Game Economist Cast
What does the new wave of open economies mean for monetization? Will negative externalities overcome cosmetics economies in the long run? What exactly does a game economist do?
Game Economist Cast is a roundtable discussion of the latest developments in mobile, HD, and crypto games through a bunch of people figuring it out using the economic tool kit.
Game Economist Cast
E32: Should more firms be like Valve? (w/Dr. Peter Klein)
Economist Dr.Peter G. Klein joins the cast to discuss Why Managers Matter: The Perils of the Bossless Company. We debate Valve's organizational structure, the evidence for manager economic impact, and Sweden's success.
Read more about Dr.Klein here and find his book below:
https://hankamer.baylor.edu/person/peter-g-klein-0
https://www.amazon.com/Why-Managers-Matter-Bossless-Company/dp/1541751043/
15:37 Why Managers Matter
22:29 CEOs
35:22 Valve
Let's start with utility.
Speaker 2:I don't understand what it even means.
Speaker 3:Everybody has some kind of utils in their head that they're calibrating.
Speaker 2:There's hardly anything that hasn't been used for money.
Speaker 1:In fact, there may be a fundamental problem in modeling that we don't want to model.
Speaker 2:Dr Klein, we're so happy to have you here and talk about why managers matter, the perils of a bossless company, which you were the co-author with Dr Foss Nikolai is he an economist as well?
Speaker 4:He is an economist and a business professor, like I am. He teaches at the Copenhagen Business School. Oh, he's out here in Scandinavia. Yeah well, he's just down the road from you. Copenhagen's way more fun. I'll tell him that, or you can tell him that you appreciate Denmark as a minor province of Sweden.
Speaker 2:It is true. I actually I've been meeting some free marketeers out here and I think this is a real problem for the libertarian wing. I think the defenses have been really weak. We just focus on like 90s privatization. They're still spending over 50% of GDP and I know what taxes are like here. I can tell you. I can tell you with excruciating specificity.
Speaker 4:Although, as when you look at measures of the economic freedom of the world index and other measures, the ease of doing business.
Speaker 2:Scandinavian countries actually do quite well. Low corporate tax rate. I will say that.
Speaker 4:Exactly. Scandinavian countries, in my mind, are not socialist economies, they're free market economies with some GDP spending.
Speaker 4:But hear me out this, this terminological issue, Right. So they're capitalist economies with a huge welfare state attached. So the state does not own all the factories. The state does not limit, does not stifle competition among entrepreneurs. The state allows capitalism to exist but then takes half of the income and redistributes it to its favored causes. That's a kind of mixed economy welfare state capitals. But my point is this is not. The success of the Nordic countries is not an argument for classical socialism, ie state ownership of the means of production.
Speaker 2:I completely agree, but I don't think that's the argument. I think that is an argument that is stuck in post-World War II libertarian thinking Like the Cold War is over, this is what you got to fight against. This is what everyone is creeping up on. Is this form of and is that? Okay, I don't think so you got to fight that. I like the classics.
Speaker 4:Of course you're right, times change. No one is proposing what millennials and Gen Zs, who identify as socialists, they want.
Speaker 4:Sweden, they want this country, but they don't understand what it is that like Sweden, denmark or Norway, desirable even by their own standards. So I agree that the framing has to be a little bit different, but I think it's still the case that success or failure depends a lot on the degree to which underlying property rights are respected. There is free movement of capital and labor, entrepreneurship is allowed and so forth, and I think some of the folks we're talking about, instinctively they think those things are harmful, but they don't realize that the societies they admire actually encourage those things. So that's where the conversation needs to go.
Speaker 2:So I guess, with that being said, can you tell me a little bit about why Austrians haven't really adopted crypto in the way at least I thought they would. They should have been really into competitive money. This is a lot. Goes back to original banking theory. This should be like the Holy Grail, and I see so little alignment between those folk and traditional crypto platforms, and I'm just curious about why we wouldn't see greater alignment between those two groups.
Speaker 4:Look, there's a lot of overlap, certainly between the Austrian community and the crypto community, in that there are a lot of people who are either trained in Austrian economics or interested in Austrian economics, who are very engaged with the crypto world, and there are a lot of people who are into crypto, who are looking for some kind of theoretical foundations to justify the idea of things like private competing currencies, and have found Austrian economics to offer some theoretical rationale. I think maybe what you're picking up on. Why is there not a closer alignment that Austrian economics, austrian monetary economics, theorizes about and has produced a lot of historical and empirical evidence for the advantages of non-state controlled money? Quote, unquote real money or hard money or whatever you want to call it, which historically has been a lot of different things, right? Seashells in Polynesia and tobacco leaves in Virginia, and typically precious metals, gold and silver in most civilizations. Crypto is a kind of crypto, is a technology or a specific product for facilitating exchange, right? Different cryptocurrencies serve today as a medium of exchange. Most forms of crypto, bitcoin, for example. I would not call Bitcoin money.
Speaker 4:The way Austrians think about money is a commonly accepted is what Mises calls it, or a generally accepted medium of exchange, meaning in almost all of my daily transactions I can pay using that commodity a seashell if I'm in Polynesia, or a tobacco leaf if I'm in Virginia, or a gold coin or a silver coin or whatever if I'm in Polynesia, or a tobacco leaf if I'm in Virginia, or a gold coin or a silver coin or whatever if I'm in most parts of the world. Or, today, a fiat paper currency. So different specific cryptocurrencies are currently being used as media of exchange for some transactions and, of course, they also are used as payments technologies and distributed ledger technology can be used for a lot of things, but a specific implementation of a specific cryptocurrency, a specific coin, is just one of many competing products in that market. So Austrian economics doesn't favor one cryptocurrency over another, any more than Austrian economics says gold is a better money than silver or gold and silver are better monies than seashells.
