Finding the Right Policy Design. The Case of Flood Protection Policies in Switzerland.

One of the challenges policy-makers face is to design policies that achieve the desired societal changes. Research can contribute to addressing this challenge with a better understanding about various policy designs and their calibrations. The present paper employs ten indicators deduced from the literature on policy design and establishes contrasting hypotheses that reflect the academic debate on policy design. For example, while some parts of the literature suggest that coercion increases policies’ likelihood to achieve outputs consistent with policy goals, proponents of market solutions advocate against regulation and for financial incentives. Further discussions in the literature include the design of sanctions, the distribution of competences, the inclusiveness of target groups, or budgetary arrangement, and more. In order to address design questions, we apply scholarship on policy design to the case of flood protection in Switzerland, a country with a long history of flood policies. We evaluate 169 years (1848-2017) of flood protection policies by following a mixed-data approach. In a first step, we employ coded data, i.e. we identified 92 flood protection measures in 29 legislative texts and evaluated those measures against ten policy design indicators using a systematic coding scheme. The scheme distinguishes five different calibrations of each indicator allowing us to establish categorical variables characterizing policy measures. In a second step, we compare the “objective” coded data with “subjective” survey data. In the survey, we asked 131 public and private flood policy experts about the success factors of flood policy designs. The paper shows policy design trends over time (e.g. high vs. low pressure on target groups). Seven out of ten policy design indicators exhibit a (clear) directionality highlighting their relative weight in comparison with those three indicators that lack a trend. We further curtail the number of indicators that particularly matter for designing policies by analyzing correlations between indicators. Finally, we compare coded data with expert opinions on policy design, and find, for example, that sanctions tend to be low in flood policy even though actors have a clear preference for stricter sanctions. Results improve our understanding of policy design and, simultaneously, enable future research at the intersection of policy process and policy design research. A policy design perspective contributes to a better conceptualization of policy outputs and thereby cross-fertilizes policy process research with its rich explanations for policy changes.