Abstract
This chapter analyses democratic acceptance of spatial planning policy measures by applying a two-level model, and using a Bayesian multilevel modelling approach. This involves analyses of 18 popular votes on spatial planning measures between 1984 and 2008 in Switzerland, implying potential acceptance determinants at the individual as well as the contextual level. The chapter opens with an overview of the applied theoretical framework for the concept of acceptance, before the theory behind individual determinants and contextual determinants, including hypotheses is discussed (Sect. 4.1). Subsequently, the data, model and methods are presented (Sects. 4.2 and 4.3), followed by the results (Sect. 4.4). The results demonstrate that determinants on both the individual and contextual level impact voters’ acceptance of spatial planning measures. At the individual level, voters’ political affiliations are an important factor for their voting decisions, as well as whether they are homeowners or not. At the contextual level, policy measures which contain incentive-based instruments have a higher probability of being accepted than ones that are based on bans and rules. Moreover, the degree of organisational capacity and conflict capability of interests concerned seems to influence voters’ decisions. The chapter closes with a discussion on the findings and resulting conclusions (Sect. 4.5).
This is a preview of subscription content, log in via an institution.
Buying options
Tax calculation will be finalised at checkout
Purchases are for personal use only
Learn about institutional subscriptionsNotes
- 1.
This chapter appeared in a modified form in a journal article, which has been published in the journal Land Use Policy (see Pleger 2017).
- 2.
In addition to proximity, Soss and Schram (2007, p. 121) propose visibility as a second dimension that captures “the degree to which a policy is salient to mass publics”. This dimension will not be discussed in further depth here as it is firstly relevant for studies on policy feedback and secondly due to the Swiss direct-democratic system both policy fields—environmental and spatial planning policy—are assumed to have a relatively high visibility.
- 3.
Due to underlying assumptions about the direction effects of the controlled variables, they are not held constant throughout the analyses, but it is controlled for them by including these variables in the analyses as potential independent variables.
- 4.
Archer and Tritter 2000, p. 1 maintain that “[r]ational choice could plausibly lay claim to being the grand theory of high modernity.” The core assumption of rational choice theories can be summarised as “acknowledging agents’ meaningful values and goals (aka ‘utilities’ and ‘preferences’), which they seek to maximize in the outside world, whose constitution attaches various ‘costs’ to their realization” (Archer and Tritter 2000, p. 5; see also Brown 2005).
- 5.
See Sect. 5 for a more detailed explanation of the explanatory notes for ballot proposals in Switzerland.
- 6.
There is a large body of literature stressing the advantages and applications of multilevel modeling. For a profound discussion of multilevel modeling see Snijders (2011), Luke (2004), Hox (1998), Greenland (2000), Steenbergen and Jones (2002). For a more detailed explanation of multilevel modeling and its application for political science see Bühlmann and Freitag 2006 and Bühlmann 2006.
- 7.
Markov Chain Monte Carlo (MCMC) simulating is also referred to as “Gibbs sampling” (Hosmer Jr and Lemeshow 2004, p. 321; see also Seltzer et al. 1996; Congdon 2005, pp. 2–6).
- 8.
See also Browne and Draper (2006); Congdon (2005), Schoot et al. (2014), Stegmueller (2013) and van de Schoot and Depaoli (2014) for more details on comparing Bayesian and likelihood-based methods and the advantages of Bayesian statistics.
- 9.
As Congdon (2005, p. 16) notes, “Bayesian analysis of discrete data follows the generalized linear model (GLM) structure but is not constrained to asymptotic normality to obtain posterior inferences.”
- 10.
Note that the results for model 1 only consist of a dataset for 16 popular votes caused by a lack of data for the two variables trust in government and location type for two votes. Due to their non-significance, the two variables were removed from the dataset for the computation of models 2–4 to enable a dataset of 18 popular votes.
- 11.
Throughout the whole empirical analyses of this book, numbers without decimal places reflect numbers rounded to the nearest whole number.
- 12.
The Wald statistic tests the null hypothesis that \(\sigma_{u0}^{2}\) = 0.
- 13.
The data is also insofar well suited for multilevel analyses as the sample sizes do not vary greatly over the 18 popular votes. Appendix A.1.1 includes the sample sizes for each ballot and Appendix A.1.2 shows the ranked second-level residuals, which have been calculated in the null model (i.e. without any explanatory variables).
- 14.
It can be claimed that the multilevel model contains a time perspective, which leads to a development in the acceptance of spatial planning measures over time. However, here I argue that no time-related acceptance development is expected due to the representativity of the survey samples. The expected importance of spatial planning proximity assumes acceptance differences based on the degree of proximity. Since the used data set consists of representative surveys, it can be assumed that the share of people being highly affected by the measure (e.g. homeowners) and those being less affected (e.g. people who neither own a house nor plan to buy one in near future) remains constant over time. Corroborating this assumption empirically, statistical tests including a time variable did not reveal any systematic influence of time.
- 15.
Throughout this book, when referring to results from Bayesian statistics, significance means that the credible interval for a variable does not contain zero, which points to a systematic relationship (see Whitener 1990, p. 317). Significant is used interchangeably with systematic relationship in the body text and the tables’ descriptions specify that bold results denote that the respective “95%-credible interval does not contain zero”, which corresponds to a systematic relationship (see also Stadelmann-Steffen and Vatter 2012).
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
Copyright information
© 2019 Springer International Publishing AG, part of Springer Nature
About this chapter
Cite this chapter
Pleger, L.E. (2019). Determinants of Democratic Acceptance: A Two-Level Analysis. In: Democratic Acceptance of Spatial Planning Policy Measures. Springer, Cham. https://doi.org/10.1007/978-3-319-90878-6_4
Download citation
DOI: https://doi.org/10.1007/978-3-319-90878-6_4
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-90877-9
Online ISBN: 978-3-319-90878-6
eBook Packages: Earth and Environmental ScienceEarth and Environmental Science (R0)