Speaker 4:What Austrian economics says is a particular community of users, a particular economy or society will tend to adopt a commodity that serves most effectively as a medium of exchange. As of today, it's not Bitcoin or Dogecoin or any other particular product. Now, maybe it will be at some point, but we don't advocate I, as an economist, I don't advocate for one specific implementation of crypto any more than I would advocate for John's Bank or Bob's Mint. Rather, in the abstract, we can say that non-state controlled currencies perform better than state controlled fiat currencies.
Speaker 3:So it wouldn't be the case that I think a lot of people in the early days of crypto, when they were talking about is this a cryptocurrency? I've often called it a misnomer using the phrase cryptocurrency. I think it just doesn't communicate really well what it is In fact. Actually, I'm really happy you said commodity, because I have writing a couple of years ago where I called it a commodity. I said the better analogy is a commodity, so it's really interesting that you say that. It's not that it's in a current state of not being a currency and will always be in that state. It's just that it doesn't abide by the characteristics of a currency. It could in the future, though, if it becomes a common currency of exchange or medium of exchange.
Speaker 4:Yeah, that's exactly right. A society, an economy might choose a particular form of crypto or multiple forms of crypto as commonly accepted medium of exchange. I just don't think we're quite there yet but we get there.
Speaker 4:Who knows, maybe gold and silver coins make a comeback and that ends up being preferred by some or all users to a specific kind of digital currency.
Speaker 4:I think what also complicates it a little bit, chris and maybe this is partly your question too is I think of blockchain technology as facilitating multiple objectives, including just it's a transaction facilitator, which is a role that can be separated from the pure money role of medium of exchange, unit of account and so forth. So a particular crypto technology performs a variety of functions, some money like functions and some transaction cost contracting functions, which I think is great. By the way, I think we should encourage entrepreneurial innovation and experimentation with all kinds of different technologies, including remote payment technology and contract enforcement technology and transaction facilitation technology, and I think the Austrian view is let's allow competition to flourish. Whatever people choose to adopt as their dominant medium, exchange will be the one that they choose to adopt, and we want a system in which you can do that I guess what falls out of the model, though with this I feel like this generally accepted medium exchange, I don't know it feels like a cop-out.
Speaker 2:What does that get you by making that assumption? So why we all work on, like individual web, three economies, and usually that involves, like managing your own token and right now, the token that in most of these games is only redeemable for goods within the game, and we usually generate that value by either enforcing prices that occur due to peer trade, having some sort of you could call it a tax Like we're essentially the government and we can mandate that to complete this exchange between these two goods, we're going to charge you a token that only we issue. So we have that government property but we really think about it as money. But if we're saying that it isn't money and it could have money-like functions and just not call it money, but is there any practical assumption that we need to revise because it's not generally accepted? I?
Speaker 4:think of it this way. Go back to Carl Menger, founder of the Austrian, who did the pioneering work on defining what money is, what function money performs right. So money facilitates trade. You can define an economy so narrow I could say in my household we only use these little pieces of paper that I myself have created. That's fine. Within that micro economy, that currency may perform the function of money. But then if I try to integrate my household with another household, then the thing's down right.
Speaker 4:You want what people living in an economy want is some way of avoiding what economists call the double coincidence of wants that you get from barter. You want an independent commodity such that prices for all goods and services that are being exchanged can be defined in units of that special commodity, which we call money. That facilitates, we can price everything in that one common unit. If that unit is divisible into little, small increments, we can facilitate all kinds of trades. If we're trading cows for horses, what do you do if you only want half of a cow? You also want something that permits what Mises calls economic calculation, right?
Speaker 4:So an entrepreneur in this economy, whether it's a mini, virtual economy, or whether it's the United States or wherever you live entrepreneurs. In order to figure out whether to engage in business activity or not, they need to be able to. They need some mechanism for comparing expected future revenues against current costs in a common unit so they can calculate profit loss right. If everyone is using a different commodity to facilitate exchange, we still have exchanges going on, but Misa' point is that economy is limited in how far it can develop right, because entrepreneurial activity is stymied, is hindered by the absence of a common unit in which to calculate revenues and costs so that you can calculate profits and losses.
Speaker 2:Yep, and you just back it into a generalized currency anyways. That's probably what you would use to nominate it.
Speaker 4:Yeah, people say, oh, we've got this other thing. But let me just translate that. And it's like when you travel abroad you might be computing, you're converting the local currency to the currency that you know to figure out should I buy this souvenir or not. We could certainly do that, but you need some. There needs to be some unit that the decision maker can use to be able to figure out how to allocate time, energy, resources and so forth, to allow that decision maker to achieve, you know, their most valued ends. Again, in a commercial sense, we can still engage in some form of rational economic activity without prices On it Robinson Crusoe on a desert Island with one with Friday to trade with or in a particular gaming platform, or just among the four of us. It's not that we can't do anything, but you can't develop a modern industrial economy with specialization and the division of labor and all that if people are using different units of account for computing prices. You need to be able to translate everything into one sort of commonly accepted medium.
Speaker 2:So let me ask you a games question. We all work on games and a lot of modern game economy development sometimes about managing currencies. So games will have multiple currencies and the reason we use multiple currencies is because we want to gate certain action, some sort of activity. So if I price something in a particular currency and you have to do jumping jacks to earn that currency, it's the only way you can get that currency. I can essentially enforce jacks for you to get the good that's priced in the currency. That you have to do when you get jumping jacks so we can gate activity and we can prevent spillover effects. So now you can't do anything, and so that's one of the techniques we kind of use here and I guess I'm curious about, because we have to run all these different units of accounts sometimes in these economies to keep them contained like price things. Do you really think we're hurting players and their planning process, Like they're struggling to form an expectation of what the economy is like?
