Managing Safety Dilemmas

Paul Balmert

Paul Balmert is a graduate of Cornell University’s School of Industrial and Labor Relations, and his career in chemical manufacturing spans 30 years. In 2000, Mr. Balmert formed Balmert Consulting (, a consulting practice that is principally focused on improving operations execution, including improving the management effectiveness in leading and managing safety performance. Mr. Balmert is the author of the best-selling book Alive and Well at the End of the Day; The Supervisor’s Guide to Managing Safety in Operations. He may be reached at

V. Scott Pignolet

July 1, 2006

dilemma: de-‘le-ma, noun-an argument presenting two equally conclusive alternatives

In the February issue of Insulation Outlook, we examined the Risk Conundrum, where we said we suspect that in many catastrophic events, the real risk-hazard times probability-is put in play when people do not follow processes or execute well. Failure to execute is the far greater risk and should be a key target for any risk management process. Also introduced was the Investigation Dilemma-the idea that finding out what really went wrong sounds easy until you try doing it. The struggle to assign responsibility sets up a dilemma of classic proportions: truth versus the consequences of being found at fault.

The Dilemma series concludes in this issue with the fourth installment of Managing Safety Dilemmas. We will examine two dilemmas that can send anyone rushing to the medicine cabinet for migraine relief: the System Dilemma and the Middle Dilemma.

The System Dilemma

"The system you have is perfectly designed to produce the results you are getting."
-W. Edwards Deming

It is very likely that no single individual in the twentieth century had a greater impact on product quality than Dr. W. Edwards Deming. Respected consultant to industry leaders in Japan and the United States, author of the Fourteen Absolutes of Quality, Deming was a towering presence in the world of manufacturing.

Except for his size-six feet, eight inches tall-nothing in Deming’s early career was exceptionally noticeable. He graduated with a doctorate in mathematical physics at the height of the Depression and then went to work for the Department of Agriculture before helping out with the 1940 U.S. Census. Soon, Deming turned his attention to the process of making things.

During World War II, Deming introduced statistical methods to improve the quality of war materials. His techniques made a huge impact. In the post-war era, he found a receptive audience among Japanese manufacturers eager to rebuild their factories and erase the perception that their products were inferior. Deming left a lasting mark in Japan: the Deming Prize for quality achievement.

By the 1980s, U.S. manufacturers were finally ready to listen to his ideas about improving their product quality. To a practicing statistician, as Deming was, the world of performance takes on the shape of a bell curve. The best and the worst-and everything clumped in the middle-are not all that different, in a statistical sense. They are all products of the same system.

What is that system? A system is best thought of as "the complex relationships between related components." The idea comes from the natural world (think eco-systems) recognizing that even small changes can have large effects on the world. When applied to the manufacturing world, the system is all the factors in play: raw materials, production equipment, methods and processes, people, and all that that implies.

Deming argued that if you wanted better results, you had to change the system that produced those results. It was crucial to stop blaming the people making poor quality products and start changing the entire system that produced those products. The argument carried the day. Those engaged in making things-from consumer electronics to chemicals and paint to cars and trucks-are starting applying statistical methods to their production. For example: change the process, move the mean, reduce variability, and tighten up the distribution curve. The results were nothing short of astounding. Product quality improved, as did cost and productivity, and, ultimately, profitability. Deming was a genius; his impact, profound.

Where’s the Dilemma?

With a success story like this, you are probably wondering where there could possibly be a dilemma, and what any of this has to do with managing safety performance.

But you have only heard half the story-and the better half at that. Remember, every good dilemma has two conditions that are equally valid, and totally incompatible. Here is the rest of the story. Deming was right: there are systems, and those systems are often a significant factor in determining results. Thus, the best way to change performance is to change the system. However, the actions of individuals also make up a significant factor in determining results. For starters, people are one of the components that make up the system. Unlike all the other components, people come fully equipped with the ability to choose how they act. That is not a minor difference.

In short, individuals are creatures of the system-but always willing creatures. That is what makes the System Dilemma.


Suppose you do not like the results and the behavior of people in the system. Who bears responsibility? Follow Deming’s logic and you might wind up someplace you would rather not be. If the system determines behavior, you would conclude that the system bears the responsibility. But how do you hold a system accountable? Are you willing to let the system excuse the behavior of individuals?

Here is a way to test your conclusions: suppose the behavior involves possible accounting fraud? Any of the headline cases from the business pages-such as WorldCom, HealthSouth, and Enron-serve as perfect examples of the dilemma. You know the story: accounting irregularities on a massive scale deprive shareholders of billions. The accountants plead guilty, saying that their bosses made them do it.

