perceived efficiency of alter-native solutions. Journal of Applied Social Psy-chology 36: 2803-2824.         [ Links ]

Louviere, J., D. Hensher, and J. Swait. 2000. Stated Choice Methods - Analysis and Application. Cambridge University Press, Cambridge, UK. 420 pp.         [ Links ]

Meyerhoff, J., and U. Liebe. 2009. Status quo effect in choice experiments: Empirical evidence on attitudes and choice task complexity. Land Eco-nomics 85:515-528.         [ Links ]

Ministerio Secretaría General de la Presidencia. 2001. Decreto 90 - Norma de emisión para la regulación de contaminantes asociados a las descargas de residuos líquidos a aguas marinas y continentales superficiales. Biblioteca del Congreso Nacional de Chile. Available online at: www.leychile.cl/N?i=182637&f=2001-03-07&p= (Website accessed June 10, 2008).

Morrison, M., and J. Bennett. 2004. Valuing New South Wales rivers for use in benefit transfer. The Australian Journal of Agricultural and Resource Economics 48:591-611.         [ Links ]

Orrego, R., A. Burgos, G. Moraga-Cid, B. Inzunza, M. Gonzalez, A. Valenzuela, R. Barra, and J. Gavilán. 2006. Effects of pulp and paper mill dis-charges on caged Rainbow Trout (Oncorhynchus mykiss): Biomarker responses along a pollution gradient in the Bio-Bio river, Chile. Environmental Toxicology and Chemistry 25(9):2280-2287.         [ Links ]

Poppenborg, P. and T. Koellner. 2013. Do attitudes toward ecosystem services determine agricultural land use practices? An analysis of farmers’ decision-making in a South Korean watershed. Land Use Policy 31:422- 429.

Rogers, R.W., and S. Prentice-Dunn. 1997. Protec-tion Motivation Theory. p. 113-132. In: Goch-man, D.S. (ed.). Handbook of Health Behavior Research I: Personal and Social Determinants, Plenum Press, New York.         [ Links ]

SINIA. 2008. Servicio de mapas. Sistema Nacional de Información Ambiental (SINIA), Ministerio del Medio Ambiente, Chile. Available online at: http://sinia.mma.gob.cl/ (Website Accessed Sep-tember 30, 2010).

Stern, P.C., T. Dietz, T. Abel, G.A. Guagnano, and L. Kalof. 1999. A value-belief-norm theory of support for social movements: The case of envi-ronmentalism. Human Ecology Review 6:81-97.         [ Links ]

Vasconi, P. 2006. CELCO el caso de la planta Valdivia. Análisis de Políticas Públicas 36, Publicaciones Fundación Terram, Chile.         [ Links ]


Received October 16, 2015.
Accepted March 10, 2016.

Creative Commons License Todo el contenido de esta revista, excepto dónde está identificado, está bajo una Licencia Creative Commons