Speaker 4:Look, I'm not a game designer. I defer to your expertise on what you're trying to make the game as enjoyable for players so they don't want to play, and so forth.
Speaker 4:So you're engaging in a little bit of social engineering. You're engaging in some Soviet-style central planning Congratulations, comrade, being head of your local pub bureau. Again, my view is what you're doing? This little mini society that you've created, that's your private property. More power to you socially engineer the heck out of it, if you want.
Speaker 4:Within the game economy, the game designer may say I'm going to impose my will on the economy because I'm the entrepreneur who runs this game and if people don't like it, they can play another game. Again, that's to me. That's similar to and now I'm starting to nudge us towards the topic on which I might actually know something is it's like a company within a business firm. You don't have a pure free market internal to that firm. You got bosses, you got owners and you got managers and you got a hierarchy and all this stuff we're going to talk about.
Speaker 4:Some people have asked me in the past and they say wait a minute, aren't you one of those Austrian econ, free market libertarian types? And here you are talking about hierarchy being beneficial? Isn't that a contradiction? How do you sleep at night? Blah, blah, blah. Okay, when people, when Austrian economists or libertarian scholars, whatever, when they talk about the advantages of liberty and competition and markets, they're not saying. That's not a denial of the fact that people sometimes voluntarily organize themselves into hierarchies because doing so allows people to achieve some objective more effectively than they otherwise would. If anybody here has kids, if as a parent, you impose some structure or order on the household where kids do not have free reign to run around and do whatever they want, that doesn't make you an inconsistent Austrian economist or libertarian right. You can say look, people are organizing themselves into families. Maybe it illustrates the point. Better to talk about spouses, maybe, than kids. I subjugate myself. My spouse subjugates herself to me. Nobody forced us to do that and we can do that while still extolling the virtues of free and open competition.
Speaker 2:Let's talk about this book why Managers Matter the Perils of a Bossless Company.
Speaker 3:So my first question is why do managers matter and what is the distinction between a flat structure and a hierarchical structure, which is basically what the whole entire book is about. I found it's not necessarily a criticism or a critique about these types of structures. It's really just talking about the benefits and the cost of these different structures and why most companies use a hierarchical structure as opposed to a flat structure. Why do managers matter?
Speaker 4:You just provided an excellent summary right there. Books like this need a hook. So the hook is the book is framed as a critique of a certain kind of popular narrative or narrative that's popular in some circles and I bet in some of your circles. That management and hierarchy is old and in the modern era the network, knowledge-based, internet economy, whatever we no longer need the traditional managerial structure. Everything can be flat and autonomous agents interacting through a platform, et cetera. You don't really need bosses, you don't need managers. That ties into the sort of cultural norms we get from Dilbert or the Office or something that you have. These middle manager types. It's like a Michael Scott. They're not really adding value. They're kind of fumbling and bumbling. They've been promoted to that position for some reason and boy, if they would just get out of the way and let the people who do the actual work do their thing, we would be better off.
Speaker 4:So that's a narrative in popular culture, it's in some of the academic literature, it's very prominent in the tech world and so forth, and so we wanted to write a book, a popular book. I want to emphasize to your audience that most of what myself and my colleague, dr Foss have written together and separately over our careers are more technical, academic articles and books. But we wanted to write something that would be to the general reader and we wanted to write a book on what management really is and what function management really performs, historically and in the modern world. But we positioned it as a critique of the idea that managers don't matter because you can eliminate all the hierarchies and in the modern world. But we positioned it as a critique of the idea that managers don't matter because you can eliminate all the hierarchies and flatten the firms and so forth. So what we argue? First we offer a little critique of this sort of flat hierarchies narrative, both theoretically and historically and using lots of case examples and so forth, and then in the second part of the book we offer an explanation of what good management is, what the value-added function of management can be, and we offer some practical guidance for managers about how to be more effective than they are.
Speaker 4:And you, chris, hinted at the punchline of the book is that how a managerial hierarchy is structured, like how tall versus how flat the firm's management structure is. There's no universal answer that says all firms should have a really tall structure or a really flat structure. Really, what economics and other social science disciplines tell us as applied to management is that all forms of organization have benefits and costs and the benefits and costs can be greater or lower under different circumstances. And the real challenge is to figure out, okay, under these specific circumstances, for my company and my industry in this country in this particular time and place, for these workers with these products whatever, what are the benefits on the margin of being a little bit more flat or a little bit more hierarchical? And we touch on things like.
Speaker 4:One of the big sort of mechanisms that we illustrate is what the technical term for it would be interdependence or interdependency.
Speaker 4:So if you think about tasks and jobs and teams or subunits within the organization, if these smaller units can perform whatever tasks are assigned to them very effectively without much interaction with the other units they can all be semi-autonomous or independent then a relatively flat structure where management sets a few general rules and make sure things are running smoothly but doesn't otherwise get involved very much, can actually be fine, because the output of the organization is just a summing up, it's just adding up all those independent parts. However, in another kind of setting where the tasks or the teams or the subunits are really interdependent like neither one. To be successful, the other one has to do something that matches, that harmonizes. You need to coordinate their behavior. Under those circumstances, a very flat structure can end up generating a lot of inefficiency, gaps. Right there you need a little bit more central direction to coordinate and make all the pieces fit together so managers can try to assess the degree of interdependence, then make an informed decision about what their role should be in terms of getting involved or kicking back.
