STRIKE PROPENSITIES AT THE BRITISH WORKPLACE*
| Author | David Blanchflower,John Cubbin |
| DOI | http://doi.org/10.1111/j.1468-0084.1986.mp48001002.x |
| Published date | 01 February 1986 |
| Date | 01 February 1986 |
OXFORD BULLETIN OF ECONOMICS AND STATISTICS, 48, 1 (1986)
0305-9049 $3.00
STRIKE PROPENSI HES AT THE BRITISH
WORKPLACE *
David Blanch/lower and John Cubbin
I. INTRODUCTION
Strikes are widely believed to be an important component of the British
economic problem both directly in terms of lost output, and because of
the poor state of industrial relations of which they are widely regarded
as a symptom. This paper draws on a new data source, the Department
of Employment/Policy Studies Institute/ESRC Workplace Industrial
Relations Survey (1980) to investigate the relative importance of
various aspects of the workplace environment in increasing the proba-
bility of industrial action in general and strikes in particular.
In Section II we set out our conceptual framework. In Section III we
give a brief description of our data sources. Section IV is a detailed
discussion of our findings; these are summarized in Section V.
II. THE CONCEPTUAL FRAMEWORK
Strikes have variously been attributed to mistakes, malice, political
opportunism, weak management, militant unions, poor institutional
arrangements and so on. This gives rise to an indigestible mass of
variables that are thought to affect strike propensities. In this section
we make use of a number of ideas developed in game theory to provide
us with a framework for classifying such variables.
Concepts borrowed from game theory have been widely employed to
analyse bargaining between workers and employers (see for example,
Coddington (1968) and more recently in Geroski and Knight (1983b)).
In such circumstances the bargaining 'game' is rather complex; it is in
fact a 'supergame', which consists of an endless (in principle) series of
constituent games, each game corresponding to a bargaining round.
Unfortunately, theoretical analysis rapidly becomes too complex to
yield useful predictions. However, insight into the nature of bargaining
games has been provided by experimental work on behaviour within
Prisoner's Dilemma settings, especially the classic experiments of
* We are grateful to the sponsors of the survey for making the data available, to Mel Slater
for advice cn the GLIM computer software, and to Neil Miliward, Paul Geroski and participants
at a seminar at the 1ER and an anonymous referee for helpful comments. All remaining errors
are the responsibility of the authors.
19
20 BULLETIN
Fouraker and Seigel (1963) on bilateral monopoly and oligopoly and
also the work of Friedman (1969), Lave (1962) and Malouf and Roth
(198 1). This experimental work has usually been set in the context of
games where no direct negotiation was involved, because bids were
transmitted through an intermediary. Although these results may there-
fore not be directly translatable to a real-world bargaining situation
they do suggest variables relevant to the present study. With incomplete
information the non-cooperative solution becomes more likely. Although
Friedman found that better information increased the frequency of
cooperative solutions, in some oligopoly experiments there was also
a greater tendency to engage in price wars. This should inject some
caution into the simple belief that more information always improves
the quality of bargains.
To obtain a bigger share of the gains from trade, each party must
understate its payoff from a settlement; in the present context, its loss
from a strike. This produces an incentive for each party to hide or
distort information, even though the chances of a perfect bargain may
be improved by better information. Indeed, strikes may occur purely
for the purpose of demonstrating willingness to take industrial action.
This contrasts sharply with the view, put forward by Siebert and
Addison (1981) that strikes are essentially accidental.
Skill in playing is generally improved by repeating the game. There
seems to be an initial learning period in experimental games before
cooperation predominates. Furthermore, the knowledge that the game
is to be repeated is important in securing cooperation, for in experi-
mental games when the last move came a large proportion of those
previously cooperating reverted to non-cooperative behaviour. (Alter-
natively, cooperation could be maintained by terminating the game at
random with a constant small probability.)
Apart from information, communication, and repetition, an impor-
tant factor seems to be the number of players. Dolbear et al. (1968),
Fouraker and Siegel (1963), and Friedman (1969) all found that in
(oligopoly) games the more persons (players) involved, the lower the
average degree of cooperation. The psychological orientation of the
parties may, however, be important. Fouraker and Seigel identify three
personality types amongst their subjects - simple maximizers, co-
operators and rivalists. As one might expect, simple maximizers and
cooperators more readily achieved outcomes than rivalists. This suggests
that the objective characteristics of the workplace may be subsidiary
to the characteristics of the main protagonists.
To summarize, in a static world, with knowledge of the payoffs,
where the game is to be repeated an indefinite number of times in the
future, the empirical literature suggests that perfect bargaining is
achievable. But the world is not static and payoffs are only imperfectly
known. Personalities, issues and the nature of the bargain can all
To continue reading
Request your trial