Monday, November 29, 2021

Shout or Discuss? Social Movements and Firm Collaborations

Firms are often targeted by social movements seeking to reform their behaviors. Think of the environmental destruction from fast-food packaging, the climate change driven by carbon emissions, and the wildlife reserves threatened by land development or mining. Often a single firm will attract attention from multiple social movement organizations that all want to change the same thing it is doing—such as air pollution—and its decision-makers may try to engage in discussions with them to find effective solutions. Will such engagement be possible, and if so, which social movement organizations will collaborate with the firm?

As Kate Odziemkowska demonstrates in research published by Administrative Science Quarterly, the answer lies in networks. If you read this earlier blog post and the one before, you have become used to networks being the answer to many questions, but this time you may wonder what kind of networks are in action. The answer is surprising but intuitive once you consider it: Because each social movement organization is fueled by its membership’s devotion to its cause, and some are more radical than others, what matters is the network of connections between the very radical social movement organizations and the moderate ones.

How does this work? First, firms prefer to collaborate with moderates, and moderate social movement organizations are much more likely than radicals to collaborate with firms. Second, social movement organizations collaborate with each other before they collaborate with firms, but moderates and radicals collaborate only in some social movements, not all. The effects of this collaboration (or lack of it) may seem counterintuitive: Moderates who collaborate with radicals in the social movement are more likely to work with firms, while moderates operating in a social movement where collaboration with radical groups does not happen are much less likely to work with firms. In the latter situation, moderates may fear that collaborating with firms will destroy their relationships and audience support within the social movement. Moderates connected to radicals are thus able to discuss, while moderates who lack that connection must shout in order to maintain their credibility in the movement.

What about the firm’s willingness to have discussions with social movement organizations? That part of the story is much simpler. To start, firms generally prefer to change nothing, especially along the dimensions of environmental and social responsibility that movements typically target. That is exactly why social movements target firms. Once they have been targeted by a social movement, firms do in fact seek collaborations. And they are interested in neutralizing the entire social movement, not just the social movement organization they start collaborating with, so for firms, it may be better to have discussions with moderate social movement organizations that are connected to radicals. This tactic is riskier for the social movement organization than for the firm, because radical social movement organizations may be critical of the collaboration.

This research brings up an interesting dilemma facing social movement organizations. To fulfill their goal of changing firm behaviors, they should engage in discussions with the firms and collaborate to find the best way forward. But many social movement organizations are built on the anger of their members, so talking to “the enemy” can weaken the movement. They need to have the strength to weaken themselves for the sake of fulfilling their goals.

Odziemkowska, Kate. 2021. Frenemies: Overcoming Audiences’ Ideological Opposition to Firm–Activist Collaborations. Administrative Science Quarterly, forthcoming. 

Thursday, November 18, 2021

How to Use Authority: Peer Pressure Works Best

Have you encountered organizations with authority structures that don’t quite match the training and status of the people holding each position? They are more frequent than you might think. Engineers and scientists doing research and development often work in teams and departments headed by managers who know much less than they do. Waiters in restaurants want their orders to arrive on time but are at the mercy of the cooks. And importantly for this blog post, police officers have years of training and experience but are moved around by dispatchers who receive all the calls for help (911 calls, if they are in the USA) but have much less training and status.

So, are there any problems when the authority and status do not match? This was the question Arvind Karunakaran explored in a paper recently published in Administrative Science Quarterly. As you might expect, the main problem is that the higher-status workers who have lower ranking in the organization (at least for a specific function) often feel free to ignore instructions or even orders from lower-status superiors. That sounds strange when the workers are police officers who are in uniform, need others to comply with their own orders, and are supposed to be highly disciplined.

Why do they ignore instruction? First of all, they think the dispatchers do not understand their work well enough, so they resent being told how and when to work. Also, the dispatchers often keep the officers busier than they would like to be, so officers may be interested in taking a break or not responding to dispatch calls that sounds unimportant. Indeed, not responding to dispatch calls over the radio is a common way for officers to avoid complying with orders.

How can an organization handle this type of situation, where the low-status superiors give instructions that sometimes get ignored by high-status workers? The research on police officers gave some useful lessons. First, organizations often have discipline procedures that allow immediate supervisors to refer problems upwards to higher-level managers. The dispatchers did that sometimes, and the results were… not much change. So at least for police officers, this conventional approach seems to be ineffective in the long run.

Second, people often use social approaches, such as trying to build a personal connection and using informal pressure in addition to the formal commands. Dispatchers could do that, and it was especially easy because they could choose to communicate more or less formally: to use the formal shared radio channel or a direct private channel. The results of the informal approach were… not much change. That didn’t work in the long run either.

So, what did work? One simple trick that dispatchers used was to talk to the non-responding officer informally, often with some humor, and to do it over the formal shared channel. When that was done, often other officers would join in, and the non-responding officer would end up responding and complying with orders both at that moment and going forward. So that worked.

The most interesting part of this research is why it worked. Informal talk over a shared channel could be heard by other officers with the same status and rank as the non-responding one. They would join in the chat to tease the officer but also implicitly to pressure the officer to respond. Coworkers often do so, because seeing someone else ignore instructions often means that they are slacking off, which can mean more work for others or ultimately that the non-responding team member is not reliable. People are usually sensitive to such problems, and same-status coworkers like the fellow officers can put pressure on more effectively than lower-status workers such as the dispatchers.

As always, peer pressure wins the day.

Karunakaran A. 2021. Status–Authority Asymmetry between Professions: The Case of 911 Dispatchers and Police Officers. Administrative Science Quarterly forthcoming.

Monday, November 15, 2021

Learning about Hardly Ever: The Role of Storytelling

Have you encountered the senior member of the organization who always tells stories? The one who is initially fascinating but who you gradually learn to avoid longer conversations with because the same stories are repeated over and over and have nothing to do with your work? What a bore. But storytelling can actually be very important in organizational learning, especially if it involves rare and important events.

This is what Christopher G. Myers examined in a paper published in Administrative Science Quarterly. The research looked at flight nurses in helicopters, the kind of people who have dramatic work that TV series love to portray, and on TV they typically do heroic stuff that calls for a lot of knowledge about how the patient is affected. How much of that is reality?

The learning problem is obvious. Helicopter patient pickups are done as rarely as possible because they are very expensive, and the patient conditions are always urgent and critical because that’s when helicopter pickups make sense. So, the need for heroic work is not an exaggeration. But the heroic work needs to be accurate too, and here the problem is that there are so many ways that a patient can become urgent and critical that it is very difficult for a flight nurse to learn on the job. They still need to learn about “hardly ever” events because these events occur, and incorrect treatment can be very consequential for the patient.

So what do they do? Tell stories. When changing shifts, they will chat, and the chat is regularly about what has happened in the previous shift, especially if it was unusual. That way they can learn on the job using not only their own experience but also the experience of others, as told through stories. If the story is dramatic enough, it will not just spread to the next shift but will also be retold a few times to different people, who will all learn about the “hardly ever” event and how it was solved.

