Resource and capacity constraints can impair health care organizations’ functioning and adversely affect the workforce, particularly during shocks. This paper examines whether organizational slack resources—surplus resources accumulated by hospitals—benefit staff by reducing absenteeism and turnover, and whether slack can buffer the negative workforce effects of a shock, operationalized here as the COVID-19 years.
The study aimed to identify (1) direct relationships between three types of slack and hospital absenteeism and turnover rates, and (2) whether effects of slack differ between COVID-19 and non-COVID-19 years.
The analysis used a nationwide dataset comprising annual reports from all Dutch hospitals covering 2017 through 2021. The study sample included 342 hospital-year combinations. The authors applied linear mixed-effects modeling to estimate direct effects of slack resources on workforce outcomes and to test moderation by COVID-19 (comparing COVID-19 years with non-COVID-19 years).
The paper distinguishes three conceptually distinct forms of organizational slack:
These categories reflect different mechanisms through which surplus resources might support organizational resilience and workforce stability.
Linear mixed-effects models were used to estimate direct effects of the three slack types on two workforce outcomes—absenteeism and turnover—and to test whether the relationship between slack and these outcomes was moderated by the COVID-19 period. The sample frame and modeling approach allowed the authors to use repeated hospital-year observations and to assess both main effects and interactions with the shock period. The abstract reports the modeling approach and sample size; specific variable operationalizations, covariates, and full model diagnostics were not provided in the abstract.
The study reports two main findings:
There was a significant curvilinear U-shaped relationship between absorbed slack (staff levels) and turnover rates. This indicates that the association between staff surplus and turnover is not strictly linear; turnover rates vary in a U-shaped pattern across levels of absorbed slack. The abstract does not provide exact inflection points or effect sizes.
Unabsorbed slack (financial reserves) and potential slack (capacity for additional loans) significantly buffered against the increasing effect of COVID-19 on absenteeism rates. In other words, hospitals with greater financial reserves or borrowing capacity experienced less of an increase in absenteeism attributable to the pandemic compared with hospitals with less of those types of slack. The abstract reports statistical significance for these buffering effects but does not supply numerical estimates or confidence intervals.
These findings differentiate the types of slack by outcome: staff-level slack showed a nonlinear association with turnover, while financial- and loan-capacity slack provided protective buffering against pandemic-related absenteeism.
Based on these results, the authors argue that incentivizing health care organizations to acquire and maintain appropriate slack resources can help reduce turnover and, during shocks such as pandemics, mitigate adverse impacts on absenteeism. The study suggests re-evaluating common policies and organizational practices that strictly constrain resources in the name of cost reduction, because such constraints may leave hospitals vulnerable to workforce instability during shocks.
The abstract frames these implications in workforce crisis management and pandemic preparedness: maintaining financial reserves and borrowing capacity appears especially valuable as a buffer during system-level shocks.
The abstract and PubMed record report study aims, dataset scope (2017–2021, Netherlands), sample size (342 hospital-year combinations), modeling approach (linear mixed-effects), and main directional findings (U-shaped absorbed slack–turnover relationship; buffering by unabsorbed and potential slack on absenteeism during COVID-19). The abstract does not report specific numerical effect sizes, p values, parameter estimates, model covariates, sensitivity analyses, or limitations. If those details are needed for operational decision-making or deeper appraisal, the full article should be consulted.