Nothing to fear, but fear itself? Exploring the importance of mortality anxiety for interest group research

Abstract

This paper explores how the concept of ‘mortality anxiety’ fits into the literature on interest groups. Most of the relevant literature has used mortality anxiety as a proxy for organizational survival. Recent studies, however, highlight that this is not a very accurate proxy, hereby questioning the overall usefulness of the concept for the literature. This paper argues that mortality anxiety may still have an important place in the literature, but not in the way it was originally intended, namely to understand the political strategies of interest organizations. Organizations which fear for their survival, we argue, should make substantially different strategic choices than organizations absent of this fear. As a consequence, mortality anxiety is still a critical variable for the interest groups and non-profit literature. We empirically illustrate our argument based on data from a survey project in five countries (N = 2904): Belgium, Italy, Lithuania, the Netherlands, and Slovenia.

Appendix: Comparative interest group survey

The project’s objective is to develop systematic knowledge about the organizational development as well as the political strategies of civil society organizations, interest groups, lobby groups, and advocacy organizations. The aim is to achieve a better understanding of the daily operations of civil society organizations across different national settings. For this purpose, the project involves various surveys and the systematic mapping of the interest group populations in different European countries, as well as at the EU level. At this moment, large surveys have been conducted in Slovenia, Belgium, Sweden, the Netherlands, Lithuania, Italy, and the EU level, in the first phase. Poland and Spain were included in the second phase, and additional surveys are planned for the Czech Republic, Germany, and the UK.

The response rate is quite evenly distributed across countries in our analysis. More precisely, response rates were as follows, from lowest to highest: Slovenia (36%); Netherlands (38%); Lithuania (40%); and Belgium (41%). While such response rates are quite good for interest groups surveys (Marchetti 2015), an important condition is that the responses are equally distributed among key group types. For this reason, all country teams dedicated a lot of resources to generating good response rates, reflected in the overall sample. For instance, respondents were called if they did not respond to requests, in order to motivate them to fill in the survey. Hereby, certain underrepresented groups were prioritized to make sure the sample was as valid as possible. The distribution across group types is listed in Table 4. Overall, the distribution among group types reflects other mapping studies, whereby business groups are the largest set of interest groups (ranging between 33 and 51 per cent), followed by NGOs (identity and public interest groups). Moreover, we also see that the percentage of business groups in the EU is 56 per cent, which is in line with other studies highlighting that business groups dominate interest group communities more at the EU level than at the national level. This is a first indication that any biases in the selection and/or response rates of group types should not be preoccupying.

Table 4 Distribution among group types across countries

Are there systematic biases in the sample we may not have picked up? Potentially yes, as in most cases it was the more wealthy organizations that declined to participate in the surveys, such as the national chambers of commerce. Importantly, however, this also applies to the more wealthy NGOs, such as Greenpeace. It therefore seems plausible that we overestimate the overall level of fear for survival among interest groups in the countries, but not specifically for a certain type of group. Hence, we are quite confident that the mechanism we describe in this paper can be generalized to the population. Moreover, in the analyses we include country dummies to account for variation across countries.

Another issue relates to the countries in our sample. On the one hand, there is a great diversity among countries in the EU, i.e. we have two Eastern European countries and two Western European countries. This means there is a large diversity in political systems and political cultures which should be beneficial for any generalization. On the other hand, there needs to be some caution related to the interest mediation system as most countries score relatively high on corporatist scales (see Jahn 2014). This could affect the generalizability as it may be slightly different in countries with a pluralist system. This should be studied in future research.

Another indication that our results are not biased is because our data are, in key elements, in line with other research, using different datasets. For instance it is more commonly known that NGOs which face more competition for resources also rely more on outside lobbying than NGOs which face fewer such obstacles, while this does not affect business groups with varying levels of resource competition (e.g. Dür and Matteo 2016; Hanegraaff et al. 2016). While we use a different indicator, namely a direct measure of fear for survival, the mechanism is the same. This strengthens our belief that we have a valid sample. In short, the samples are certainly not perfect, but the best available. Moreover, we have no indication that there is a systematic bias in our data to invalidate the claims made in this paper compared to findings in other projects. We are therefore confident that the claims we make in this paper can be generalized to the populations of the respective countries under study.