This is the same social mechanism as the boring senior worker who you may have encountered, but it is a great way to learn fast in work that has much variation. Come to think of it, even that senior worker could be a source of learning because the stories most often told are usually about something from the past that does not happen often these days. Are these stories relevant? Sometimes they turn out to be. It is nearly two years ago that airlines encountered a sharp drop in the passenger traffic because of Covid. Did any of them benefit from stories about the drop in passenger traffic following the 9/11 attack? Possibly so.

Stories often involve mainly socializing and bonding, and they are ways that people form ties in organizations by sharing their experiences, whether useful or not. But in some forms of work, and on some occasions, storytelling is also a crucial learning process that helps the organization deal with “hardly ever” events.

Myers, Christopher G. 2021. Storytelling as a tool for vicarious learning among air medical transport crews. Administrative Science Quarterly, forthcoming

Friday, October 29, 2021

Networking for Success: Is It Only for the Successful?

Those who do research on social networks, and most of those who are in occupations where contacts with others give ideas, opportunities, and deals, know what a network engineered for success looks like. They are brokers at the center of a web that reaches far, where many of the people they know do not know each other. In such a network, the person at the center can combine knowledge, connect opportunities, and make deals.

So how can one get such a network? The answer is simple – by being successful. Those who have proven success become attractive to others, so they can pick and choose among those who want to connect with them. But of course, here lies the problem for those who want to build a network for success because they don’t have success yet. Can it be done? This is the question answered by Yonghoon G. Lee and Martin Gargiulo in a paper published in Administrative Science Quarterly. In addition to solving a very practical problem, this research is done in a context that we all know, directly or indirectly: songwriting in the Korean pop (K-pop) industry. Yes, this is research with BTS, not BS.

The problem for people early in their careers, or even later in their careers with no proven success yet, is that it is very difficult to get the spider-web network that combines knowledge. If they cannot benefit from such a network because they are not prominent enough yet, instead they should have a close-knit network of friends who help each other, right? And naturally, these friends will also be unsuccessful – otherwise they would have formed the looser brokerage network. That means having a network of helpful people who cannot offer much help. How can they make the transition into a better network?

The songwriters in this research were more likely to improve their networks in two circumstances: if they had peers who became successful or if they were stuck in a rut of writing unsuccessful songs that were quite similar to each other. Seeing your network contacts achieve success in collaborations not involving you is a signal that you are—or your network is—not good enough. Imagine how it feels to collaborate with someone, maybe even multiple times, without writing a successful song. But then suddenly this collaborator has another project not involving you, and it becomes a success. Feels like a failure? Absolutely – and Lee and Gargiulo show that songwriters exposed to the success of peers would reach out to new distant collaborators. Getting stuck doing the same thing over and over again with no success is also a clear signal that you are—or your network is—not good enough. Songwriters in this research who kept writing songs in the same style with no success would reach out to new distant collaborators as well.

The transition to a larger network with brokerage occurred faster for songwriters who eventually became successful than for those who never succeeded. Songwriters who get these signals that either they or their networks are not good enough—and who more quickly conclude that it is their networks’ fault—seem to have a better chance to make it. It is not easy to become successful in the music industry, and many songwriters fail even after reaching out to new collaborators. But hope springs eternal, and adding a good network seems to help. 

Lee, Yonghoon G. and Martin Gargiulo. 2021. Escaping the Survival Trap: Network Transition among Early-Career Freelance Songwriters. Administrative Science Quarterly, forthcoming.

Thursday, October 28, 2021

Risk Taking in Venture Capital: Chasing Performance, using Networks, or Something Else?

Venture capital is all about risk. The firms find young ventures, or more often select from many ventures seeking funding. These ventures have in common that they are at such an early stage that their market and offerings are unproven, and they cannot succeed unless they receive the funding, and often also advice and instructions. In other words, they are all risky.

There are still ways of adjusting the risk. Some ventures are more risky than others, and even the same venture has different risk levels depending on whether the venture capital firm participates in the first round of funding or a later round of funding. But if venture capital is essentially about risk, and they need to take this risk to gain the high returns they want, what determines the risk level they take at each funding opportunity? This is what Songcui Hu, Qian (Cecilia) Gu, and Jun Xiac studied in research just published in Organization Science. The answer, or should I say answers, are very informative.

The first part is that venture capitalists chase performance: if their performance has been disappointing, they will take greater risk than if it has been good. This is not just how venture capitalists act – firms make more changes when their executives fall behind their goals, and often these changes are risky. Indeed, the most special thing about venture capitalists is that they control the risk very easily by choosing more first-stage investments when their performance is disappointing. Other firms also try to control risk, but sometimes their risk taking is unplanned.

The second part is that venture capital firms form networks with each other through participating in funding syndicates. These syndicates divide up investments and spread risk, but they also produce collaboration, information exchange, and friendship. They result in networks that look different for each venture capital firm. Some place themselves in the center of spider-web networks that spread out widely. Others find positions between different groups of venture capitalists, becoming brokers of information. And here is the main finding from their research. These networks not only determine the information available; they also influence how venture capitalists think about opportunities available. So, what is the result?

Broker venture capitalists see more investment opportunities, and greater variety of opportunities, giving them practice assessing and experimenting with risk. As a result, they can make big adjustments of risk taking according to performance. Spider-web venture capitalists also get many offers, but they are more similar to each other, and this compromises their ability to adjust risk. The result is clear: All venture capital firms try to adjust risk to reach their performance goals, but the firms that are brokers in their networks are much more able to do so.

So, risk taking is a result of chasing performance goals, and of having the right kind of networks. Some decision makers are half aware of these adjustments, but many do not know that these factors influence risk. If they were, would they more carefully consider how goals are constructed? Would they pay closer attention to how their networks are built? I think the answer to both questions is yes. Gaining and spreading this knowledge is why we do research and teach the results.

Hu S, Gu Q, Xia J. 2021. Problemistic Search of the Embedded Firm: The Joint Effects of Performance Feedback and Network Positions on Venture Capital Firms’ Risk Taking. Organization Science forthcoming.

Tuesday, October 5, 2021

How Conservative is the CEO? Two Answers with Different Implications

Researchers have spent some time documenting that political ideology creeps into CEO thinking, and through that, influences firm decision making. This is especially true for decisions that are political to begin with, such as whether to counter economic problems with downsizing or whether to retain staffing and instead let investors take more losses. But we also know that a wide range of decisions are influenced by CEO political ideology, even those that do not look so political.

What we have not known until now is that political ideologies are more complex than the liberal versus conservative spectrum that defines US politics. A paper by M.K. Chin, Stephen X. Zhang, Asghar Afshar Jahanshahi, and Sucheta Nadkarni published in Academy of Management Journal has now given a vivid demonstration of how this complexity influecnes firms. In most other nations, conservative can mean either economically conservative, as in the type of person who values competition as a source of economic strength, or socially conservative, as in the type of person who values traditional judgment and intuition over analytical thinking. And, the same person can be either economically or socially conservative, or both, or none of them.

What does this mean for the decision making? The authors looked at corporate entrepreneurship, which is a strategically important decision that is difficult to justify analytically because the uncertainty is so high. After all, corporate entrepreneurship means diverting resources from the main business of the firm in order to enter a less familiar form of business. Corporate entrepreneurship is sometimes successful, and spectacularly so, but often it leads to losses. So how do firms decide when to engage in corporate entrepreneurship?

