The first source of social influence one might consider is word-of-mouth: friends telling friends about their new car. But car purchases are relatively rare. One may, however, see hundreds of new cars a month: convertibles, pickups, sedans, and SUVS passing before one’s eyes in a stream of gleaming metal.

Measuring Visual Influence
To determine whether this visual influence could be prompting people to buy cars, McShane and his colleagues started with a database of 1.6 million car sales and leases made over the course of nine years in 1,000 zip codes across the country. Running a few numbers showed that in areas where more cars had been bought in the past month, more cars were also being bought a month later. In order to determine whether this trend had to do with visual influence—and not, say, with how much money people made in those zip codes or the presence of a sale—they dug deeper.

The authors reasoned that if visual influence were at play, only the cars seen in the immediate vicinity on a regular basis would influence car sales in a county. They examined whether, for a given zip code, vehicle purchases in closer zip codes influence future purchases more than those in distant zip codes. They organized the zip codes into groups: each zip code was grouped with an adjacent zip code, a zip code 10–30 miles away, one 30–60 miles away, and one more than 100 miles away. (Key demographic information, like income, was kept the same across all zip codes grouped together to control for non-visual effects on car buying.)

Analyzing how much each zip code’s purchases affected its partners, they found that, indeed, the effect of past purchases on future purchases decayed with increasing distance between zip codes. Within the focal zip code, the effect of past purchases on later purchases was large; the effect of cars bought in the adjacent zip code on those bought in the focal zip code was moderate; from the zip code10–30 miles away, the effect was even smaller; and with the zip codes even farther away—whose inhabitants would almost never be seen in the focal zip code—the effect was not noticeable. This pattern is “at least consistent with visual influence,” McShane says.

Digging Deeper
To make a stronger case, the team looked for other variables that would be consistent with visual influence. For instance, when they saw the amount of variability in the strength of the effect, they checked to see what differed among zip codes. “What could possibly explain the fact that [the effect in] Westchester, NY, is bigger than Washoe, Nevada, which in turn is bigger than Wake, North Carolina?” McShane asks. “It turns out that, among other things, we find that in places where people commute more by car, there's a greater effect.”

This makes sense, because more commuters on the road every day means more exposure to cars. And, indeed, according to weather data the team obtained, areas with more sun showed a stronger effect as well. There were no relationships between population size or number of vehicles per household and effect size.

The team also found that there were differential impacts in the effects of car purchases by gender.  Building on prior work demonstrating that people from one social group are more likely to avoid products associated with other social groups in product categories relevant to identity, the team hypothesized—and found—that, for example, prior sales to men had a greater visual influence effect than prior sales to women among car categories more strongly associated with men than women (e.g., pickup trucks).  “It doesn't show that's it visual. But it shows that it's consistent with the literature on identity, making it much more likely social influence is at play.”

According to the team's analyses, for every seven cars that at are sold, another one is sold as a result of local inhabitants seeing the others on the road. That is, there is a spillover effect where direct purchases generate further purchases due to visual influence. These additional purchases, McShane suggests, need to be taken into account when assessing the impact of an advertising campaign: a dealer may be getting more bang for his buck than he thinks.

A significant part of the value of a house depends on what the next generation will be willing to pay for it, explains Sergio Rebelo, a professor of finance at the Kellogg School of Management. “So we are always making a guess about how well the next generation is going to do.” And the rosier our view of the future, the more we are willing to pay for existing homes.

Still, economists who tried to understand booms and busts in housing markets have had a notoriously difficult time of it. Economists assume that, while the future is uncertain, people generally agree about the likeliness of different future outcomes and operate rationally on this knowledge. And in a world where everybody has the same expectations, Rebelo points out, learning good news about the future today should drive prices up, well, today. But in practice, home prices change gradually; on average booms last six years and busts five. (See Figure 1.)

Figure 1: Housing price indices from U.S., Japan, Switzerland, and U.K. from 1970 to 2010.

Complicating matters further, Rebelo explains that “in many episodes we have a hard time finding variables that correlate well with large house price movements.” That is, factors like rental prices and employment rates, which people might use to determine whether buying is a good idea, are simply not strong predictors of house prices.

Disagreement About the Future
So Rebelo, along with colleagues Craig Burnside, a professor of economics at Duke University, and Martin Eichenbaum, a professor of economics at Northwestern University, made a key assumption: even though we are all rational beings with access to the same information about the economy, we may nonetheless interpret the information differently. In other words, we can all look at the same predictors of future U.S. growth, but still “agree to disagree” about just what those predictors indicate.

