Unions and the economic basis of attitudes

• Author: Michael White,Alex Bryson
• Published date: 01 July 2016
• DOI: http://doi.org/10.1111/irj.12145
• Date: 01 July 2016

Unions and the economic basis of attitudes

Alex Bryson and Michael White

ABSTRACT

Unions make differences to employee satisfaction that correspond to their effects on

individual economic advantage. Panel data reveal how changes in economic circum-

stance and changes in job satisfaction are linked to changes in union coverage. When

individuals move into a union covered job they receive a wage mark-up and express

enhanced pay satisfaction. Conversely, those moving from a union covered job on av-

erage lose any mark-up and have significantly reduced satisfaction. Similar findings

emerge for working hours. On average individuals prefer shorter hours, something

they tend (not) to achieve on moving (out of) into a unionised job, resulting in higher

(lower) satisfaction. Switching into union coverage lowers satisfaction with job secu-

rity, even though coverage has no effect on the risk of unemployment. This is because

covered employees suffer greater costs of re-employment for a given level of unem-

ployment risk, partly because of loss of the union mark-up.

1 INTRODUCTION

In their seminal contributions to research on the attitudes of unionised employees,

Borjas (1979) and Freeman and Medoff (1984) (FM) pointed out that union mem-

bers often expressed dissatisfaction, even though they could be expected to have

better terms and conditions than non-members. FM labelled this ‘the anomaly

of dissatisfied union workers’. At the same time, F M suggested the nee d ‘to com-

pare the effect of unionism on specific aspects of the conditions of work with the

effect of unions on expressed satisfaction with those conditions’(Freeman and

Medoff, 1984: 140).

In recent years research on unionism and attitudes has continued to grow. How-

ever, the suggested comparison between attitudes and union effects on real employ-

ment conditions has somewhat dropped out of view. Instead, much of the recent

work considers the possibility that union members have unmeasured characteristics

that bias their replies in a negative direction, and tests whether dissatisfaction remains

after applying statistical techniques to remove this suspected bias. Of course, much re-

search also continues into the union wage effect (‘union mark-up’), and other union

effects on employment conditions, but with rare exceptions this work has been discon-

nected from the issue of unionism and attitudes.

In this article we seek to re-establish the linkage between unionism and satisfaction

(‘union attitudes’) and unionism and employment conditions (‘union economic out-

comes’). The key initial step is to conceptualise the main economic outcomes for em-

ployees that unionism might or does bring, and then to select union attitudes that

correspond to these outcomes. Having set up this framework, we proceed to an

❒Correspondence should be addressed to Alex Bryson, University College London, Department of Social

Science, 20 Bedford Way, London WC1H 0AL, UK; email: a.bryson@ucl.ac.uk

Industrial Relations Journal 47:4, 360–378

ISSN 0019-8692

© 2016 John Wiley & Sons Ltd

analysis in which (like other recent contributions) we purge the results of bias from

unobserved, persistent individual characteristics that may be associated with union

coverage. Further, in developing the analysis we differentiate between various condi-

tions under which employees can enter, leave, or continue in a job and/or in a

unionised setting. This permits us to test whether employees’attitudes vary not only

according to the economic (dis)advantages that unionisation brings, but also accord-

ing to the specific circumstances in which that advantage or disadvantage shows itself.

The results reported here make several new contributions. Contrary to much of the

literature, we show that union attitudes are broadly positive both with regard to pay

and with regard to hours of work. On the other hand, they are broadly negative with

regard to job security. We also show that these effects vary according to whether in-

dividuals are continuing in an existing job or moving externally to a new job, with im-

plications for continuity or change in union coverage. Movement into coverage tends

to produce positive effects, while negative attitudes are associated with moves out of

coverage. These attitudinal results correspond to a high degree with union economic

outcomes. We find, as in most of the existing literature, a positive wage mark-up, and

shorter hours can still be gained by moving to a unionised job. We also break new

ground in estimating the effect of unionisation on the individual probability of unem-

ployment and on the individual cost of re-employment through wage shrinkage.

The research covers the period 1990–2007, a period that witnessed profound

changes including in the employment scene (Bresnahan, 1999; Castells, 2000; Dreher

et al., 2008). In the private sector, coverage and membership declined though they

held up in the public sector (Cully et al., 1999; Kersley et al., 2006). We do not address

the period since 2008, in part because of a lack of suitable data and in part because

there is still a long way to go in understanding what are the key labour market

changes of the ‘austerity’era. However, we believe our findings will be of value as a

baseline against which union effects in the subsequent period can eventually be

assessed.

The structure of the article is as follows. Section 2 conceptualises union economic

effects and how these map onto what employees seek. Section 3 briefly reviews recent

evidence on union attitudes and on union economic effects. Section 4 describes the

data, variables, and analysis methods used in the research, and Section 5 presents

the results. Section 6 offers a summary of findings and conclusions.

2 UNIONS’ECONOMIC EFFECTS

We focus on ‘economic’conditions because these are important to (most) employees,

they are (often) the subject of formal negotiations, and they are fairly well-defined and

measurable. On these criteria we select as economic effects (i) pay, (ii) hours of work,

and (iii) job security. Of course unions have other effects that are important but not

‘economic’. For instance, a negative union effect on employee satisfaction with super-

vision has often been found (Freeman and Medoff, 1984: 141–2); however, it is diffi-

cult to obtain independent measures of supervisory behaviour and effectiveness to set

alongside employee attitudes toward supervision.

Our assumption that pay bargaining is a central economic role of unions is unlikely

to be contested. Even with the declining strength of unions in the private sector, by

2004 pay was still at the top of the bargaining agenda, with three-fifths of union work-

places (61 per cent) negotiating (Kersley et al., 2006: 194). The same source reports

that hours of work came second—53 per cent of union workplaces were bargaining

361Unions and the economic basis of attitudes

© 2016 John Wiley & Sons Ltd