This is where CEO ideology comes into play. The economically conservative CEO wants to see analysis that justifies corporate entrepreneurship and wants the corporate entrepreneurship sponsors to win resources through competition with the rest of the firm. But how can they? The rest of the firm already delivers results; they just have ideas and hopes. This swings decision making away from corporate entrepreneurship.

The socially conservative CEO relies more on intuition, including the intuition of other decision-makers in the leadership team, and will be less driven by numbers and unaffected by ideas of resource competition. This is exactly the winning formula for a firm to take the risk of entering the uncertain world of corporate entrepreneurship.

These propositions make sense when considering economic and social conservativism separately, and the only problem has been to realize that these are different forms of conservative (versus liberal) ideologies. Both exist, and they have independent effects. Indeed, the research spurred by these ideas found solid support for these effects when examining how firms in Iran made their decisions on corporate entrepreneurship. Using data from Iran, which is rarely studied, is another merit of this research, but it seems fair to guess that many other nations will show exactly this division between different types of political ideology.

Chin MK, Zhang SX, Jahanshahi AA, Nadkarni S. 2021. Unpacking Political Ideology: CEO Social and Economic Ideologies, Strategic Decision-Making Processes, and Corporate Entrepreneurship. Academy of Management Journal 64(4): 1213-1235.

Wednesday, September 1, 2021

Jack of All Trades, Masterful One

I know that the title abuses the old adage “jack of all trades, master of none,” but it does so for a reason. First, as a small sidenote, the full expression used to be “jack of all trades” and was meant as a compliment. “Master of none” was added to make it less flattering. Second, I want to talk about some old research on how the old version, “jack of all trades,” might be more accurate provided the knowledge of each trade is not superficial.

In research published in Academy of Management Journal, Alva Taylor and I analyzed the collector values of old comic books. You know, the type of products that today don’t come in print, but instead appear as movies based on Marvel or DC characters. I am sure you have seen some of them. Comic books are interesting for research because we can measure the quality and innovativeness easily: high quality means high average value; high innovativeness means great variation in the value. Why the latter? Because anything new and surprising can fall flat but can also become a massive hit.

We had many findings, but I am particularly interested in the effect of creators having worked in multiple genres before making a new comic book. That’s the same as learning multiple trades because each genre has its own styles and conventions, so learning a new genre is difficult. But also, it can give fuel for innovation because knowledge of multiple genres helps the creator make novel combinations. And indeed, experience with multiple genres resulted in more innovative comic books. (It also increased quality, but that’s not the point I want to emphasize today.)

Here is the part I did not tell you yet. Comic books can be created by individuals or by teams, so we can talk about one person’s experience with genres, or the sum of genre experiences by a team. Is there a difference in which one becomes most innovative? Yes. Experience with more than three genres means that an individual will become more innovative than a team although individuals start out being less innovative. Clearly, individual creators have an easier time integrating genres.

Now there is research suggesting something similar happens not with knowledge integration, but with cultural integration in new ventures. This is important because many new ventures seek to combine the organizational cultures known by their founders into something new and unique, but often they end up adhering to an industry standard instead. Recent work by Yeonsin Ahn shows that cultural integration is helped by broader cultural experience, but only if this experience is held by individuals.

I can’t help but think that there is an interesting parallel here. It is so much easier to build up knowledge by sharing the work across individuals and forming a team. But, if the goal is to combine what has been learnt, individuals are better at it.

Taylor A, Greve HR. 2006. Superman or the Fantastic Four? Knowledge combination and experience in innovative teams. Academy of Management Journal 49(4): 723-740.

Monday, August 23, 2021

One of Us: How Women’s Inclusion Hurts Women

Have you heard stories or seen TV shows about how surgeons are the bossiest of doctors, ruling operating theatres like emperors, except that everything they do is more urgent than any imperial demands? That’s an exaggerated stereotype, but some surgeons fit it, and some forms of surgery are so exacting in process and speed that surgeons cannot tolerate slack. Many surgeons get away with bossiness because their work requires it—and because doctors are at the peak of the hospital pecking order and surgeons at the peak of the doctor pecking order. Everyone looks up to them.

Except that surgeons who are women are a little less looked up to than surgeons who are men. The usual mechanisms are at work, such as men (and some women) thinking of surgery as an activity that fits manly men better (so much cutting and bleeding...) and women generally having difficulty getting accepted in the top tiers of any occupation that has traditionally been held by men. But research by M. TeresaCardador, Patrick L. Hill, and Arghavan Salles published in Administrative Science Quarterly has found another source of difficulty: nurses.

Why are interactions with nurses problematic for female surgeons? Ironically, the source of the problem is that most nurses are women, and they interact differently with other women than with men. Nurses tend to act according to the script when the surgeon is a man: he orders, they obey. He does not need to chat or be friendly to get precise and timely work done, so the only benefit of being a friendly male surgeon is that he is seen as a nice guy. The same tends to be true when male nurses interact with female surgeons: they act according to the script.

But female nurses want – even demand – to include a female surgeon in the club of womanhood, where friendly chatting is required, members must know each other’s children’s names and ages, and work is rarely done exactly according to script. “After all,” they may think, “the female surgeon is one of us. That means she should also share some of the burden of the nursing tasks in addition to her work as a surgeon. That’s only fair. If she does not accept our requirements for inclusion and instead acts bossy, we can slow down our responses to her needs and make her job more difficult.” This is what precisely happened in the hospital the authors studied.

The result is extra work for the female surgeon and the loss of some of the special position that a surgeon has in the hospital pecking order. Maybe that’s OK because hospitals are too status conscious and hierarchical to begin with. But the problem is that only women surgeons face these demands for inclusion and the extra work accompanying it. It is discrimination against women, by women.

Is this something that women who are not surgeons should worry about? It probably is. What happens in the hospital is that different occupations interact to produce a result, and the higher-status occupation depends on the lower-status occupation for its success. That should sound familiar to many workplaces: higher-status workers are expensive, so organizations become effective by leveraging them through having lower-status workers do supportive tasks. If the supportive tasks are done differently depending on the sex of the lower-status and higher-status workers, this is an important source of workplace discrimination we need to better understand.

Cardador, M. Teresa, Patrick L. Hill, and Arghavan Salles. 2021. Unpacking the Status-Leveling Burden for Women in Male-Dominated Occupations. Administrative Science Quarterly, forthcoming.

Monday, August 16, 2021

How Smart are Strategists? Looking at Others to Find Your Own Failure

Both in research and in practical life we understand why innovations spread gradually, and why firms copy each other. Whenever an innovation is introduced, it could be good or bad, and this uncertainty makes managers hesitant to adopt the innovation until they have seen that others use it and benefit from it. There are many successful innovations in this world, but also many failures.

What about managers assessing the success or failure of their current strategy? That sounds like a much simpler problem because they know the market share, the revenue, the profit – everything they need to know in order to decide whether to stay with the strategy or adopt it. But, in research I published in Administrative Science Quarterly long time ago, I found that it is not quite that easy. Even when assessing their current strategy, managers copy each other, except in that case they are copying abandonments, not adoptions. If you had the same strategy as me and you abandoned it, I might just decide that mine is not good enough either.