So, just how realistic is the researchers’ assumption that knowledgeable people might disagree? Consider that in 2005—just one year before housing prices began their decline and three years before financial markets imploded—the president of the Federal Reserve Bank of St. Louis suggested that house prices could potentially rise another 40 percent. In reply, the president of the Federal Reserve Bank of Dallas said to Alan Greenspan,“Mr. Chairman, I’d like to propose that he buy my house in Washington.”

Rebelo and his colleagues put their assumption to work in a model that describes both gradual boom–bust cycles and permanent price increases. In the model, some people believe that the long-term fundamentals of the economy will change for the better, and that house prices should rise, whereas other people believe that the fundamentals of the economy should stay the same, so prices should not change. (A third group of people is agnostic about future house prices.)

People meet randomly. When two people with different views meet, the one who is more certain of his viewpoint convinces the other that his viewpoint is correct with some probability.  Thus, via word of mouth, optimism about the housing market can gradually spread through a population—much as would a socially contagious disease. (Indeed, the model borrows mathematical techniques from Swiss mathematician Daniel Bernoulli, who used them to describe the spread of smallpox.)

So What Happened in 2006?
If the fundamentals of the economy are revealed to have changed for the better—after a strong uptick in employment, for instance, or an important international trade agreement—everyone agrees about market prices, and any increases in housing prices are permanent.

But where there is disagreement about market prices (with optimists believing them to be low and wanting to buy, and pessimists believing them to be high and wishing to sell), the model can produce boom–bust cycles. In a boom–bust cycle, an increase in the number of optimists drives prices up. But, Rebelo explains, “what makes the price come down is there’s not as many people excited about buying a home all of a sudden."

These dynamics can be useful for understanding what happened to the U.S. housing market in 2006, says Rebelo. “Some people say, ‘Oh, prices crashed because there was new regulation imposed on the financial industry. But this argument gets the causality all wrong. Nothing happened in 2006. Absolutely nothing. We simply ran out of new home buyers."

Keep Calm and Carry On
That is the counterintuitive theory that Sandeep Baliga of the Kellogg Graduate School of Management and Jeffrey Ely of Northwestern University’s Department of Economics advance in their paper “Mnemonomics: The Sunk Cost Fallacy as Memory Kludge.” The authors argue that human beings—even rational ones—have a limited capacity to remember the original reasoning behind their decisions. If that capacity is exceeded, the information could be lost—so we need a mental placeholder that can remind us of why we decided something, just as tying a string around your finger reminds you that you need to pick up milk on the way home from work. This kind of ad hoc “memory device” is called a mnemonic. Mnemonic devices often encode some aspect of the relevant information as well, just as the letters in the mnemonic name “Roy G. Biv” stand for the first letters of the colors in the rainbow.

“It’s a shorthand way of remembering something that you might otherwise forget,” Baliga explains. “You look at it and you say, ‘Ah ha!’ Our theory about sunk costs is like that. You made a decision early on, and it cost you some money; later, you forget the details of why you made the decision, but you remember the cost. That's the mnemonic.” The PhD student trudging ahead to complete his studies, for example, may not remember why getting a degree felt so important to begin with, but he does know that it has cost him thousands of dollars and years of effort so far. Therefore, it must make sense to complete the degree, because that “sunk cost” is a mnemonic device: it both reminds him of his original motivation and encodes some information about it. “He’d say, ‘It cost me this much time and money, so it must be important, and I'm going to carry on,’” Baliga says.

Inducing Forgetfulness for Fun and Profit
As economic theorists, Baliga and Ely could have laid out their intriguing argument (the title of their paper is a portmanteau of “mnemonic” and “economics”) and let their empiricist colleagues create models to test it. Instead, “we thought it would be fun to see ourselves if the theory held any water,” Baliga says. To simulate the experience of making decisions and then losing the memory of why they were made, Baliga and Ely devised a simple two-stage investment game. In the first, or initiation, stage, the player is presented with two numbers, each representing a possible value of a project to invest in. The game then presents the player with a numerical cost to initiate the project and asks if he wants to invest at that cost or pass. The second, or completion, stage shows the player the same information—the two possible value outcomes of the investment and the initiation cost that was already sunk in the previous round—along with a new number representing the cost to complete the project. The player can then choose to continue investing in the project or not.

