Paul O’Neill, former CEO of Alcoa, remains somewhat of an icon to the profession of occupational safety and health leaders, many years after his retirement. This is because O’Neill—in so many ways—pioneered safety leadership at the executive level starting in 1987, with his introductory speech to shareholders. O’Neill’s remarks that day shocked those in the audience, many of whom moved quickly to sell shares of the aluminum titan. The focus of the incoming CEO’s talk? Worker safety. O’Neill believed in the opportunity to change Alcoa by establishing zero injury work environments for the global manufacturer. The secret to this was a firm commitment to changing deeply ingrained attitudes influencing habits.

Q & A With Paul O’Neill

Q: You must have some viewpoints around government's role in regulating safety in the workplace.

A: When I was at the Treasury I got in trouble for being outside of what historically had been considered to be the portfolio of the Secretary of Treasury, and so after I had been there after about a month, I got invited, because of my previous incarnation, to come and give a speech to a conference in Washington, it was a safety conference, I think it was organized by OSHA. And so I got up and said to the Secretary of Labor that I believe if we in the United States did safety correctly, we would say to ourselves, memorialized as the law, that after maybe a year or 18 month's grace period, that any institution, public, private, or nonprofit, that didn't have a lost workday case rate under one would have their permission to operate taken away from them.

You are experts in these things so you know about the regulatory book for OSHA, so you find things in the OSHA book like how long step ladders need to be, or how short they need to be, you understand different circumstances, and you find admonitions about broken windshields in forklift trucks.

And if you got to the typical OSHA inspection, you go through the safety numbers, see the signs on the wall about hearing protection required in the right place and the right color, you can be a wonderfully compliant organization and still have a dismal safety performance.

And so I'm a great believer that because it is possible for every kind of institution to have a lost workday case rate under one, that we shouldn't tolerate something else. We ought to give people time to get there. We out to give them coaching, counseling, instruction, and all the rest of that, but at the end of the day we should not let institutions operate that violate a certain code of ethics and morality about how we care about human beings and our society.

But you should know this about me, I believe there are certain things that need to be regulated, but I'm in favor of regulations that in effect are easy to recognize in terms of performance or non-performance, easy to observe whether people are doing the right thing.

So for months I have been saying about the economic crisis that had an unfortunate beginning in the housing business, that we had to pass a law that makes it illegal for people to give or take mortgages in the United States without a 20% equity down payment component. Full stop. I don't need a legion of people to go inspect what the mortgage lenders are doing or what individuals are doing, what banks are doing, I just want a simple standard that will protect our system from the mischievous collapse that we're in the midst of now by having a clear and easy to follow rule, and the same for safety.

I suspect you would all know this, it is not possible to write enough standards and regulations to govern every aspect of human behavior that will permit you to achieve your goal. And let me say this to you, even if you could do it, no one could read it all. It's impossible. If you go into an organization and if you find they've got manuals across the shelf that tell you what to do ... nobody reads the manual, folks. You know what it's for? The lawyers like it. The lawyers like it because if there's a problem you can go look and say, we told you so.

The things we care about need to be habitual. It's like I said about breathing so that in really great places ... when a group of people are going down a stairway and there's a handrail, and you're going with the CEO, if the CEO doesn't hold the handrail, you pull on his sleeve or her sleeve and say, hold the handrail. That's what we do here. We all do it. It doesn't matter what your rank is, it's habitual.

And sometimes people will forget it is true, that human beings will forget things they know, and that's where the compatriot part of this comes in – it’s an important belief that we need to be our brother and sister's keeper. This is not about one-off Wonder Woman or Jack Armstrong. This is about individual responsibility. But when people forget, which they will sometimes, you help them remember, and they don't take it as a criticism, they thank you for saying to them buckle up your seatbelt, or hold onto the handrail, or in an office environment with that sliding drawer down there that you left out while you got called away on the phone or something, somebody could come around the corner and be blindsided and trip over it.

I mean, it's all those kinds of things that make up a zero-incident rate. It's individuals being personally accountable but all people who work with them being constantly remindful in a positive way.

Q: How long did it take you from your start of the CEO position until the employees on the floor were habitual or had the behavior of zero loss?

