The Koslowsky Model of Commuting Stress

            In an effort to synthesize the various strands of commuting stress research, Koslowsky (1995), p. 119 developed a model linking stressor and strain variables. His model is a sequential one in which a series of stages provides the pathway of influence from predictor to outcome variables. The stages are as follows:

1.      Potential stressors such as time, distance and traffic congestion (objective impedance)

2.      Perceived stress (subjective impedance)

3.      Physiological reactions and symptoms of various types

4.      Emotional and attitudinal strain responses that are influenced by physiological responses

5.      Job avoidance inclinations by workers

6.      Behavioral/performance indicators

Alternatively, this six step sequence can be described as follows. The stressors of physical impedance influence subjective impedance which in turn leads to physiological stress reactions such as increased heart rate. These physiological responses then influence emotional strain responses such as mood change, fatigue, anxiety, depression, and burnout, as well as attitudinal changes involving decreased job satisfaction. These strain responses reflect general distress at work and at home and lead to job avoidance

inclinations such as missing days of work, arriving late for work, or looking for another job. Finally, performance on the job is negatively affected to include chronic medical conditions or increased accident rates as well as various behavioral changes at work and at home (Koslowsky et al., 1995).

            The model of Koslowsky (1995) is robust enough to explain the negative effects of long commutes for both work and home and makes sense of the various streams of research on commuting. However, there are certain deficiencies in the model which are worth mentioning. First, the model is sequential in tracing the lines of influence from independent variable to dependent variables. While this has the advantage of being rather parsimonious, it ignores possible feedback loops which would be more consistent with stress models of cognitive therapy. While acknowledgment is given to the reciprocal relationship of stages three and four in the model, Koslowsky and colleagues do not carry the reciprocity far enough. For example, they illustrate their conception of feedback loops by referring to the common experience of commuters in which a particularly harrowing experience during the ride to work (subjective impedance) is translated into physiological responses such as increased heart rate that remains elevated long after the person is conscious of why he or she is anxious. This state of anxiety then effects attitudes towards the job while at work.

While this makes sense as far as it goes, their model limits them from recognizing the obvious implications of their example for prior stages in their model. In other words, the increased heart rate of anxiety responses not only influences stages which are later in their proposed sequence (e.g., job attitudes), but also influences subjective impedance, a rather early stage in the sequential model of predictor and outcome variables. In fact, the literature they cite suggests such a feedback loop, at least in the case of women, whose greater stress reactions to long commutes result in greater subjective impedance (Novaco et al., 1991).

Second, even the first stage in their proposed sequence is potentially affected by emotional symptoms such as burnout. Objective impedance variables such as traffic congestion remain objective in so far as baseline driver skill levels are assumed. The negative effects of later stages in the sequential model, such as burnout, can decrease the actual skill level of the driver even if the assumed potential skill level remains the same.

Third, the model is inadequate for explaining potentially complex family effects such as mood changes in the children at various phases of development. To be fair, Koslowsky and colleagues intended their model for use in an organizational research context and so family effects are confined to the commuter’s home experience as measured by self-report.

The model could be strengthened by maintaining the same structure but adding to it a series of multiple feedback loops. Stages three through six all have potential for feeding back into stage two (subjective impedance). Physiological reactions and symptoms (stage 3) may heighten the perceived stress. Emotional and attitudinal strain responses (stage 4) can also affect the perception of stress. In other words, the second traffic jam perceived while one is still distressed over earlier traffic congestion may create a spiraling effect for perceived stress. Likewise, job avoidance inclinations can be both a result of and a cause of perceived stress, especially if the tendency to avoid work creates unnecessary time urgency by virtue of leaving late to begin a morning commute. Finally, negative behavioral/performance indicators (stage 6) at work may feed back to the second stage of perceived stress. A commuter whose performance at work is suffering, may have less of a sense of the worth of the commute and thus perceive it to be more of a strain.

Despite these deficiencies in Koslowsky’s model, it serves to illustrate the interdomain transfer effects demonstrated empirically by Novaco and colleagues (1991). Long commutes have negative home consequences for the mood of parents at home. Does it effect the young adolescents of these parents indirectly? In order to answer this question we must examine more closely the characteristics of early adolescent depression.