One of the arguments that is sometimes made in the literature on leadership is about whether it is a generic skill that can be applied in any setting, or if having deep past experience -- and success --at doing the work of organization is required to be a successful leader. We talk about this quite a bit in The Knowing-Doing Gap. We argue that leaders aren't as easily fooled by hollow smart talk, and are more likely to talk in ways that help their followers succeed (rather than simply sound impressive), when they have past deep experience in the industry along with years of experience doing, managing, and succeeding at the kind of work their people do.
I talked about this in my recent post on managers vs. leaders too, and argued that the most successful leaders either understand the work they work that they are leading (think of Steve Jobs, Bill Gates, John Lasseter of Pixar, or George Zimmer of the Men's Warehouse) or they take the time to understand the work they lead (I used the example of Bill George at Medtronic). My arguments on this point were questioned, and rightly so, by a commenter who pointed out that I needed to be careful about just using success stories to "prove" my point -- as it is possible that just as many or more people who have little or no experience succeed at the same or higher rate. Or it is even possible that, if I looked at the population of leaders who have experience doing the work and doing it well, that they would actually perform worse than people who simply had generic leadership skills but did not understand the work that their people do.
This "success bias" is, in fact. a big methodological problem with Good to Great. Collins talks about 11 Level 5 leaders out of a sample of 1435 companies. He argues that this difference in leadership is one of the main reasons that these 11 companies made the leap from good to great. But he considers no data on Level 5 leaders who led merely good -- or downright lousy -- companies. It is entirely possible that, if someone carefully examined the full population that Level 5 leaders actually do worse on average. I don't believe that this would happen, but I don't have the evidence to support this hunch (nor does Collins). Rather, I see this critique (and others, see The Halo Effect) as a reason to see Good to Great as a book with some interesting but untested hypothesis rather than rigorous research. And, as this commenter pointed out after my post, I made pretty much the same error that Collins does in my earlier post -- so to come clean, I think that my assertions about experience are best viewed as hunches or hypothesis that require additional research.
I can report, however, that I ran into a story in the Atlantic last night about research on basketball teams that is consistent with this hypothesis. You can find the story here (scroll down a bit). In short, three researchers -- Amanda Goodall, Lawrence Kahn, and Andrew Oswald -- from Cornell did a study of 15,000 National Basketball Association games between 1996 and 2004. They found that coaches who had played in the NBA won a lot more games than those who hadn't played, and that teams who had former coaches who not only played, but who had also been NBA all-stars too, did even better. As The Atlantic summarized:
"On average, teams with former all-stars as coaches placed six spots higher in league rankings than teams with coaches who had never played in the NBA, a huge bump-up in a league with only 29 total teams during the years studied."
This research is interesting for two reasons. The first is that is does bolster the hypothesis that, when people have deep understanding of the work they are managing, and have a history of success, they are more likely to be successful leaders (assuming all other things being equal.. I am not sure that Dennis Rodman has the other qualities to be a good coach). The second is that it shows the dangers of placing too much weight on a single vivid case. Phil Jackson has the highest percentage of any NBA coach, winning multiple championships at Chicago and Los Angeles. Jackson played in the NBA, but was a bench-warmer rather than a superstar.
If you are curious about this paper and want to dig into deeper, you can download a pdf of "Why Do Leaders Matter? The Role of Expert Knowledge."
P.S. Steve Levitt of Freakonomics fame did an analysis of Collins' "Great companies" over the summer, and found that most now have ordinary or poor performance -- and if you had bought a portfolio of these companies at the time the book had come out, and held it, Levitt concludes it would have performed worse than the S&P 500. Of the "Greats," two that have fallen most include Circuit City and -- I had forgot this -- Fannie Mae. Check out the article. As Levitt points out, predicting the future is a tough thing to do.
Andrew, I like the bricklayer/environment analogy. This got me thinking about a few things.
My current company wants to hire software project managers who have a programming background. I have so far unsuccessfully argued that a programming background is not needed for a good software project manager. I have worked with many effective project managers who have never programmed. However, I have found that the effective project managers partner with key technical people who know the details.
I would think that a bricklaying league commissioner who never layed a brick would make a point of having key advisors who know more about bricklaying. I am sure that the CEO of any major company cannot have experience in every aspect of running their company, but they have key advisors who know more about each area.