But Baliga and Ely inserted a catch: each player had to repeat the initiation stage on twenty different projects before moving on to the completion stage. “When you come back to the completion stage on the first project, you’ve made so many other decisions on other projects that you’ve forgotten the specific reason why you chose to initiate it,” Baliga explains. “However, you can see exactly what you already spent.” After conducting their experiment on 100 first-year MBA students at the Kellogg School, the authors found that a significant number of them used their prior sunk costs as a mnemonic reminder of their initial investment strategy. If the sunk cost was low, they declined to complete the project; if it was high, they continued to invest. “As a player, you infer from the high initiation cost that the value of the project must have been really high—and so you figure it makes sense to keep paying in order to complete it,” Baliga says.

Prorated Fallacies
But the authors’ experiment also uncovered a novel effect of limited memory on economic decision making: a subtle variation on the sunk-cost fallacy that Baliga calls the “pro rata bias.” This bias is a similarly “irrational” tendency to figure sunk costs into current and future decisions, with the imagined goal of amortizing those past sunk costs with inflated variable costs in the present. “Let’s say you’re a pharmaceutical company who spends a lot of money on R&D,” Baliga explains. “Now let’s also say that a new drug discovered by that R&D process costs pennies to produce. That R&D money is gone and not coming back, but because of the pro rata bias, my students want to charge $5 per pill because of those sunk costs.”

Baliga and Ely were surprised to find that this pro rata bias appeared as a baseline effect in their experimental results—that is, even when the investment-game players were furnished with enough information during the completion stage to cancel out the limited-memory effects, the pro-rata bias persisted. To Baliga, this suggests that sunk-cost thinking may be “hardwired” into the human mind as a useful evolutionary adaptation. “We're programmed by instinct—it's like wanting to eat fatty food or meat whenever you see it,” he says. “We have to exert an enormous amount of self-control to avoid those tendencies. In the same way, even if memory were plentiful, we would still commit these sunk-cost fallacies.”

An Evolutionary Explanation?
So what of the fallacy itself? If it is a useful mental shortcut passed down to us by our savannah-dwelling ancestors, how can it also be “a mistake,” as Baliga calls it in no uncertain terms? “We call it the ‘theory of the second best,’” he says. “The more you can remember, the better it is, and you don't have to commit this fallacy at all. But if it's a common feature of human beings that they do forget stuff, then the sunk cost fallacy is an optimal response to that forgetfulness. Sunk costs can encode information about decisions you made in the past, and if that's the case you should take them into account, because if you didn't, you'd make even worse decisions.”

Meanwhile, what about corporations, with “memories” limited only by their employees’ prowess at record keeping? According to Baliga, companies and businesses follow sunk-cost biases as often as individuals do. So why would a company like IBM or Apple need the same mnemonic “kludge” against forgetting that a single fallible person does?

“The memory-limitation problem is actually bigger in organizations,” Baliga asserts. “Costs on a balance sheet are easy to keep track of. But strategic vision—‘why was this a good idea in the first place’—is not. As managerial turnover occurs, an organization forgets. You may inherit a previous CEO’s projects, but you can’t directly access his vision or strategic reasoning. What you can do is say, ‘The previous guy was smart, and I'm going to carry on because I can see what costs he sunk into it.”

Baliga cautions that the sunk-costs-as-mnemonic theory does not imply that every incoming CEO should blindly march ahead on whatever plans they inherit. Steve Jobs, for example, famously scrapped most of Apple’s then-current product lines when he retook the company reins in 1997 and led the organization to record profits. “In that situation, though, Jobs had clear evidence of what wasn’t working in the business, so he changed course,” Baliga explains. In contrast, he says, current Apple CEO Tim Cook is continuing on the path set by Jobs, even though he may not literally share the vision that started it all. “It's up for interpretation, because that type of ‘soft information’ can never be fully recorded,” Baliga says. “Our research suggests that it may be prudent not to ignore the decisions made by your predecessors. They might encode things you don't know, and may never know.”

As with bees, so it is with business. Researchers have long held the conceit that innovative entrepreneurship is impossible. “They assert that ‘entrepreneurs can’t do anything new in the economy,’” says Daniel Spulber, a professor of management and strategy at the Kellogg School.

This is because established firms have several advantages when it comes to taking inventions to the market. “They have all kinds of assets that are complementary to innovation,” Spulber says, including established corporate structures, marketing channels, an existing customer base, and access to capital. To go it alone, on the other hand, the inventor must undertake the cumbersome effort of setting up a new firm and dealing with the uncertainty that any new invention faces in the market. Even Joseph Schumpeter, perhaps the greatest advocate of entrepreneurs, suggested in his classic book Capitalism, Socialism, and Democracy that only large companies have the resources and market power necessary for innovation.

