Work engagement or burnout: Which comes first? A meta-analysis of longitudinal evidence

Abstract Currently, burnout (BU) and work engagement (WE) and are considered different forms of workplace well-being, negatively related that might even co-occur, or as direct opposites and mutually exclusive. These contrasting views generate difficulties regarding the true nature of the relationship between the two concepts. In the present paper, we aim at clarifying this issue by testing the cross-lagged effects between BU and WE. We conducted systematic database searches using keywords relevant for WE, BU and design type (e.g., longitudinal), and we found 25 eligible research studies (N total  = 13271 participants). The selected papers a) reported a longitudinal research study; b) included measures of BU and WE, and c) reported the correlation matrix between BU and WE at all measurement moments. First, we used meta-analytical formulas to compute the averaged correlations between BU and WE. Second, we used the averaged effects to complete a correlation matrix, which was used to test the cross-lagged effects between BU and WE, using structural equations modeling. On the entire sample of studies, we found insignificant cross-lagged effects between BU and WE. However, when the time-lag between the two measurement moments was used as a moderator, significant reciprocal cross-lagged effects were found between exhaustion and WE, at 12-month time lag. Notably, it appears that the validity of causal perspective depends on the size of the time lag.