This paper presents a model of final-offer arbitration that distinguishes
between the union rank and file and their negotiator. If the union negotiator
has better information than the rank and file with regard to the bargaining
enviroment and the negotiated wage depends not only on this enviroment but also
the effort exerted by the negotiator, then the rank and file may not be able to
tell whether a poor wage outcome resulted from a poor bargaining enviroment or
because the negotiator was shirking. This is the classic principal—agent problem
with asymmetric information.
Through contract design the union rank and file could elicit the correct
behavior from the negotiator without resort to arbitration. But, as is shown in
this paper, under certain circumstances the rank and file could do better by
having the union negotiator go to arbitration some of the time. In a two state,
model it is shown that arbitration will occur only in the ‘bad’ state (where the
bargaining enviroment is unfavorable to the union). Arbitration is more likely
to serve a useful purpose in contract design the less risk averse the rank and
file, the smaller the direct costs of arbitration to the union, the more likely
the ‘good’ state of nature and the more difficult it is to induce 'truth
telling‘ in the absence of arbitration.
arbitration
This paper is a non-technical survey of the results of recent quan-
titative analyses of interest arbitration systems operating in the U.S.
It contains a review of the broader context in which arbitration has
become a feature of public sector wage determination, and surveys of
quantitative studies of arbitrator selection and decision—making in
simulation experiments and in practice. For reasons that still remain
unclear, simple statistical analyses continue to confirm a very stable
set of operating characteristics for these systems. The data suggest
that the variability in the outcomes that exists across arbitration
systems is a product either of constraints placed on arbitrator
decisions by the institutional setup or of differences in the behavior
of the parties in response to different institutional setups, and not of
differences in arbitrator behavior.
In final offer arbitration the decision of the arbitrator provides the parties
with information about the preferences of the arbitrator that were not
available prior to the award. A union (employer) victory tells the parties
the fair wage belief of the arbitrator lies above (below) the mean of the
parties’ final offers. With inter-arbitrator reliability and temporal
stability in the characteristics of the bargaining pair, the award will alter
the parties’ expectations about the preferences of an arbitrator in the next
bargaining round and change negotiated settlements. The evidence from
Wisconsin teacher and school board negotiations supports this hypothesis. The
change in the negotiated wage increase from the round prior to an award to the
round after an award is about 2 percentage points greater when the union's
final offer is chosen than when the employer's offer is selected. In the
round following arbitration the variance in negotiated settlements also
declines and the structure of negotiated settlements converges to the
estimated structure of arbitrator beliefs.
Arbitration systems are often used to resolve labor disputes
because on-going employment relationships are likely to contain specific
(human capital) investments. Recent research indicates that the ex ante
acceptability of arbitration to the parties must depend, in part, on the
unpredictability of the arbitrator's award. It is shown that the usual
selection process for arbitrators does imply that arbitrator decisions
should be statistically exchangeable (in the limit), and the evidence
available to date supports this hypothesis.
Most research on arbitrator decision-making has used a model that assumes
there is only one disputed issue. This study shows this model is not
appropriate in multi—issue disputes under a final—offer by package law. Using
data from Wisconsin, I find arbitrators do give substantial weight to non-wage
issues in multi-issue disputes. However, the written awards of arbitrators
substantially understate the weight given to non-wage issues in the final
offer selection.
A data set composed of teachers' contracts in the Canadian province
of British Columbia is used to examine the effect of arbitration on
collective bargaining outcomes. The data set is unique in that it
spans 35 years of arbitration experience.
I do not find any evidence of an effect of arbitration on wage
rates. Strong positive state—dependence exists in the data: Bargaining
pairs that used arbitration last period are more likely to do so this
period. However, the probability of going to arbitration is only
marginally increased by arbitrations that occured more than one year in
the past. Fixed effects associated with bargaining pairs were not as
important as theories which link dispute rates to characteristics of
individual bargainers would suggest.
No relationship was found between the percentage of arbitrations and
levels of macro-economic variables. Some evidence is presented that
the moving average of the percentage of arbitrations is high when there
is upward pressure on teacher's wages. The coefficient of variation of
the Canadian CPI was positively correlated with the percentage of
arbitrations, which suggests that economic uncertainty increases the
number of disputes.
Finally, a simple model of arbitrator behavior proposed by
Ashenfelter et al., is remarkably consistent with these data.
In particular, I find that arbitrated settlements can be predicted solely
on the basis of the negotiated settlements reached in the nine month
period prior to arbitration. I also find that the variance of
arbitrated settlements is significantly lower than the variance of the
preceding negotiated settlements in 13 of 21 years. On the other hand,
the estimated coefficients in the prediction equations for arbitrated and
negotiated settlements are significantly different, with the arbitrated
settlements giving more weight to lagged wage outcomes.
This paper reports the results of a systematic experimental comparison
of the effect of alternative arbitration systems on dispute rates. The key to
our experimental design is the use of a common underlying distribution of
arbitrator "fair" awards in the different arbitration systems. This allows
us to compare dispute rates across different arbitration procedures where we
hold fixed the amount of objective underlying uncertainty about the
arbitration awards.
There are three main findings. First, dispute rates are inversely
related to the monetary costs of disputes. Dispute rates were much lower in
cases where arbitration was not available so that the entire pie was lost in
the event of a dispute. This confirms the empirical importance of the
so-called "chilling effect" on bargaining that has been conjectured is
produced by the adoption of arbitration systems. Second, the dispute rate in
a final—offer arbitration system is at least as high as the dispute rate in a
comparable conventional arbitration system. Contrary to the usual argument,
we find no evidence that final-offer arbitration eliminates the chilling
effect. Third, dispute rates are inversely related to the uncertainty costs
of disputes. Dispute rates were lower in conventional arbitration treatments
where the variance of the arbitration award was higher and imposed greater
costs on risk-averse negotiators. Our results can also be interpreted as
providing tentative evidence that the negotiators were risk—averse on
average. Finally, we find general agreement between the dispute rates in our
experiment and dispute rates found in the field in comparable settings.