Points-Based Attendance Systems Associated With Presenteeism Despite Paid Sick Leave Protections

Abstract

Presenteeism, or attending work despite being ill enough to warrant absence, is endemic among working Americans. Although research suggests that paid sick leave mandates reduce presenteeism, it has not considered whether workers exposed to points systems are more likely to work while sick or how points systems may interact with paid sick leave mandates. We used 2024 data from the Shift Project on service workers employed at sixty-three large firms to study the relationship between points systems and presenteeism and to determine whether it varied by employee coverage under a paid sick leave law. We found that exposure to a points system was associated with an 18.9-percentage-point increase in presenteeism, with effects observed even among workers who had not incurred points. Paid sick leave mandates did not moderate exposure to a points system or the association between exposure to points systems and presenteeism. Results suggest that points-based attendance systems undermine the public health benefits of paid sick leave mandates. To reduce presenteeism in jurisdictions with paid sick leave mandates, outreach to covered workers and employers and stronger enforcement efforts may be needed.

Presenteeism, or attending work despite being ill enough to warrant absence, is endemic among working Americans¹ and has an array of negative consequences for workers, including increasing the risk for future illness² and cardiovascular disease,³ worsening mental well-being,⁴ and increasing the likelihood of labor force exit.¹ The negative effects extend beyond the individual worker by increasing the spread of disease,⁵ thus threatening population health.⁶

Many factors shape presenteeism, including understaffing of work sites,^7,8 employment insecurity,^3,7 and workplace culture.⁹ Many private sector employers, particularly those in the service sector who employ shift workers, aim to discourage workers’ absence to ensure sufficient staffing, and supervisors have pressured workers to come to work even when sick.^10,11 To help achieve those ends, employers increasingly have established points-based attendance (points) systems, in which employees receive points or occurrences for missing work that lead to disciplinary consequences up to and including termination, regardless of the reason for missing work.^11-13 Studies have found that large companies have disciplined workers using points-based systems and that workers have reported to work when sick¹² or needing to care for sick children¹³ out of fear of job loss. Advocates have raised concerns that points-based attendance systems may encourage presenteeism. To address this issue, New York State recently passed the Lawful Absence Law,¹⁴ which makes it illegal for employers to discipline employees for legally protected absences, including clarification that improperly assigning points under an attendance policy is a form of illegal retaliation.

Paid sick leave with job protection guarantees workers’ ability to stay home when ill. The United States lacks a federal guarantee for paid sick leave,¹⁵ but fifteen states and four additional cities and counties had mandated it by 2024.¹⁶ These mandates require employers to provide workers with one hour of job-protected sick leave for every thirty to forty hours worked, depending on the jurisdiction and firm size.¹⁶ Accrued leave can be used when a worker is ill, needs to care for an ill family member, or must attend medical appointments. Paid sick leave mandates have been found to be effective in reducing presenteeism^17,18 and transmission of communicable disease.^19-21

However, even when mandated by law, workers’ access to paid sick leave often remains constrained, and as with many labor standards,²² employer noncompliance remains an issue. Recent survey evidence from California suggests widespread noncompliance with its paid sick leave mandate, with 35 percent of workers reporting that they would not be paid if sick from COVID-19 and could not work for two weeks.²³ In more recent work, Daniel Schneider and colleagues found that 12 percent of service sector workers lacked access to any kind of paid sick leave and 29 percent reported conditions that likely violated California’s standard, such as being disciplined for use of paid sick leave.²⁴

The sanctions associated with points systems may be more salient to employees than the job protection that paid sick leave mandates provide, such that points systems could prevent their public health benefits from being fully realized. Previous research has not directly examined whether workers exposed to points systems are more likely to work while sick, whether points systems are less common in jurisdictions that have adopted paid sick leave mandates, or whether the presence of a paid sick leave mandate moderates any association between points systems and presenteeism.

