Raffaella Sadun on Effective Management
While it seems intuitively obvious that good management is important to the success of an organization, perhaps that obvious point needs some evidence given how so many institutions seem to muddle through regardless. Enter Raffaela Sadun, the Charles E. Wilson Professor of Business Administration at Harvard Business School and co-leader of the Digital Reskilling Lab there. Working through several managerial mega-projects she co-founded, Sadun can both identify traits of successful management and even put a quantitative value to what good management can bring to a firm (spoiler alert – as Sadun will explain, it’s a big number).
In this Social Science Bites podcast, Sadun discusses her research findings with host David Edmonds, who open his inquiry with a very basic question: What, exactly, do we mean by ‘management’?
“It’s a complicated answer,” Sadun replies. “I think that management is the consistent application of processes that relate to both the operations of the organization as well as the management of human resources. And at the end of the day, management is not that difficult. It’s being able to implement these processes and update them and sort of adapt them to the context of the organization.”
In a practical sense, that involves things like monitoring workers, solving problems and coordinating disparate activities, activities that ultimately require someone “to be in charge.” But not just anyone, Sadun details, and not just someone who happens to be higher up. “The most effective managers are the ones that are able to empower and get information and reliable information from their team, which is fundamentally a bottom-up approach rather than a top-down approach.”
If that sounds a little different from the adversarial relationship many expect between workers and managers, well, good management is a little different, she continues. “I can see how you can think of this as being a trade-off (profit versus well-being of workers), but if you look at the type of practices that we measure, as I said, they’re not exploitations, but they are ways to get people engaged and empowered to sort of participate into the work. It’s always possible that there are organizations that push so much on one side of the equation that make people very unhappy. In my experience, these type of situations are not sustainable.”
Good people – the ones employers prize — won’t put up with too much garbage. “Talented people are attracted–to the extent that they want to work for somebody else—they’re attracted to places where their life is not miserable.”
Sadun came to her conclusions through projects like the World Management Survey, which she co-founded two decades ago. “We spoke with more than 20,000 managers to date—around 35 countries, [and ..] collected typically [by] talking with middle managers.” Other big projects include the Executive Time Use Study, and MOPS-H, the first large-scale management survey in hospitals and one conducted in partnership with the US Census Bureau. In her native Italy, Sadun was an economic adviser to the Italian government in the early 2020s, earning the highest honor possible from the government, the Grande Ufficiale dell’Ordine “Al Merito della Repubblica Italiana.” In the United States, serves as director of the National Bureau of Economic Research Working Group in Organizational Economics, and is faculty co-chair of the Harvard Project on the Workforce.
To download an MP3 of this podcast, right-click HERE and save. The transcript of the conversation appears below.
For a complete listing of past Social Science Bites podcasts, click HERE. You can follow Bites on Twitter @socialscibites and David Edmonds @DavidEdmonds100.
TRANSCRIPT
David Edmonds: I find it difficult to manage myself, let alone anybody else. But a vital key to success in organizations, both private and public, is good management. That at least is the evidence from research carried out by Raffaella Sadun. Italian-born Professor Sadun teaches at Harvard Business School. She focuses on how management affects productivity and growth.
Raffaella Sadun, welcome to Social Science Bites.
Raffaela Sadun: Thank you so much for having me. It’s a pleasure being here.
Edmonds: Well, Social Science Bites has been going over a decade now, and to my shame, we haven’t done a single interview on management. So, I don’t normally ask biographical questions, but I’m intrigued to know how you became interested in the subject.
Sadun: Great question. So, as you know, I’m an economist by training, and I actually did my PhD at the London School of Economics. I got interested in management because I was very interested as a PhD student in growth and in understanding why certain countries were able to grow consistently more than others. Gradually, my interest shifted from a country focus to a firm-level focus. What happens in countries depends on the activities of these units, which are firms and organizations. And then eventually, I realized that if I wanted to understand the productivity of organizations and firms, I needed to go deeper and understand what happened inside the black box. And that’s how I got interested in managing.
Edmonds: Okay, well, it’s obviously a very important subject. All of us at some stage will have experienced being managed, and obviously, a smaller proportion of us will have had to manage others. Can we define our terms? First, what counts is management?