“Yet we see innovative entrepreneurs all the time—Larry Page and Sergey Brin at Google, Pierre Omidyar at eBay, Mark Zuckerberg at Facebook, and Jeff Bezos setting up Amazon.com,” Spulber points out. Plainly, then, the negative view of entrepreneurship’s contributions to the economy has as little credibility as early bumblebee aerodynamics. “So if you have a great idea that would result in an innovation in the marketplace—why not just transfer it to an existing firm?” Spulber asks rhetorically. One reason may be tacit knowledge: the fundamental understanding that inventors possess about their own creations.

The Entrepreneurial Advantage
As part of a broad project to discover why we need innovative entrepreneurs, Spulber applied game theory to model the economic importance of tacit knowledge. Such know-how “is difficult to transfer to others,” he says. “I conclude that tacit knowledge can be an explanation for why inventors become entrepreneurs.” Just as composers can often conduct their own works with more authority than professional conductors, entrepreneurial inventors can translate their own findings into marketable products more effectively than established organizations that buy or license the rights to the findings.

“The main idea is that the inventor and the existing firm have to choose to compete or cooperate,” Spulber explains of his model. In the first of three stages, “inventors invest in R&D and produce an invention. Existing firms may themselves invest in invention, but they also invest in capacity to absorb the inventions of others.” In the second stage, he continues, “inventors decide whether they want to share that invention through licensing or sale of patents. Existing firms have to decide whether they want to stick with what they’ve got or buy ideas from others.” In the third stage, collaboration or competition commences. The model attempts to specify how much inventors will invest in developing their tacit knowledge and how much existing firms will invest in absorbing the knowledge of others.

“The inventor’s tacit knowledge is important because it can overcome the competitive advantages in technology implementation that existing firms derive from complementary assets,” Spulber writes. That is, the difficulty of transferring tacit knowledge puts the existing firm at a disadvantage, and however effective the transfer, existing firms cannot possibly know as much as the inventors about their inventions.

And there are other benefits to an innovative entrepreneur. Says Spulber, “having entrepreneurship available means that inventors will invest more in developing inventions, because they may be able to develop them themselves.” He continues, “This persuades them to invest more in the entire effort of inventing. It is the payoff from more original invention.”

In addition, innovative entrepreneurship tends to result in greater returns on increasing R&D investments. This is because the innovative entrepreneur must cut prices to compete with existing firms, thus driving up sales. Greater sales then lead to greater returns from improved products and more efficient production processes.

Finally, Spulber’s research indicates that the character of any particular invention determines the route it takes to the market. “If you have a very high-quality invention”—that is, one with great market potential—“and the associated tacit knowledge is very difficult to transfer, you’re more likely to become an entrepreneur,” he explains. “Lower-quality inventions are more likely to be transferred for a given amount of tacit knowledge.” The development of online book sales, for instance, exemplifies a high-quality invention. “Jeff Bezos must have contemplated the possibility of working with traditional booksellers Barnes & Noble or Borders,” Spulber speculates. “But he struck out on his own and created Amazon.com.”

Implications for Public Policy and Management
Spulber, who is completing The Innovative Entrepreneur, a book slated for publication next year, notes that his work has public policy implications. “The important purpose of patent systems is to provide inventors with the right to exclude access to their invention in return for asking the inventor to disclose details to others,” he says. “Recognizing the tacit knowledge of the inventor indicates that patents don’t contain everything. That fact makes patents even more important. By protecting inventors’ intellectual property, we want to encourage them not only to disclose their knowledge in the patent but also to commercialize their invention and hence share their tacit knowledge, or apply it through entrepreneurship. Entrepreneurs benefit from patent protection.”

Corporate executives should also take note. Spulber refers to Shuji Nakamura, the engineer who invented blue, green, and white light-emitting diodes (LEDs) and the blue laser. Nichia Corporation, the small Japanese company that employed him, gave him very little help. “Nakamura purchased the machine needed to manufacture the LEDs, took it apart, and reconfigured it until he made it do what he wanted it to do,” Spulber points out. “He achieved what many others with more funding and personnel were unable to do. The information he published—even though his company did not want him to—had a lot of value. But his effort must have involved a lot of tacit knowledge known to him alone.”

Then, after Nichia gave him a bonus of just $180 for his invention, Nakamura not only sued the firm, eventually settling on a $9 million payment; he also decamped to the University of California, Santa Barbara. “Tacit knowledge can be quite important to corporations,” Spulber observes, “as this ‘asset’ has legs.”

