Rationalizing the Irrational: Inconsistencies Among Union Members and Non-Members

A fundamental workforce dichotomy has informed the industrial relations literature since its emergence: union members versus non-members. Increasing the number of affiliates has become the raison d'être of most unions and it has been considered the main measure to assess union strength, especially during the decline of union membership in the last decades. Because of the importance accorded to union membership, the focus of scholars has been mainly directed towards explaining why some wage-earners join unions, why other ones never join them and why some of them leave unions after a certain amount of time.

In this paper, we would like to go a step further and also examine the neglected heterogeneity existing among wage-earners union members and among wage-earners non-members. We focus on a particularly puzzling dimension of this heterogeneity. Why do some wage-earners become members even though they do not have confidence in unions? Conversely, what are the reasons leading some employees to remain non-members even though they have a positive attitude towards unions? Are there countries in which these paradoxical behaviors are more likely to occur? In other words, the goal is to analyze the propensity to show a consistent (member of a union while having a positive attitude towards unions or non-member while having a negative attitude towards unions) or inconsistent (member of a union while having a negative attitude towards unions or non-member while having a positive attitude towards unions) union membership behavior.

Using the data of the World Value Survey between 1981 and 2014, combined with a series of macro-level databases, we explore these questions by focusing on fourteen OECD countries. To identify wage-earners showing a consistent or inconsistent behavior, we combined the information on the membership status of every wage-earner with the level of confidence (low or high) he declares towards unions. This leads us to create four categories of individuals: consistent or inconsistent union members and consistent or inconsistent non-members. Using a series of multi-level regression models, we explain the propensity to show a consistent versus inconsistent membership behavior in both groups of wage-earners. Preliminary results show that, while some individual-level predictors show a significant impact, they do not contribute to explain the differences existing between countries. These can be accounted for only by including in the models a set of contextual factors related to the industrial relations system and to the economic features of each country.