Call centers (whether in one physical location or a remote configuration of workers from home) are finely tuned operations whose economic outcomes often depend on very precise optimization of staff levels against anticipated call volume.[1] Other similar operations include retail stores and restaurants. When an employee is unexpectedly absent in a call center, it may mean that calls are missed, that other workers must adjust and will do their jobs less effectively, or that a buffer of extra workers must be employed or kept on call to offset the effects of absence. What is it worth to reduce such absences? What costs can be avoided, and what is the likely effect of organizational investments designed to reduce the need or the motivation of employees to be absent?
A first reaction might be, “We should cut absences to zero because employees should be expected to show up when they are scheduled.” However, as discussed in this chapter, the causes of absence are highly varied, so cutting absence requires a logical approach to understanding why it happens. In fact, an increasing number of jobs have no absenteeism, because they have no real work schedule! They are project-based, and thus are accountable only for the ultimate results of their work. In such jobs, employees can work whatever schedule they want, as long as they produce the needed results on time. For many jobs, however, adhering to the work schedule is an important contribution to successful operations.
Sometimes it is cost-effective just to tolerate the absence level, and allow work to be missed or employees to adjust. In other situations, it is very cost-effective to invest in ways to reduce absence. It depends on the situation.
Particularly when employees are absent because they are taking unfair advantage of company policies (such as claiming more sick leave than is appropriate), it is tempting to conclude that such absence must be reduced even if it takes a significant investment. It seems “unfair” to tolerate it. Upon further reflection, however, it’s clear that absence is like any other risk factor in business. How we address it should be based on a logical and rational decision about costs and benefits. We need a logical understanding of the consequences of absence to make those decisions. We provide that logic in this chapter.
Let us begin our treatment by defining the term absenteeism. Absenteeism is any failure to report for or remain at work as scheduled, regardless of reason. The use of the words as scheduled is significant, for this automatically excludes vacation, personal leave, jury-duty leave, and the like. A great deal of confusion can be avoided simply by recognizing that if an employee is not on the job as scheduled, he or she is absent, regardless of cause. We focus here on unscheduled absence because it tends to be the most disruptive and costly of the situations where an employee is not at work. Scheduled or authorized absences are more predictable. The employee is not available to perform his or her job as expected. This often means that the work is done less efficiently by another employee or is not done at all. This chapter describes in detail the potential costly consequences of absence.
Although the definition of absenteeism might leave little room for interpretation, the concept itself is undergoing a profound change, largely as a result of the time-flexible work that characterizes more and more jobs in our economy. A hallmark of such work is that workers are measured not by the time they spend, but by the results they achieve. Consider, for example, the job of a computer programmer whose sole job is to write or evaluate computer code. The programmer is judged by whether the program runs efficiently and whether it does what it is supposed to do reliably. It doesn’t matter when the programmer works (9 to 5 or from midnight to dawn) or where the programmer works (at the office or at home).
If the work schedule doesn’t matter, the concept of absenteeism has no meaning. If workers never “report” for work, and if they are allowed to vary their work time, and are accountable only in terms of results, the concept of absenteeism ceases to be relevant. Many teleworkers fit this category. Before attempting to assess the costs of employee absenteeism, therefore, it is important to identify where, within the organization, absenteeism is a relevant concept.
Of course, absenteeism remains relevant for the millions of workers who are scheduled to report to a central location, such as a factory, an office, a retail store, or a call center. In fact, even those who can work from home in a call center, like Jet Blue’s airline reservations agents, have to be at home and on the phone at certain times to make the scheduling work. More broadly, the growing importance of location-specific or time-specific customer-service operations, such as the millions of employees who are engaged in repairs (of cars, appliances, or plumbing systems) or delivery (of pizzas, newspapers, or mail), makes employee absence a very real and potent issue for many organizations.
At the outset, let us be clear about what this chapter is and is not. It is not a detailed literature review of the causes of absenteeism, such as the characteristics of jobs,[2] gender, age, depression, smoking, heavy drinking, drug abuse, or lack of exercise.[3] Nor is it a thorough treatment of the noneconomic consequences of absenteeism, such as the effects on the individual absentee, coworkers, managers, the organization, the union, or the family. Instead, the primary focus in this chapter is on the economic consequences of absenteeism, and on methods for managing absenteeism and sick-leave abuse—in work settings where those concepts remain relevant and meaningful.
The logic of absenteeism begins by identifying its causes and consequences. To provide some perspective on the issue, we begin our next section by citing some overall direct costs and data that show the incidence of employee absenteeism in the United States and Europe. Then we focus more specifically on causes and consequences, and present a high-level logic diagram that may serve as a “mental map” for decision makers to help them understand the logic of employee absenteeism.
How much does unscheduled employee absenteeism cost? In U.S. workplaces in 2005, the direct cost (excluding opportunity costs) was about $660 per employee per year.[4] For every 100 employees, that’s $66,000, and it may cost some large employers more than $1 million per year. In the United Kingdom, absences are also costly, as the following figures demonstrate:[5]
Across all companies, £11.6 billion ($20.18 billion) was paid out to staff who were absent and to other employees to cover for absent staff.
Average cost of sickness was £476 per employee ($828).
Average of 7.1 days taken in sickness per worker each year.
Figures for average days off are higher in the public sector (10.1) than the private sector (6.7).
The total days lost due to absenteeism each year in the United Kingdom are 40 million.
