![Sample our Politics & International Relations journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/5947/TOC-Subject-Sample-2021-Politics-International-Relations.jpg"})
Abstract
This article starts from the remarks by Peter Mair on the growing gap between responsiveness and responsibility – or middle-run responsiveness – and the declining capacity of parties to bridge that gap. It focuses on the empirical analysis of the association between economic and substantive democratic dimensions and responsiveness, which are highly relevant to the way in which parties compete and govern within contemporary democracies. Following an introduction of the topic, the second section puts forward key concepts and hypotheses; the third presents the operationalisation of the variables and the applied method; the fourth and primary empirical section of the article analyses the non-procedural determinants of political and economic responsiveness, including freedom and equality as well as several key economic structural factors. The concluding remarks recapitulate the main empirical findings and submit a number of aspects that party leaders ought to take into account when addressing the thorny issue of responsiveness.
Notes
1. This section develops the analysis begun in Morlino (Citation2011: ch. 8).
2. In addition, the empirical analysis also shows the existence of an internal, core mechanism that we refer to as the mutual convergence of qualities, be it towards a strengthening or a weakening of all of these different qualities, which in turn better explain the meaning of these connections.
3. The issue on how to more precisely measure the two types of responsiveness will be addressed in the following sections.
4. It would be possible to analyse the indicators separately in order to look at the differences in the effects. However, for parsimony we decided to follow this strategy.
5. For the latent variable model, we employ independent normal priors for the model parameters. We estimate the model using Markov Chain Monte Carlo and Gibbs samplings and, to assess their precision, we extract from the posterior distribution the mean and the 90 per cent credible interval.
6. The indicators from which we build this dependent variables have slightly different scales (0–10 and 1–10). To make them fully comparable we rescaled them to make both range from 0 to 10.
7. In particular, we use separate cluster imputation (Van Buuren Citation2011), which consists of including the grouping factors in the imputation.
8. Three is considered a sufficient number to get reliable estimates. We ran the imputation for each of our four datasets.
9. We indicate countries with c and years with t.
10. For the multilevel models we use non-informative priors for all the parameters. Standard diagnostics were used to show that the models do not present any convergence problems.
11. As we use different data sources, we also run the models controlling for them including a categorical variable which distinguishes among the several surveys. However, the results are very similar to those presented here.
12. Predicted values are computed by holding the variables at their means.