Where does job embeddedness collide?

Mitchell, Holtom, Lee, Sablynski, and Erez introduced job embeddedness in 2001 in Why people stay as the whole web of forces that keeps an employee in a job, built from three parts: the links a person has to other people and activities, the fit between the person and the job and its surroundings, and the sacrifice that leaving would cost. In this conception the construct is a composite, an aggregate summed across many concrete facets rather than one overall feeling, and it was built to predict who quits beyond job satisfaction and perceived alternatives. This is the foundational model that the other conceptions on this page reorganize, split, or replace, so a study that just says job embeddedness usually means this three-part composite unless it says otherwise.[1]

Unit

One employee's attachment, aggregated from links, fit, and sacrifice

Recovery means

Not applicable

Measured by

A composite index summed from roughly forty links, fit, and sacrifice items

Lee, Mitchell, Sablynski, Burton, and Holtom cut the same construct along a second seam in 2004: each of fit, links, and sacrifice can arise on the job, inside the organization, or off the job, in the surrounding community. Their study The effects of job embeddedness reported that the two domains predict different things, with off-the-job (community) embeddedness tracking staying and attendance more closely and on-the-job embeddedness tracking performance and citizenship. This conception treats job embeddedness not as one lump but as two separable domains that behave differently, so it warns that a single overall score, of the kind the composite model produces, can hide which domain is actually doing the work.[2]

Unit

An employee's attachment separated into organization and community domains

Recovery means

Not applicable

Measured by

The composite items grouped into an on-the-job total and an off-the-job total

Crossley, Bennett, Jex, and Burnfield argued in 2007 that summing many specific facets into a formative composite makes the construct hard to validate, and proposed instead a global measure: a short set of reflective items that ask directly about an employee's overall attachment to the organization, such as feeling tied to it or finding it hard to leave. In their study Development of a global measure of job embeddedness the reflective scale predicted turnover and added information beyond the original composite. The conception shifts the construct from an aggregate assembled out of parts to a single underlying attachment that the parts merely reflect, which is a different measurement model, not just a shorter questionnaire, and it sits in direct tension with the composite and two-domain conceptions.[3]

Unit

An employee's overall attachment to the organization, taken as one latent quantity

Recovery means

Not applicable

Measured by

A short reflective scale of global attachment items rated by the employee

A fourth conception treats job embeddedness less as a particular scale and more as a retention force: the reason people stay, which is not simply the mirror image of the reasons people leave. Holtom, Mitchell, Lee, and Eberly's 2008 review Turnover and retention research and Jiang, Liu, McKay, Lee, and Mitchell's 2012 meta-analysis place embeddedness inside turnover theory as a web of on-the-job and off-the-job forces that hold an employee in place, distinct from dissatisfaction or a better offer elsewhere. In this framing the construct answers a different question from the classic leaving models, namely what binds someone who might otherwise go, and the term names that theoretical role in a retention account rather than any one instrument used to capture it.[4][5]

Unit

The bundle of forces holding an employee in place, as a theoretical role

Recovery means

Not applicable

Measured by

Meta-analytic and review estimates of embeddedness predicting staying, over and above leaving factors