Studies have found that the average employee in the United States misses 2.3 percent of scheduled work time, or an average of 5.5 unscheduled absences per year (slightly more than one workweek).[6] In comparison, the average European employee misses an average of 10 days per year, or two workweeks.[7] Despite its cost, consider that Norwegian workers are absent, on average, almost 3 months of the year, compared to 4.2 weeks, on average, for Swedish workers, 1.8 weeks for Italian workers, and 1.5 weeks for Portuguese workers.[8]
In the United Kingdom, the reasons given for absence are widespread, but generally fall into one of five categories.[9] By far the most popular (cited by 93 percent of employees) is a cold or flu. The other popular reasons are stomach upset and food poisoning, headaches and migraines, back problems, and stress. In reality, according to U.K. health-care consultants IHC, at least half of all workplace absence has absolutely nothing to do with health. Instead, people decide to stay away from work for a variety of personal or domestic reasons, such as responsibility for child or elder care, demotivating jobs, low pay, bullying in the workplace, or excessive drinking the night before.[10]
In the United States, the leading cause of absenteeism is personal illness (35 percent), while 65 percent of absences are due to other reasons.[11] The five most common causes cited by employees for being absent are shown in Figure 3-1.
The decision to invest in reducing absence requires a consideration of the payoff. What consequences of absence will be avoided? We’ve noted that absence occurs only in jobs where employees are required to be at work at specified times. So, the consequences of absence directly relate to the fact that an employee is unavailable to work as scheduled. Absence is more “pivotal” (changes in absence affect economic and strategic success more), when the situation has these characteristics:
Others have to perform the work of the absent employee.
A process must be stopped because of the absence of an employee.
Activities must occur at a certain time, and are delayed or missed because an employee is absent.
At a general level, four categories of costs are associated with employee absenteeism. We elaborate each of these categories more fully in the sections that follow. For the moment, let us describe these categories as follows:
Costs associated with absentees themselves (employee benefits, and, if they are paid, wages, too)
Costs associated with managing absenteeism problems (costs associated with supervisors’ time spent dealing with operational issues caused by the failure of one or more employees to come to work)
The costs of substitute employees (for example, costs of overtime to other employees or costs of temporary help)
The costs of reduced quantity or quality of work outputs (for example, costs of machine downtime, reduced productivity of replacement workers, increased scrap and reworks, poor customer service)
In computing these costs, especially the costs of managing absenteeism problems, and revenues foregone, researchers commonly use the fully loaded cost of wages and benefits as a proxy for the value of employees’ time. However, as we cautioned in Chapter 2, “Analytical Foundations of HR Measurement,” although this is very common, keep in mind that it is only an approximation, and the assumption that total pay equals the value of employee time is also not generally valid.
Figure 3-2 presents an illustration of the ideas we have examined thus far.
In the context of absenteeism, analytics refers to formulas (for instance, those for absence rate, total pay, supervisory time) and to comparisons to industry averages and adjustments for seasonality. Analytics also includes various methodologies used to identify the causes of absenteeism, and to estimate variation in absenteeism across different segments of employees or situations. Such methodologies might comprise surveys, interviews with employees and supervisors, and regression analyses.
Measures, on the other hand, focus on specific numbers (for example, finding employee pay-and-benefit numbers, time sampling to determine the lost time associated with managing absenteeism problems, using the pay and benefits of supervisors as a proxy for the value of their time). Keep these important distinctions in mind as you work through the approach to costing employee absenteeism that is presented next, even though we present both measures and analytics together here because they are so closely intertwined.
At the outset, it is important to note an important irony—namely, that even in organizations or business units where the concept of absence is relevant, the incidence, and therefore the cost of employee absenteeism, is likely to vary considerably across departments or business units (and across times of the year). With respect to seasonal variations in absenteeism rates, for example, surveys by the Bureau of National Affairs (BNA) in the United States have shown over many years that the incidence of employee absenteeism is generally higher in the winter months than it is in the summer months.[12] The costs of absenteeism are therefore likely to covary with seasonal trends, yet it is paradoxical that such costs are typically reported only as averages.
With respect to the cost of employee absenteeism, the following procedure estimates that cost for a one-year period, although the procedure can be used just as easily to estimate these costs over shorter or longer periods as necessary.[13]
Much of the information required should not be too time-consuming to gather if an organization regularly computes labor-cost data and traditional absence statistics. For example, absenteeism rate is generally based on workdays or work hours, as follows:
Absenteeism rate = [Absence days / Average work force size] × working days, or |
Absenteeism rate = [Hours missed / Average work force size] × working hours |
In either case, getting the right data will involve discussions with both staff and management representatives. Figure 3-3 shows the overall approach.
To illustrate this approach, we provide examples to accompany each step. The examples use the hypothetical firm Presto Electric, a medium-sized manufacturer of electrical components employing 3,000 people.
Determine the organization’s total employee-hours lost to absenteeism for the period for all employees—blue collar, clerical, and management and professional—for whom the concept of absenteeism is relevant and for those whose jobs are pivotal to the overall success of the organization. Include both whole-day and part-day absences, and time lost for all reasons except organizationally sanctioned time off, such as vacations, holidays, or official “bad weather” days. For example, absences for the following reasons should be included: illness, accidents, funerals, emergencies, or doctor appointments (whether excused or unexcused).
As a basis for comparisons, Figure 3-4 illustrates monthly job absence rates as reported by the BNA. Although the time-series data shown in Figure 3-4 clearly reflect a seasonal component (higher rates in January and February, lower rates in June, July, and August), average monthly rates by year do not indicate a long-term trend, at least for the years shown.