How does this even begin to make sense? They have all the information they need, one may think, and do not need to look at each other. But social influence is so powerful that people copy all sorts of things (as anyone who follows fashion in clothes will know), and managers do the same when making decisions that affect the profit of their firms.

There is in fact a justification for copying abandonment. A strategy should not just be assessed based on how good it is right now, but also on how good it will be in the future. If other firms abandon because they think it is failing, then using that information is smart strategizing. Of course, it is unclear whether managers copy abandonment because of sheer social influence or because they are letting other assess the future.

I did research on how radio stations changed their format (what kind of music and other content they broadcast). To understand what was happening, I also interviewed program directors, who make the decisions, and announcers, who actually create the programs. Interestingly, many of them thought that abandoning an old format was a smart thing in general, because its market share was gradually decreasing, but they also named specific radio stations that had abandoned too soon, and without having a good alternative ready. So what managers is a mixture of social influence and smart strategizing.

This research was done a while ago, but the conclusion has become a theme in much of the research I do on managerial decision making. The smartest story of why they make decisions is not true. The dumbest story is not true either. All decision making is a mix of different influences, and managers are simply trying to balance different considerations to end up with decision that makes sense.

Greve HR. 1995. Jumping ship: The diffusion of strategy abandonment. Administrative Science Quarterly 40(September): 444-473.

Thursday, July 15, 2021

Entrepreneurship Failure: Poor Skills or Bad Luck?

We spend way too much time focusing on success. How much space in popular press is spent on the centi-billionaires and the firms they founded? How much academic research is drawn from successful enterprises, those who founded and financed them, and the CEOs currently leading them? Let’s talk about failure for a little while.

We understand that entrepreneurship success, the founding of enterprises that survive and grow, in most cases has a big skill component, though luck is needed too. Is the same true for entrepreneurship failure? Probably not, as Diego Zunino, Gary Dushnitsky, and Mirjam van Praag point out in research published in Academy of Management Journal. Skill is so important for success that we can be pretty sure it is present along with some luck. But by the same token, bad luck can sink an enterprise regardless of skill, so failure does not mean that skill is absent. It does, however, raise the possibility that skill is absent. 

Why is this important? Well, the successful entrepreneur often does not form any new enterprises because managing growth and ensuring continued success is already plenty of work, and it is rewarding work too. Failed entrepreneurs often wants to form a new enterprise, because they naturally believe that they are highly capable and just got unlucky. After all, entrepreneurship does go along with a high self-image and willingness to risk other people’s money, and these days “serial entrepreneur” is something of a badge of honor.

But what about the investors who are asked to fund enterprises? Do they look at the track record of the entrepreneur? How do they assess it? First, we need to understand that very few investors face the situation of those who were asked to help fund Amazon. Jeff Bezos told them that they had a 70 percent chance of losing their money, which is fairly realistic (actually the percentage is higher). More importantly, he had no past failures because he had never founded an enterprise – he had been an employee. Most entrepreneurs asking for funding will have a short or long track record of dead or moribund enterprises.

One simple and incorrect decision rule is to view any failure as a sign to stay away. Clearly that will exclude many skilled founders and promising enterprise. Another is to ignore past failures. Clearly that means not seeding out some entrepreneurs who really ought to get a job instead. But can potential investors thread a reasonable middle path?

Fortunately, the researchers found that they can. When assessing a potential venture investment, how promising people found it and how much they could be willing to invest was influenced by past failure, but not so much that past failure ruled out investment. Instead, past failure made the potential investors more sensitive to clues about whether they entrepreneur had skills that would help the venture. So, neither of the simple and incorrect decision rules are at work, but instead some form of middle path. This is what we want to see.

So, does that mean all is well? Not quite. We have to remember that the research shows average investor reactions, and averages are usually smarter than individuals when making judgments like this. This means that entrepreneurial failure does not cut off funding for new ventures. It does not mean that all individual investors avoid the simple and incorrect decision rules. Good news for entrepreneurship, less so for investment. 

Zunino D, Dushnitsky G, Praag Mv. 2021. How Do Investors Evaluate Past Entrepreneurial Failure? Unpacking Failure Due to Lack of Skill versus Bad Luck. Academy of Management Journal forthcoming.

Friday, July 9, 2021

Why do people discriminate?

We know that discrimination is common in organizations, in the economy, and in our social life. People are treated differently depending on a broad range of criteria, starting with race and gender, and there seems to be no form of training, qualification, or accomplishment that can help people escape discrimination. A classic example are Asian-Americans, who are a so-called “model minority” with a well-known taste for higher education. They suffer discrimination first through the accusation that they somehow do not deserve the education they have earned and then, more nastily, through violent attacks following the Covid-19 pandemic.

The fact of discrimination is well known, but the reasons are less clear – in part because there are too many explanations, and they contradict each other. Two well-known ones are taste discrimination and statistical discrimination. Taste discrimination is simple: people discriminate because they dislike, usually because others (parents? friends?) have told them who to dislike. Statistical discrimination is more complicated because the idea here is that some of those who are discriminated against should be assessed negatively, but it is hard to tell who, so the safe option is to discriminate against all. For example, an employer may think that some young women will get pregnant and quit soon and may decide that all young women should be thought of as short-term employees who do not need to be trained for promotion.

To many of us, statistical discrimination sounds like an excuse that may be true occasionally, but we assume most discrimination is based on cultural beliefs. But is that really so? Bryan Stroube has some interesting findings in research published in Administrative Science Quarterly. The findings were based on the discovery of transactions that offered reasons for statistical discrimination in one period, but these were removed later. In a peer-to-peer lending platform, there is always the concern that the loan may not be repaid, so statistical discrimination could be used to fund loans only to the most trusted social group. If the platform issues repayment guarantees, this motive for discrimination goes away. That is exactly what happened in the platform he studied.

So, what happened to the discrimination? This was a platform in China, where discrimination against women is common in economic arenas, even though women are thought to be reliable in paying back loans. You can probably suspect what happened. Women were discriminated against before the loan guarantee. After the loan guarantee, the economic security of women as lenders was no longer an issue, so women were even more strongly discriminated against.

Where does that leave the explanation of discrimination? Clearly people are capable of considering economic consequences and adjusting to them, and this affects the degree of discrimination. But at its core, discrimination is based on distaste and is culturally determined. Money is no excuse.

Stroube, Bryan K. 2021. Economic Consequences and the Motive to Discriminate. Administrative Science Quarterly, forthcoming.  

Thursday, July 1, 2021

Creative Sparks: Innovating by Moving and Processing Knowledge

Does knowledge help innovation? This is a simple question that is difficult to answer. In science, training people well enough to build on the knowledge of others is essential for advancing knowledge. But also, knowing too much forces thinking into established streams, making incremental additions easier but radical innovation harder. In business, most firms will place their bets on knowing more, to the extent of locating R&D in places with expertise, such as Los Angeles for video games or Silicon Valley for electronics and software more generally. Some firms even scatter their R&D around to have multiple listening posts to capture local expertise.