Speaker 1:To jump on that for a bit. I think you mentioned how interconnected are the parts that this organization is producing, and I think games are particularly interconnected and multidisciplinary. You've got art, you've got animation, you've got game design, sound design, engineering, software marketing, and I think it's exactly what you described. When I see these kind of decentralized organizations, sometimes the art team goes ahead and they make 5,000 bushes, and then the design team oh, actually, we decided we want to be set in a city instead of in the jungle, right? Oh, we just wasted all that time. And that top-down central control stereotypically, has been successful when you've got this like autocratic design director who has like the vision for what they're producing.
Speaker 4:I think you're right and again, like everything else, it's a balance. You need this person to do to design the trees. Well, if this person is really an expert in writing the code to make these trees beautiful or waving in the wind or whatever, you don't necessarily need to look over their shoulder and write the code for them or make sure that they're designing the tree in exactly the right way. But you've got to figure out how many trees do we need? When do we need the trees? Do the trees have to match this? Do have to integrate with that? That's where the supervision comes into place. It's not micromanaging. It's making sure that the right people are assigned to the right tasks at the right time and that you allow them to use their own individual creativity, motivation to fill in the details.
Speaker 2:But you need some kind of a structure to make all the pieces fit together do we have any good empirical evidence that, like measures the counterfactual on the value of a good manager? That really should I think of them as operators on capital, like they have an immediate effect around the people, like they multiply their productivity rate by some constant. How should I think about the value add?
Speaker 4:Sure, I don't know that it's possible to quantify it in exactly that way. There are lots of different ways we can think about it right. We can certainly do-.
Speaker 2:Random assignment in the military maybe.
Speaker 4:Yeah, so there are plenty of sample studies that use some kind of experimental or quasi-experimental design where, if you can, if you military is an interesting case. There are a few cases in academia where students are randomly assigned to an instructor or subordinates are randomly assigned to a supervisor and when you look at the variation in performance across units you try to see what fraction of that is attributed to the performance of the team members and how much is attributable that can be attributed to the value add from the supervisor. And there are all kinds of estimates of positive effects depending on the particular context. Right Could be anywhere from 5, 10, 15, even 20% of the value add. In some cases can be attributed to the performance of the supervisor. The problem is in nature. We typically don't get that kind of random assignment. We have a kind of a sorting where good employees want to work with good managers, which makes it hard to tease out the incremental composition of the manager.
Speaker 4:Something else, another kind of study that's been done so probably won't frustrate you. Looking at unexpected death, for example, there are several studies illness, or a CEO has cancer. What happens to the firm because people are making decisions and adjusting, moving in and out of roles, anticipating the future role of the CEO. But there are several studies that look at CEO dies in a plane crash, right. What is the immediate impact on the productivity or different output efficiency measures of the firm, aside from just going into mourning for a few days?
Speaker 4:And a lot of those studies show that who the CEO is matters a lot, that productivity and firm value often fall sharply following the unexpected, unanticipated departure of the CEO. There are others if a close family member of a CEO dies in an accident unexpectedly, the performance of the firm falls. So in that case it's not that the employees are necessarily they probably don't even know the relative of the CEO but the CEO himself or herself is not able to perform at the normal level because of dealing with an unexpected tragedy and that those events can have significant impact on the performance of the firm. So again, it's hard to generalize and of course there are plenty of cases of bad management, ineffective managers or owners who are not adding value to their enterprises. But there are lots of examples also where good management does seem to matter in terms of organizational performance.
Speaker 3:So talking about this kind of qualitative difference between certain managers. We know this is actually common findings in the education literature is that if you put a really good principal in charge of a school, that school is going to do great and then the moment they leave the school goes back to its previous state. So we know that good managers are good and bad managers are bad. You also talk in the book about Mark I authority and Mark II authority, which I really thought were interesting concepts that I think are seminal in this book. But would you mind explaining those terms?
Speaker 4:Yeah, absolutely, and I'm glad that you asked, because this is something that can be explained in a straightforward way that I think will make sense to your listeners. Right, Because in the gaming world, right, everybody understands the difference between designer, who writes down the rules of the game, and the people who actually play the game, right? So there's a literature in economics and in law, political science, on institutions. In fact, we're recording this on what? The 16th of October, just a couple of days ago, Monday, they announced the 2024 winners of the Nobel Prize in Economics, and it's three economists who have worked extensively on institutions. Their contribution is described as having improved our study of the impact of institutions on economic growth. Institutions can be described as the rules of the game right, the two different types of authority. Mark I authority refers to and we borrowed this terminology from Herbert Simon of the game right, the two different types of authority. Mark one authority refers to and we borrowed this terminology from Herbert.
Speaker 4:Simon, by the way, we didn't make it up, but this refers to what many people think of. When you hear of authority, supervision, hierarchy, what typically comes to your mind is command and control. I'm a soldier in the army and my supervisor says march over here, take that hill, do this, do that, and my job is to obey. Yes, sir, yes, ma'am, you tell me what to do, give me orders, I carry them out. That sounds like a hierarchical boss-employee kind of relationship. But of course that kind of relationship has a lot of disadvantages. Right, it doesn't allow, it's demotivating to the person being bossed around. And it assumes, for that to work well, it normally assumes that the person giving the orders is at least as well-informed as the person receiving them and in fact has this sort of superior big picture understanding of battlefield tactics or whatever, and the grunt's job is just to obey and go along. The supervisor, the leader, the director knows really what's, knows what's going on and knows where to put all the different pieces on the board.