In our example, assume that Presto Electric’s employee records show 102,900 total employee-hours lost to absenteeism for all reasons except vacations and holidays during the last year. This figure represents an absence rate of 1.75 percent of scheduled work time, about average for manufacturing firms. Begin by distinguishing hours scheduled from hours paid. Most firms pay for 2,080 hours per year per employee (40 hours per week × 52 weeks). However, employees generally receive paid vacations and holidays, too, time for which they are not scheduled to be at work. If we assume two weeks vacation time per employee (40 hours × 2), plus 5 holidays (40 hours), annual hours of scheduled work time per employee are 2,080 – 80 – 40 = 1,960.
The total scheduled work time for Presto Electric’s 3,000 employees is therefore 3,000 × 1,960 = 5,880,000. Given a 1.75 percent rate of annual absenteeism, total scheduled work hours lost annually to employee absenteeism are 102,900.
Compute the weighted average hourly wage/salary level for the various occupational groups that claimed absenteeism during the period. If absent workers are not paid, skip this step and go directly to step 3.
For Presto Electric, assume that about 60 percent of all absentees are blue collar, 30 percent clerical, and 10 percent management and professional. These are broad categories, but if your organization has a computerized absence-reporting system, it can compute the exact average hourly wage/salary level for absentees. For purposes of illustration, we will also assume that all employees are paid for sick days taken under the organization’s employee benefits program. The average hourly wage rate per absentee is estimated by applying the appropriate percentages to the average hourly wage rate for each major occupational group. Table 3-1 does just that.
Estimate the cost of employee benefits per hour per employee. The cost of employee benefits (profit sharing, pensions, health and life insurance, paid vacations and holidays, and so on) currently accounts for about 40 percent of total compensation.[14] One procedure for computing the cost of employee benefits per hour per employee is to divide the total cost of benefits per employee per week by the number of hours worked per week.
First, compute Presto’s weekly cost of benefits per employee. Assume that the average annual salary per employee is $23.29 per hour × 2,080 (hours paid for per year), or $48,443.20. Let us further assume the following:
Average annual salary × 40 percent = Average cost of benefits per employee per year |
$48,443.20 × 0.40 = $19,377.28 |
Average cost of benefits per year per employee / 52 weeks per year = Average weekly cost of benefits per employee |
$19,377.28 / 52 = $372.64 |
Average weekly cost of benefits per employee / hours worked per week = Cost of benefits per hour per employee |
$372.64 / 40 = $9.32 |
Compute the total compensation lost per hour per absent employee. This figure is determined simply by adding the weighted average hourly wage / salary per employee (item 2 in Figure 3-3) to the cost of employee benefits per hour per employee (item 3 in Figure 3-3). Thus:
$23.29 + $9.32 = $32.61 |
Of course, if absent workers are not paid, item 4 in Figure 3-3 is the same as item 3.
Compute the total compensation lost to absent employees. Total compensation lost, aggregated over all employee-hours lost, is determined simply by multiplying item 1 by item 4.a or 4.b, whichever is applicable. In our example:
102,900 × $32.61 = $3,355,569.00 |
Estimate the total number of supervisory hours lost to employee absenteeism for the period. Unfortunately, because existing records seldom provide the information necessary to compute this figure, it will be more difficult to estimate than wage and benefits estimates. As a first step, estimate the average number of supervisory hours spent per day dealing with all the problems stemming from employee absenteeism, matters such as production problems, instructing replacement employees, checking on the performance of replacements, and counseling and disciplining absentees.
Such an estimate can be accurate only if the staff member making it talks to the first-line supervisors and higher-level managers who deal directly with employee-absence problems. Interview a representative sample of supervisors using a semistructured interview format to help them refine their estimates. Areas to probe include the effects of typically high-absence days (Mondays, Fridays, days before and after holidays, days after payday). No published data, no industrywide averages, are available to determine whether these estimates are reasonable. However, it is true of estimates in general that the more experience companies accumulate in making the estimates, the more accurate the estimates become.[15]
Methodologically, it is difficult to develop an accurate estimate of the amount of time per day that supervisors spend, on average, dealing with problems of absenteeism. That time is most likely not constant from day to day or from one month to the next. In fact, the time per day, on average, that supervisors spend managing absenteeism problems is likely to vary considerably across departments or business units. Although structured interviews are quite common, diary keeping may actually be more effective. Time sampling for diary-keeping purposes is particularly important, for as we noted earlier, absenteeism may vary across shifts of workers (time of day) and by day of the week. These are by no means the only methods available, and others might also prove useful. Careful consideration of these issues when costing employee absenteeism will yield measurably more accurate results.
After you have estimated the average number of supervisory hours spent per day dealing with employee-absenteeism problems, compute the total number of supervisory hours lost to the organization by multiplying three figures:
Estimated average number of hours lost per supervisor per day
Total number of supervisors who deal with problems of absenteeism
The number of working days for the period (including all shifts and weekend work)
In our example, assume that Presto Electric’s data in these three areas is as follows:
Estimated number of supervisory hours lost per day: 0.5 hours
Total number of supervisors who deal with absence problems: 100
Total number of working days for the year: 245
Based on these data, the total number of supervisory hours lost to employee absenteeism is as follows:
0.5 × 100 × 245 = 12,250 |
Compute the average hourly wage rate for supervisors, including benefits. Be sure to include only the salaries of supervisors who normally deal with problems of employee absenteeism. Typically, first-line supervisors in the production and clerical areas bear the brunt of absenteeism problems. Estimate Presto Electric’s cost for this figure as follows:
Average hourly supervisory salary: | $28.90 |
Cost of benefits per hour (40 percent of hourly salary): | + 11.56 |
Total compensation per hour per supervisor: | $40.46 |
Compute total supervisory salaries lost to problems of managing absenteeism. This figure is derived simply by multiplying total supervisory hours lost on employee absenteeism (step 6) times the average hourly supervisory wage (step 7), as follows:
12,250 × 40.46 = $495,635.00 |
If an organization chooses to replace workers who are absent, the key considerations are how many substitute employees will it hire and at what cost? Sometimes, the total cost is a combination of these two elements, as when some additional workers are hired to replace absentees, say from an agency that supplies temporary workers, and other, regular workers are asked to work overtime to fill in for the absentees. Alternatively, a very large organization, such as an automobile-assembly plant, might actually retain a regular labor pool that it can draw on to fill in for absent workers. At Presto Electric, let’s assume that the firm incurs total costs of $350,000 per year for substitute employees.