Your comment about environmental factors reminds me of a business article that was stating that CEOs pay is not directly reflective of performance. The article was critical of high paid CEOs at companies that were losing money. However, the analysis did not seem to take into account environmental factors.
I think that if an oil company had a very incompetent CEO and a US airline had a highly skilled CEO, the oil company would still outperform the airline in every financial aspect this year. The environmental factors are well beyond what a CEO can affect.
Thanks for some thought-provoking posts.
Posted by: Kevin Rutkowski | September 16, 2008 at 12:11 PM
Bob and Kevin,
this is very interesting and I've spent a lot of time thinking about this. And a wonderful form of procrastination it has been. Let me see if I can articulate my thoughts effectively and my concerns with what one measures to determine a leader.
If you're a brick layer, it's very easy to determine who is the best brick layer. Count the bricks. Some smart, clever brick layer may come up with a more effective way to lay bricks and be deemed the "leader". This leader is the best brick layer.
If we want to improve things, we might decide that it's a good idea to have someone coach a team of brick layers. We might say that a coach should work with 11 brick layers to make them all more effective. Determining the most effective coach is pretty easy. Count the total number of bricks laid by each team. You'll find that some coaches are more effective than others. If the coach has previously worked as a brick layer, it's reasonable to assume that they'll be more empathetic and effective. Good coaches will also have different ways of organizing their teams, different methods of laying bricks, different attitudes towards laying bricks, etc. Sophisticated coaches will use different techniques to maximize the their teams, but it's still pretty easy to measure coaching effectiveness.
Now, let's say that we want to create a larger organization. We might have a division. A division consists to 10 teams of brick layers. In order to resolve disputes and make things more organized, we might say that we have a division head. This is necessary because everyone is trying to improve so there are competitions between the teams.
Now, the skills necessary to be an effective division head are not necessarily the same skills necessary to be a good brick layer or even a good coach. Also, complicating the issue, how do you measure who is an effective division head? Looking at the total number of brick laid might be effective, but they are pretty far removed from the brick laying process and the things they handle on a day-to-day basis are probably not similar to the brick layers or the coaches.
If we elevate this one more level and create a league, the problem becomes even more obvious. The league is composed of 10 divisions. Some of these divisions are in the mountains, others are in the dessert and others have to deal with great lakes. The types of bricks that each division can get are different just as laying bricks at 2,000 feet above sea level presents different challenges from laying bricks in the dessert, which is completely different from laying bricks in the middle of a Wisconsin winter.
How do we determine if our league commissioner is effective? A brick layer or a coach may have innovative ideas and it's easy to asses their effectiveness. But determining the effectiveness of a division or league commissioner is more problematic.
We might come up with a thermometer that looks at different parameters. Temperature, barometric pressure, humidity etc. Or in the case of leadership: strategy, execution, deep past experience, etc. And these things might give us a lot of insight into the weather or into the effectiveness of our division heads or league commissioners. However, no amount of studying a thermometer is going to change the weather. Likewise, someone taking the lead in the introduction of a new order of things, is so consumed by the daily details and so effected by the changing environment in which they work, that knowledge of these things doesn't predict effectiveness. Being familiar with these things might make someone more effective or it might make them less effective. The major determinant is going to be the influences of the outside environment.
Environmental conditions (weather, altitude etc.) will affect brick laying more than strategy, past experience or culture. Whether we call someone a manager or a leader and the degree to which these ideas have an effect at different levels of the organization is an interesting question, but will it determine their effectiveness? No. The environmental conditions and how peoples' effectiveness is measured at different levels of the organization is too abstract.
Posted by: Andrew Meyer | September 16, 2008 at 09:25 AM
Andrew, I would say that even a mid-level manager is generally a leader. Leaders can lead a group of people chosen by others within a predefined set of rules.
This is true even at the highest levels of leadership. For example, when a new CEO starts at Toyota, he or she inherits employees who where chosen by many other people and has to perform within the well established rules of the Toyota Way.
A leader can make something out of nothing, but I don't think that's a requirement of leadership.
It also seems that there are many creative people who make something out of nothing and create order out of ambiguity who are not generally thought of as leaders. The arts, engineering, and architecture are full of people with these talents who are not traditionally considered leaders in business literature.