A: We began to see progress right away ... and I was saying when there's a new CEO there's kind of a Hawthorne effect, you all know about the Hawthorne effect where you get the result even though it's not related in any continuous way to what you say, but everybody gets the benefit of the doubt. You get the Hawthorne effect and no matter what people really think, they say, oh yeah, that's absolutely right, we really believe in that, we're going to do that, but they don't really believe it and they don't understand it.

So there is a curve, a learning curve, I believe this, and it's about all kinds of systems, health and medical care, primary and secondary education, about everything that I know about, once you get started on a decline curve of continuous improvement, you can see something between 30 and 45% year-over-year improvement even though it's a jagged line. At some point you get discouraged.

I believe looking at endpoints that are 12 months apart ... I want to look at the data everyday but you should expect to see measurable, statistically significant progress on a 12-month to 12-month basis as you go through time, and this is our experience at Alcoa. We had some periods where I would say we plateaued. It felt like we were not making progress and then we would have what would look like a catch-up function. For those of you who are statistically inclined, it was fascinating to look at this data and to see how the upper and lower statistical bounds adjusted downward, or closed overtime, so that we were actually reducing variability.

But I'll tell you another war story, and I must tell you some of these things are really gratifying when you didn't plan them and they just kind of happen. We had a huge plant in Davenport, Iowa, it's still there, that manufacturers most of the wing planks for aircraft, especially for Boeing when they were around and all the rest of that, I think it was about 1995, so I've been there eight years.

There was a Wall Street Journal reporter who wanted to write a story about the manufacturing of wing planks for aircraft around the world, so he flew to the tri-city, he rented a car, he drives to the Davenport plant which is a huge plant, 3,000 people on the Mississippi River, and when he got there it was pouring down rain. And so he turned off the car, he didn't have an umbrella so he threw the door open and he jumped out of the car. He started running and a voice yelled out to him, stop, and he was so startled he stopped and this burley guy came over with an umbrella and he said, "I don't know who you are but we really care about safety here. It's wet, the pavement's slick, you're likely to fall and hurt yourself, and we don't let that happen here at Davenport. So let me take you into the building."

It was an hourly worker who had the understanding so ingrained in him. He didn't care who this person was, he was not going to let that happen, let this person slip and fall, and I have to tell you The Wall Street Journal reporter called me and said, "I think your safety thing is working." That was a really gratifying day. When the reach of ideas has an impact like that, that is a real satisfying thing.

Now, this is, again, something that I learned overtime that we incorporated in Alcoa, a notion that we had the responsibility for every human being that ever entered our property, not just people who worked for us, which meant contractors, delivery people. This was a really important thing because we found the contractor injury rates, when we began to see and measure them and care about them, were off the wall because they had a lot of casual employees.

They didn't really respect important safety ideas about how you back a truck into a loading dock, you almost have ... I don't know whether you own your own trucking fleet or not, but you need to include in your understanding of the responsibility for human beings, everyone you come in contact with, everyone you touch, and you need to insist on that.

Once you have figured out how to avoid or prevent injury you can't let adherence to those ideas be discretionary. It's not that you need a bunch of police, you need the individuals in the organization to line up and play, when we have a better idea we will all practice it and we will practice it all the time without failure.

I'd say this is one of the great dilemmas in working in health and medical care ... I told you about people getting infections, you know why people get infections in hospitals? Lots of different reasons, but principle one is a lack of adherence to hand hygiene policy that we've known about as human beings since we learned in 1866 that doctors were killing women because they weren't doing hand hygiene between delivering babies.

I mean, it's just staggering to know that we've known it all that time ... in subsequent times, our brilliant scientists and engineers have figured out the biology of all this so it's not just a statistical speculation, it's a proven scientific fact, and yet ... I'll tell you a particular place, it's a fairly famous hospital ... I was there a few years ago and I asked them to share with me their hand hygiene adherence policy, and it was 28%.