When fully productive, regularly scheduled employees are absent, chances are good that their work either is not done or, if it is, that there is a reduction in the quantity or quality of the work. The key considerations in this case are how much of a reduction is there in the quantity or quality of work and how much does it cost? Such costs might include items such as the following:
Machine downtime
Increases in defects, scrap, and reworks
Production losses
Here is an example. Suppose a small organization that is operating at full capacity has 100 salespeople in the field calling on accounts and soliciting orders every day. If the typical salesperson generates, on average, $1,000 worth of orders per day, and 10 salespeople are absent on a given day, the business lost to the organization (revenue foregone) due to employee absenteeism on that single day is $10,000.
The standard level of quality or quantity of work might also be compromised through the reduced productivity and performance of less-experienced replacement workers, as when customers are served poorly by employees who are stretched trying to “cover” for their absent coworkers, and potential new business is lost as a result of operating “under capacity.”[16]
As in step 6, some of these estimates will be difficult because many of the components are not reported routinely in accounting or HR information systems. Initially therefore, determination of the cost elements to be included in this category, plus estimates of their magnitude, should be based on discussions with a number of supervisors and managers. Over time, as the organization accumulates experience in costing absenteeism, it can make a more precise identification and computation of the costs to be included in this category. At Presto Electric, assume that production losses and inefficient materials usage as a result of absenteeism caused an estimated financial loss of $300,000 for the year.
Compute the total estimated cost of employee absenteeism. Having computed or estimated all the necessary cost items, we now can determine the total annual cost of employee absenteeism to Presto Electric. Just add the individual costs pertaining to wages and salaries, benefits, supervisory salaries, substitute employees, and the costs of reduced quantity and quality (items 5, 8, 9, and 10). As Figure 3.4 demonstrates, this cost is more than $4.5 million per year.
Compute the total estimated cost of absenteeism per employee per year. In some cases, this figure (derived by dividing the total estimated cost by the total number of employees) may be more meaningful than the total cost estimate because it is easier to grasp. In the case of our hypothetical firm, Presto Electric, this figure was $1,500.40 per year for each of the 3,000 employees on the payroll.
102,900 | |
2. Weighted average wage/salary per hour per absent employee | $23.29 |
3. Cost of employee benefits per hour per absent employee | $9.32 |
4. Total compensation lost per hour per absent employee
| $32.61 |
5. Total compensation lost to absent employees (Total employee-hours lost × 4.a or 4.b, whichever applies) | $3,355,569.00 |
6. Total supervisory hours lost on employee absenteeism | 12,250 |
7. Average hourly supervisory wage, including benefits | $40.46 |
8. Total supervisory salaries lost to managing problems of absenteeism (Hours lost × Average hourly supervisory wage; Item 6 × Item 7) | $495,635.00 |
9. Costs of substitute employees | $350,000.00 |
10. Costs of reduced quantity and quality of work | $300,000.00 |
11. Total estimated cost of absenteeism (items 5, 8, 9, 10) | $4,501,204.00 |
12. Total estimated cost of absenteeism per employee (Total estimated costs / Total number of employees) | $1,500.40 |
As noted in Chapter 2, the purpose of the process component of the logic, analytics, measurements, and process (LAMP) model is to make the insights gained as a result of costing employee absenteeism actionable. The first step in doing that is to interpret absenteeism costs in a meaningful manner. To do so, begin by evaluating them—at least initially—against some predetermined cost standard or financial measure of performance, such as an industrywide average. This is basically the same rationale organizations use when conducting pay surveys to determine whether their salaries and benefits are competitive. Unfortunately, absenteeism-cost data are not published regularly, as are pay surveys. Very little data are available to help determine whether the economic cost of employee absenteeism is a significant problem. The costs of absenteeism to individual organizations occasionally do appear in the literature, but these estimates are typically case studies of individual firms or survey data from a broad cross-section of firms and industries, rather than survey data from specific industries.
Is it worth the effort to analyze the costs of absenteeism to the overall organization, and more specifically, to strategically critical business units or departments where the concept of absenteeism is relevant? The answer is yes for at least two compelling reasons. First, such an analysis calls management’s attention to the severity of the problem. Translating behavior into economic terms enables managers to grasp the burdens employee absenteeism imposes, particularly in strategically critical business units that are suffering from severe absence problems. A six- or seven-figure cost is often the spark needed for management to make a concerted effort to combat the problem. Second, an analysis of the problem creates a baseline for evaluating the effectiveness of absence-control programs. Comparing the quarterly, semiannual, and annual costs of absenteeism across strategically critical business units or departments provides a measure of the success, or lack of success, of attempts to reduce the problem.