Would you excuse the behavior of the accountants? After all, they are just creatures of the system, right? If you are not buying that logic, you are in good company. Neither did the prosecutors, who argued successfully that accountants at these three companies knowing violated the law. You can find a couple dozen of them now serving time in the federal penitentiary. If you were a shareholder, you would probably like to see the corporate bosses sitting in the pen with them.

The Wrong Stuff

Now for the finishing stroke to the System Dilemma: its relationship to safety.

You have seen the scenario: everybody knows there are problems, but they all just go along because nobody wants to rock the boat. That’s the power of the system. Then, something really bad happens: a horrible accident, a major loss, and people get seriously hurt. History is replete with accidents that are now household names: Longford, Challenger, Bhopal. These were all situations where systems failed, and lives were lost. All the dirty laundry gets hung out in the accident investigation-and there is always a full load. Systems fail because people fail. Viewed as a past event, it is always obvious that any number of people had plenty of opportunity to prevent the accident. But they did not do it.

Why, if it’s wrong, did so many go along for the ride? As the saying goes, no snowflake in an avalanche ever feels responsible.

Sam Rayburn of the House of Representatives famously said, "If you want to get along, go along." The pressure to "go along" is enormous and often undesirable. We tell our school-age children to think for themselves and not do what everyone else is doing just to fit in with the crowd they want to impress. But what happens at work? We turn around and do exactly the opposite. It is called running with the herd. The system can be cruel to those who push from within to change it. Pick your favorite martyr: There are many to choose from.

Maybe there is a safety problem. Maybe we know something is wrong. But we act just like everyone else in the system: it is not our problem. We do not own the system and could not change things if we tried. So we do nothing.

The System Dilemma

There is the System Dilemma in full relief: a system exists that determines performance, and individuals determine their own performance. Both statements are true.

Ignore the first, and results will not change for the better. Ignoring the second is an invitation for irresponsible behavior. People do not have to go along if they truly do not want to. Deciding to falsify financial records is simply a matter of pleasing the boss, getting a bonus, and keeping the job instead of doing the right thing.

The same holds true for safety issues. Taking shortcuts, signing waivers, and ignoring the warning signs of unsafe equipment are all choices. The simple solution to the System Dilemma is this: Nobody has to make the wrong choice.

The Middle Dilemma

"He who walks in the middle of the road gets hit from both sides."
-George Shultz

Credit Bob DuBrul and Dr. Barry Oshry with inventing the term "Middle Dilemma" some twenty years ago. As systems consultants, Oshry and DuBrul had an uncomplicated way of looking at organizations. No matter what the nature of the organization, there were only three roles that mattered: tops, middles, and members.

Oshry and DuBrul’s principal interest was in the role played by those in the middle, who link members with tops. Their appreciation of the middle role came from their work with a wide range of middles: waiters; camp counselors; church pastors; and, yes, supervisors and managers in the world of industry. You can see by this list that "top" and "member" describe a wide variety of roles that are not limited to our traditional view of leaders and followers.

The model may seem simple (the best ones always are), but that does not mean it is useless-particularly as it explains the difficulties faced by those functioning in the middle of a system.

It’s Lonely in the Middle

A waiter, who links the customer and the kitchen, provides the perfect illustration of the difficulty of life in the middle. The waiter takes the order, the kitchen staff prepares the food, and the waiter serves the meal. When the food does not meet expectations, guess who bears the brunt of the criticism from the customer? It is not the chef. He is back in the kitchen, far removed from that heat.

The waiter has no control-and very little influence-over what goes on in the kitchen. The kitchen staff is completely insulated from the customer: they never have to deal with their own failures. That is a duty left to the waiter (who, by the way, is working for tips). How much of a tip do you think an unhappy customer leaves for the waiter?

You can see how DuBrul and Oshry were on to something. Middles play a vital role, but it is one that leaves them feeling powerless and all too often caught in the crossfire between the two parts of the system they connect. It is a frustration that most of us have experienced at one time or another. When it comes to managing safety performance, however, the Middle Dilemma can be more than frustrating: it can be downright dangerous.

The Middle Dilemma and Managing Safety

By now, if you play any sort of middle role in your organization, you have probably jumped ahead to the connection between Middle Dilemma and managing safety performance.

Just like our waiter, those working in the middle in the world of operations feel the brunt of the decisions and actions made by those at the top of the organization. They hear about it from the members they manage; sometimes they even get to witness firsthand what happens when things go wrong.

