How much do people in Scotland value characteristics of marine and coastal areas

Annex 1: Statement from CSERGE at the University of East Anglia

The Centre for Social and Economic Research on the Global Environment (CSERGE) at the University of East Anglia provided support and capacity building in designing and administering the Choice Experiment (CE) study in this project which is aimed at eliciting the value of the Scottish marine environment. Expert advice and support focussed on three key stages of the research:

• Study design and research timeline management: this process aimed at ensuring robustness of the methodological approach and definition of the key research milestones. CSERGE contributed to define the different steps of the study from consultation with experts to selection of attributes and levels, planning of the pilot design and preparation of the final questionnaire. For these steps, CSERGE supported the discussion with the steering group, the analysis of the pilot survey data, and the preparation and revision of the questionnaire structure to make the CE more relevant to the Scottish context. This stage was crucial to bridge scientific and socio-economic considerations and develop a CE with clear links to the national policy environment, in particular through reframing the restriction and biodiversity attributes and adding an area-based attribute. CSERGE also advised on experimental design options and coding in Ngene, on the pilot survey distribution including liasing with online sample providers and on the survey coding in Qualtrics.
• Data collection, sampling and data handling: this process aimed at supporting the survey administration through Qualtrics as well as preparing the data for the econometric analysis. For the data collection, a special support was given in setting and monitoring quotas and coordinating the data collection with the survey company. In particular, advice was provided on merging the data collected with the CE experimental design, cleaning the final dataset in preparation for the econometric analysis of choice data, and defining rules to identify protesters and speeders.
• Data analysis: this process aimed at ensuring robustness of the modelling and the analytical approach. Capacity building was provided regarding the modelling of CE data using the R package mlogit for multinomial and mixed multinomial logit models. In addition, support was provided to interpret modelling outcomes including estimated coefficients, willingness-to-pay values, confidence in estimated results, and implications for potential policy scenarios, and considerations around uncertainty.

The CE has been developed following the best practices in environmental valuation and stated preferences, therefore we are confident that results can be robustly used to inform Scottish policies on marine management, including impact assessments and cost-benefit analyses. A balanced trade-off between the detailed, science-informed definition of attributes and levels (e.g., biodiversity changes) and the need to simplify complex issues to facilitate respondent’s understanding was required. This resulted in an inevitable generalisation of some aspects of the CE. Nonetheless, the flexibility attained reflects the main characteristics of possible marine management policies (e.g., restrictions). Therefore, the use of results should focus on matching CE outcomes (in particular willingness-to-pay values) with characteristics of potential policy implementation scenarios and keep in mind the resulting uncertainty.