If we return to the logical elements of absence cost, we can consider the process you can use to relate those costs to ongoing budget and strategy issues in an organization:
Cost of payments for nonwork time of absentees. Signals that this is an issue to be found in performance-management and budgeting processes. To connect absence to tangible process issues for business leaders, look for evidence that levels of paid time off are higher than standard, or benchmarks. Managers and other leaders will often signal their interest in reducing the costs paid for nonwork time by noting that sick leave or unscheduled vacation days are higher than they expect. This is an opportunity to take the logic noted above and suggest how much this might change if absence changed.
Cost of payments for time of those who manage absence. The process signals here will be when the number of supervisors is excessive, or where supervisors note that they are spending a great deal of time on “nonproductive workforce-management” issues. To connect this to the budgeting or management process, it may be useful to look at supervisor or manager goal-setting or performance-management conversations. Are statements like these common when setting goals or evaluating performance issues? Do supervisors and managers often suggest that they could be more effective if they spent less time managing around absent employees? What would they be doing if they did not have to manage employee absence? Answers to these questions allow you to connect absence reductions to tangible changes in supervisor behavior.
Cost of time of replacement workers. Signals that this is an important cost element that will emerge when business units see the total labor costs or headcount levels as higher than other similar units or benchmarks. Leaders may complain that they often don’t have enough work for all of their employees, but that they must keep the extra employees around to fill in. From a process standpoint, you can use the logic we have described to engage in a discussion about just how much pay for lost time would be reduced if some of the extra employees could be diverted elsewhere or even removed from the work force.
Cost of reduced work quantity or quality. The signals here will likely not be found in headcount numbers or labor-cost numbers. Instead, the process for unearthing this evidence will require looking at the performance numbers for operations themselves. Managers and executives may be wrestling with stubborn problems of fluctuations in work output or quality. That will be a first clue to possible absence-related productivity issues. Alternatively, they might note very specific connections between the fact that when a particular worker fails to be at work, there are specific things that don’t get done, customers that don’t get served, or teams that have to operate with less-than-full contributions. You can take these examples, and use the logic above to determine how much of the problem is due to absence, and how much investing in absence reduction might change them.
In the next section, we present a case study that moves beyond the calculation of absenteeism costs to illustrate how awareness of those costs led a health-care clinic to address a critical operations issue.
A large, multispecialty health-care clinic was experiencing high absence rates among employees with direct patient-care responsibilities. In terms of costs, the absenteeism problem was impacting the satisfaction of patients with the care they received (and influencing their perceptions of quality). No wonder; fully 25 percent of patient-care work went undone, and 67 percent of non-patient-care work went undone. Remaining workers suffered from burnout and strained relationships with their supervisors. Of course, employee absenteeism was only one of several possible causes of these problems. Focusing only on reducing absenteeism per se might not address important, underlying employee-relations issues.
With the help of a consultant, the clinic sought to identify the root causes of employee absenteeism for the segment of the work force that had direct patient-care responsibilities. It found that a majority of the absentees were parents who had young children. In many cases, those parents were unable to find emergency or sick-child care, and this caused last-minute staffing shortages due to unscheduled absences. Moreover, the Family Medical and Leave Act permits employees to use their own sick time to care for ill children (and requires employers to grant employees up to 12 weeks of unpaid annual leave).[17]
Based on this information, management of the clinic made the decision to provide sick-child care and backup childcare facilities both for patients, when using the clinic, and for employees to use in emergencies. Doing so yielded payoffs in attraction and in retention of members of this critical segment of the clinic’s work force. One year later, the unscheduled absence rate for employees using the backup childcare facility was 70 percent less than that of employees who were eligible but did not use the facility.[18]
This finding was certainly good news in terms of the overall employee-absence rate, but it suggests the need for further diagnostic information to uncover reasons why employees who were eligible to use the sick-child and backup childcare facilities chose not to do so. That is the nature of HR research: Addressing one problem (in this case, excessive employee absenteeism) helps to identify additional ones that require management attention.
In the final part of this chapter, we present two other approaches to managing absenteeism and sick-leave abuse that may prove useful, depending on the diagnosis of root causes. These include positive incentives and paid time-off policies. We hasten to add, however, that organizationwide absenteeism-control methods (for example, rewards for good attendance, progressive discipline for absenteeism, daily attendance records) may be somewhat successful, but they might not be effective in dealing with specific individuals or work groups that have excessively high absenteeism rates. Special methods (such as flexible work schedules, job redesign, and improved safety measures) may be necessary for them. It is the careful analysis of detailed absenteeism-research data that can facilitate the identification of these problems and suggest possible remedies.[19]
This approach focuses exclusively on rewards; that is, it provides incentives for employees to come to work. This “positive-incentive absence-control program” was evaluated over a five-year period: one year before and one year after a three-year incentive program.[20]
A 3,000-employee nonprofit hospital provided the setting for the study. The experimental group contained 164 employees who received the positive-incentive program, and the control group contained 136 employees who did not receive the program. According to the terms of the hospital’s sick-leave program, employees could take up to 96 hours—12 days per year—with pay. Under the positive-incentive program, employees could convert up to 24 hours of unused sick leave into additional pay or vacation. To determine the amount of incentive, the number of hours absent was subtracted from 24. For example, 24 minus 8 hours absent equals 16 hours of additional pay or vacation. The hospital informed eligible employees both verbally and in writing.
During the year prior to the installation of the positive-incentive program, absence levels for the experimental and control groups did not differ significantly. During the three years in which the program was operative, the experimental group consistently was absent less frequently, and this difference persisted during the year following the termination of the incentives. The following variables were not related to absence: age, marital status, education, job grade, tenure, or number of hours absent two or three years previously. Two variables were related to absence, although not as strongly as the incentive program itself: gender (women were absent more than men) and number of hours absent during the previous year.
