Taking the guesswork out of HR

Metrics service provides insights into absenteeism, resignations, turnover
By Ian Cook
|Canadian HR Reporter|Last Updated: 10/25/2011

Are we spending enough on HR? Is our absenteeism rate a problem? Are we better at filling vacancies than our competitors? All of these are important questions for HR but, for a long time, it has been hard to get clear answers — success has been largely based on faith.

The HR Metrics Service answers questions such as these and puts hard facts behind HR effectiveness. Now in its third year, with participation from coast to coast and a database of more than 160,000 employees, the service is starting to generate valuable patterns and trends that are shaping HR strategy and practice across Canada.

What makes the data particularly valuable is the common set of standards and quarterly reporting frequency. This means organizations can keep track of results on a 90-day cycle and trust they are comparing apples to apples. At the macro level, it allows for analysis of trends and patterns of behaviour from quarter to quarter as well as year to year.

Absenteeism patterns

For example, absenteeism has followed a similar pattern for the last two-and-a-half years. It is high in quarters one and four (winter and fall) and lower in quarters two and three (spring and summer). This outcome may be intuitive but, when it is represented in black and white, there is no room for doubt or debate.

A more concerning pattern with regard to absenteeism is a steady increase. Comparing the second quarter of 2011 to the second quarter of 2009, the median has increased by 25 per cent and in some quarters 35 per cent. This increase is adding to HR workloads and is a clear indicator that the HR challenges anticipated with an aging workforce and a workforce that is more inclined to move jobs regularly are no longer conceptual.

When absenteeism is plotted against average age, employers with a higher average age experience more absenteeism than those with a lower average age. These challenges bring costs to organizations through rising benefit costs associated with absences, higher overtime costs to cover the lost work hours and increased management time to handle the absences.


Absenteeism has also been linked to turnover or, more specifically, resignations, where people voluntarily leave an organization. As absenteeism has been increasing, so too has the median resignation rate. Last year saw a 53-per-cent increase in resignations compared to 2009, and 2011 looks to be on a similar trend compared to 2010.

Even as the unemployment rate in Canada started to rise again in the fall of 2010 and the spring of 2011, it did not significantly impact the median number of resignations — they increased through both time periods.

Another trend worth noting is the seasonal pattern to resignations. They tend to be lower in quarters one and four and higher in quarters two and three. Some of this can be explained by the “bonus effect” — employees waiting for a year-end bonus payout before moving on — but other factors may be at play.

Calculating HR’s value

The metrics service also has a data source that can begin providing answers to some of the more challenging HR questions, such as how much of a difference does HR make and is it worth the money invested in the function?

Relatively straightforward analysis of a couple of key measures — costs per employee and resignation rate — proves employers achieve a different outcome when they spend above a base level on HR.

When resignation rates are plotted against HR costs per employee, those that spend little on HR have significantly higher resignation rates than those that spend more.

If the costs of resignations are factored in, which average 1.5 times an individual’s salary, those that spend little on HR are worse off than those that spend in the mid-range. What they lose in replacement costs is more than they would have to spend to properly resource HR.

However, the relationship is not linear and at a point above about $600 per employee per quarter, the incremental improvements in resignation rates are very slight. Therefore, spend on HR does make a difference up to a point but more spend does not necessarily bring greater returns related to resignation rates.

This is an important marker for anyone looking to benchmark the HR spend and ensure the budget allocated will meet the expectations being set.

Looking to the future, the data set being developed through the HR Metrics Service is of significant value when it comes to academic research into HR practices. There is evidence to support the hypothesis HR practice and HR investment are correlated to improved performance outcomes, according to Joe Schmidt, assistant professor at the Edward School of Business at the University of Saskatchewan in Saskatoon, and his colleagues. By analyzing patterns in the data, they were able to find the following correlation: Organizations that have high labour costs and difficult-to-fill positions will experience lower absenteeism and a lower resignation rate when they invest more in HR.

However, the direction of causality has yet to be proven. It may be those with better performance outcomes can afford to spend more. However, Schmidt and his team are confident that as the data set increases, so too will their ability to prove definitively which aspects of HR practice make a difference.

The growth in HR measurement in the last three years has been significant. Supported by changes in technology and demand from the executive suite, HR has quickly come to grips with deriving value from data. With the continuing economic volatility and laser-like focus on effectiveness, the drive to act on evidence rather than faith is here to stay — and will continue to get better.

Ian Cook is director of the HR Metrics Service and director of research and learning at the British Columbia Human Resources Management Association (BC HRMA) in Vancouver. He can be reached at (604) 694-6938 or ijcook@bchrma.org. For more information, visit www.hrmetricsservice.org.

Add Comment

  • *
  • *
  • *
  • *