Sadun: Great question. Actually, it’s a complicated answer. I think that management is the consistent application of processes that relate to both the operations of the organization as well as the management of human resources. And at the end of the day, management is not that difficult. It’s being able to implement these processes and update them and sort of adapt them to the context of the organization.
Edmonds: So that’s interesting. You mentioned resources, but doesn’t management necessarily have to involve power over other people?
Sadun: I think this is a misconception in the sense that the biggest role that managers and management doubt is actually a coordinating role. Monitoring this part of the story, that’s where the power element comes through, because you have a hierarchy. And typically, we think about a manager as somebody who has responsibility over others. But the monitoring role is only one piece of the equation. Why do you monitor and why do you collect data over the activity of others is because you want to first of all understand what’s happening. So, there is an element of problem-solving that goes into there. But you also want to coordinate the activities of the team and of the team inside the organization. Of course, there is a hierarchical element because somebody has to be in charge. But actually, the most effective managers are the ones that are able to empower and get information and reliable information from their team, which is fundamentally a bottom-up approach rather than a top-down approach.
Edmonds: OK, well, we’re definitely getting on to effectiveness. But managers will have multiple roles. Can you just give a sense of the variety of things that managers have to do?
Sadun: [Laughs.] Well, OK, if you want, I actually can give us some very specific data, because I did a time-use analysis of CEOs who are like, you know, the top managers. If you look at their calendars—we collected data on more than 1,000 CEOs in six countries—what do they do with their time? They meet other people. It’s a crazy number, it’s like 70 percent of their time goes into these interactions, which again, speaks to the coordinator role. I think of this type of occupations, there is an internal component interacting with the team members and employees of the firm. But when you talk about CEOs, there is also an element of interacting with the outside world; they are the ambassadors for the firm. When you go down the hierarchy, there is still a lot of work that goes into interactions to collect information or pass information through, but it will be, I think, as you move down gradually more and more towards the inside of the organization, rather than the outside of the organization.
So that’s like in terms of how time is spent. What happens in these meetings. It can be very heterogeneous. There are meetings where you just collect information, there are meetings where you do problem solving, you actively try to find a solution to a problem eliciting information from others or giving instructions, and then there is the most symbolic element of management, which is motivating people. For example, giving feedback or even showing your presence. That’s also part of the job. But, you know, there are so many other things. I think these are, in my mind, at least these are the main ones.
Edmonds: Are you equally interested in management in the public sector and the voluntary sector or is it the private sector you’re primarily focused on?
Sadun: I am fascinated by organizations in general. And I have done a lot of work both in the private sector as well as in the public sector, in particular, in healthcare. There is a very specific reason why I’m interested in management in the public sector. On the one hand, I see a lot of potential for good management to help the performance of the public sector. And on the other hand, I also see a lot of misconceptions. Often, unfortunately, it has happened, the temptation to simply port what works in the private sector in a public sector context.
And so, I think that this is a real challenge: understanding what is management in the public sector. Is it fundamentally different? The answer to that: my view is no. There are some basic principles that apply. And we see that in the data too. But more fundamentally, how do you get a public-sector organization where people have very different incentives, very different types of training and experiences, to buy in the basic principles of management? That’s a really important question that is often ignored.
Edmonds: And the research you’ve just cited was research that you’d conducted with CEOs and meetings. But as I understand it, most of your research is at a lower level of management—at a middle level.
Sadun: That’s correct. So, I was very lucky when I was in the UK to be part of our research group that was based at the LSE and eventually sort of became a multinational itself. That research project — it’s called the World Management Survey — and what we do with this project is collect information that we think is quite reliable on the adoption of basic management practices across organizations. We spoke with more than 20,000 managers to date—around 35 countries. And as you correctly say, this type of information, which is under the adoption of basic management practices, is collected typically talking with middle managers.
Edmonds: 20,000 sounds like an enormous data set. You mentioned you think it’s reliable. How can you be so sure that you can extrapolate from this number to management generally?
Sadun: Actually, when we started this research, which was almost 20 years ago, we started the project being very skeptical about the role that management could play for productivity, for example. So, we wanted to get really reliable data precisely because we didn’t want to fall into the trap of case studies, which are very selected organizations, or this notion of management practices, which is often kind of esoteric—a bit fluffy, to be honest. So, we wanted to be concrete.
