Are Contingent Rules Ever Used in Close Relationships? 457T 1 Fairness .61 T 2 Fairness***(.51***)T 1 Conflict T 2 Conflict.74(.80).63(.64).13(.02).26*(.20)
.06(.38)
.07(.14)
.77*** (.80***)Time 1 - Time 2Figure 18.2 Links between relationship conflict and perceptions of fair-
ness in relationships across time.We are suggesting that people in general start off their
close relationships believing in a communal norm and doing
their best to follow it. Such a norm requires mutual respon-
siveness to needs. However, it is inevitable that needs will be
neglected. When this happens or when it is perceived to have
happened, distress will occur, as we have argued. The dis-
tress, in turn, sets the stage for a possible switch to a contin-
gent, record-keeping norm such as equity or exchange. Thus,
it is likely that it isdistressthat—in most relationships some
of the time and in some relationships very often—leads to
record keeping. Such record keeping will, however, necessar-
ily be retrospective at first. As such, it is very likely that it will
be biased in such a manner as to result in evidence of inequity.
Perceptions of inequity will, in turn, lead to judgments of un-
fairness. Then, in an iterative fashion, these perceptions will
lead to further distress. Note that this is the reverse of what
has typically been argued in the past, which is that record
keeping and calculations of equity come first and that distress
results when inequities are detected (Walster et al., 1978).
Is there any evidence for our proposal that distress pre-
cedes perception of unfairness in close relationships (rather
than vice versa)? The answer is yes (Grote & Clark, 2001). In
a recent study we tracked both conflict and perceptions of
unfairness in a sample of about 200 married couples. These
couples were enrolled in the study at a time when the wife
was in the third trimester of her first pregnancy. Marital con-
flict (distress) was tapped at that time, again a few months
after the baby was born, and a third time when the baby was
about 1 year old. We also asked many questions about the di-
vision of household labor at all three points in time and about
how fair the husband and wife felt that division of labor to be.
(Notably, the division of labor was almost always judged to
be unfair, with the wife performing more whether she stayed
at home, worked part time or worked full time, and with both
spouses agreeing that this was unfair.)
The longitudinal panel design of this study allowed us to
conduct path analyses on the data in order to ascertain
whether conflict at Time 1 predicted perceptions of unfair-
ness at Time 2 (controlling for perceptions of unfairness at
Time 1). Our theoretical position led us to the prediction that
it would. We were also able to test whether perceptions of un-
fairness at Time 1 would predict conflict at Time 2 (control-
ling for conflict at Time 1). Traditional perspectives would
lead to the prediction that it would. However, our theoretical
perspective led us to predict that that would not necessarily
be the case. That is, we believed that the division of labor
could be inequitable and could be judged to be unfair when a
social scientist came along and asked about it but still might
not disrupt the relationship if both partners felt that their
needs were being met and did not feel stressed.
The results, which are shown in Figure 18.2, were as we
expected. Conflict at Time 1 (which we felt was indicative of
situations in which at least one person was feeling that his or
her needs were not being met) prospectively and significantly
predicted perceptions of unfairness at Time 2 controlling for
perceptions of unfairness at Time 1. Perceptions of unfairness
at Time 1, however, were not significant prospective predic-
tors of conflict at Time 2 controlling for perceptions of con-
flict at Time 1. This occurs, we assert, because in low-stress
times when partners’ needs are being met (as we suspected
was the case for most couples prior to the birth of an eagerly
anticipated first child), people are not keeping track of inputs
and outcomes day to day and are not calculating fairness.
Whereas they can report on inequities in housework when a
social scientist asks them to do so, we believe that most of
our couples were not doing this on their own. That is why our
measures of perceived unfairness did not predict conflict. In
contrast, the early measures of conflict, we suspect, did pick
up on those couples including at least one member who felt
that his or her needs were being neglected. It is among these
couples, we suspect, that record keeping (much of it ret-
rospective and biased) emerged, resulting in perceptions of
unfairness.
Once record keeping does emerge and unfairness is per-
ceived, we have predicted that those perceptions of unfair-
ness will increase unhappiness further. Evidence for this
subsequent process emerged in the Grote and Clark data as
well. Specifically, when changes in the patterns of data from
Time 2 until Time 3 were examined, it was found that per-
ceptions of unfairness at Time 2 (shortly after the baby had
been born) until Time 3 (when the baby was about 1 year old)
did significantly predict increases in conflict, controlling for
conflict at Time 2. This occurred, we believe, because once
couples were stressed and record keeping commenced, find-
ing evidence of inequities increased distress still further. One
interesting result was that conflict measured at Time 2 did
not predict further increases in perceptions of unfairness as
Time 3 (controlling for perceptions of unfairness at Time 2).