Posted by: Kevin Rutkowski | September 15, 2008 at 07:32 PM
Interesting post, Bob. When I read it, I pulled up a list of the top NBA coaches of all time, ranked by wins. I indicate which coaches played in the NBA and which were All Stars.
1. Lenny Wilkens - 1,332-1,155 (.536) NBA All-Star
2. Don Nelson - 1,243-929 (.577) NBA
3. Pat Riley - 1,200-632 (.655) NBA
4. Jerry Sloan - 1,048-697 (.601) NBA All-Star
5. Larry Brown - 1,010-689 (.558) ABA All-Star
6. Bill Fitch - 944-1,106 (.460) Neither
7. Red Auerbach - 938-479 (.662) Neither
8. Dick Motta - 935-1,107 (.479) Neither
9. Phil Jackson - 931-487 (.657) NBA
10.Jack Ramsay - 864-783 (.525) Neither
Posted by: Wally Bock | September 15, 2008 at 01:35 PM
Thanks for this post on leadership and the post scriptum on Good to Great ... and the comments from Michael Maccoby. Assuming that Leadership is a good thing in this world, how can we help develop it? Recommending to build experience and successes will contribute in many situations. Following on the comments from Michael Maccoby, isn’t it also with identifying more objectively and sharing about the way each of us can develop leadership his or her own way, and is capable of understanding and sharing about how his or her followers can develop their leadership their own way? Are not we done yet with the quest of a unique standard approach of leading?
Posted by: Frederic Lucas-Conwell | September 15, 2008 at 08:15 AM
Bob,
this is very interesting. Is an NBA coach a leader or a mid-level manager? They work with people someone else chose, usually the President of Operation or GM. They work within structures laid down by the NBA and further refined by their individual team. Finally, they operate within the approach set up by the team owners and at the pleasure of one or more of those people.
Doesn't an NBA coach succeed because of their ability to manage the players, plays and practices within a fairly confined realm? This is not to take anything away from their intelligence, insight, empathy, talents or fame, but does it really make them a leader?
When an entrepreneur starts a company they face ambiguity. Part of their ability to succeed is their ability to create order where others are stymied by ambiguity. When Bill Gates started creating personal software and then operating systems for personal computers, he created order out of ambiguity. He created a new marketspace or rather changed an existing marketspace so much that the reordering created vast new opportunities. Isn't this ordering and opportunity where others only saw ambiguity and chaos similar to what Jobs did at Atari and then Apple, NeXT, Apple, Pixar etc.?
In 1917 when Lennin returned to Russia, he saw power laying on the streets and incarnated the ideas of Marx. That had never been done before (successfully). Similarly when Hitler applied Keynes' ideas and recreated the German economy in the mid 30s when the rest of the world was sinking ever deeper into the Great Depression, he created opportunity where there was only chaos and ambiguity. And he did this without ever reading the great Keynes' work.
I'm not commenting on the rightness or wrongness of what any of these examples did. There is no reason to believe a leader does good things. I can't remember for sure, but I believe Peter Drucker resisted using the term "leader" because the term had such negative connotations.
My reason for this rant, is that as an entrepreneur, I believe there is an element of leadership which involves dealing with ambiguity. The first person who ever coached a basketball team was a leader in creating a new element to the game. Phil Jackson, however successful and famous he may be, is a manager or a coach. Within the confines of that role, he may have some leadership aspects, but he is the coach. When they talk about a leader on the team, they talk about Koby Bryant. Under pressure from the other team and the clock, he must be creative and find ways for himself or his teammates to score. There is an element of ambiguity which gives him an opportunity to be creative and a leader. That is why he is referred to as the team leader. Phil Jackson is the coach.
Isn't part of leadership the ability to wake up in the morning and make something out of nothing? To create the structures, processes, culture and in the case of business, the business models to make the enterprise succeed? Isn't that different from someone whose successful operating within the structures, processes, culture and models that have already been set up?
Are there aspects of management in leadership? Yes. Are there aspects of leadership in management? Yes. I believe Chairman Mao wrote about this. He was a leader. Maybe I prefer being an entrepreneurial manager.
Posted by: Andrew Meyer | September 14, 2008 at 10:20 PM