Had the incentive program been expanded to include all 3,000 hospital employees, net savings were estimated at $42,000 per year. This is an underestimate, however, because indirect costs were not included. Indirect costs include such things as the following:
Overtime pay
Increased supervisory time for managing absenteeism problems
Costs of temporary labor
Intentional overstaffing to compensate for anticipated absences
Cautions: A positive-incentive program may have no effect on employees who view sick leave as an earned “right” that should be used whether one is sick or not. Moreover, encouraging attendance when a person has a legitimate reason for being absent—for example, hospital employees with contagious illnesses—may be dysfunctional.
In and of itself, absence may simply represent one of many possible symptoms of job dissatisfaction. Attendance incentives may result in “symptom substitution,” whereby declining absence is accompanied by increased tardiness and idling, decreased productivity, and even turnover. If this is the case, an organization needs to consider more comprehensive interventions that are based, for example, on the results of multiple research methods such as employee focus groups, targeted attitude surveys, and thorough analysis and discussion of the implications of the findings from these methods.
Despite the potential limitations, the study warranted the following conclusions (all monetary figures are expressed in 2006 dollars):
Absenteeism declined an average of 11.5 hours per employee (32 percent) during the incentive period.
Net dollar values to the organization (direct costs only) are based on wage costs of $26.60 per hour (composed of $20.46 in direct wages plus 30 percent more in benefits).
Savings were $50,168 per year (11.5 hours × Average hourly wage [$20.46] × 164 employees).
Direct costs to the hospital included 2,194 bonus hours, at an average hourly wage of $20.46 per hour = $44,889.
Net savings were therefore $5,279 per year, for an 11.76 percent return on investment ($5,279 / $44,889).
This approach to controlling absenteeism and the abuse of sick leave is based on the concept of consolidated annual leave. Sick days, vacation time, and holidays are consolidated into one “bank” to be drawn out at the employee’s discretion. Employees manage their own sick and vacation time, and are free to take a day off without having to offer an explanation. If the employee uses up all of this time before the end of the year and needs a day off, that time is unpaid. What about unused sick time? “Buy-back programs” allow employees to convert unused time to vacation or to accrue time and be paid for a portion of it.
Employers that have instituted this kind of policy feel that it is a “win-win” situation for employees and managers. It eliminates the need for lying by employees (that is, abuse of sick leave), and it takes managers out of the role of enforcers. Employees typically view sick leave days as a right—that is, “use them or lose them.” Paid time-off (PTO) policies provide an incentive to employees not to take unnecessary time off because excessive absence is still cause for dismissal. PTO continues to rise in popularity, from 59 percent of employers using them in 2003 to 67 percent in 2005. Fully 58 percent of employers combined PTO with buy-back programs in 2005. Together, employers rated them as the most effective absence-control programs.[21]
A comprehensive review of research findings in this area revealed that absence-control systems can neutralize some forms of absence behavior and catalyze others.[22] Although the positive-incentive program described earlier was effective in reducing absenteeism over a three-year period, one study showed that absence-control policies could actually encourage absence.[23] In the firm studied, employees had to accumulate 90 days of unused sick leave before they could take advantage of paid sick leave (for one- to two-day absences). The policy suppressed absences only until employees reached the paid threshold, at which time they took sick leave ferociously.
Other studies have shown that punishments, or stricter enforcement of penalties for one type of absence, tend to instigate other forms of missing work.[24] This is not to suggest, however, that absence-control policies should be lenient. Unionized settings, where sick-leave policies are typically more generous, are clearly prone to higher absenteeism.[25] Such policies convey a relaxed norm about absenteeism, and research evidence clearly indicates that those norms can promote absence taking.[26]
Slack productivity from ailing workers is sometimes called “presenteeism.”[27] Like absenteeism, presenteeism is a form of withdrawal behavior. It often results from employees showing up but working at subpar levels due to chronic ailments.[28] This is not a new category of costs, but rather an illustration of our fourth cost category, namely, the costs of reduced quantity or quality of work. For example, the estimated productivity loss associated with depression is 6.4 percent. It is 4.9 percent for arthritis, and 3.2 percent for obesity. In the aggregate, economists estimate that presenteeism costs U.S. businesses $180 billion annually, or an average of $255 per year in reduced productivity for each “presentee.” Surprisingly, it may actually be a much costlier problem than its productivity-reducing counterpart, absenteeism. Unlike absenteeism, however, presenteeism isn’t always apparent. Absenteeism is obvious when someone does not show up for work, but presenteeism is far less obvious when illness or a medical condition is hindering someone’s work. Researchers are just beginning to address presenteeism and to estimate its economic effects.
Logic. Research on presenteeism focuses on such chronic or episodic ailments as seasonal allergies, asthma, migraines, back pain, arthritis, gastrointestinal disorders, and depression.[29] Progressive diseases, such as heart disease and cancer, tend to occur later and life and tend to generate the majority of direct health-related costs for companies. In contrast, the illnesses people take with them to work account for far lower direct costs, but they imply a greater loss in productivity because they are so prevalent, so often go untreated, and typically occur during peak working years. Those indirect costs have largely been invisible to employers.[30]
Analytics. To be sure, methodological problems plague current research in this area. For example, how does one quantify the productivity loss associated with chronic ailments? Some researchers have used detailed questionnaires that include self-reports of perceived productivity loss. Others ask respondents how much productive work time they think they have lost because of various medical problems. Different research methods have yielded quite different estimates of the on-the-job productivity loss—from less than 20 percent of a company’s total health-related costs to more than 60 percent.[31] Without a standard tool or research method, it is difficult to develop a common understanding of what “productivity loss” actually means. Loss of productive work time and impact on company health-care costs are very different outcomes.
