Absenteeism

Abstract

Absenteeism, which MeSH classifies under industrial psychology, is the habitual or repeated absence of an employee from scheduled work. It is costly and difficult to study, because the same record of days and spells away from the job can be assembled into measures that behave differently and correlate only weakly. The organising framework is the Steers and Rhodes process model, which locates attendance in the joint product of an employee's motivation to attend and their ability to attend; antecedents include job attitudes, absence culture, the social contagion of absence, and health and work strain. Its mirror image, presenteeism — attending while ill — is now a research field in its own right. This article covers what absenteeism is, how attendance is modelled and absence measured, its causes and correlates, and presenteeism, with three interactive demonstrations.

Keywords: absenteeism, attendance, presenteeism, absence culture, work motivation

What Absenteeism Is

Absenteeism is the failure to report for or remain at scheduled work when attendance is expected. The definition is deceptively simple, because absence is heterogeneous: a single record of days away conflates certified sickness, unexplained non-attendance, authorised leave, and family emergencies, which have different causes and warrant different responses. Researchers therefore distinguish involuntary absence, which is largely outside the employee's control (genuine illness, a transport failure), from voluntary absence, which reflects a choice not to attend. The distinction is conceptually essential and empirically fraught, because an organisation observes only the absence, not the motive behind it, and employees have reason to recode voluntary absence as involuntary.

Absence also has a characteristic shape across a workforce. It is not distributed normally but is strongly skewed: most employees are absent rarely or never in a given period, while a small minority accumulate the majority of the lost days. This J-shaped or Poisson-like distribution means that averages mislead and that the behaviour of the high-absence tail, rather than the modal employee, drives an organisation's total. It is also why the two broad families of measure — counting the amount of absence versus counting its frequency — can rank the same workforce differently, a problem taken up below.

Models of Attendance and Absence

The dominant framework is the process model of attendance proposed by Steers and Rhodes, which reframed the question from why are employees absent? to what determines whether an employee attends? (#ref-steers-rhodes-1978). In their account, attendance is the joint function of two proximal causes: an employee's motivation to attend and their ability to attend. Motivation to attend is itself shaped by satisfaction with the job situation and by internal and external pressures to attend — economic conditions, incentive systems, work-group norms, personal work ethic, and organisational commitment. Ability to attend captures the constraints that override motivation entirely: illness, family responsibilities, and transportation problems. The model's lasting contribution is its insistence that high motivation does not guarantee attendance when ability is compromised, and that neither variable alone predicts the behaviour. The first demonstration below makes this joint dependence manipulable.

Nicholson offered a complementary reframing in his conceptual synthesis of absence behaviour and attendance motivation, arguing that absence is best understood not as a stable trait of the person but as an adjustment to the work situation — a negotiated, dynamic accommodation between the employee and the constraints and inducements of the job (#ref-nicholson-1977). Where the process model foregrounds the decision to attend on a given day, Nicholson foregrounded the regularities of attachment and withdrawal that build up over time.

A third strand treats absence as one member of a family of withdrawal behaviours unfolding in time. Harrison and Martocchio's twenty-year review organised the field around the origins, offshoots, and outcomes of absence, and argued for treating it as a time-based construct whose meaning changes with the interval over which it is aggregated — a day, a quarter, a career (#ref-harrison-martocchio-1998). On this view absence, lateness, and turnover are not independent problems but related expressions of an underlying withdrawal process.

Measuring Absence

No question in the field has proved more consequential than how absence is counted, because the construct has no single natural metric. Two broad families dominate. Time-lost measures sum the total duration away — days or hours absent over a period — and answer the question of how much working time was foregone. Frequency measures count the number of distinct absence spells regardless of their length, on the logic that a spell reflects a fresh decision to be absent and that frequent short absences are more disruptive and more diagnostic of voluntary withdrawal than a single long certified illness. The two indices correlate only modestly, because a long illness inflates time-lost while contributing one spell, and a pattern of frequent single days inflates frequency while contributing few total days. Steel's analysis of the operational issues showed that this is not a technicality: the reliability and the construct validity of absence variables depend heavily on the index chosen and the period over which it is aggregated (#ref-steel-2003).