Speaker 4:There's another way to think about authority too, authority also. We can also think about authority as creating a situation or a scenario in which the subordinate can actually use his or her own and values and motivation and so forth. Think of this, think of Mark 1 authority as telling somebody how to play the game. Mark 2 authority is designing the rules of the game and then letting people play right. So in a typical work environment, right, the manager needs somebody who I think this eric's example can create the trees. Okay, the manager doesn't have to create the trees, but the manager is the one who knows that trees should be created. The manager's job is to find somebody who's good we think will be good at building the trees, give them the tools and resources that they need to be successful in tree building and then let them get out of the way and resources that they need to be successful in tree building and then get out of the way and let them use their skills to make the trees.
Speaker 4:So that is also a kind of authority, but it's not command and control. It's not looking over somebody's shoulder and telling them what to do. It's creating the situation in which that person is then placed and allowed to use their own judgment, their own knowledge, their own whatever to figure out how to do the job. That's also a kind of authority, and in many cases that's the more effective kind of, because it takes advantage of the skills, knowledge, beliefs, motivation that the employee may have. But it's also not just, it's not bossless right? Somebody's got to be coordinating the activity. So creating the, designing the rules and letting people play how they want is also a kind of authority. Authority is not just making people play the game.
Speaker 3:a certain way. In games, I think the analogy is probably emergent gameplay. It's a consequence of the environment, the physics within the digital environment, and some of the most successful games of all time have very strong emergent gameplay. So I think that's really interesting. You can create the perfect designed experience and it's over in 10 hours and you go home. The most successful projects have been those where people are inventing new ways to play the game 10, 20 years later.
Speaker 4:Yeah, no, you're exactly right, and part of being a good manager in this Mark II sense is also knowing when the rules need to be revised and when you see something emergent that you didn't anticipate, figure out how to encourage it or how to assist with it. An example that is useful to me and may be useful to some of your listeners too, just from my own experience as a professor professors are highly autonomous. Professors don't want to be told what to do in the classroom or what to do in their research and so forth. But the unit of analysis commodity that we're selling in most colleges and universities is not an individual course or a syllabus. We're offering a degree. So I'm a department chair, I'm chair of the Department of Entrepreneurship and Corporate Innovation. We issue a degree. We have an entrepreneurship major, an entrepreneurship minor, we have a graduate program in entrepreneurship and in order to get that major, to get the diploma, you've got to have a set of courses entrepreneurship one and entrepreneurship two, and then all these specialty courses and some foundational courses and so forth. I can't tell the professor of a specific course exactly what to do in the classroom, nor do I want to. That person may know more about that topic than I do. But the pieces have to somehow match. The second course is supposed to build on the first course and you have to do the foundations before you do the electives or whatever. So you need somebody like me, for example.
Speaker 4:Just okay, let me make sure I have the right people, the right professors, assigned to each class. Let me make sure that the general scope of what they're covering fits in with our curriculum. You don't want too much duplication. I get this all the time from students. They'll come to me and say, hey, I'm in Professor So-and-so's class and he presented this idea. And oh my gosh, that's the fourth time I've seen it. I've seen that in three other classes. I'm tired of it.
Speaker 4:Okay, that means that's a coordination failure right Within the degree. Or students will say Professor Jones said we're supposed to know X, y, z from our previous classes, but we don't actually know it. Can you help us? What do we do? That's another kind of a coordination failure too. That's a gap. They are gaps. There are redundancies. A skilled department chair or dean or whatever knows how to minimize those problems but yet doesn't know how to actually do all the work. You got to hire good people to do the work and another part of Mark II authority is okay. Something's not working. This professor is not a good fit for our curriculum. This professor maybe needs to move on to some other assignment, or we need to hire somebody new, or our curriculum is no good. Yeah Right, we need to revise our curriculum. Some in some cases. Some of those things can happen organically, spontaneously, but in the typical case it takes someone with kind of a bird's eye view to try to affect those changes, sometimes with a little bit of nudging and with carrots and sticks.
Speaker 2:We've been talking about institutions the last couple of days. I think of a company as having a constitution. It's having a set of rules and agents are optimizing under those rules and, like you said, there's a lot of really interesting dynamics that come out of this. But wouldn't economics have something to say? If this is a very high leverage decision, how you internally organize your company, like what rules you put in place, have a great effect on output, and I think that's really the output that's echoed in public choice is like hey, we talk about the power of the agenda setter, we talk about all these little hacks that happen. If that is the case, doesn't economics have more to say about how a company should be organized, not just that different organizations exist? Let's talk about Valve. Valve, to me, could they add more? I don't know what they're optimizing against. I have no idea what they're optimizing against, but doesn't economics have something to say about how they're organized? That isn't just. This is how they are, this is how they ought to be.
Speaker 4:Sure, economics provides a framework, a set of principles. We're saying, for example, what I was describing before as a contingency framework. We can flesh that out a little bit. So, for example, where tasks, jobs, sub activities are relatively independent, but also where employees respond very strongly to higher levels of autonomy, maybe there's some kind of performance-based pay that's built into that as well. Where the designer is relatively uninformed about some of the technical details compared to employees, those are all examples of contingencies that would push us toward a flatter structure.
Speaker 4:Under those circumstances the benefits of flatness exceed the costs and the reverse would be true. Employees don't care that much about autonomy, they want security, they want safety. They prefer flat pay to incentive-based pay. The designer actually is relatively well-informed about technical details and their high degrees of interdependence. Then theory tells us that organization will be more profitable when it is organized in a more hierarchical structure compared to a flat one. So economics gives us these sort of principles that we can apply to different situations on the ground. Of course you've got to figure out, okay, is Valve really like scenario A or scenario B? That takes some in-depth, that takes some deep knowledge of what's really going on at that specific company and that specific industry with their specific technology.