Beyond that, how does one quantify the relative effects of individual ailments on productivity for workers who suffer from more than one problem? The effects of such interactions have not been addressed. Nor has the effect on team performance been studied in cases when one member has a chronic health condition that precludes him or her from contributing fully to the team’s mission.
Measures. A key question to address is the link between self-reported presenteeism and actual productivity loss. If that link cannot be established, it will be almost impossible to convince skeptics that the phenomenon is real and that it deserves management’s attention. As an example, consider that some of the strongest evidence of such a link comes from several studies involving credit card call-center employees at Bank One, which is now part of J. P. Morgan Chase.[32]
There are a number of objective measures of a service representative’s productivity, including the amount of time spent on each call, the amount of time between calls (when the employee is doing paperwork), and the amount of time the person is logged off the system. The study focused on employees with known illnesses (identified from earlier disability claims) and lower productivity scores. One such study, a good example of analytics in action, involved 630 service representatives at a Bank One call center in Illinois. Allergy-related presenteeism was measured with such objective data as the amount of time workers spent on each call. During the peak ragweed pollen season, the allergy sufferers’ productivity fell 7 percent below that of coworkers without allergies. Outside of allergy season, the productivity of the two groups was approximately equal.
Process. The next step, of course, is to use this information to work with decision makers to identify where investments to reduce the costs of presenteeism offer the greatest opportunities to advance organizational objectives. One way to improve productivity is by educating workers about the nature of the conditions that afflict them, and about appropriate medications to treat those conditions. Companies such as Comerica Bank, Dow Chemical, and J. P. Morgan Chase are among those that have put programs in place to help employees avoid or treat some seemingly smaller health conditions, or at least to keep productive in spite of them.[33] To ensure employee privacy, for example, Comerica Bank used a third party to survey its employees and found that about 40 percent of them said they suffered from irritable bowel syndrome (IBS), which can involve abdominal discomfort, bloating, or diarrhea. Extrapolating from that, the company estimated its annual cost of lost productivity to be at least $8 million a year. Comerica now provides written materials for its employees about IBS and has sponsored physician seminars to educate workers about how to recognize and deal with it through their living habits, diet, and possible medications.
Education is one thing, but getting workers to take the drugs that their doctors prescribe or recommend is another. The Bank One study found that nearly one quarter of allergy sufferers did not take any kind of allergy medication. The same study also concluded that covering the cost of non-sedating antihistamines for allergy sufferers (roughly $18 a week for prescription medications, less for generics) was more than offset by the resulting gains in productivity (roughly $36 a week, based on call-center employees’ wages and benefits, which averaged $520 a week).[34]
These results raise a tantalizing question, namely, might a company’s pharmacy costs actually be an investment in work force productivity? Certainly companies should monitor and control corporate health-care expenditures. It is possible, however, that by increasing company payments for medications to treat chronic diseases that the companies might actually realize a net gain in work force productivity and eliminate the opportunity costs of failing to address the presenteeism issue directly. One obvious example of this is the flu shot. Numerous studies have shown that the cost of offering free flu shots is far outweighed by the savings realized through reductions in both absenteeism and presenteeism.[35] Another simple approach to reducing presenteeism is to offer paid time off, as discussed earlier. This might be an appropriate strategy, because in the United States about half the work force gets no paid sick days whatsoever.[36] Implementing even a modest program of sick leave may well offset the reduced productivity associated with chronic presenteeism.
Software that calculates answers to one or more of the following exercises can be found at www.shrm.org/publications/books.
1. | Consolidated industries, an 1,800-employee firm, is faced with a serious, and growing, absenteeism problem. Last year, total employee-hours lost to absenteeism came to 119,808. Of the total employees absent, 65 percent were blue collar (average wage of $23.15 per hour), 25 percent were clerical (average wage of $17.80 per hour), and the remainder were management and professional (average salary $34.60 per hour). On the average, the firm spends 38 percent more of each employee’s salary on benefits, and, as company policy, pays workers even if they are absent. The 45 supervisors (average salary of $26.35 per hour) involved in employee absenteeism problems estimate they lose 25 minutes per day for each of the 245 days per work year just dealing with the extra problems imposed by those who fail to show up for work. Finally, the company estimates it loses $529,500 in additional overtime premiums, in extra help that must be hired, and in lost productivity from the more highly skilled absentees. As HR director for Consolidated Industries your job is to estimate the cost of employee absenteeism so that management can better understand the dimensions of the problem. (Use the worksheet provided to record your answers.) |