Compounding the measurement problem, self-reported absence is systematically inaccurate. A meta-analysis of the reliability, validity, and accuracy of self-reported absence found that employees substantially under-report their own absence relative to organisational records, so that studies relying on self-report attenuate the very relationships they seek to estimate (#ref-johns-miraglia-2015). Table 1 summarises the principal families of measure and what each is suited to.

Causes and Correlates of Absence

The oldest hypothesis in the field is that dissatisfied employees stay away. Hackett and Guion's re-evaluation of the absenteeism–job-satisfaction relationship synthesised the accumulated studies and found the association real but weak — far smaller than the folk model assumes — with correlations rarely exceeding the low tens (#ref-hackett-guion-1985). Farrell and Stamm's meta-analysis of the broader correlate literature confirmed the pattern across many predictors: personal, attitudinal, and organisational variables each account for a modest share of absence, and no single antecedent dominates (#ref-farrell-stamm-1988). The weakness of these correlations is itself informative, because it points away from the individual attitude and toward the social and situational context.

That context is captured by the concept of absence culture. Martocchio demonstrated that the shared, normative understanding within a work group about how much absence is acceptable shapes individual absence over and above personal attitudes — employees calibrate their own attendance to the perceived norm (#ref-martocchio-1994). The social mechanism has since been made explicit: ten Brummelhuis and colleagues showed that employees imitate the absence of their coworkers, and specified when the contagion is strongest, establishing absence as a partly socially transmitted behaviour rather than a purely individual one (#ref-tenbrummelhuis-2016).

Health and work strain supply the involuntary side of the ledger. Darr and Johns's meta-analysis of work strain, health, and absenteeism mapped the pathways by which psychological strain and physical illness translate into time away, distinguishing an illness pathway from a withdrawal pathway and showing that strain relates to absence partly through genuine ill-health and partly through attitudinal routes (#ref-darr-johns-2008). This dual character — absence as symptom and absence as choice — is what makes the behaviour resistant to any single explanation.

Presenteeism: The Other Side of Attendance

Attending work is not the simple complement of being absent, because an employee can be present while ill and impaired — presenteeism. Johns's review established presenteeism as a research construct in its own right and set an agenda for studying it, arguing that attending while sick carries a productivity loss that absence accounting ignores and that the choice between absence and presence is a single decision problem (#ref-johns-2010). The empirical core arrived with Miraglia and Johns's dual-path meta-analysis, which showed that presenteeism is driven by two distinct routes: a health-impairment path, on which poor health and job demands push a strained employee to attend, and a work-attitude path, on which satisfaction and commitment draw a motivated employee to attend — so that the same behaviour can signal either strain or engagement (#ref-miraglia-johns-2016). The third demonstration traces these two pathways.

The construct has consolidated quickly. Lohaus and Habermann's review organised the field's definitions, antecedents, and consequences and flagged the conceptual confusion that different measurement traditions had introduced (#ref-lohaus-habermann-2019). From occupational health psychology, Kinman synthesised the evidence on sickness presenteeism's prevalence, costs, and management, underlining that presenteeism often predicts future absence and poorer health, so that suppressing absence without addressing its causes can simply convert it into presenteeism at a later cost (#ref-kinman-2019).

Worked Example

Consider a five-person team over a 20-workday month, so that scheduled time is 5 × 20 = 100 person-days. Their absence records for the month are:

- Employee A — 0 days absent, in 0 spells - Employee B — 1 day absent, in 1 spell - Employee C — 2 days absent, in 2 separate one-day spells - Employee D — 3 days absent, in 1 spell (one continuous illness) - Employee E — 6 days absent, in 4 separate spells

The team's time-lost rate is the total days absent over scheduled time: (0 + 1 + 2 + 3 + 6) / 100 = 12 / 100 = 12%. Its frequency rate is the mean number of spells per employee: (0 + 1 + 2 + 1 + 4) / 5 = 8 / 5 = 1.6 spells.

