Been There, Bottled That: Are State and Behavioral Work Engagement New and Useful Construct “Wines”?

DANIEL A. NEWMANTexas A&M UniversityDAVID A. HARRISONPennsylvania State UniversityMacey and Schneider (2008) have providedan informative treatise on the psychologicalconstruct of employee engagement. Webelieve their conceptual work is a thought-provoking exemplar of how intuitive con-structs (such as engagement) can begin tobe legitimized for the academic audience,hopefully strengthening communicationbetween scientists and practitioners. Theyattempt to specify the meaning of the popu-lar concept vis-a`-vis more long-standingconstructsofjobsatisfaction,organizationalcommitment, job involvement, positiveaffect and affectivity, and proactive and cit-izenship behavior. The engagement label issummarily applied to describe psychologi-cal states, traits, and behaviors. Much effortis spent parsing among established con-structswithregardtotheirpartialconceptualoverlap with the newer label.We offer three reactions to Macey andSchneider’s paper. First, because employeeengagement comprises no new conceptualcontent,butratherablendofoldcontent,itismost appropriately specified as a higherorder latent construct. Second, the utility ofa state engagement construct depends uponevidence for its discriminant validity fromrelated higherorderconstructs,namelyover-alljobattitude.Third,thetermengagement—even as typically used by practitioners andlaypersons—canbeparsimoniouslyconcep-tualized as a second-order factor of widelystudied work behaviors, including focal jobperformance, withdrawal behavior, and citi-zenship behavior. That term and constructhave already been forwarded and expli-catedinpriorwork,includingameta-analyticsummary of attitude–behavior relationshipsinvolving more than 500 original studies(Harrison, Newman, & Roth, 2006). Thisprevious work supported a broad attitude–engagementmodelthat unifiedseveralindi-vidual-level constructs in both the predictorand the criterion spaces and offered analternative view on employee engagement,suggesting that (a) engagement should bedistinguished from job attitudes and (b)engagement can be both intuitivelyand par-simoniously modeled as a higher orderbehavioral construct.Is Engagement a New Construct?Welaudthegeneralgoalofcoalescingideasand evidence into abstractions that can