2. | Inter-Capital Limited is a 500-employee firm faced with a 3.7 percent annual absenteeism rate over the 1,960 hours that each employee is scheduled to work. About 15 percent of absentees are blue collar (average wage $24.96 per hour), 55 percent are clerical employees (average wage $18.25 per hour), and the remainder are management and professional workers (average salary $42.50 per hour). About 40 percent more of each employee’s salary is spent on benefits, but employees are not paid if they are absent from work. In the last six months, supervisors (average salary of $27.55 per hour) estimate that managing absenteeism problems costs them about an hour a day for each of the 245 days per work year. It’s a serious problem that must be dealt with, since about 20 supervisors are directly involved with absenteeism. On top of that, the firm spends approximately $390,000 more on costs incidental to absenteeism. Temporary help and lost productivity can really cut into profits. Just how much is absenteeism costing Inter-Capital Limited per year per employee? (Use the software available at www.shrm.org/publications/books.) |
3. | As a management consultant, you have been retained to develop two alternative programs for reducing employee absenteeism at Consolidated Industries (question 1). Write a proposal that addresses the issue in specific terms. Exactly what should the firm do? (To do this, make whatever assumptions seem reasonable.) |
| 1. | Fraser-Blunt, M. (Jan. 2007). Call centers come home. HRMagazine, 52(1), p. 85-89. See also Rouzer, P. A. (June 2000). May I help you? A more complex customer service relationship demands superior training. Retrieved from http://www.shrm.org/hrmagazine/2000index/0600/0600rouzer.asp on February 12, 2007. |
| 2. | Rentsch, J. R., & Steel, R. P. (1998). Testing the durability of job characteristics as predictors of absenteeism over a six-year period. Personnel Psychology, 51, 165-190. |
| 3. | 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. See also Johns, G. (1997). Contemporary research on absence from work: Correlates, causes, and consequences. In C. L. Cooper & L. T. Robertson (Eds.), International Review of Industrial and Organizational Psychology (Vol. 12, p. 115-173). NY: Wiley. |
| 4. | Armour, S. (Oct. 21, 2005). Companies tell sick workers to go home, take flu germs with them. USA Today, p. B3. |
| 5. | your business. (April 19, 2004). Absenteeism, lost output and bullying in the workplace. Can it be managed? Retrieved from the World Wide Web at http://www.bized.ac.uk/current/mind/2003_4/190404.htm. See also Smale, W. (March 24, 2004). The continuing cost of absenteeism. BBC News online, retrieved from http://news.bbc.co.uk/1/hi/business/3563609.stm. |
| 6. | 2005 CCH unscheduled absence survey: Costly problem of unscheduled absenteeism continues to perplex employers. (2005). Downloaded on September 28, 2007 from http://hr.cch.com/press/releases/absenteeism/default.asp. |
| 7. | On the job. (April 24, 2006). Business Week, p. 13. |
| 8. | Alvarez, L. (July 25, 2004). Oil riches spoil Norwegians’ work ethic. The Denver Post, p. 24A. |
| 9. | Smale, op. cit. |
| 10. | Ibid. |
| 11. | 2005 CCH unscheduled absence survey, op. cit. |
| 12. | BNA. (2005). Bulletin to Management. Washington, D. C.: Bureau of National Affairs. |
| 13. | This method is based upon that described by F. E. Kuzmits in “How Much Is Absenteeism Costing Your Organization?” Personnel Administrator, June 1979, 24:29-33. |
| 14. | Cascio, W. F. (2006). Managing Human Resources: Productivity, Quality of Work Life, Profits (7th ed.). Burr Ridge, IL: Irwin/McGraw-Hill. |
| 15. | Cascio, W. F., & Aguinis, H. (2005). Applied Psychology in Human Resource Management (6th ed.). Upper Saddle River, NJ: Prentice-Hall. |
| 16. | Cyboran, S. F. (2006). Absence management: Costs, causes, and cures. Workshop presented at Mountain States Employers Council, HR Best Practices Conference, Denver, CO, April 13. |
| 17. | Clark, M. M. (March 16, 2006). Managing absenteeism legally: The news isn’t all bad. Retrieved from www.shrm.org/hrnews_published/archives/CMS_016227.asp on May 9, 2006. |
| 18. | Cyboran, 2006, op. cit. |
| 19. | Miners, I. A., Moore, M. L., Champoux, J. E., & Martocchio, J. J. (1995). Time-serial substitution effects of absence control on employee time use. Human Relations, 48 (3), 307-326. |
| 20. | Schlotzhauer, D. L., & Rosse, J. G. (1985). A five-year study of a positive incentive absence control program. Personnel Psychology, 38, 575-585. |
| 21. | 2005 CCH unscheduled absence survey, op. cit. |
| 22. | Harrison & Martocchio, 1998, op. cit. |
| 23. | Dalton, D. R., & Mesch, D. J. (1991). On the extent and reduction of avoidable absenteeism: An assessment of absence policy provisions. Journal of Applied Psychology, 76, 810-817. |
| 24. | Miners et al., 1995, op. cit. |
| 25. | Drago, R, & Wooden, M. (1992). The determinants of labor absence: Economic factors and workgroup norms across countries. Industrial and Labor Relations Review, 45, 764-778. |
| 26. | Harrison & Martocchio, 1998, op. cit. |
| 27. | Goetzel, R. Z., Long, S. R., Ozminkowski, R. J., Hawkins, K., Wang, S., & Lynch, W. (2004). Health, absence, disability, and presenteeism cost estimates of certain physical and mental health conditions affecting U. S. employers. Journal of Occupational and Environmental Medicine, 46(4), 398-412. |
| 28. | Rubinstein, S. (Jan. 18, 2005). Nursing employees back to health. The Wall Street Journal, p. D5. |
| 29. | Hemp, P. (Oct. 2004). Presenteeism: At work—but out of it. Harvard Business Review, p. 1-9. |
| 30. | Ibid. |
| 31. | Goetzel et al., 2004, op. cit. |
| 32. | Hemp, 2004, op. cit. |
| 33. | Rubinstein, 2005, op. cit. |
| 34. | Hemp, 2004, op. cit. |
| 35. | Ibid. |
| 36. | Armour, 2005, op. cit. |