Now apply the Bradford Factor, B = S² × D, which squares the number of spells S and multiplies by total days D — a deliberately frequency-weighted index used in HR policy to flag disruptive patterns:

- A: 0² × 0 = 0 - B: 1² × 1 = 1 - C: 2² × 2 = 8 - D: 1² × 3 = 3 - E: 4² × 6 = 96

The team's total Bradford score is 0 + 1 + 8 + 3 + 96 = 108, of which Employee E contributes 96 — about 89% — despite accounting for only 6 of the 12 absent days (50%). Employee D, with a longer single illness (3 days), scores just 3, less than Employee C's two one-day absences (8). The example shows concretely why the choice of measure is not cosmetic: a time-lost view treats D as a bigger problem than C, while any frequency-weighted view reverses that ranking and isolates E as the pattern an attendance policy is designed to catch. The same five records, three defensible measures, three different pictures.

Figure 1

The Steers and Rhodes Process Model of Attendance

Discussion

The study of absenteeism is a case study in how a construct that seems trivially observable becomes hard the moment it is operationalised. The behaviour is real and costly, but it is heterogeneous in cause, skewed in distribution, ambiguous between the voluntary and the involuntary, and measurable in several non-equivalent ways — and each of these properties has defeated a generation of simple models. The weak and inconsistent correlation between job attitudes and absence, once treated as a disappointment, is now read as a finding: absence is not mainly an attitude expressed through the feet but a behaviour shaped by ability constraints, group norms, and social imitation that a focus on the individual employee misses (#ref-martocchio-1994; #ref-tenbrummelhuis-2016).

For the practitioner the implications are concrete. An attendance policy built on a single metric will optimise that metric and distort the others; a policy that counts only days lost rewards the frequently-absent employee relative to the genuinely ill one, and a frequency-weighted index does the reverse. The process model's lesson — that motivation and ability must both be addressed — argues against purely punitive schemes, which raise the cost of absence (and so motivation to attend) without touching the ability constraints that produce much involuntary absence, and which risk converting suppressed absence into presenteeism (#ref-kinman-2019). The most defensible programmes measure absence on more than one index, distinguish its kinds, and treat attendance and presence as the single decision the evidence shows them to be (#ref-johns-2010).

Current Directions

The liveliest front is presenteeism, which has moved in little more than a decade from a footnote in the absence literature to a field with its own theory and measurement debates. The dual-path model reframed the central question from how much presenteeism occurs to why it occurs, distinguishing the strained employee who attends out of health-driven pressure from the committed one who attends out of engagement, and showing that the two carry different consequences (#ref-miraglia-johns-2016). The practical stakes are high, because evidence that presenteeism predicts later absence and worse health implies that the two behaviours are substitutes over time, so an intervention that reduces one may simply inflate the other (#ref-kinman-2019). A second front is conceptual housekeeping: reviews have documented how divergent definitions and instruments have fragmented the presenteeism evidence base, and have pressed for a common measurement framework before the accumulated findings can be compared (#ref-lohaus-habermann-2019). Across both, the unifying move is to treat attendance as a continuum of states — absent, present and impaired, present and productive — rather than a binary, and to ask what organisational conditions push employees toward each.

Common Misconceptions

Absenteeism is mostly a matter of lazy or dissatisfied employees.

The correlation between job dissatisfaction and absence is real but small, rarely exceeding the low tens (Hackett & Guion, 1985). Much absence is involuntary — illness, caring duties, transport — and group absence norms predict individual absence over and above personal attitudes (Martocchio, 1994). The folk model survives because it blames a visible person rather than an invisible situation.

A single absence rate captures how bad a team's attendance is.

Time-lost and frequency indices capture different behaviour and correlate only modestly, so they can rank the same workforce differently (Steel, 2003). A long certified illness and a pattern of frequent single days are opposite problems that one number conceals.

Reducing absence is always a good outcome.

Suppressing absence without addressing its causes can convert it into presenteeism — attending while ill — which carries its own productivity cost and predicts later absence and worse health (Kinman, 2019). Attendance and presence are one decision, not two separate goods.