I'll tell you one more healthcare related war story. We've been working with Beth Israel Deacon, which is a Harvard teaching hospital in Boston, and in the last two months I finally got the CEO to put in his blog his personal dedication of the idea of an incident-free workplace for the people who work there, maybe six months ago. So in the last couple of months they had 61 incidents where people were exposed by a splash of the blood and body fluids—this is really dangerous today—and they've recognized that it's a problem that exists all over modern medical care, people getting a splash of blood and body fluids, and there is a protocol for protective equipment. With these particular 61 cases of exposure, the people who had been trained and knew about the protective equipment didn't use it.

So it is really important that once you know and have a demonstrated root cause analysis of how to do things right, you really need to be insistent with your people, after you've done the training, that there isn't any room for making a different decision. I've often said that health and medical care is the only discipline that I know about in society where scientifically based principles are taken as advisory.

Q: I know that you mentioned when safety became more of a precondition rather than a goal at Alcoa, it really changed things, but I'm curious…once you did that, were you able to take safety goals off of what we would call a plan scorecard? So much a percent of the bonus for plants is tied to safety reduction ... or, how did that change when you did that?

A: Yeah, this is a subject that's open to debate but I'll tell you my own view which I practiced when I was at Alcoa, I don't think you should pay people for adherence to values, you should expect them to adhere to values.

Here's a broader principle, and this is a really important principle, I think a lot of our society operates as though all people are what's called in the literature, economic men, which means with the right set of incentives and disincentives you can cause human behavior to be what you want it to be. I don't think that's right.

You may choose to get people more money for them being part of a class or something, but I think that's really a mistake because it diminishes the ideas of we're all the same and we're all in this together, and I don't really want to pay people for safety performance. I want to provide a leadership umbrella that people can get under and feel good about and not feel that that had anything to do with compensation. But as I said, for me it's just part of a bigger story. I think compensation systems generally leave a lot to be desired.

Here are some of the principles I tried to follow when I was at Alcoa, I believe in order for a corporation to be as successful they need to be to pay people competitive compensation, and we can do that, we can figure out the current market in the relevant area where we're hiring people, figure out how to do that. And I believe in a really good corporation ... beginning at the bottom level of compensation, that people ought to be able to count on 100% of that competitive compensation all the time. But very close to that lowest level, I think people should participate in positive variable compensation related to the financial success of the broader institution.

And the variable compensation, by me, should start at the bottom of the organization so that if things are not going great for some reason related to outside economic conditions, the last people who got to get paid variable compensation are the highly paid people.

Now, my belief, which I also practiced when I was at Alcoa, is that the CEO should be paid a little bit below the medium level for comparable organizations in direct cash and that the only way you do better than slightly below the medium was for the organization to not only earn the cost of capital, but to have a spread between the individual corporation's financial success and the relevant competitors.

So you shouldn't get paid for the market lifting. You ought to get paid for the differential you create because of your leadership component, and that's how I think compensation ought to work ideally.

Now, there are individuals that do really outstanding things, and by me it's okay to recognize that financially, but I think these are the principles that I'd apply ... I think, just from observation, it's something that I discovered when I was responsible for strategic analysis of planning for international paper by looking across the landscape of economic activity around the world, no matter how difficult the economic conditions are, there are always one or two companies in the world in any particular industry that do better than everybody else by quite a bit.

But they're not always the same, they change overtime, but there are always one or two, and I believe really great companies should aspire to be part of that one or two that, in good times and bad, because of the energy that their people bring to the subject, produce value, produce economic value, that other people can't understand how did they do that.

So it's interesting, it reminds me of when we bought Reynolds Metals company the last year I was at Alcoa, which was a big company. And when we began integrating them into Alcoa the former CEO said to me, "I've read all the stuff you said about safety all the time and about your performance level, I didn't think it could possibly be true,” because they were 10 times worse than we were, and so their rationale was they're lying about the numbers. It was interesting.

They just could not believe that you're doing something so extraordinary, that you must not be telling the truth. Actually, it's something you really have to guard against, and I don't have time for this war story, but when people violate the trust that you place in them as employees of a great institution, you need to be really prompt and direct in dealing with that violation of trust because it is a hinge for all rest of this. If you don't have mutual trust and confidence in each other it's really hard to make any of rest of this happen.

Learn more about safety training online.