What Kind of Change to the Civil Service System Is Really Needed? (Part 3)

Actions are accomplished through a process. A process is a series of actions to achieve a result. A system is a collection of processes that share a common aim. Here is how government operations can improve their organizational operating methods.

One of the advantages that I’ve gained by having a parallel career as civil servant and Army Reservist is the opportunity to experience the contrast in organizational operating methods.  The military has standardized and continually improves its doctrine, training, operational planning and decision-making processes.  In contrast, the organizational operating methods used in the civil service sector vary by organization, can be personality based and more compartmentalized along functional lines.

By applying the Baldrige Framework for Performance Excellence, government organizations can improve their organizational operating methods.  The underlying strategy in the Baldrige Criteria is identifying and reducing variation.

Understanding variation requires explicit knowledge of the three principles of variation, the two causes of variation, the two types of processes, and the two types of mistakes.  The three principles of variation are:

  1. Everything varies. No two things are exactly alike, no matter whether the “things” are two people, two plants, two products, two processes, or two events.  For example, the time it takes you to commute to work every day varies. 
  2. Individual things are unpredictable.  For example, you can’t predict the exact amount of time it will take you to commute to work tomorrow.
  3. Groups of things from a constant system of causes tend to be predictable. For example, you typically can predict the range of your commuting time.  In other words, “If you always do what you always did, on average, you will usually get what you always got.”  This condition is referred to as a habit, routine, standard operating procedure (SOP) or stable process.

While working at Western Electric, Dr. Walter Shewhart was challenged with developing the most economical ways for assessing and predicting the performance of a process using numerical information.  In 1924, he developed the Statistical Process Control (SPC) chart.  The SPC chart was more than just a new tool. It represented a new way of thinking.  In 1998, Donald Wheeler proposed updating SPC related terminology. His suggestions included calling the SPC chart a process behavior chart to support a broader understanding and application of the concepts behind it.     

So, what are those concepts?  Here’s an abridged explanation: Actions are accomplished through a process. A process is a series of actions used to achieve a result. A system is a collection of processes that share a common aim. The system can determine over 90% of the outcome.  Shewhart’s SPC chart helps identify the two causes of variation in a system’s processes, the two types of processes, and the two types of mistakes.

The causes of variation fall into two categories: common and special. Common causes bring about variation that’s considered usual, ordinary, and/or expected. Common cause variation is inherent in processes. Special causes bring about variation that’s unusual, out of the ordinary, temporary, and/or unexpected. Special cause variation is attributed to an assignable cause.

Processes are either stable (in control) or unstable (out of control). A stable process contains only common cause variation. Unstable processes contain both common cause and special cause variation.

The two types of mistakes are:

  1. Treating a common cause as a special cause 
  2. Treating a special cause as a common cause 

W. Edwards Deming estimated that the lack of knowledge of common and special causes of variation resulted in situations where 95% of the actions taken to improve a system or process resulted in no improvement. Government Accountability Office (GAO) reports include many government-related examples of changes that resulted in no improvement.


Commuting Example – What You May Not Know You Know

If you have a successful routine for getting to work on time, this would likely indicate a stable (in control) or predictable process.  However, no one gets to work at exactly the same time every day.  The time varies due to the volume of traffic, number of stops, length of the stops, road conditions, departure time, etc.  This is common cause variation because the variation is usual or expected. It’s a normal part of the process.

If you were unusually late one day, it might have been due to a traffic accident, road construction or a car breakdown. This is special cause variation due to a temporary or unusual event. Your process for that day would be considered unstable (out of control) or unpredictable because it contains both common cause and special cause variation.  Typically, though, you wouldn’t change your normal commuting process because of a special cause that you don’t expect to be repeated the next day.

If you change your process (e.g., allow for more time) because of a special cause that may not recur, this would be an example of Mistake 2.  If you change your route because it has too many traffic lights that cause you delays and the changes don’t result in an improvement, this would be an example of Mistake 1.  Shewhart realized that never making Mistake 1 or Mistake 2 was impossible. His aim was to regulate the frequency of the two mistakes to achieve minimum economic loss.


What You Can Do Now — Plot Points

To better understand and help identify variation, you can develop a trend chart on a personal process, such as commuting to work.  Collect the data on a daily basis and develop a trend chart.  After the trend chart has 25 to 30 data points, calculate the average and add the average line to it.  To assess the degree of predictability in your process, apply an interpretation standard.  One of the more common standards for identifying special causes is seven consecutive data points in a row above or below the average line, seven consecutive points trending up or down, or any points that appear farther away from the average line than the others.  Once you get comfortable with plotting points and assessing the predictability of your processes, you can try developing a trend chart for a work process.

If your commuting process (or any other process) is predictable (i.e., has only common cause variation), a possible goal might be to minimize the resources it takes to sustain it. If you want to improve the process, make a change and plot points to assess if the change is resulting in improvement.

When identifying and reducing variation in a process, you need to realize that virtually no process stands alone—it’s part of a broader system.  For example, your commuting process is part of the broader transportation system. It’s interdependent on other people’s commuting processes. In addition, it’s interdependent on organizations’ processes in industries such as construction, transportation, automotive, energy, environmental, safety, communication, all levels of government, justice, law enforcement, and healthcare.  For example, if your city government adopts a goal to improve pedestrian and driver safety by installing more traffic lights, improving your commuting process becomes a little more challenging. A common understanding of the variation paradigm among the stakeholders would help to ensure that improvement in one part of the system does not make things worse in other parts.    

In summary, the answer to “What kind of change to the civil service system is really needed?” is twofold. First, government organizations need to understand the variation paradigm. Second, they need to apply the Baldrige Criteria for Performance Excellence because it’s one of the better methods for identifying and reducing variation.

As government organizations continue to be required to meet higher expectations with fewer resources, new approaches are needed.  As I mentioned in the last article, the U.S. Army Armament Research, Development and Engineering Center (ARDEC) was the first federal government organization to be recognized with the award in the nonprofit category in 2007.  The VA Cooperative Studies Program Clinical Research Pharmacy Coordinating Center was recognized in 2009.  There have been 14 nonprofit/government organizations that have applied for recognition for the 2011 Award. It’s a start.

The civil service system is part of the American system of government, so any suggested changes to the civil service system will need to be supported by the American government system. This requires all citizens to understand, identify, and support efforts to reduce variation.  In the next and final article in this series, I’ll discuss a strategy for providing this support.

Perspectives and opinions presented are solely those of the author.

About the Author

Timothy J. Clark retired from the federal government with 35 years of service. He is a former enlisted soldier in the U.S. Army. He retired at the rank of Colonel, with over 30 years of combined service in the U.S. Army, National Guard and Army Reserve. He is a strategic analyst with the American Center for Quality Leadership and is active in economic and community development in a small rural county in Indiana.