Abstract:
The international pattern of strike activity has been a source of
interest to both policy makers and researchers. The worldwide decline
in strike actitvity over the last decade or two has prompted researchers
to explore, among other things, the extent to which international
changes in strike activity have been mirrored locally. While different
researchers have taken different views on the importance of
international influences on local activity, little has been done to
explore the individual linkages between countries within a VAR
(vector autoregression) framework. This article attempts this.
We address the issue of interdependence in quarterly strike
activity for eight OECD countries (Australia, USA, UK, NZ, France,
Japan, Italy and Canada) by firstly reporting simple correlation
coefficients for strike rate variables for the period 1960-98. We note
the preponderance of positive correlation coefficients between the
various countries.
Second, we test for bivariate Granger-causality, finding a
considerable degree of interdependence - though these findings may be
affected by the assumptions made about lag lengths (a not uncommon
finding associated with this methodology).
Thirdly, and perhaps most importantly, we examine the
cointegration characteristics of an 8-country (8-variable) VAR system
using the Johansen approach for identifying the number of
cointegrating equations (vectors) amongst the variables in the system.
The results of this exercise suggest the presence of a small number of
cointegrating vectors within the system. This in turn suggests that
the system has a tendency to 'wander' from some potential stead state equilibrium value. However, it is found that when global labour market
'agitation factors' (the estimated GECO unemployment rate
and nominal GOP growth rate) are included in the system, a greater
degree of relationship binding is in evidence. This finding is
consistent with a view that global labour-market agitation factors are
a major driving force in explaining 'global' patterns of strike activity.
Next, we employ a generalised impulse response analysis to
gauge the expected effect on 'global' strike activity of changes is the
labour-market agitation factors. Finally, some concluding thoughts
are proffered.