Speaker 2:I'm not asking Is your analysis of Valve? This is just how they are Like. This is just how they function Like. This was also Yannis's contribution. We also are lovely, honest. He's like our. He's the first game economist. He is forever. He is forever on this industry's tombstone. He's going to be numero uno, whether you like them or don't, and he writes this blog post about Valve, which I'm sure you've read where he's basically, is this a spontaneous order?
Speaker 2:And exactly to your point, the Austrian response is this is about government power and the state is a monopoly, so who cares? Whatever comes afterward, I guess I don't know how to think about Valve. I don't understand. Do you think Valve is set up the way it is? You don't think it's scalable? You don't think it. You're familiar with supercell because they've just pivoted into exactly your book.
Speaker 4:Basically you went back to an hierarchy. There are a couple different ways to get at that. First of all, does everyone really want valve? I think people want what they think valve represents, but I don't think everybody really wants the actual structure of valve during.
Speaker 2:I think they want the rollable desks. You're not. Everyone talks about the rollable desks, everyone loves that shit. So they have these desks which you just unplug, and then you can form these groups. It feels very hayekian.
Speaker 4:Yeah, of course, and this is part of the holacracy model, which is a particular interpretation which certain valves model is not. It does not use the language of holacracy, but others, like Zappos of course have, and part of that model is the idea that these subunits that we were talking about before can spontaneously form and dissolve. I can figure out which team I want to be in. I can physically move my desk I want where I want. Look, those are gimmicks. Right, those are hacks. Yep, those are tricks. Do they work? Again, the evidence on valve is pretty mixed in. That valve has certainly been a great success see it's mixed.
Speaker 2:How is it mixed? It makes hand over fists as well as a valuable company like per employee. Right, but what? Valve is today is not the valve of the storybook right, you mean that it's just a lot of shadow structures. I guess we could call them like the prices If we were going to very back here. There's a spontaneous hierarchy that's emerged. Is that how I should think about it?
Speaker 4:So there are two points. First of all, even Valve in its heyday had a hierarchy. According to all analyses and investigations it was just an informal, it was a kind of a shadow hierarchy, if you like, and that hierarchy sometimes worked well and sometimes it did not work well. But of course Valve has changed its business model. You guys are industry experts, I'm just an observer. Valve is not producing original content, right. Valve is a distributor and a platform. It's not the same kind of company that it was 20 years ago in terms of the functions that are actually the output that the company is producing. So does the valve model for a creative company as opposed to a sort of a light touch platform, you know, mostly a distribution company? I don't know. That's one question. But I want to come back to the shadow hierarchy, because that's one of the points that we make in the book is that even in the absence of formal job titles or whatever you have, you can use whatever. You can use some version of that Robert Michel cliche, whatever you want about organizations abhorring a hierarchy vacuum, right, Hierarchy, thin, just. But the hierarchy is often one of. Again, some people have more informal authority, People are more respected, People's voice carries more weight.
Speaker 4:The problem is that I think the challenge is that informal hierarchies, because they're less transparent, they're harder to adjust when things go wrong.
Speaker 4:Again, if you look at all these insider accounts and say, working at Valve was like being in high school. There's the cool kids who have all the power and then there's the nerds who get beat up and, metaphorically, stuck in their locker, slammed into their locker or whatever, but you don't always know who's who. You don't know when someone is coming towards you. Are they a cool kid? Are they a jock? Are they a nerd? Are they high in the informal power structure? Are they low in the informal power structure? It's better to be transparent. It's better to know what people can and cannot do with some degree of formality and transparency than to have it fuzzy and amorphous. I would not recommend to most companies that they try to replace their more formalized hierarchical structure with an informal sort of emergent hierarchical structure that's not transparent. That can be very demotivating for some employees and, of course, not everyone. Valve lost a lot of people. Some people want to work at a company like Valve, but not everyone does.
Speaker 1:So I think what you're saying about shadow hierarchy is like totally on the nose. I worked at a company that was pretty flat and you could tell these are the cool kids and these are the people who people listen to, even though their title didn't say that, I think. As a counter argument, I think sort of what the shadow hierarchy allows for is much more flexible rearrangement of power, whereas in the formal structure, joe is the manager and even if joe sucks, he's the manager. And until joe's manager tells him, hey, you're not the manager anymore, he's there, whereas in the shadow, the person who is getting bullied presumably are the low performers, the people who aren't contributing, the people who are creating more bugs in the code than they're fixing. Yeah, I think that's the argument for the flatter structure. Now, whether or not and arguably the putting things into a formal, rigid structure prevents fluid reorganization the oh John was the best for this project, but he's not the best for this other project.
Speaker 4:Yeah, eric, I agree with you Again, there are no perfect solutions, they're only trade-offs. So you've just identified a potential benefit of the more emergent, more organic and perhaps less formalized kind of hierarchical structure that it can be. It's more easily adjustable right now. Ideally we would want it to evolve and adapt in a way that promotes strong performance, but it could also drift in the opposite direction. Right Again, think about from sociological theory you can get embeddedness, or you can get certain kinds of powerful people to be more institutionalized. Right, they could end up taking advantage of the situation to become entrenched.
Speaker 4:Yeah, sure, a fluid and rapidly adapting informal structure sounds better than a more rigid formal structure that can't adapt to necessary changes Absolutely. But don't forget there can be some drawbacks of that informal structure too, and that maybe it isn't evolving towards, maybe it's adapting towards entrenchment of people who hold and are able to retain power. Again, one would have to look at the details of the personnel. One would have to. A good manager would want to look very carefully. Okay, I see that these people over here seem to be listened to a lot. Are they being listened to a lot because they have something useful to say or is it because they are good, they know psychological tricks, they're just charismatic.
Speaker 4:They know how to manipulate other people. They're less scrupulous, and I mean, who knows that? That's why a good manager has to know some psychology and some sociology, as well as economics.