So, first step was, let’s define…for us, the definition of management is essentially a set of 18 practices. It’s obviously just a reduction and simplification. But the challenge was, let’s measure the adoption of 18 practices, which we can define fairly well. And we have a good sense of what adopting this practice means and what not adopting this practice means in reality.
Edmonds: Give us an example.
Sadun: For sure. Yeah. So, for example, there are some practices that relate to the monitoring of performance. And so one of the questions would be, how do you know if you do a quality check? How do you know if you’re not having too many defects, for example, in your production process? And, you know, a place where you have a real good process to identify defects will be a place where you have data collection, and this data collection is actually reviewed, and it’s used for problem solving. So, there is a continuous improvement process working in the background.
A place where these processes are not adopted are places where you have very rare data collection, for example. Or, even worse, you may have the best possible data collection process, but nobody looks at the data. For human resources, a typical example would be, How are people promoted in this organization? Believe it or not, when you start talking to people, you get the whole range. You have places where the middle manager describes a process that makes it evident that there is a meritocratic selection process. So, we try to evaluate people for a clear range of criteria, we sort of tried to get our best to identify stars and promote the stars. And then you go to organizations where the cousin of the CEO gets promoted, because whatever, right?
This gets me to another point. The first step is getting a clearer sense of what you want to understand about your organizations. Then the second piece is: how do I get this information reliably? We do that through phone interviews that last about 45 minutes. They are double blind in the sense that the analyst that we train and select doesn’t know anything about the company, and the manager doesn’t know that we are effectively getting a sense of what is adopted in their firm, and every interview is actually validated by a second person. Every answered on this 18 questions is scored on a range of growth from one to five.. One means that the practice is not adopted. Five means that the practice is fully adopted. And we have ranges in the middle. But to get to that conclusion, it’s an open-ended conversation that gets scored by an analyst, and there is always a second person that listens to the interview and debriefs with the analyst to get a reality check on that score. So, this is a very labor-intensive process. And we’ve done it more than 20,000 times. We become a little firm. When this data collection gets started, at some point, that LLC, we had 40 people and, you know, I was personally—I was the manager of one of these operations. We select the students, we meet with them, we’re very involved in the data collection when these things happen.
Edmonds: So, 18 criteria, but presumably, these are instrumental means to achieve something else. So for example, in the private sector, they’re instrumental means to make profits, right? I mean, they’re not ends in and of themselves.
Sadun: And by the way, you know, they are things that we didn’t define ourselves at the very beginning, these 18 criteria. Initially, the project started almost as a bet, because a partner of McKinsey & Company told my adviser, “I can tell you what a well-managed organization looks like.” My adviser was telling him, you know, how do you know this? Can you prove it? So, we started with some insights from people who actually worked with organizations. They had a definition of what one manager looks like, as you say, presumably, this is also something that they thought was good, because it led to superior financial performance.
But the one thing I want to say is also that we’re not looking at practices that are just geared towards financial performance, because there is a ton of what we measure that deals with how you manage people. And the idea is to measure how human capital, how talent is managed, not just for financial performance, but also for getting people sort of engaged with their work and productive at work.
Edmonds: But in that sense, you can imagine being very successful in terms of profits, but it could be a very unhappy company. And then you’ve got some kind of tension between, “Well, do we count that as a well-managed company?” or not.
Sadun: Wonderful question. And you know, one of the spin-offs of this work, we actually also had measures of well-being, very rough. But from what I’ve seen from this data, and from what I know also about other research, I can see how you can think of this as being a trade-off (profit versus well-being of workers), but if you look at the type of practices that we measure, as I said, they’re not exploitations, but they are ways to get people engaged and empowered to sort of participate into the work. It’s always possible that there are organizations that push so much on one side of the equation that make people very unhappy. In my experience, these type of situations are not sustainable. And you’re not going to be able to retain the type of talent that is needed to make these practices work.
Edmonds: I was going to ask you about that. Because I would have thought there were short-term/long-term payoffs. I mean, I’ve had managers who were very brutal, and you worked very hard, because you were terrified of them. But you didn’t want to stay.