That example carries a message for companies that encourage ‘intrapreneurship’—entrepreneurial behavior within the corporate environment. “Companies need to understand that just having access to products invented by their employees is not the whole story,” he says. “Talented employees have plenty of knowledge in addition to the inventions they make available to their employers. Employers should consider involving those employees in the process of developing their inventions rather than just taking the inventions and running with them.”

Yael Hochberg, an assistant professor of finance at the Kellogg School who has previously researched cooperation strategies between venture capital firms, was intrigued by the challenge of measuring competition between VC firms in markets where cooperation was common versus markets where cooperation was uncommon. She collaborated with Michael Mazzeo, an associate professor of management and strategy at the Kellogg School, who has researched competition in other industries.

Cooperative Ties and Networking
Venture capital funds receive money from investors to nurture start-up companies in need of capital. These start-up companies are typically in the types of industries where there is initially plenty of growth but not necessarily much profit, which makes borrowing money from a bank difficult or impossible. The VC funds essentially buy pieces of the company, and investors from the VC firm often serve on the start-up company's board of directors, or they help to direct and advise the company.

Mazzeo says there are two ways that venture capital funds can be successful. The first is to identify potentially successful companies at a very early stage, because this allows them to buy pieces of the company while they are still relatively inexpensive. The second way is to provide the industry experience or insider knowledge to help make strategic decisions within the start-up after making an equity investment. "This is one of the reasons that the industry is geographically segmented, because there is an important informational component associated with getting the right deals up front," Mazzeo says. "Also, the VC wants to be involved after the investment is made, and that requires co-location."

Here is where the cooperation kicks in. "There is an element where sharing information and resources between firms proves to be useful, both in finding the good deals that are out there, and in managing them successfully to the point where they grow into good, independent companies," Hochberg says. If a firm finds a deal that is a good fit for its expertise, but does not quite have the capital to cover a particular investment, or vice-versa, cooperation can be beneficial. Alternatively, firms may want to share the risk of an investment to diffuse potential losses. "This type of cooperation between firms also operates largely on a quid pro quo basis because the expectation is that if I invite you to come invest with me when I find eBay, then you'll invite me to come invest with you when you find Google. Which is why we tend to see the same groups of investors co-investing on many different deals together.”

Measuring Competition
Hochberg and Mazzeo used a standard database of U.S. venture capital firms to trace the cooperative ties between firms via their co-investments. From these ties, they created a map of networks within each geographically based market. They then broke their sample into two halves: networked markets and non-networked markets. Next, they developed a model that allowed them to measure competition in both samples.

"We used a method that allowed us to understand how competitive different firms are … based solely on how many firms are in the market and how many firms are operating relative to the market size," Mazzeo says. Essentially, the researchers used the amount of additional demand needed to support one more firm in the market as a proxy for competition. Says Mazzeo, “The more demand that is needed to support an additional firm, the more competition exists in that market, because the price effect”—the downward pressure that increased competition imposes on prices—“from the additional firm requires more quantity to generate the same amount of revenue." Mazzeo and Hochberg looked at whether competition would be affected by how many VC firms operated in the market, as well as the extent to which each firm specialized in a particular industry (though the effects of specialization are beyond the scope of this article.)

A Pattern Reversal
Mazzeo and Hochberg found that the pattern of competition in networked VC markets looked very different from what was observed in other industries. "In other industries what you see is that the first competitor that is similar to you to enter the market hurts you a lot, and the second competitor hurts you a little less, and the third even less," Mazzeo says. "But that flips around in the venture capital industry, where the first competitor that is similar to you to enter the market doesn't hurt you very much, but the second competitor hurts you a little more and the third hurts you even more."

The reversed pattern is entirely consistent with the existence of cooperative networks, Hochberg says. "This makes sense because there is a beneficial element to the first competitor in the market if you are working together and sharing resources. But that benefit begins to go away with the second competitor, and it's even less with the third."
Interestingly, other research by Hochberg finds that VC firms sometimes penalize other firms for letting additional competitors into what had been a tightly networked market. "Sometimes they were shut out for a very lengthy period of time," Hochberg says. The networks also work to effectively exclude new competitors from markets, as start-up companies in need of capital and expertise are rapidly serviced by the existing network.

The weakening of competition seen in networked VC markets goes to show that it is not just whom you know that matters in this business, it is whom you choose to build bridges with.