It is exactly this practice of multiple R&D teams that has helped us learn more about knowledge and innovation. In a paper published in Administrative Science Quarterly, Alex Vestal and Erwin Danneels analyze breakthrough innovations in the nanotech industry. This industry has multiple places with expertise (“hotspots”), such as San Jose, Boston, and Los Angeles, and firms have a blend of R&D teams that are in these hotspots or in places with less concentrated expertise.

So, does it help to be near expertise? It turns out that being too close to a hotspot with the same expertise as the firm is a drawback, just as scientists believe, but if the hotspot has slightly different expertise, the firm is more likely to produce a technological breakthrough. If the hotspot has expertise that is too different, a breakthrough is much less likely. The insight here is that one learns the most by being near, but not too near, the expertise of others. Maybe this is because being too close to the outside expertise means that there is little outside knowledge that needs to be moved inside the firm?

The explanation is not so simple as that. Instead, a hotspot with the same type of expertise as the firm may generate so much knowledge that it becomes difficult to process internally. But some firms had very close personal networks within their R&D team in the hotspot, which makes processing and integration of knowledge easier. For firms like that, there is no cost to being in a hotspot with the same expertise as the firm, because this makes technological breakthroughs much more likely.

Close networks among the local R&D team are not all good, however. Closely connected R&D teams are prone to ignoring knowledge gained from R&D teams in other locations, so they can fail to move knowledge that is already inside the organization but outside their specific location. As a result, the teams with close networks are less likely to make technological breakthroughs based on knowledge from outside their local hotspot.

This is interesting because it shows how the creative spark leading to innovation depends on how knowledge is moved around and processed. We have long known that hotspots for technology and innovation have knowledge moving quite freely, so firms can locate there to detect interesting knowledge and move it inside. Getting knowledge into the firm is not the same as using it effectively though. It needs to be moved to the right place in the firm, and it needs to be processed effectively.  

The key to gaining advantages is the social network inside the firm. Location relative to a hotspot of knowledge looks like an easy solution to the problem of facilitating innovations, but the firm also has to be able to move knowledge internally and process it internally. That means having employees who are willing and able to share knowledge.

Vestal, Alex and Erwin Daneels. 2021. Technological Distance and Breakthrough Inventions in Multi-Cluster Teams: How Intra- and Inter-Location Ties Bridge the Gap. Administrative Science Quarterly, forthcoming. 

Tuesday, June 8, 2021

Nepotism and Good Management Do Not Combine – or Do They?

Here is a piece of conventional wisdom that I and many others firmly believe: Nepotism is the enemy of good management because it places untested and often unqualified people in important positions simply because of who their parents are. Most of us do not work in organizations owned and controlled by families, and even some of us who do are working in organizations owned by families but controlled by professional managers who have been carefully selected. Then there is the rest of us, who worry about the capabilities of the owner family child in an executive role, and who have read the news about scandals such as the Samsung family ownership.

But every now and then some evidence appears that gives us reason to rethink our beliefs. In a paper published in Strategic Management Journal, Guoli Chen, Raveendra Chittoor, and Balagopal Vissa look at CEO pay in family firms in India, comparing those run by CEOs from the family versus those run by non-family CEOs. Their findings contained some surprises.

Here is an unsurprising finding: CEOs from the family get paid more. Sure, we all know about favoritism and about extracting resources from a firm (which also has non-family owners) to put into the family’s pockets. This is an annoying finding, but it is no surprise. Oh and by the way, the increase in pay depending on the firm performance is nearly the same for low and high performance provided the CEO is a family member, but a nonfamily CEO does not benefit from higher performance.

Here is a surprising finding: When they checked the data more carefully, they found that the family CEOs actually got rewarded very much when the firm had very high performance, less so when it had average or low performance. This is exactly how one would design an incentive scheme for CEOs, because disproportionate rewards at the high end are necessary to compensate for their reluctance to take risk. But why is the incentive scheme especially well designed for family member CEOs?

One more interesting surprise that might be telling: The relation between high performance and higher pay is particularly strong if the firm is named after the family. So, exactly the kind of firm that would embarrass the family if the performance were low has a family CEO with good incentive pay. Interesting way for the family to control their (younger) CEO member, right?

You have probably read through this and concluded that it does not matter. Having a good incentive scheme is not the same as getting high performance. After all, poorly qualified spoiled brats are not going to accomplish much regardless of how they are rewarded. True, except for one thing. The firms managed by family CEOs had higher performance on average than those managed by nonfamily CEOs. This is certainly a paper to make us reconsider our beliefs.

Chen, G.,R. Chittoor, B. Vissa. 2021. Does nepotism run in the family? CEO pay and pay-performance sensitivity in Indian family firms. Strategic Management Journal 42(7) 1326-1343.

Monday, June 7, 2021

Are There Morals in the Stock Market? Reactions to Firm Misconduct

People dislike fraud, so much in fact that they see it as immoral and are willing to make some sacrifices to punish the fraudsters. This is even true when they are trying to earn money through investments. The best evidence on this is from research on individual people managing their personal money, but in today’s stock market most of the money is moved around by professional fund managers who are simply looking for returns. Does the market still have any morals?

Research by Ivana Naumovska and Dovev Lavie published in Administrative Science Quarterly suggests there are some morals left, but they are… selective. They looked at firms involved in financial misconduct, which is the kind of misconduct that investors care the most about because it hurts them directly. (They might be more forgiving of pollution.) It is already known that investors react by withdrawing money held in the firm accused of misconduct—and also from similar firms, because the stigma of misconduct places similar firms under suspicion.

This research goes one step further by asking whether investors are not just reacting to stigmatization but are also strategic in how they respond. An interesting feature of similar firms is that many of them don’t just resemble each other; they also compete with each other. A firm engaged in misconduct is weakened by money withdrawals and other punishments, so shouldn’t that strengthen its competitors? If it does, then that could be a reason to bet money on the competition, even if it is similar. Again, a selective form of morals.

In fact, the research went even further. Recognizing that investors differ in how well they understand how firms compete, Naumovska and Lavie distinguished between the detailed analysis of firms done by mutual funds and hedge funds and the coarser understanding of other investors. All investors will react to stigmatization and competition, but the more sophisticated investors will be less sensitive to stigmatization and more sensitive to competition. More forgiving and more strategic, in other words.

Were they? Absolutely. Measuring stock market returns, it was easy to show that both stigmatization and strategic investment took place. The authors found a U-curved relation between stock market returns and the product market overlap of each firm with the firm accused of misconduct, such that intermediate levels of similarity were the worst. Importantly, less sophisticated investors punished firms more if they were more similar to the firm accused of misconduct, showing no strategic investment. The more sophisticated investors also reacted negatively to any level of misconduct but were significantly more forgiving if the two firms were so similar that the damage suffered by the accused firm might turn out to be profitable for the other firm.

So does the stock market still have morals? Some morals, and selective morals. A somewhat disturbing conclusion is that those of us who prefer to let others invest our money through mutual funds and ETFs (not hedge funds, I hope) actually make the market less of a moral place, because those who manage our money are less willing to deal out punishment for misconduct.

