Doing HR Analytics: How To Calculate Workday Absence Rate

Introduction

There is an old saying that “80% of success is just showing up”. Regardless of who actually said it first (Woody Allen?), HR professionals know that employee absenteeism is a big deal. Sometimes people just don’t show up for work.

In some cases, the absence is scheduled, in others unscheduled but either way that lost time ultimately means lost revenue.

In today’s Doing HR Analytics post, we’ll show you some simple measures of employee absenteeism (workday absence rate) that you can implement at work today to put some numbers and “umphh” behind your intuitions and challenges (and frustrations) with employee absence and missed workdays.

Weekly? Monthly?

Before we get started, we need to decide on a time period.

In previous posts we discussed the relationship of measurement period length and noise in the data. For those new to the topic, the bottom line is that measuring more frequently decreases your signal to noise ratio…which translates into poorer data, screwy metrics, and misinformed decision making.

For this reason I generally recommend measuring HR metrics on a monthly basis. It’s a natural reporting time unit and it hits the sweet spot between timeliness and data quality.

In the calculations below we’ll use monthly measurement.

You are of course free to choose a different time period as different needs arise.

Basic Definition and Calculation

The basic definition of Absence Rate is the total number of absence days as a percentage of workdays.

Ignoring for the moment the complexities of new hires, departures, and time period, the basic calculation is the following:

which then becomes

As a reminder, we multiply by 100 at the end to convert our value to a percentage. This makes it easier to understand and communicate our result.

Example 1

Let’s suppose we have 22 work days in April (our offices are closed on the weekend).

Let’s also suppose we we are a small office with only 7 employees throughout the entire month.

Putting these two pieces of information together, we get a total of 154 employee workdays in April (7 employees X 22 workdays = 154 employee workdays). This gives us our denominator.

For the numerator on top, we simply add up the total number of workday absences. In this case, we’ll say we had a total of 5 workdays missed.

To get our Absence Rate for the month of April then, we just divide the 5 lost days by 154 employee workdays possible and then multiply the result by 100.

Beyond the Basic

The basic example should help build intuitions around the measure: divide the number of absence days by the total number of workdays possible.

In a larger organization, however, we are dealing with hundreds, thousands, or tens of thousands of employees and calculations around the number of employees at a given time can get complicated.

Fortunately, we can sidestep the messiness of figuring out exactly how many employees we had on what day by calculating an average headcount for the month.

This is a common technique in HR Analytics and should be familiar to any one who has calculated employee turnover; [see here](https://www.hranalytics101.com/turnover-why-do-we-divide-by-the-average/).

As a refresher, it’s calculated as follows:

Updated Calculation with Average Headcount

The updated Monthly Absence Rate calculation using the average headcount is the following:

Note that this updated calculation has the same structure as our basic calculation above: dividing the total number of absence days by the total number of employee workdays.

The only difference here is that we are now using the Average Headcount in our period of interest instead of counting the individual number of employees on every single day of the month. That’s easy to do with 7 employees but not with 700 or 7000.

Example 2

It’s May 1st and our April data is in!

Number of Employees on April 1st: 1007
Number of Employees on April 30th: 988
Number of Workdays in April: 22
Number of workday absences: 1024

Let’s get to work with our updated formula.

The numerator is straightforward, just counting up the number of total workday absences (here given as 1024).

For the denominator, we first need to calculate the average headcount. We started April with 1007 and end with 988, giving us an average headcount of 997.5.

Next we just plug in the number of workdays in April (22), completing our denominator.

All together, our updated calculation now looks like this:

Next Steps

With these calculations in hand, we’re ready to start measuring workday absences.

But then what? What are you supposed to actually DO with these measures.

Like most things analytics, it’s critical to provide some context around the numbers first to really know what they mean and what actions (if any) you might need to take.

As some first steps, I would suggest the following:
Step 1. Calculate your workday absences on a monthly basis for each of the last 2-3 years and then PLOT YOUR DATA. That will help you identify any long term trends up or down. It will also help you spot seasonal spikes or those due to predictable national events and holidays.
Step 2. Break down your absence rates by geographical region, department, or function to identify any meaningful differences.
Step 3. Break down your absences by day of the week to see if particular days are typically worse or better than other.

If you see something interesting, start a conversation and learn as much as you can about that employee population.

Don’t assume high absences mean bad hiring or bad management. It might be but it might also be something else like unexpected difficulties in commuting, unexpected school closures, a local round of the flu, or other issues that can be addressed without pinning blame on workers or leadership.

Above all else, look at the numbers and see if they are driving a measurable business problem. Numbers are great, but just because you can measure it doesn’t mean it’s critical.

If workday absences are creating big staffing problems at your company, then your numbers can help identify solutions. If not, use your analytic acumen to dig into another problem. I’m sure there are plenty to go around.