Speaker 3:So would you say, Peter, that the primary driver or the primary variable is people or industry. When it comes to whether this type of structure works the flat structure, yeah, some of both.
Speaker 4:Clearly both play a very important role. I think it's hard to totally separate them because the kind of people that you get often have a lot to do with the kind of industry that you're in.
Speaker 3:Yeah, that's true.
Speaker 4:Right Offends to anybody on this call, but coders are not always, they're not always conformists and they don't want to wear. They don't want to wear a gray suit and tie and they don't want to sit behind a desk.
Speaker 2:You should see game developers.
Speaker 1:Lower your expectations. Let me tell you, even the business people are sloppily dressed. Yeah, let me tell you, even the business people are sloppily dressed.
Speaker 4:Yeah, in my industry you just if you had a more formal hierarchical structure with more Mark one authority rather than Mark two authority, you're either the talent would not be willing to work and would not work well in that kind of environment. And probably the same is true in a lot of creative industries. But in manufacturing and services and transportation and entertainment and a lot of other industries, again, there may be plenty of people who are happy to, people who will enjoy and can thrive under a slightly more rigid environment. One other thing too, because I teach entrepreneurship, sometimes people ask me if one of your students who takes your entrepreneurship class ends up working for Citibank as a mid-level employee, working for the man, do you consider that a failure of your program? And the answer is absolutely not, because, much as I mean, look, every human being to some extent likes autonomy and wants to be able to exercise their creativity and so forth. But if you ask a lot of people, to what extent are you willing to sacrifice some of that for a steady paycheck? Some people will say yes. Even I myself.
Speaker 4:I'm an entrepreneurship scholar and I would like to think of myself don't laugh but I would like to think of myself as a creative person who comes up with new ideas, like writing this book and all that. But I'm technically an entrepreneur, at least not very much in my daily life. I'm a salaried employee. I'm a professor with tenure. I have about the least risky employment profile of anybody that you can imagine. I have friends who think like me, who said I can't deal with the race of academia, it's so bureaucratic and it's this and it's that, and they've quit. I could quit my job at the university and I could just be a scholar, public intellectual. I could make my living from selling books, giving lectures, consulting, hustling, whatever, and maybe I don't know maybe I would be good, maybe I would be okay at that, maybe I'd be able to earn an okay living. I just don't want to. I don't want the anxiety. I don't want this to wondering where the next paycheck is coming from.
Speaker 1:I am totally in the same boat.
Speaker 4:Sacrificed some upside potential, upside get rid of the downside, and to me that's just purely a matter of personal preference. So what I'm getting at is what people sometimes forget in these conversations is they want a very loose and fluid, organic kind of structure, but they also want the compensation system that goes along with that. They want strong pay for performance right. They want a kind of entrepreneurial maybe. There's like a bidding system. People bid against each other for jobs and how much you take home at the end of the month depends on how you performed in a kind of internal competition.
Speaker 4:I think that if people want to do this, fantastic. But I think managers need to recognize that not all employees want that. Not the case that all employees want to be competing in a market-like environment every day. Some don't want the hassle, some just want to write code in your industry and they want to know that they're getting paid every two weeks and how much they're going to be paid. And even if they're a little bit constrained in what they can do, that's still a better option for them and there's nothing wrong with that.
Speaker 1:People always say they want the upside risk but they don't want the downside risk. For example, one of the theories in this recent wave of like flat company popularity was that communication technology like the internet, like cloud computing, like cloud collaboration or Google Docs and video meetings and stuff would increase communication and that might somehow reduce the need for hierarchy. And then now we have remote work is increasingly popular. I'm curious historically, have technologies affected levels of hierarchy or how they're arranged?
Speaker 4:Yeah, it's a great question. The answer is not as much as we thought, and I think the reason is everything that you just described about communications technology, for example, making it easier for us to coordinate at a physical distance and also allows more independent, autonomous decision-making, because now it's easy, for the local decision-maker has access to a lot more information than he or she did before, and so some of that kind of base coordination can happen for the bottom up because we have all this new communications infrastructure All true. However, at the same time, right, think of what else all this information technology does. Also allows for supervisors to monitor the activities of their team more closely, right, I don't know what you guys all do at work, but there's all kinds of project management software, their ERP systems, that allow managers and their managers and so forth to keep very close tabs on what's going on daily basis. The question is okay, which of those effects outweighs the other, right? So think of this. If, if you like to use-.
Speaker 1:Is the idea here that improved monitoring technology increases the value of managers because they can? What's going on?
Speaker 2:better. It's almost like they're party whips. They're getting more effective at whipping it's not part of that principal agent problem it?
Speaker 4:potentially it increases the potential value of management, but there's a downside to that right. In fact, something we discussed briefly in the book is a study that Foss and I and another colleague did, looking at some data on these European companies. It was based on a survey of senior management how they were performing their roles once the firms had invested in more and better information technology. So what these top level decision makers said is first, we did eliminate some middle management layers once we'd implemented these new systems, but we found ourselves micromanaging more than we did before in a way that we regret. Right, the temptation to go onto the shop room floor and stick your nose in things is greater than it was before, when you had those middle management layers. Because, again, I use an example in the book that sent out a infamous email once where he talked about trying to flatten the structure at tesla and he said he said any employee at tesla can talk to any other employee directly, one-on-one, without having to go through a filter, without having to go through, and what he meant was I'm the lowest, I'm the janitor at the Tesla plan. I have a problem or an issue? In principle, I can speak directly to Elon Musk without going through a bunch of hierarchical layers and that sounds great. If you're that, you're the janitor who has a problem he won't solve. But if you flip that around, do you want to be doing your janitorial work? And at any moment your phone could ring and it's Elon and he's not happy about the cleanliness of the bathrooms or whatever. I don't think I would want that. I think most people would not want that.