How much of a problem is that? Arguably punishing firms that are similar to one that commits any form of misconduct may be unfair because one should be presumed innocent until found guilty. But that overlooks some important details. First, people are presumed innocent, but we don’t need to hold firms to the same standard. Second, the stock market is not a court, and it is perfectly acceptable to move money away from the possibility of future misconduct. I am perfectly comfortable with stock market morals through stigmatization, and I am uneasy about the implications of this research.

Naumovska, Ivana and Dovev Lavie. 2021. When an Industry Peer is Accused of Financial Misconduct: Stigma versus Competition on Non-accused Firms. Administrative Science Quarterly, forthcoming. 

Sunday, June 6, 2021

Women’s Work? Firms May Penalize Social Impact Work by Male Employees

An important part of the move towards firms showing social responsibility is the spread of social impact work programs. These programs let employees offer part of their time to various initiatives with social impact. For example, my school has built a playground in an area of need. (I agree that a playground built by professors is a scary thought, but most people working for us are not professors.) Social impact work programs are thought to be very useful because they connect organizations and society much better than money donations, and many firms encourage them.

What if the same firms penalize the workers who take part in them? That would make no sense, but now we know it is happening. This is thanks to research by Christiane Bode, Michelle Rogan, and Jasjit Singh published in Administrative Science Quarterly. They looked at a major consulting firm with a social impact program that employees could volunteer for, and they checked what happened to the promotions of those who did or did not volunteer.

The good news is that it is fine to work for social impact if you happen to be a woman. Then there is all the bad news. First, the promotions of men who worked for social impact were delayed. Bad for them, who thought that the firm’s promise to encourage and reward social impact would be followed up. Bad for the firm, which claimed to want a robust social impact program.

Second, the promotions of men who worked for social impact were delayed. Bad for the firm’s (and the world’s) idea that professional workers are generally evaluated based on merit, not gender, because the difference between men’s and women’s consequences clearly demonstrates a workplace with some sexist views.

Third, the promotions of men who worked for social impact were delayed. Perhaps one of the more disturbing findings from this research was that this effect was not specific to the consulting firm. It could also be shown in a random sample of people acting as evaluators for a potential promotion. Their evaluations were very informative, and not in a good way. First, it was clear that the negative evaluation of men doing social impact work was entirely due to men making promotion evaluations. The male evaluators were the ones who acted as if they thought that men – but not women – should stay away from social impact.

Moreover, when reporting the reason for their negative evaluations of men, they were clear that a lower fit to the job was the problem. Doing good outside the job does not lower a woman’s fit to the job, but it lowers a man’s fit to the job. What is going on here? The findings raise a clear suspicion that the evaluators either know that women do lots of non-work work anyway (like most family work) or think that a dual work and society focus is fine for women but not for men.

The findings are discouraging because they suggest that a lot of education is needed to make social impact work a safe and gender-balanced effort for firms. The good news is that it is not too hard to accomplish because promotions are decided by a well-defined group of people, and it is easy to track how well they make decisions. In fact, the main reason the consulting firm findings were so clear is that they had good performance statistics for their employees, so it was easy to check whether the promotions were fair or not. This is exactly the foundation needed to make sure that promotion evaluators act in ways that match firm priorities, not oppose them.

Bode, Christiane, Michelle Rogan, and Jasjit Singh. 2021. Up to No Good? Gender, Social Impact Work, and Employee Promotions. Administrative Science Quarterly, forthcoming. 

Thursday, May 27, 2021

Creativity and Diversity: More Lessons from Management

The idea that creativity is stimulated by combining different kinds of information has been shown to be true many times, most recently in research showing that network brokerage of different groups is most effective when there is also instability of membership. Interestingly, the simplest way to combine different kinds of information is the hardest: bringing together a diverse group of people to work as a team is sometimes good for creativity, sometimes bad, and sometimes there is no effect.

How can we make sense of this? First, we can be patient. Teams trying to be creative quickly face a difficult task that many of them will fail. Instead, looking at creative efforts over time, by multiple teams, gives clearer results. Second, separate different types of diversity. Knowledge diversity creates creativity, but many other kinds of diversity have no effect on creativity but can create discomfort and difficulty working together. Unfortunately, people get along more easily with those who look like them and talk like them, so any team of people who work together can be divided by gender, race, or nation of birth. This is the reason that team diversity often has unclear effects: those who would benefit the most from working with each other often have difficulty doing so.

Alina Lungeanu and Noshir Contractor looked at the effects of knowledge diversity and cultural diversity in teams of scientists involved in the ultimate creative task: the generation of the new scientific field of oncofertility. Creativity in science is demanding because it not only requires new ideas; the ideas also must be objectively correct. Science is stricter than art in assessing the value of creativity. Creativity in science is demanding also because it requires time; it takes many years and publications to produce useful knowledge.

So, what did they find? For scientists collaborating, knowledge diversity means that they draw on different knowledge of past research, which happens to be easily measurable. More diversity produced more creativity. Cultural diversity produced less creativity. As added evidence that difficulty working together held back collaboration, they also found that scientists were especially likely to repeat collaborations with prior collaborators and were also more likely to collaborate with friends of friends than with total strangers.

This repeats a lesson that is worth repeating because it is so often ignored. The creative spark comes from encountering different knowledge, ideas, or norms. Different forms of thinking help creativity. But to make that encounter happen, people need to open up to each other and communicate freely. That requires some level of comfort with each other, so team-building efforts may be needed when team members come from different backgrounds. So first, facilitate communication, then let the communication generate creativity.

Lungeanu, A., N.S. Contractor. 2014. The Effects of Diversity and Network Ties on Innovations: The Emergence of a New Scientific Field. American Behavioral Scientist 59(5) 548-564.

Tuesday, May 18, 2021

Hiring Mental Disorder: How Employee Mobility Spreads Depression, Anxiety, and Stress

Mental disorders should be taken seriously because a condition such as depression, anxiety, and stress takes a significant toll on those who suffer it and their families, and it affects their work performance. Particularly tragic events occur when stressed operators of hazardous equipment (such as trucks, trains, or planes) make mistakes that threaten safety. The only good thing about mental disorders compared with other diseases is that they are not contagious.

Except, asdiscovered in new research by Julia M. Kensbock, Lars Alkærsig, and Carina Lomberg published in Administrative Science Quarterly, they are.

Focusing on the workplace, the authors found that mental disorders are contagious. The behavior of a person suffering from a mental disorder affects others by making their interactions and the work more problematic. The end result is that some coworkers end up with the same disorders of depression, anxiety, and stress. This is especially problematic because mental disorders spread through behaviors, so an undiagnosed patient in the organization is particularly threatening as no treatment or adaptation is possible. In a way, this is similar to how patients with contagious diseases can transmit even when they are undiagnosed or asymptomatic.

But this is not the worst part of the story. Organizations in which many employees have mental disorders – maybe because they have spread within the organization – become unhealthy, so hiring one of their employees carries a similar risk of hiring mental disorder as hiring an employee who has a mental disorder diagnosis. Of course, unhealthy organizations are exactly the places that employees want to leave in order to escape depression, anxiety, and stress, but they do not realize that they may be bringing it to their next workplace.

