Wednesday, September 21, 2016

Wide Research, Narrow Effects: Why Interdisciplinary Research – and Innovation – is Hard

Interdisciplinary research is seen as very valuable for society and economy. Some of that could be hype, but there are some good examples of what it can do. You have probably noticed that oil is no longer 100 dollar per barrel, and the US is no longer a big importer. This is a result of fracking, a result of interdisciplinary research. And if you don’t like fracking, a good alternative is photovoltaic energy, which comes from the sun, and from interdisciplinary research.

So some interdisciplinary research has been good for society. Is it also good for the scientists who are supposed to do it? The answer to this question is very interesting, and is reported in an article in Administrative Science Quarterly by Erin Leahey, Christine Beckman, and Taryn Stanko. The start is easy to explain: interdisciplinary research is less productive, but it gets more attention. The answer got more complicated, and more interesting, when they started looking at why that happened. 

The first step was to look at whether interdisciplinary research is more difficult to do, or whether it is because it is harder for it to gain acceptance and get published. The answer is clear: it is not harder to gain acceptance, but it is harder to do, especially early on. The second step was to look at why this research got more attention. Here many factors played a role, but one stood out to me: Actually what increases especially much is the variation in how much attention interdisciplinary research gets, and that helps explain the increased average. So interdisciplinary research is related to fracking in one more way – few reap the awards from it.

This paper doesn’t really result in career advice for scientists, because everyone will be interested in different kinds of research, and have different ideas on how much risk to take on. But has important insights on how innovations are made. Building on closely related ideas is much easier to do, so no wonder much of what scientists – and companies – do is incremental. And this is true even though we often tell stories of the great successes of interdisciplinary research and integrative innovations, while forgetting all those who tried and didn’t succeed. Whether that means we cross-fertilize knowledge too little, too much, or just enough is hard to tell.


Thursday, September 15, 2016

Facing Ragnarok: How Community Diversity Helps Disaster Recovery

Let’s start with religion. Ragnarok is a series of disasters on earth, ending with the great battle of the Norse Gods against Giants and beasts. For those who don’t read sagas (or play the computer game Ragnarok), the Norse Gods are very diverse, which helps them defeat their enemies. So, is diversity also a good thing when facing adversity in our world?

A recent paper in Administrative Science Quarterly by Sunasir Dutta tells us that the answer is yes and no. Let me explain. He examines the effect of natural disasters on communities in California, which of course is one of the few states with enough disasters to do such a study. He is interested in whether the communities can found new human services organizations to help disaster recovery. The answer is that communities with more diversity of voluntary associations are better able to recover from disasters, and this effect is bigger for more unexpected disasters and more complex disasters. An unexpected disaster would mean something that the community does not expect, like a flood in the Southern California counties that often get hit by wildfires instead. A complex disaster is when multiple events happen in the same year, like an earthquake and a wildfire (I am not making this up – it happens). So part of the answer is yes, and it is a good illustration of how communities build organizing capacity that can help them later on.

But the answer is also no. Political diversity makes a community less capable of founding new human services organizations to help disaster recovery, possibly because it is related to disagreement and polarization that complicates the unified effort needed to form human services organizations.

This is an interesting contrast because it illustrates how diversity can have many different effects. Voluntary organizations provide a community with models of organizing, trained volunteers, and networks of people who help each other. They are a form of organizing capacity that gets stronger the more kinds are present. I have written more about this in a blog post on how research shows that communities are imprinted with the memory of past organizing. On the other hand, political views are markers of ideological boundaries. They also represent different views of who are responsible for community help, how it should be organized, and how it should be led. No wonder these forms of diversity have opposite effects.


Wednesday, September 7, 2016

Republicans in the Board: Leadership Trumps Collective Accomplishment

Facebook is the quintessential small company with high value that are becoming more common in the modern economy. How small is small? As of December 2015, it employs 12,000. What about Google? 57,000 employees. And if we turn to the other side of the modern economy, services, we find that Walgreens has 240,000 employees – I am deliberately not showing Walgreens instead of Walmart because we all know that Walmart is a giant firm. Taking the step back into industry, the formerly dominant part of the economy, Ford Motors has 199,000 employees (again I am not picking the largest, which would be GM).

When these firms do well – or poorly – who do we credit or blame? The answer is surprisingly political. In a recent study of executive compensation published in Administrative Science Quarterly, Abhinav Gupta and Adam Wowak examine whether the different views of leadership held by Republicans and Democrats affect how Chief Executive Officers get paid. The idea is simple – members of the board of directors determine how much the CEO gets paid, and they reveal their political views through donations to the parties (naturally they can donate to none, one, or both). If the board members believe that CEOs are largely responsible for what ten to hundred thousand people do, they will pay the CEOs more on average, and they will also make the pay more sharply dependent on the company's recent performance.