Glossary

Ability to attend.

In the process model, the constraints that can override motivation — illness, family responsibilities, transport — determining whether a willing employee can in fact attend.

Absence culture.

The shared, normative understanding within a work group of how much absence is legitimate, which shapes individual absence beyond personal attitudes.

Absence frequency.

The number of distinct absence spells in a period, regardless of their length; treated as an index of voluntary withdrawal.

Absenteeism.

The habitual or repeated absence of an employee from scheduled work, encompassing both voluntary and involuntary non-attendance.

Attendance motivation.

An employee's drive to report for work, shaped by satisfaction with the job situation and by internal and external pressures to attend.

Bradford Factor.

An absence index equal to the number of spells squared times total days (S² × D), weighting frequent short absences heavily; used in HR policy to flag disruptive patterns.

Involuntary absence.

Non-attendance largely outside the employee's control, such as certified illness or a transport failure.

Job satisfaction.

An employee's evaluative attitude toward the job; a weak but genuine correlate of absence and a proximal input to attendance motivation.

Presenteeism.

Attending work while ill or impaired; the mirror of absenteeism, carrying its own productivity loss and predicting later absence.

Process model of attendance.

Steers and Rhodes's framework locating attendance in the joint product of motivation to attend and ability to attend.

Spell.

A single continuous episode of absence, from the first missed day to the return to work, regardless of its duration.

Time-lost rate.

The total duration absent as a share of scheduled time; sensitive to long illnesses and used to cost lost productivity.

Voluntary absence.

Non-attendance reflecting a choice not to attend when attendance was possible; conceptually central but hard to observe directly.

Withdrawal behaviour.

The family of related behaviours — absence, lateness, turnover — understood as expressions of an underlying disengagement process unfolding in time.

Key Researchers

David A. Harrison

(University of Texas at Austin, McCombs School of Business). Co-author of the twenty-year review that organised absence research around its origins, offshoots, and outcomes, and a leading voice on withdrawal as a time-based construct. [Faculty Page]

Gary Johns

(Concordia University, Montreal). The dominant modern scholar of workplace attendance; authored the review that defined presenteeism as a research construct and co-developed the critique of how absence is measured. [Google Scholar]

Gail Kinman

(Birkbeck, University of London). Occupational health psychologist who synthesised the evidence on sickness presenteeism's prevalence, costs, and management. [ORCID]

Joseph J. Martocchio

(University of Illinois at Urbana-Champaign). Demonstrated that a work group's absence culture shapes individual absence, and co-authored the twenty-year review; a central figure in modelling absence as a social event. [Faculty Page]

Mariella Miraglia

(University of Liverpool Management School). Co-author of the dual-path meta-analysis of presenteeism, establishing that attending while ill is driven by both strained-health and work-attitude routes. [ORCID]

Nigel Nicholson

(London Business School). Author of the conceptual synthesis of absence behaviour and attendance motivation, which recast absence from a trait into a negotiated adjustment to the work situation. [Faculty Page]

Richard M. Steers

(University of Oregon, Lundquist College of Business). Co-author of the process model of attendance, the field's organising framework, which located attendance in the joint influence of motivation and ability to attend. [Faculty Page]

Frequently Asked Questions

What is absenteeism?

Absenteeism is the habitual or repeated absence of an employee from scheduled work. It covers both involuntary absence, such as certified illness, and voluntary absence, which reflects a choice not to attend, and it is studied as a costly and heterogeneous workplace behaviour.

What causes employee absenteeism?

There is no single cause. Job dissatisfaction correlates with absence only weakly; much stronger roles are played by health and work strain, by an employee's ability to attend, and by the shared absence culture of the work group, which calibrates individual attendance to a perceived norm.

Is absenteeism caused by low job satisfaction?

Only partly. The satisfaction-absence correlation is real but small, rarely exceeding the low tens. The weakness of this link is itself a finding, pointing toward situational and social causes rather than individual attitude as the main driver of absence.

How is absenteeism measured?