Middles also provide a second function that is not always entirely beneficial: they serve as a layer of insulation between members and the top. It might make life at the top a little bit easier, but it can also mean that important information gets bottled up. When that happens, the tops do not know what the middles know about what is really going on, which is not a good thing. It is not necessarily the fault of those at the top. Those in middle roles-with job titles like front line supervisor, area superintendent, and process engineer-live where the real work of the organization takes place. They are familiar with all the important details that matter about safety performance: information like the real qualifications of those performing the work, the true condition of the equipment, how well policies and procedures are being followed, and what the real performance data look like. Said another way, the middles know reality; the tops may very well not.

"You Can’t Handle the Truth."

If those at the top always acted as though they understood this and were hungry for the truth, there probably would not be a safety version of the Middle Dilemma. But tops do not always want to hear the details about organization reality. It is messy, confusing, and sometimes contradictory to what they would prefer to think is reality.

That puts the middles on the horns of the dilemma. A middle can tell management all and be branded as an alarmist and obstructionist. Or, a middle can drive reality underground and be viewed as a "can-do" type. The path of least resistance is always easier-at least until there is a serious problem. When that happens, the tops are shocked to learn what is really going on. "How could you let that situation exist?" they ask the middle. There is never a good answer to that question. Know the feeling? If you do, you’ve got plenty of company.

A famous midnight conference call in 1986 between NASA’s space shuttle management team and their rocket contractor illustrates the problem as well as any ever documented. The contractor offered the sound, engineering-based reasons why the shuttle should not be launched at a temperature lower than fifty-three degrees. Frustrated by the implications, a senior NASA official blurted out, "When do you want me to launch, next April?"

Remarks like that, coming from tops, usually have a significant effect. In this case, the contractor put on its "management hat" and decided that in this situation, perhaps the science was not quite that important, to disastrous results.

The Middle Dilemma

Now that we have the Middle Dilemma out for public inspection, what can we do about it? It is a serious dilemma-one that has proven to be fatal on more than one sad occasion. It is not enough to describe it here. We feel obligated to offer some ideas about how to manage the dilemma.

In the Middle Dilemma, most of the important action happens between the middle and the top. The two principal factors to be dealt with are the restricted flow of important information up the chain of command and the insulation of the top from organization reality. It may well take all the diplomatic skills of Henry Kissinger to manage in the middle, but it is well worth trying. Besides, what other options do you have?

1. Don’t hide the truth.

NASA recognized that a big part of its cultural problem stemmed from the fact that top management had become cut off from science and engineering. Read that as "removed from the hard facts about reality."

Let’s face the truth. As middles, much of this is a problem of our own making. Our desire to make our operation look good to those at the top causes us to act in very predictable ways: cleaning up the place before the big visit; showing the boss the newest and best part of the operation instead of the oldest and worst; and putting the best spin on the hard data about conditions and performance. Given two versions of reality, we are inclined to report the better case.

Try tilting in the direction of the center. Disclose some of the bad with the good. Air a little bit of your dirty laundry. In the short term, you might not look as good; but in the long term, you probably will be better off with the top understanding reality.

2. Present reality better.

Yale professor Edward Tufte has built a successful career teaching how to present reality better. A master of the chart, he teaches that the conventional means of communication, such as PowerPoint slides, do not present reality well. Tufte says, "?the popular PowerPoint templates (ready-made designs) usually weaken verbal and spatial reasoning, and almost always corrupt statistical analysis."

Having sat through countless technical and management presentations to the tops, it is clear that most of us middles do not explain things very well. We know reality, but all too often it gets lost in a flood of acronyms and confusing data presented in a rapid succession of PowerPoint slides. Take a lesson from those in the advertising business: keep the message simple and do not be reluctant to repeat it.

If all else fails, try communicating the old fashioned way. Talk to people. When Louis Gerstner became president of IBM, he sent a powerful message to his organization when he turned off the projector and said, "Let’s just talk about your business."

3. Remind those at the top what is really at stake.

We have all gotten into the practice of using sanitized language to describe serious problems. It is no longer an emergency, it is a "non-routine event." Things are going haywire, and we call it an "abnormal situation."

That might be a good way to quiet the hysteria and avoid offending. It also can lull us into ignoring a serious problem. The CEO at Alcoa, after hearing a report of a fatal accident involving a twenty-year-old employee, turned to his senior management team and said, "We killed him."

Sometimes a little bit of blunt language is just what is needed to inject a healthy dose of reality into the situation. Bottom line: proceed with caution. But by all means, proceed.