We leveraged novel data on hourly workers employed at large firms in the retail, big-box, grocery, pharmacy, delivery and fulfillment, fast food, and casual dining sectors. In these sectors, points systems are common,¹¹ and the stakes of presenteeism are high. Service sector jobs are customer-facing and often involve food handling, preparation, or service, and presenteeism among service workers can increase transmission of communicable disease and create a public health risk.^25,26 Voluntary provision of paid sick leave is uncommon among service sector employers,²⁷ and given the prevalence of low wages, unstable schedules, and limited hours, paid sick leave mandates may be an important tool to enable workers to stay home with pay when ill.

Our study explores three hypotheses: that the presence of points systems is associated with presenteeism, that employers subject to paid sick leave mandates are less likely to use points systems, and that paid sick leave mandates moderate the association between points systems and presenteeism.

Study Data And Methods

Data And Sample

We used data collected by the Shift Project²⁸ from hourly workers employed at sixty-three large firms in the service sector, using targeted advertisements on Facebook and Instagram. The Shift Project used Meta’s advertising platform as both a quasi-sampling frame and survey recruitment channel. This approach enabled timely data collection on novel topics fielded to relevant samples of workers. We weighted the data on age, sex, race and ethnicity, employer, and state population size to represent the population of service sector workers in the American Community Survey (2014–23).²⁹ We further weighted the sample to “right-size” the percentage of respondents employed at each of the sixty-three firms to each firm’s national employee headcount based on Data Axles’s Infogroup–Business Academic estimates.³⁰ Demographic characteristics of respondents in paid sick leave and non–paid sick leave jurisdictions are presented in online appendix exhibit A1.³¹ The geographic distribution of respondents is presented in appendix exhibit A2.³¹

We developed a survey module to identify workers exposed to points systems and learn more about their experiences (appendix B).³¹ This module was included in the Shift Project survey wave fielded during September–November 2024, along with a standard set of questions on job quality, demographics, economic security, health outcomes, and parenting. The survey was self-administered online in Qualtrics. Of the 6,700 respondents who began the survey, 63 percent completed the module on points-based attendance policies. We further restricted the sample to respondents with complete data, yielding an analytic sample of 3,018 hourly workers.

Measures

Extent Of Exposure To Points-Based Attendance Systems:

Our key predictors included whether a worker was exposed to a points system and the extent of that exposure in the previous twelve months. Workers self-reported whether, to their knowledge, their workplace had a points system, which was defined in the survey as a policy, system, or bank where employees lose or earn points, occurrences, demerits, or disciplinary charges (points) for being late, leaving early, or taking time off. Workers who reported being subject to points systems were then asked whether they had accrued any points and, if so, whether they had incurred any consequences as a result. We first constructed a dichotomous indicator of any exposure to a points system. We then constructed a four-category variable that contrasted those who were not exposed to points systems with their counterparts who reported either exposure but no points accrual; exposure, some points, but no consequences; or exposure, some points, and experience of consequences.

Presenteeism:

To measure presenteeism, respondents were asked, “In the past month, did you ever work at [name of employer] even though you were feeling sick?” Respondents could select “Yes,” “No, I was sick but I stayed home,” or “No, I haven’t been sick in the past month.” We constructed three measures of presenteeism from this variable. First, using all respondents, we coded a measure of presenteeism equal to 1 for those who went to work while sick or 0 for respondents who reported feeling sick and staying home and for respondents who did not feel sick. Second, we limited the sample to those who reported being sick in the past month (n = 2,190) and coded presenteeism as 1 for those who went to work while sick or 0 for those who were sick and stayed home. Among workers who reported working while sick, we constructed a third variable coded as 1 if a respondent reported working while sick “in order to avoid losing or earning a point” and 0 otherwise.

Paid Sick Leave Mandate:

Workers employed in one of fifteen states or four localities with active paid sick leave mandates during data collection in 2024¹⁶ were coded as being exposed to a paid sick leave policy.

Control Variables

Models controlled for individual factors and employment characteristics. Individual characteristics included age, sex, race and ethnicity, cohabitation status, educational attainment, and presence of children in the household. Employment characteristics included the usual number of weekly hours worked, wage, union membership, and manager status.