Chiefly in two ways: time-lost measures sum the total days or hours absent, while frequency measures count the number of distinct absence spells. The two correlate only modestly, and composite indices such as the Bradford Factor weight frequent short absences more heavily.

What is the difference between absenteeism and presenteeism?

Absenteeism is being absent from work; presenteeism is attending while ill or impaired. They are mirror images of a single attendance decision, and presenteeism carries its own productivity cost that simple absence accounting overlooks.

Why is self-reported absence unreliable?

Employees systematically under-report their own absence relative to organisational records, because of imperfect memory and social-desirability bias. Studies that rely on self-report therefore tend to understate the relationships they are trying to measure.

What is absence culture?

Absence culture is the shared, normative understanding within a work group of how much absence is acceptable. It shapes individual absence over and above personal attitudes, and helps explain why absence spreads among coworkers.

Does reducing absenteeism always help an organisation?

Not necessarily. Suppressing absence without addressing its underlying causes can convert it into presenteeism, which predicts later absence and poorer health, so a narrow focus on attendance figures can shift the cost rather than remove it.

References

Darr, W., & Johns, G. (2008). Work strain, health, and absenteeism: A meta-analysis. Journal of Occupational Health Psychology, 13(4), 293–318. https://doi.org/10.1037/a0012639

Farrell, D., & Stamm, C. L. (1988). Meta-analysis of the correlates of employee absence. Human Relations, 41(3), 211–227. https://doi.org/10.1177/001872678804100302

Hackett, R. D., & Guion, R. M. (1985). A reevaluation of the absenteeism–job satisfaction relationship. Organizational Behavior and Human Decision Processes, 35(3), 340–381. https://doi.org/10.1016/0749-5978(85)90028-7

Harrison, D. A., & Martocchio, J. J. (1998). Time for absenteeism: A 20-year review of origins, offshoots, and outcomes. Journal of Management, 24(3), 305–350. https://doi.org/10.1177/014920639802400303

Johns, G. (2010). Presenteeism in the workplace: A review and research agenda. Journal of Organizational Behavior, 31(4), 519–542. https://doi.org/10.1002/job.630

Johns, G., & Miraglia, M. (2015). The reliability, validity, and accuracy of self-reported absenteeism from work: A meta-analysis. Journal of Occupational Health Psychology, 20(1), 1–14. https://doi.org/10.1037/a0037754

Kinman, G. (2019). Sickness presenteeism at work: Prevalence, costs and management. British Medical Bulletin, 129(1), 69–78. https://doi.org/10.1093/bmb/ldy043

Lohaus, D., & Habermann, W. (2019). Presenteeism: A review and research directions. Human Resource Management Review, 29(1), 43–58. https://doi.org/10.1016/j.hrmr.2018.02.010

Martocchio, J. J. (1994). The effects of absence culture on individual absence. Human Relations, 47(3), 243–262. https://doi.org/10.1177/001872679404700301

Miraglia, M., & Johns, G. (2016). Going to work ill: A meta-analysis of the correlates of presenteeism and a dual-path model. Journal of Occupational Health Psychology, 21(3), 261–283. https://doi.org/10.1037/ocp0000015

Nicholson, N. (1977). Absence behaviour and attendance motivation: A conceptual synthesis. Journal of Management Studies, 14(3), 231–252. https://doi.org/10.1111/j.1467-6486.1977.tb00364.x

Steel, R. P. (2003). Methodological and operational issues in the construction of absence variables. Human Resource Management Review, 13(2), 243–251. https://doi.org/10.1016/s1053-4822(03)00015-9

Steers, R. M., & Rhodes, S. R. (1978). Major influences on employee attendance: A process model. Journal of Applied Psychology, 63(4), 391–407. https://doi.org/10.1037/0021-9010.63.4.391

ten Brummelhuis, L. L., Johns, G., Lyons, B. J., & ter Hoeven, C. L. (2016). Why and when do employees imitate the absenteeism of co-workers? Organizational Behavior and Human Decision Processes, 134, 16–30. https://doi.org/10.1016/j.obhdp.2016.04.001