And there is an even worse part of the story too. To see why, ask yourself who in the organization affects coworkers the most. The answer is obvious – managers do. A manager interacts with many, influences many, and has the power to affect the work and rewards of many others. Hiring a manager with a mental disorder or from an unhealthy organization means that the organization now has a person who is fully capable of transmitting depression, anxiety, and stress to others, and the researcher team found that managers indeed have disproportionately high effects on the mental health of the organization.

We already know how the climate of a workplace, and the work done in it, is negatively affected by hiring jerks, especially jerk managers. The damage from having a manager with a mental disorder is similar, or possibly worse. But, the takeaway here is not that organizations should shun employees and managers with mental health disorders. Given their prevalence, trying to do so would have negative consequences for everyone involved. Instead, this research should be a wakeup call for any organization that is not already educating its people about mental disorders and working to improve their mental health. Mental health disorder is a treatable condition (unlike most jerks, I suspect). Given the dangers of contagion, there is no time to waste.

Kensbock, J. M., Alkærsig, L., & Lomberg, C. (2021). The Epidemic of Mental Disorders in Business—How Depression, Anxiety, and Stress Spread across Organizations through Employee Mobility. Administrative Science Quarterly, forthcoming. doi:10.1177/00018392211014819

Tuesday, May 4, 2021

Organizations Rule! When Self-Organization Became Structured

We often overlook an interesting contrast between the workplace and our private lives. In the workplace, we have an organizational structure in which people have fixed authority, are grouped together in units, and have specified processes for regular operations and for how to handle many exceptions. In our daily life, we have none of these (unless we are the von Trapp family), but we still get things done. To the curious mind, this begs the question of what organizations are for, apart from controlling people who cannot be trusted on their own.

Felipe G.Massa and Siobhan O’Mahony have published research in Administrative Science Quarterly that gives a nice answer to that question. They examined the self-organized Anonymous movement, which started off radically different from normal organizations in structure (they had none), processes (they improvised them), and ethos (freedom of information and action is paramount). Anonymous earned a reputation as an unpredictable group of activists that could suddenly descend on targets through protests and hacker attacks, seemingly organized through little else than internet forum conversations.

The only problem with this reputation is that it is only true of early stages of the Anonymous movement. They did indeed organize through shared forums and used shared norms and jargon to define boundaries and direct action. But Anonymous attracted so many newcomers that these mechanisms were no longer enough to maintain the identity of the movement and the cohesion of their actions, resulting in chaos. Reacting to this, senior members of the movement sought to use norms both for integrating new participants and for directing the protest actions they had become famous for, which were becoming less systematic and predictable.

Norms work well, right? After all, the Catholic church has applied strong norms and has been in operations for a couple thousand years. But churches are organizations too, and they use structure and processes just like any other organization. And the time came when Anonymous started looking more and more like an organization.

Anonymous now has well-defined roles, with different levels of experience and expertise determining what role a member fills. Anonymous has a hierarchy, with decisions made centrally and communicated outwards to the peripheral members. Anonymous has training of new members, manuals for how to act, and tests that allow promotion into higher ranks. In short, Anonymous is an organization.

Is this controversial or surprising? My first guess is that the most surprised people share one of two very different beliefs. One is the belief of economists that coordination of many is a simple matter of aligning preferences through some simple device, such as a price. This belief is correct, but it turns out that prices and mass-market goods are one of the few contexts in which it holds. The other is the belief of ideologists that mass movements can be coordinated by shared beliefs and norms. That is also correct, but only for short periods of time, as Anonymous found.

My second guess is that organizational theorists are the least surprised. We should not be surprised because what we have learned time and time again is that organizations are unbeatable for coordinating the actions of many, whether they be friends, strangers, or in between. Just ask Anonymous, if you can find the right person to ask.

 Massa, F. G., & O’Mahony, S. (2021). Order from Chaos: How Networked Activists Self-Organize by Creating a Participation Architecture. Administrative Science Quarterly, forthcoming. doi: 10.1177/00018392211008880

Friday, April 30, 2021

Church or Science? Moral Authority in Stem Cell Research

Who decides what can be done in society and business? The conventional answer in Europe is that it was the church in medieval times, the state after that, and now it is a free-for-all with business having a prominent role. All parts of the conventional answer are inaccurate, and we know that the correct answer depends on the context. So, let us consider a context with very strong dividing lines: stem cell research. Stem cells are “primitive” cells that have not yet specialized into specific types, so they can be used to cure a wide range of medical conditions provided they can be coached into becoming a type of cell that needs to be replaced or repaired. They are also controversial cells because a primary source of stem cells for research and production is fetal tissue – early-stage embryos.

Church beliefs on what embryos are for and scientists’ ambitions to cure degenerative diseases like Parkinson’s and Alzheimer’s clashed, and this conflict was examined by Joelle Evans in research published in Administrative Science Quarterly. In her study, a research laboratory encountered outside pressures against the stem cell research and reacted by having internal debates and forming a response. In so doing, the scientists took on a second role as creators and marketers of a moral stance explaining why stem cell research was valuable and how it should be done.

What she documented is an unusual role for scientists. More commonly, science is thought to be a free exploration of questions for which there is no moral judgment until the time comes to use the insights. This type of scientific handoff is very common, although it has been known to create complications in some cases – such as among the scientists who developed the atom bomb, with full knowledge of the purpose of their research.

Stem cell research is not nuclear physics and has no weapons application, but the question of what kind of raw materials can be used, and how they can be used, is fraught with moral problems. These moral problems have practical implications. For example, the USA has rich access to surplus blastocysts (pre-embryos) because in vitro fertilization (IVF) procedures create more blastocysts than can be safely inserted. IVF clinics routinely destroy excess blastocysts because they are barred from turning them over for stem cell research. As of 2019, a few hundred stem cell lines had been approved for use, and these are called stem cell lines because each originates in a single blastocyst with cells that keep being reproduced.

The researchers in this study faced two debates. Externally, they faced criticism for their use of stem cells and calls to account for it morally. Internally, they differed in their views on what could and should be allowed, with the internal lines of contention being shaped by the external pressures. The need to make an external account for their work was unfamiliar for researchers and made complicated by their internal divisions.

How did they respond? Interestingly, the combination of external pressure and internal fissures helped the lead researchers formulate a set of moral values that they could justify through connecting with accepted forms of ethical reasoning and explain externally and internally.

This served two purposes. Externally, the scientists gained a role in defining the value of their work and the constraints on how it should be conducted. Internally, they unified an organization that could easily have become divided, maintaining motivation for the team members troubled by the apparent conflict of moral values. They achieved strength through unity while embracing the diversity of beliefs within their laboratory walls.

Evans J. 2011. How Professionals Construct Moral Authority: Expanding Boundaries of Expert Authority in Stem Cell Science. Administrative Science Quarterly Forthcoming.

Thursday, April 22, 2021

Do People Governed by Algorithms Improve or Quit?