Empirical Strategy

We first documented the prevalence of workers’ exposure to points systems and estimated a series of linear probability models to study the relationship between our dichotomous measure of exposure to a points system and each of our three measures of presenteeism. We then reestimated these models, substituting our categorical measure that captures exposure to a points system, accrual of points, and experience of consequences. Next, we examined whether exposure to a points system was less common among workers covered by paid sick leave mandates. Finally, we examined whether the relationship between exposure to a points system and presenteeism varied by paid sick leave mandate coverage. All models included employer fixed effects, and standard errors were clustered at the employer level.

Limitations

Our study was subject to several limitations. First, the sample generated from our social media recruitment method was a nonprobability sample. Prior methodological work shows that the data set closely replicates findings from two gold-standard probability data sources—the Current Population Survey and National Longitudinal Survey of Youth (1997)²⁸—but it is possible that our estimates of exposure to points systems might not be representative of the service sector. Second, the data do not allow for analyses that provide causal estimates. However, our test of whether paid sick leave mandates were associated with the presence of points systems provides evidence on the relationship between these attendance systems and paid sick leave mandates. Third, our measures for exposure to a points system captured experiences during a twelve-month period, whereas our measure of presenteeism captured experiences in a one-month period. Relatedly, our data collection occurred before the peak season for respiratory illness, and thus might not have fully captured peak communicable disease levels. These timing-related limitations may have led to an underestimation of presenteeism.

Study Results

Forty-nine percent of respondents reported exposure to points-based attendance systems (appendix exhibit A1),³¹ with the greatest use seen in the Midwest and Southwest (appendix exhibit A3).³¹ Exposure to points systems was associated with increases in presenteeism (exhibit 1). Overall, exposure to a points system was associated with an 18.9-percentage-point increase in presenteeism (model 1). We observed a gradient in the association between the extent of exposure to a points system and the probability of presenteeism. Workers who were exposed to points systems but did not themselves have points were 9.3 percentage points more likely to work while they were sick than their peers who were not exposed to a points system. Those who reported having received points and those with both points and consequences were 29.5 percentage points and 38.7 percentage points more likely, respectively, to work while sick (model 2).

Exhibit 1:

Exposure to employer points-based attendance policies and likelihood of working while sick in 63 large firms, 2024

	Worked sick		Worked sick in past month, conditional on illness		Worked sick to avoid punishment
	1	2	3	4	5	6
Points system	0.189****		0.118****		0.181****
Points system exposure
Unexposed		ref.		ref.		ref.
Points system, no points		0.093**		0.108***		0.141**
Points system, points		0.295****		0.136****		0.049
Points system, points + consequences		0.387****		0.130***		0.338****
Constant	0.586	0.488	0.912	0.897	0.546	0.528
No.	3,018	3,018	2,190	2,190	2,190	2,190

SOURCE Authors’ analysis of data from The Shift Project, 2024. NOTES Models 1, 3, and 5 test the association between working while sick and exposure to any points-based attendance system. Models 2, 4, and 6 test the association between working while sick and the intensity of exposure, using no exposure to any points system as the reference. “Points system, no points” refers to those exposed to a points system who have not themselves received a point. “Point system, points” refers to those who have received a point but not experienced a related consequence. “Points system, points + consequences” refers to respondents who have received both a point and a consequence related to that point. Demographic controls: age, gender, race and ethnicity, educational attainment, cohabitation status, and presence of children in the household. Work controls: usual weekly hours, wage, union membership, and manager status. Models also include employer fixed effects. All standard errors clustered by employer. Weighted by age, sex, race and ethnicity, employer, and state population size, as per the 2014–23 American Community Survey.²⁶

^*p < 0.10

^**p < 0.05

^***p < 0.01

^****p < 0.001.

