Heterogeneous preferences for recreation-oriented management in commercial forests: the role of citizens’ socioeconomic characteristics and recreational profiles

Abstract

The management of forests for multiple benefits, such as recreational services alongside timber production, can greatly benefit from the knowledge of public preferences for management-related forest attributes. This paper investigates citizens’ recreational use and preferences for recreation-oriented management in the case of state-owned commercial forests in Finland using data from a choice experiment study. We focus on attributes related to the typical management practices applied to enhance recreation: scenic buffer zones along waters, game bird habitats, and the quality of scenery along hiking trails. Recreational use and its frequency were found to be related to citizens’ background and specific outdoor activities. The scenic buffer zones were the most important attribute to the citizens. While the recreation-enhancing practices were generally valued, respondent segments with distinct preferences were found. Preference heterogeneity was related to citizens’ socioeconomic characteristics and recreational profiles.

Keywords:

• forest recreation
• multiple-use forestry
• sample-selection model
• choice experiment method
• preference heterogeneity

Acknowledgements

This research was funded by Metsähallitus and Finnish Forest Research Institute (Metla). The authors are grateful to Tapio Pouta, Pauli Wallenius, Martti Aarnio, Johanna Leinonen, Lauri Karvonen, and Mervi Väisänen of Metsähallitus for co-operation and for helping them to choose the attributes and their levels for the CE. Irene Murtovaara at the Natural Resources Institute Finland is acknowledged for carrying out the layout of the mail questionnaire, and Keijo Kallunki at Metsähallitus for programming the web-based survey.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1. The level that ensures protection against nutrient run-off is about 5 meters, which is the lowest attribute level for buffer zones used in this study.

2. For a recent application with individual-specific status quo levels, see Ahtiainen, Pouta, and Artell (2015).

3. The sample-selection ordered probit model was estimated with LIMDEP 10.0 software (Greene 2012a).

4. Basically, the potential explanatory variables for the models used in the analysis were selected based on theory (e.g., McFadden 2001) as well as results from previous studies (e.g., Birol, Karousakis, and Koundouri 2006; Campbell et al. 2013; Tolvanen, Juutinen, and Svento 2013). However, as there are many potential variables which represent different indicators of the theoretically justified factors, we included only statistically significant variables in the final models and discuss the results regarding other tested variables. We systematically tested all outdoor activities, because there is no theory indicating which are the most relevant in determining recreation demand and the values of management attributes.

5. Interestingly, an alternative specification with the number of outdoor activities practiced as an additional explanatory variable (not reported) suggested that some of the activities are ‘main activities’ that serve as the main reason for the visits, while others are complementary or ‘bundled’ activities practiced alongside other activities. However, the result should be interpreted with caution due to possible two-way causation between the annual numbers of visits and activities.

6. We chose to use a 3-class model, based on a balanced assessment of the adjusted r², AIC and BIC statistics (Birol, Karousakis, and Koundouri, 2006). The 3-class solution provided better fit with the data than a 2-class solution. Models with more than three classes did not converge. The LCM was estimated using NLOGIT 5.0 (Greene, 2012b).

7. Given that effects coding was used, the value of the current state (omitted level) is obtained as the difference between the altered (non-status quo) levels (Bech and Gyrd-Hansen, 2005): 18.62 = –1*(8.08 – 26.70).