Sadun: Yeah. And actually, you know, I think that one of the most interesting aspects of the research, at least for me, is that in the data, we see that there is a strong complementarity between good management and talent. Talent, meaning the fraction of people that have a college degree, the education level of the workforce, and the education level of the managers. Talented people are attracted–to the extent that they want to work for somebody else—they’re attracted to places where their life is not miserable. [Laughs.] But they can actually contribute.
Edmonds: Don’t run through all 18 of them, but mention a few more of the 18 criteria.
Sadun. Yeah, so we mentioned already some of the operational ones. One of my favorite ones. at least. is questions related to continuous improvement. For example, how meetings are run, whether the meetings are used in a punitive way to review performance, which is a trap in which many, many managers unfortunately fall. You know, you get the data and things are going badly, your first instinct is to pick up your finger and point it at somebody else and say, “It’s your fault,” right? So that would be a place where there isn’t so much continuous improvement, there is a lot of punishment, but not so much continuous improvement.
Other places where, in fact, these interactions are used for problem solving are places where we measure if the data is shared, is reviewed, and what happens after the meetings whether there is a follow up plan, and there is a continuous way of making sure that the problem that arose in the meeting doesn’t present itself and you really try to go to the root cause. I think that this is fundamental. And this is where even good people, well-intentioned people, fall: taking the underperformance as a personal…blaming people rather than thinking about the process.
And then the other piece that I did not mention is more forward-looking and speaks to how targets are set. Every organization needs to look at the future, not just the present. You typically would have targets that understand how to allocate your attention in their short, medium and long run. These targets need to be stretched, because you want to be ambitious. But if they are too stretch, you just frustrate people. And if they are too lacks, you know, you’re getting complacency. And so, getting at this very difficult balance of having stretch targets that are achievable is something that we tried to measure with our data.
Edmonds: Presumably, you also weighed these factors, because you would expect that some good management practices were just more vital than others.
Sadun: You know, at the end of the day, what you see when you look at this data is that all these 18 questions, which are—think about it—very, very rough, I’m not saying that we capture all of management, maybe just a part of it, the operational part of it—they’re very correlated with each other, which means that there is a lot in factor—call it good management, if you want—that sort of permeates the processes of the organization, both on the operational side and on the human capital side.
For this reason, our simplest approach is actually to take a basic average across all these 18 factors. And we’ve checked whether different weighting schemes or different portions of the survey are more correlated with external measures of performance, and we don’t find that that’s the case. So we go just go with the simplest approach, which is taking the low average.
Edmonds: So here’s the $64 million question. Well, there’s been inflation. Let’s call it the $74 million question: what difference does good management make?
Edmonds: And it sounds like that ought to be extremely easy, right? You’ve identified these 18 wheels, you just need to get them spinning in the right direction: you go into a company, tell them what needs to be done, and hey, presto, you’ve got your productivity increases.
Sadun: OK, so this is where, unfortunately, we hit the limit of economics, and we start going into psychology, sociology, and organizational behavior, because clearly, that’s not the case. That would be the case if organizations were machines and you could just have pull up a lever and tell people, “OK, from now on, we’re going to be Toyota.” And clearly organizations don’t work that way. And people don’t work that way. Why is that? Well, because following a routine is very beneficial for the organization. But it can be very costly and very painful for the individual worker, especially if that worker doesn’t know why they’re doing it, and what’s coming for them.
There is this case study about General Motors wanting to become Toyota in the United States, at some point. They were saying that “Toyota was able to be very productive through lean management, we want to do lean,” and they were not able to do it. Why not? Well, because the unions and the workers more generally had zero trust that the organization was going to adopt these practices to improve their well-being. They thought that this was going to lead to productivity improvements that eventually was going to lead to layoffs.
Edmonds: So you’re an economic doctor who can diagnose a problem but not address it.
