Forest fire prevention in mediterranean forestsan environmental economics approach

Resumen

Forest fires raise a high concern amongst general citizens, being one of the more relevant environmental problems for society. A general consensus exists among experts of the necessity to increase fire prevention and biomass control measures at the landscape scale, overcoming the focus of resource allocation on fire suppression. Forest fire prevention in Mediterranean forests has traditionally been based on the construction and maintenance of lineal preventive structures or fuel breaks. These structures have a two-fold objective: the most obvious one is stopping low intensity surface fires, although providing access to fire fighters in their suppression activities is an objective with much greater importance. The fuel breaks may produce high impacts on the landscape and their construction and maintenance is costly for public administrations. New ways of managing these structures to reduce costs and negative impacts include the use of controlled grazing or increasing the landscape integration of these fuel breaks by shifting towards shaded structures. These decisions are usually based on technical and financial criteria. However, the high social concern of the population towards forest fires and their prevention means these management changes are likely to affect social welfare. Incorporating social preferences for these changes as another criterion would lead to a more balanced decision-making process, considering the non-market costs and benefits that the different management alternatives produce. Environmental economics has developed a number of methods that allow measurement of social preferences for environmental or alternative management changes. These may indeed have a value for society, but because they lack a market they cannot be directly incorporated into the financial analysis of these changes. Applying stated preference methods to estimate these social preferences constitutes an initial step in the cost-benefit analysis of the different policies or management options that influence the provision of environmental goods and services provided by Mediterranean forests. This PhD thesis presents an environmental economic valuation study developed in the province of Malaga, where controlled grazing in fuel breaks for fire prevention was pioneered. Later on these experiences would spread, constituting the network of grazed fuel breaks of Andalucía (RAPCA). Introducing controlled grazing as a complementary tool in fire prevention as well as the required changes in the fuel breaks needed to host livestock inspired the set up of this economic valuation study. The study applies a multiattribute method, the contingent ranking, , with the objective of estimating the social preferences for changes in three relevant attributes for the management of fuel breaks: i) cleaning tools for biomass control, ii) fuel break design, and iii) density of fuel breaks in the landscape. For that purpose, a valuation questionnaire was designed. A sample of 510 respondents, representative of the Malaga population was surveyed though face-to-face interviews in December 2009. The three scientific papers that constitute this thesis analyze the results obtained through different perspectives, but they are all rooted on the economic valuation of the goods and services not traded in markets. The first paper analyzes the results of the valuation survey, exploring the results of contingent ranking through a random parameter logit model. This model also considered a series of socioeconomic covariates that contribute to explain the willingness of the population to support a series of management change scenarios in fire prevention. The results show that the population is sensitive to the proposed changes in fuel break management. The density of the fuel breaks is, out of the three management attributes, the one that more clearly determines social preferences. Controlled grazing and light machinery follow in terms of their importance in shaping positively social preferences; prescribed burning on the other hand produces a disutility on the respondents. Finally fuel break design plays a minor role in defining social preferences. The socioeconomic covariates analyzed show that urban dwellers and recreationists are more prone towards changing scenarios. On the other hand, unemployed respondents are more willing to stick to the status quo scenario. This paper also includes consumer surplus scenarios that amongst other things show that a hypothetical increase in the payments to the shepherds that participate in the RAPCA would be justified from the social demand point of view. The second paper further explores the use of econometric models that allow analysis of the observable components of utility with the objective of exploring the heterogeneity of social preferences for fire prevention. This work adopts a prospective approach, based on the hypothesis built upon the findings in the focus groups and pilot tests conducted and justified due to the scarcity of related literature in the Mediterranean context. Three different econometric models are estimated: a model of random parameters that estimates preferences from a continuous perspective, and two additional models that do so from a discrete perspective, either at the attribute level (discrete mixture models) or at the population level (latent class model). The results allow identification of extreme preference patterns, both at the attribute and population level. Therefore, the discrete mixture model shows that social preferences for the increase in the density of fuel breaks are not homogeneous, with some segments of the population rejecting such an increase. A latent class model with four classes shows a good fit and reveals the presence of two extreme groups in terms of preference patterns. It also allows identifying hypothetical and yea-saying biases. Socioeconomic covariates are also considered in this model along with attitudinal covariates related to forest fires. These reveal that being more concerned about forest fires as an environmental problem does not necessarily imply having a better knowledge of the causes of forest fires. The third and last paper has a methodological approach. It analyses the influence that a ¿cheap talk¿ script, introduced before the contingent ranking exercise together with a short opt out reminder in the middle of the ranking exercise, may have in reducing the hypothetical bias that has been proved to exist in several stated preference studies. The results show that the presence of the cheap talk affects the preferences of the respondents, while the effect of the reminder is not significant. Based on these findings, recommendations for future studies point towards reallocating the reminder towards the end of the exercise where it can be more effective or either removing it completely.