It is not a secret that Democrats attribute a lot of credit to the collective effort of workers and managers, and less to the CEO. What has become clear recently is that there is a clear authoritative streak in the Republican Party, and accordingly a tendency to credit the results to the CEO. Do the results of the study bear this out? Abundantly. A Republican dominance of a board leads to higher CEO pay. It also leads to more dependence of the CEO compensation on accounting profit and stock market value increase. Even more revealing: these relations are stronger when the analysis is limited to the compensation committee, which is the subgroup of the board that determines CEO pay (usually with the help of consultants).

So now we know that CEO pay is political, in additional to the earlier findings showing that it is customary and performance dependent. Does that make CEO pay unfair? Well, actually the answer to that is political too. Consider how you feel about this issue; it probably fits your political views in general.


Wednesday, August 31, 2016

Adam Smith Meets Machiavelli: When in Markets, Connect to the State

We have heard a lot about lobbying in the US, and its effects on law writing, law enforcement, government contracts, and subsidies. We also know that in authoritarian governments, firms can act as the arm of the state, and firms and state can be mixed up to such an extent that it is hard to tell who is running who. A classic example is the People’s Liberation Army in China, which owned 20,000 companies 10 years ago (and was ordered to sell them). If state relations help firms in both market economies and authoritarian states, what happens when an authoritarian state gets a market economy? Again China offers an example.

A forthcoming article in Administrative Science Quarterly by Heather Haveman, Nan Jia, Jing Shi and Yongxian Wang looks at whether firms profited from connections to the state. Of course we expect the answer to be yes, but the interesting part is when the firms profited most. Here are some interesting highlights. First, market development reduced profits, so competition works. Adam Smith would be pleased. Second, state connections had little overall effect on profits – surprising, right? But remember that this was a period with changing market development over time, and the interesting finding is this: State connections increased profits when the market became more developed. Machiavelli would be pleased – even when markets are strong, an authoritarian state can Trump them.

How does this happen? The short answer is that we don’t really know the details; there are simply too many ways that a state can support specific firms that it likes, and can place other firms at a disadvantage. That’s exactly why state connections are suspicious; their results are hard to trace. At least one advantage with an authoritarian state is that it is not subtle, so some of the things it does are easy to trace. Haveman and her coauthors found that resources were directed to the state-connected firms, which explained some of their advantage, but far from all.

Will this pattern continue? Well, I have also done research on China with Cyndi Man Zhang, and we found that Chinese managers were still heavily state-oriented in their actions. Maybe that is why the ones that could also use the state did well. But the advantage is so large that it is hard to see them stop doing it, especially when they got better at using the state the further along the transition to a market economy had gone. So Adam Smith has met Machiavelli and been defeated.


Wednesday, August 24, 2016

We are the Uber of Hype: When Entrepreneurs Crowd

Teaching entrepreneurship at a business school has its annoyances, and prime among them is students visiting to seek opinions and support for their business idea, which they describe as “We are the [insert famous company] of [insert market]”. That is a quick and catchy way of summarizing the idea, but there are problems with it. The main one is that it also summarizes that the idea is copied from something else. Copied ideas are not always bad, but very often they are. The first problem is that copying is easier than originality, so a copy faces competition. The second problem is that copies of successful firms (and yes, they are more common than copies of failed firms) involve too much excitement to understand the source of success. I have earlier written about DoorDash, which can be described as the Uber of meal delivery. Except that Uber gets its competitive advantage from knowing where all its cars are, always. It is not so hard to know where a restaurant is, so DoorDash is uber-like without a key uber advantage.

Well, going against logic and annoying business school professors are not decisive problems of copying successful firms. After all, none of them say what will happen in the market. So what happens, really? Now we know, thanks to a forthcoming paper in Administrative Science Quarterly by Elizabeth Pontikes and Bill Barnett. Their idea is simple. Look at spectacular successes in the financing of ventures, as well as agonizing failures of firms, and see if they result in crowding around successes and flight from failures. Then, see if the success crowd fails, and the entrepreneurs who stay near failures succeed.

Their answers are in support of the annoyed business school professors, and against the “uber of something” entrepreneurs. Yes, entrepreneurs crowd success and flee failures. But also, venture capitalists react exactly like business school professors, and they simply turn off the funding for entrepreneurs who crowd success. Also, firms that crowd success are more likely to fail, whereas firms that follow failures are less likely to fail. Clearly the impulse to chase success is a trap, whereas trying near a failure calls for thoughtfulness, which is often a step toward good outcomes.

So does this mean that entrepreneurs should be thinking about how to make a Blackberry variation, given its recent failure? Well, not so fast. Unthinking copying of anything is a problem, so they should not unthinkingly enter near a failure. Instead, they should think carefully about the elements of success, and those that led to failure. Blackberries had devoted followers because of the keyboard, and were also well liked for their security. They were hit by the sheer breadth of smartphone capabilities, and also by the market power by the largest high-price firms (Apple, Samsung) and the race upward in the market by the low-price firms. Is it still possible to make a device that keeps the Blackberry strength and survives the current market? Asking that question is the way to get from failure to success.