Wait, what does it mean to be governed by algorithms? If you do not know that yet, you are not working for any of the gig contracting platforms (Uber, TaskRabbit) or employers using algorithms to assess employees and predict and manage training and promotion. The increase in data processing capacity and machine learning tools means that algorithms have crept into a multitude of organizations and influenced how they manage people. Importantly, in some places the use of algorithms is acknowledged, and the results are shared with the employees, but in other places it is secret. Some workplaces make the algorithm transparent to employees, and others make it opaque. Because people usually learn how to game transparent algorithms to get high scores, opaque algorithms are becoming increasingly common and are currently the most important to understand.

So, what do opaque algorithms do to people? That is the topic of research by Hatim A.Rahman published in Administrative Science Quarterly. He focused on a labor platform that matches freelance workers with clients. The platform implemented an opaque evaluation routine that produced a new type of quality score for freelancers that was visible to them and to potential clients. How do people react to such scores? We know that scores become goals, and people commonly try to improve their performance by making changes. That is exactly why transparent algorithms result in inflated scores after a period of adapting to the algorithm. But opaque algorithms do not tell people how to improve, making the scores produced by those algorithms less useful as goals.

Instead of targeted improvements, opaque algorithms can produce experiments to find out what elements of the algorithm affect the score, and how. Many freelancers tried to change how they worked with clients through simple actions such as changing the type of work, length of contract, procedure for closing the contract, and so on. But these changes were unlike the changes made when decision-makers face goals that are more easily understood. As has been documented in research on performance feedback, it is very common for people facing low performance relative to a goal to react by making changes to improve the performance. That happened with the opaque algorithm too, but it was much more selective.

First difference: Not everyone tried to make changes. Many individuals who were not highly dependent on the platform responded by quitting it. And this was true whether they had high or low performance, so even many high-performing freelancers (according to the algorithm) simply left.

Second difference: Not everyone’s likelihood of making changes was a result of the algorithm score. Low-performing individuals were experimenting with different approaches regardless of whether they had setbacks in their scores or not. That was important because in the platform, a score below 90% was considered low, so the result was continuing turmoil in how freelancers were working.

Third difference: Among those who performed best and were dependent on the platform, those who experienced setbacks made changes to how they worked. So far so good, especially if those changes actually improved how they worked. But what about those who did not experience setbacks in the score? They tried to limit their exposure, including by not working with new clients on the platform. Having a high score was valuable, and accepting new work on the platform might endanger it, so they preferred to stick with existing clients or to find new clients that would let them work outside the platform.

Clearly, the opaque algorithm produced scores that made it easier for clients to distinguish between freelancers, and it also governed the freelancers by changing how they behaved. Were these changes improvements? Normally performance feedback on a meaningful goal results in improvements, but it is far from clear that an opaque goal has the same effect. Indeed, the three differences in how these freelancers reacted suggest that the opaque algorithm was a poor governance tool. 

Rahman, Hatim A. 2021. "The Invisible Cage: Workers’ Reactivity to Opaque Algorithmic Evaluations." Administrative Science Quarterly, Forthcoming.

Tuesday, April 13, 2021

CEO Power: Who is it Good For?

The CEO of a firm is given power to control the firm. That is because the purpose of the CEO role is to create a position that has centralized control, to enable consistent formulation and execution of strategy. The CEO is in turn governed by the board of directors, who also have a say on the strategy and who assess the results of the strategy execution. So far so good, but then there is the question of how much power the CEO should have, and what happens when the CEO has too much power.

Here is a classic example of too much power: The CEO can influence how the board of others assesses the CEO by choosing the reference group of other firms and CEOs that this firm is compared against. Do some CEOs really have this power? Does it benefit them? Does it harm the firm? Research by Pino G. Audia, Horacio E.Rousseau, and Sebastien Brion published in Organization Science gives the answers: Yes, yes, and yes.

The key here is that a firm can be compared either against a standard reference group such as an industry average or a reference group that is tailor-made to suit someone in the firm. Remarkably, even though the SEC guidelines strongly warn against tailor-made reference groups, 30% of the firms in the data they analyzed had them. Interviews indicated that although this choice of reference group should be done by the Chief Financial Officer (CFO), the reality was that the CEO was often deeply involved. So yes, at least in some firms CEOs had the power to change how they were assessed.

Did they benefit from this power? The simplest and arguably most meaningful measure of benefit is how much the CEO is paid. Controlling for everything else that influences pay, CEOs of firms with tailor-made reference groups were paid more when the firm performance was lower. So, clearly the point of a tailor-made reference group was to get paid more when firm performance indicated that they did not deserve it. It was an insurance policy.

Were firms harmed? The question of harm is difficult to answer, but it is easy to discover whether others tried to penalize the firm. This they did. Securities analyst coverage and rating of a firm’s stock is very important for the firm value, and securities analysts quite systematically downgraded firms to a lower rating or even stopped following them if they had tailor-made reference groups. The firm’s owners also reacted, as they saw an increase in governance resolutions filed by shareholders at the annual meetings.

The CEO use of power to shape the assessment of the firm’s performance was consequential for the CEO and the firm, and as we might expect, in opposite directions. The CEOs used their power to benefit themselves in ways that hurt their firms. This, at least, gives one answer to the question of what it means to give a CEO too much power to control the firm.

Tuesday, March 23, 2021

How Much to Know? Entrepreneur Experience and Venture Failures

We are well aware of the advantages of knowledge, both when doing something conventional and when making innovations. People who teach entrepreneurship do it because knowledge helps the foundation of a successful venture. But is there such a thing as too much knowledge?

When I work alone, more knowledge is better – it generally gives performance at a higher level, and it gives the creativity that can yield outstanding outcomes. My earlier research on innovations in the comic book industry is one of many research projects showing this effect. Things get more complicated when multiple people come together to do something complex like founding a new venture. The great benefit of having much knowledge and many kinds of knowledge is the potential for combinations that others cannot think of. The problem with doing this in a team is that combining knowledge requires communication, and communication gets harder when each member has different types of knowledge.

So how to resolve this dilemma? Research by Florence Honoré published in Academy of Management Journal shows a nice path.
The idea is simple: variety in knowledge is great as long as it stays within one person, but teams should also have shared knowledge that transfers directly to the problem they are solving.

This makes sense if we think of specialized knowledge as roughly equivalent to languages. People with shared experience speak the same language, while people who have different experiences need to translate. A person with many kinds of knowledge is a polyglot, and that is OK because we are all good at talking to ourselves.

So what did the research show? New ventures were at great risk of failing if their venture team had a mixture of experiences across team members, and this problem got worse the more founders had earlier been employed by multiple firms. On the other hand, more shared experience among the founders helped the firms survive, and this was especially true if the venture had multiple founders who transitioned from the same prior employer to founding a venture together.  

The implication for entrepreneurship is obvious. A team with experience in the same industry that they are forming a venture in will perform better, and they will do especially well if they make sure to also have one member with a variety of experiences. And, they should listen carefully to that person, because it is exactly the integration of knowledge from this polyglot member that can benefit the venture most. In other words, this research is another win for diversity in business.

Honoré, Florence. 2021. "Joining Forces: How Can Founding Members’ Prior Experience Variety and Shared Experience Increase Startup Survival?" Academy of Management Journal forthcoming.