We found a similar pattern when we conditioned working while sick on being sick in the past month. Those exposed to points systems were significantly more likely to work when sick (model 3), and the relationship was relatively stronger for those who had accrued points and experienced consequences (model 4). Exposure to a points system was also associated with an 18.1-percentage-point increase in the likelihood of reporting working while sick to avoid punishment (model 5). Workers who were exposed to points systems but had not accrued points were 14.1 percentage points more likely than those who were not exposed to a points system to work while sick to avoid punishment. Among workers who were exposed to points systems who had accrued points and experienced consequences, the likelihood of working while sick to avoid punishment was 33.8 percentage points higher than for those not exposed to points systems. Results are robust to a sample of only those workers with at least one year of job tenure (appendix exhibit A4)³¹ and model specifications that vary the inclusion of controls and fixed effects (appendix exhibit A5).³¹

We found that coverage under a paid sick leave law was not associated with a lower likelihood of exposure to a points-based attendance system: nearly half of firms had points systems regardless of coverage (appendix exhibit A6).³¹ Finally, our investigation found no statistically significant differences in the association between points systems and presenteeism by whether respondents worked in jurisdictions with paid sick leave mandates.(exhibits 2-4) However, exhibits 2 and 3 show insignificant but suggestive (p < 0.1) evidence that points systems might be associated with higher levels of presenteeism in places that have paid sick leave laws in place.

Exhibit 2. Likelihood of working while sick in 63 large firms, by exposure to paid sick leave mandates, 2024.

Source/Notes: SOURCE Authors’ analysis of data from The Shift Project, 2024. NOTES Demographic controls: age, gender, race and ethnicity, educational attainment, cohabitation status, and presence of children in the household. Work controls: usual weekly hours, wage, union membership, and manager status. All models include employer fixed effects. All standard errors clustered by employer. Weighted by age, sex, race and ethnicity, employer, and state population size as per the 2014–23 American Community Survey.²⁶

Exhibit 4. Likelihood of working while sick to avoid punishment among those who were ill in 63 large firms, by exposure to paid sick leave mandates, 2024.

Source/Notes: SOURCE Authors’ analysis of data from The Shift Project, 2024. NOTES Demographic controls: age, gender, race and ethnicity, educational attainment, cohabitation status, and presence of children in the household. Work controls: usual weekly hours, wage, union membership, and manager status. All models include employer fixed effects. All standard errors clustered by employer. Weighted by age, sex, race and ethnicity, employer, and state population size as per the 2014–23 American Community Survey.²⁶

Exhibit 3. Likelihood of working while sick among those who were ill in 63 large firms, by exposure to paid sick leave mandates, 2024.

Source/Notes: SOURCE Authors’ analysis of data from The Shift Project, 2024. NOTES Demographic controls: age, gender, race and ethnicity, educational attainment, cohabitation status, and presence of children in the household. Work controls: usual weekly hours, wage, union membership, and manager status. All models include employer fixed effects. All standard errors clustered by employer. Weighted by age, sex, race and ethnicity, employer, and state population size as per the 2014–23 American Community Survey.²⁶

Discussion

We found that exposure to points-based attendance systems was associated with service sector workers’ decisions to come to work when sick, even among those who had not received points or demerits because of absence. Presenteeism increased with exposure to points plus consequences. Our findings suggest that workers change their behavior in response to the threat of points to avoid accruing them; thus, points systems appear to create an incentive structure that increases presenteeism.

Prior studies^17,21 demonstrate that paid sick leave mandates reduce presenteeism, suggesting that paid sick leave mandates, where available, might be expected to moderate the effect of points systems on presenteeism by offering workers job security when taking leave for health-related reasons. However, we found that the association between points systems and presenteeism remained even in jurisdictions that require employers to provide paid leave for health reasons. Our findings suggest that points systems encourage workers to work while sick, even in jurisdictions where they are covered by paid sick leave mandates, thus limiting these mandates’ effectiveness in reducing presenteeism. Given that food and retail workers are susceptible to communicable disease infection and transmission because of the public-facing nature of their work,^23,24,32 our findings indicate that points-based attendance systems have negative implications for public health. Providing information to workers and employers about workers’ rights to use their earned sick leave³³ in jurisdictions with paid sick leave requirements may help reduce the influence of points-based attendance systems on presenteeism.

Conclusion

We observed a consistent relationship between exposure to points systems and presenteeism that persists even when paid sick leave mandates are in place. These results suggest that the use of points-based attendance systems in the service sector undermines the public health benefits of paid sick leave mandates. To reduce presenteeism in jurisdictions with paid sick leave mandates, outreach to covered workers and employers and stronger enforcement efforts may be needed.