Pontikes, Elizabeth G., and William P. Barnett. 2016. "The Non-consensus Entrepreneur: Organizational Responses to Vital Events." Administrative Science Quarterly forthcoming.

Wednesday, August 17, 2016

Please Lobby Someone Else: Social Movements Can Make Firms Unpalatable

Under fundraising rules following the Citizens United v. FEC Supreme Court, funding needs for politicians are so great that a member of congress is expected to raise at least $18,000 per day, every day of the year. For new members, that can mean as much as 30 hours per week on fundraising calls, an amazing use of time for members of a legislative body that is supposed to deliberate to govern and make laws. Equally important, many of those calls have to be to firms, who can afford higher donation amounts than individuals. Firms also like this arrangement, naturally, because lobbying for favors involves donations as an important tool.

So is it possible for a firm to lose touch with politicians? Mary-Hunter McDonnell and Timothy Werner has a forthcoming article in Administrative Science Quarterly showing that this can happen. The idea is simple, but powerful. Politicians are stuck between those who fund their reelection and those who actually vote. Often voters don’t pay close attention to the funding, so politicians can take money from a broad range of firms, but social movements can single out firms for attention. This does not mean that voters will also notice whether these firms have political connections, but politicians cannot take the risk that they will notice – so they stay away from the firms that get singled out.

How do we know? It is actually quite easy to see how well connected a firm is politically. It can donate money (but politicians can return the donation), it can be invited to testify to congress, and it can simply get contracts to do work for the government. All of these connections get weaker if social movements target firms for boycott actions, they showed. And even more importantly, the research found that the weakening of connections was especially pronounced if the social movement issue had much voter attention, or if the firm was well known to begin with. This makes sense because the risk that voters will notice is especially high if either the firm or the issue is well known.


What does this mean for firms, and for society? Well, for firms the lesson seems clear – they need to avoid the attention of social movements in order to keep the attention of politicians. Not a very ethical implication, but a clear one. And for society? Well, it may seem nice that social movements can isolate firms from politicians through boycotts, but is it really? The main reason it is possible is that other firms are happy to pick up the slack through lobbying more when some firms are isolated. And those firms are not necessarily better. We live in a world of B2C (business to consumer) and B2B (business to business) commerce, where B2B is much bigger and less exposed to boycotts. If there is a boycott of Victoria’s Secret (which has happened) and defense contractor Lockheed Martin picks up the slack, do we have a better society?


Thursday, August 4, 2016

Keeping Threesomes Stable and Creative

Just to halt any false expectation you may have, this is a blog post about how sets of three (or more) firms can collaborate for innovating over time. It is a very interesting and important topic because the approach for doing innovations has changed a lot recently, and especially in fast-moving industries involving information technology: apps, smartphones, internet servers, automotive computer controls for example. We have long done research on firms being innovative (on their own), and we are beginning to know a lot about firms pairing up to make innovations. Unfortunately, the new approach is to do innovations in groups of three or more, and to keep working in different configurations of firms. That is something we know little about.

Fortunately, a forthcoming paper in Administrative Science Quarterly by Jason Davis has important insights. He looks at the central problem of how these groups of firms can stay together and maintain the high trust and low conflict needed to be innovative over time. Using data on collaborative projects from firms in the computing industry, he found a key insight that is only possible by understanding how different interfirm collaboration is from interpersonal collaboration.

People can form stable collaborative relations if one of them acts as a leader and is recognized as such, or if they form a cohesive team with all members staying close. That’s not what makes firm collaborations stable and innovative. Firms cannot have one leader over time because repeated projects typically involve different configurations of interests and capabilities. They cannot be cohesive because each project typically has greater complementarities between one pairing (dyad) of firms than between the other two. As a result, the configurations that work for people fall apart in conflict or in distrust when firms try to repeat collaborative innovations.

Instead, what is special about firm is recognition that firms are interdependent and that they need to work together in the future, even if they do not collaborate closely at a specific time point. This leads to a form of collaboration called group cycling: each pair of firms that are most interdependent works together closely, with the third at a distance, but they remain in touch and aware of the need to reconfigure with a different focal pair later on. Conflict is avoided because the most complementary firms work together. Trust is kept because all partners see current collaborations as preparations for future collaborations.

Interfirm collaborations is a special kind of network among organizations. Part of its special status comes from the realization that, unlike many other networks, what works for firms is different from what works for people. This is a good insight in management because we often think that intuition carries over from one level to the next.

Davis, Jason P. 2016. The Group Dynamics of Interorganizational Relationships: Collaborating with Multiple Partners in Innovation Ecosystems. Administrative Science Quarterly, forthcoming.