Speaker 4:So it's middle management plays the way it's been described in the literature is. Middle management plays two different functions brokerage and okay. So brokerage is passing information up and down and helping to filter and consolidate. Okay, what messages do the people at the bottom need to hear from the top and vice versa? The buffering role is providing some protection in both directions, right? So again, I'm a middle manager in my job at the university as a department chair, and sometimes there'll be a message from the president or the provost or the dean, somebody above me. We're not happy about XYZ and the faculty need to step it up. Okay.
Speaker 4:So what do I do? I don't just forward that email to the faculty, to my faculty. I send my faculty a separate email and I say guys, look, here's the situation. I'm getting this chirping from above. They're not happy about this. We could do this a little bit better to try to make those people happy.
Speaker 4:I'm watering down the message. I'm interpreting the message in a language that I think my people will want to hear. And likewise, when the faculty are mad about something that the dean or the president is doing, they don't just email me. They shouldn't just email the dean or the president. They should come to me and tell me what their problem is. I'll say oh okay, I know a way to put that so that the dean will respond favorably. Or I could say you don't want to go there. Trust me, I can't tell you why you did not want to send that email.
Speaker 4:So I'm helping to manage the flow of information in a way that it's both ends to be more successful in getting what they want, and sometimes I'm offering some protection. Now again, I might do that badly. So the drawback of that model is what if I'm this man? I'm shaping the messages up and down in a way that benefits me. That's a risk, but it's not always the case. It's not always the case that having those middle layers is reducing the quality of the information. It could be improving the information flows as they go up and down.
Speaker 2:So if we've been talking about institutions, we've been talking about there's. This company structure operates on productivity in certain ways, like the formal and informal rules. It feels like this is also something that affects the whole GDP of the company. One set of rules has a certain outcome, a certain growth rate, a certain set of characteristics with it, and it would be hard for me to imagine a completely separate institutional setup having the same growth rate. This is something that touches a lot of different parts, touches a lot of how things are organized with one another. Why do companies invest so little in figuring out what their optimal structure is? Normally, what happens in the company reorg is that some executive gets promoted. They think, hey, I think it should be like this. It ends up somewhat different reporting lines are traded. You have so many parallels of this in the book. Right, it's. A new person comes in with a big idea. They founded the company on this. But if these things really do matter so much, why don't firms invest more in figuring out what their optimal structure is?
Speaker 4:Look, that's a fair question. The flippant answer would be because they haven't all read my book, and shame on them. They need to. But no, what you ask is an example of a much broader question. Look, why do some firms perform better than others? If there's something that the most successful firm is doing that other firms in the industry are not doing, why don't they copy that right?
Speaker 2:It's almost like you're leaving money on the table. That is the evolutionary process, though right If we think about that is like the Hayekian view is there's a natural selection process and the winners should have something. They should tell us something.
Speaker 4:It should be like a statue that's being etched every iteration. You need an evolutionary process, because doing this is hard. Figuring out what's the best thing to do is not easy. And look, you want my personal off the record diagnosis of why firms are not. Why do firms get it wrong in terms of their organizational structure? There are multiple reasons. One they don't know that org structure is a big deal. They haven't read my book to know how big a deal it is. If they would only hire me as a consultant, I could fix their problem. That's one possible answer. Another is even if you know that org structure matters, there are a lot of other things that matter too. Strategy matters, being able to attract and retain the best human capital and other resources. All those things matter too. Managers have limited time and attention. They just have not devoted as much effort as they could to getting the org structure right. That's a second kind of explanation. A third is they are trying like crazy to figure it out and they just can't figure it out, these contingencies that I describe.
Speaker 4:It's easy for us as outside observers, as analysts, to go in and do a postmortem on some failed company, and so clearly they didn't understand this contingency or that contingency. But it's not so easy to do it in real time, otherwise everybody would do it. It's much harder to and there would be no. There would be no, there were no leftover, no differentials to explain. So sometimes life is hard, and especially figuring out these managerial and organizational issues. They take some skill. There's trial and error. Don't forget also that we don't live in a free market. We, all of these industries, are some amount of subsidies and regulations and other things that distort that kind of evolutionary process. Or, if you want to use Hayekian terms, your discovery process of firms, figuring out what strategy, what org structure, what capital structure et cetera, is the best for them. So that might also be part of the explanation.
Speaker 3:I think a lot of it has to do with unobservability. It's very difficult for other firms to observe the organizational structure. Versus in games it's very easy to observe the gameplay, so I can make a perfect copy. Especially in the mobile gaming industry, this is extremely common. You'll have a hundred, a thousand of the same exact game with a different skin laid over top of it. So that type of that type of replication is very easy. But replicating the internal structure of a company is so much more difficult because it's not observed.
Speaker 4:Yeah, I agree, and even and sometimes it's hard for the company itself to figure out what its own org structure is, solve the problems of making its org structure better. It's just a difficult.
Speaker 2:It's just a Okay. We've kept you for a long time. We want to appreciate your time and we really thank you for taking the time to be on. We all read your book and loved it and we love we love seeing more conversation about Valve. I hope you're able to get a Supercell in the next version. It'd be great to hear your thoughts on Supercell. Let's see how they do this year before we crown the hierarchy. You take a shit on hierarchy or things that are not hierarchical based for the first part of the book and this was how damn Supercell started out flat and now they're hierarchical.
Speaker 4:I would love to explore that case in more detail in a follow-up article or book. So yeah, I appreciate the questions guys, good, interesting discussion and look forward to continuing the conversation at a later date.