Sadun: Oh, come on. I’m trying. No, look. [Laughs.] I’m not a psychologist, I’ve observed a lot of good managers. But I’m not a management guru from that perspective, you know, I’m not gonna write a book that gives you, “These are the four things that you have to do to improve your productivity.” But I think that actually the best thing I can do to help in practice is to make sure that people understand that organizations are made of human beings. And that whereas there is a promise that management can deliver in terms of performance, I think that the best I can do at least is to also highlight that the reason why we have variation in management in the first place is not because people are just ignorant and don’t know what they should be doing, but because it’s very hard to move collectively organizations to collectively believe that there is a set of practices that could be beneficial and should be adopt. You know, the reason why I look at time use, especially time use of CEOs, is because as much as I believe that a lot of the action happens at the middle manager level, I think that fundamentally the reason why we see so much disengagement and so much punishment and wrong dynamics in an organization comes down to the fact that the CEOs and the leaders of organizations may not have a good sense of what it means to have a well-run organization.
Edmonds: But I’m wondering if, given it’s so difficult to implement the lessons in practice, whether we should be skeptical of the whole McKinsey industry—the industry of people going into private companies and telling them how they can restructure for the good of the shareholder or the staff or whoever it is.
Sadun: Look, there is an element of what needs to be done that is purely cognitive, and there’s an element of benchmarking where you are relative to other information that firms don’t have and consultancies can have. And that’s valuable. So that’s like the starting point. However, if you really get to implementation and execution, I’m very skeptical of situations in which you want to turn around your organization and you want a consultant do it for you, because that’s not credible.
Edmonds: They’re paid huge sums of money to do just that.
Sadun: What can I say? I mean, [Laughs] look, for CEOs that are like, deeply involved with this type of the other approach that was mentioning, which is being close to the action, it’s a huge commitment, because it’s a commitment in terms of time—maybe you don’t do as many interviews with the BBC, or, you know, with the public press—but you spend more time internally. That’s one. You have to have tremendous discipline, and you have to be credible and sort of not deviate from the course of action that you set for the organization. And maybe one of the reasons why people delegate that implementation and execution piece, it’s because it’s costly. You know, McKinsey is costly in terms of money. But the other approach is costly in terms of your time and your attention.
Edmonds: Let me ask you one question about the international aspect of this, that big data set, you mentioned comes from a variety of countries, it’s not just from the US. I would have thought there were major cultural differences and that hierarchies in Japan are not the same as they are in Italy or the US and that would skew the data somehow.
Sadun: Let me first say that it’s absolutely true that we seek cross-country variation, both in terms of the adoption of management practices, as well as what you were mentioning, the organization of the firm and in particular how decentralized the company is and how much autonomy is given to the middle manager. And, exactly in line with what you were saying, there are different cultures and these different cultures seep through the boundaries of the firm, I think it’s fascinating to see that Japanese firm will have a very specific hierarchical culture structure in Japan and outside of Japan. So, think about the impact that culture has on firms. I know I’m a nerd, but this thing is so—so I find it so extremely interesting, right.
It’s also true, however, that most of the variation, especially when it comes to the adoption of management practices, but also organizational structure, is within countries and within industries. Let me take India because India has an interesting distribution, they’re sort of shifted to the left relative to the United States, in terms of the adoption of management practices. Remember that the left means that there isn’t a lot of adoption, but there is a distribution, there is a bell curve everywhere. And the extreme right tail of the Indian distribution of management practices is actually overlapping with the right tail of the US distribution, which means that in India, although the distribution is very heterogeneous and the average has shifted to the left, there are some amazing organizations in India. And you can find this pattern pretty much everywhere we’ve seen. Then the other pieces in every country we’ve looked at within countries, greater adoption of management practices, is associated with better performance in terms of productivity, profitability, and growth, which sort of cuts against this notion that what’s good in one country may be terrible in another country. On average, for the type of things that we measure, which are super basic, that’s not true.
Edmonds: That’s fascinating and surprising. Let me just end by asking another question about you. You said you had to manage some of the process to gather this huge data set. What kind of manager were you?
Sadun: [Laughs] A learning manager. You know, I think that actually, researchers, more and more are managers, especially if you do empirical research, it’s not you sitting in your room building your model, but you manage two, three, four…I think right now I have like six arrays and a lab and a virtual team. And so, I’m deflecting the question from what type of manager I am. I think that all researchers should try to get some management training because our job is changing. Our job has a lot to do with coordination and empowerment where I think management can really play a role.
Edmonds: Very well deflected. Raffaella Sadun, thank you very much indeed.
Sadun: Thank you. A pleasure being with you.
