DAVOS, Switzerland — Many economists have good reasons to be more dismal than usual. After all, their academic discipline took a bashing in the wake of the financial crisis, and the practical relevance of economic insights has been repeatedly questioned.
It’s all the more surprising, then, that ever more companies are hiring chief economists. Slowly, a whole series of industries that were not traditionally known for a focus on economics have begun to see the benefits of economic expertise, ranging from retail to airlines to human resources management.
Let’s call them “tech economists,” a new generation of microeconomists operating at the intersection of economics, big data and computer science. Well-known examples include Hal Varian (Google), Preston McAfee (Microsoft), Patrick Bajari (Amazon), Steve Tadelis (eBay) and Paul Milgrom (Auctionomics). It’s a small but increasingly sought after group that gives companies an important edge over their competitors, sometimes worth hundreds of millions of dollars.
And it’s an alliance proving attractive to both sides, with economists gaining access to resources and opportunities to refine academic hypotheses normally out of reach in universities.
But how did economists suddenly become so popular?
When economists use models to explain and predict, they have to make assumptions. The accuracy of these assumptions depends on the information available and in turn determines the usefulness of the models. Right now, the total amount of information in the world is expected to grow by a factor of 300 between 2005 and 2020, from 130 to 40,000 exabytes. That is roughly equal to a 60-minute episode of your favorite TV series running continuously for over 4 billion years. And it is good news for economists.
“The rise of big data, along with an increasing sophistication of tools to analyze it, has improved the ability of economists to make good assumptions,” says Silvia Console Battilana, CEO of Auctionomics.
Microeconomists benefit from this development more than macroeconomists because they can zero in on an individual market. This narrower focus allows them to define market participants more accurately and better understand how they will react to different incentives.
The main questions that economists are concerned with are “what if” questions. What if you raise the price of a product? What if you increase the number of market participants? What if you change a rule when bidding for your product? How will the buyers, sellers and competitors react?
Microeconomists are well placed to answer those questions for several reasons. One is that they are taught to think conceptually about aspects relevant to a marketplace, including considerations for demand elasticities and business models.
Another reason economists are so useful in the data era is that they have been trained in the mathematical models and statistical analysis necessary to make sense of big data. Of course, they are not the only group that has statistical and mathematical expertise. So do, for example, mathematicians and physicists. But the crucial difference is that economists tend to be less abstract in their thinking than their peers with similar skills, which also has to do with the close links that exist between economics departments, business schools and the private sector.
Armed with data, economists can now answer questions about pricing and competitor behavior much more accurately. In the best case, data eliminates the need to make assumptions altogether. After all, why speculate about how people might behave when you can find out how they do behave?
You don’t need tech economists for data aggregation or simple statistical analysis, but they are almost irreplaceable in a couple of vital areas.
One of them is machine learning, roughly understood as the ability of computers to program themselves via a particular algorithm. We see this in Google’s predictive searches or how eBay with time seems to show you more interesting products. These sites’ algorithms combine data on what is trending with other users with data they have on your preferences. Most of the econometric modeling that predicts what link on a webpage is going to get clicked is done by economists.
Tech economists also have an edge in providing strategic advice to chief executives. This is a question of marketplace design. Let’s say your business model is based around auctions. Should you have more or less bidders? What should be the rules of bidding? When mining or drilling rights, government contracts and public goods are auctioned off—or when tech companies such as Google and eBay consider changing their auction algorithms—such questions might be worth millions and sometimes even billions of dollars.
This is what Battilana’s company, Auctionomics, does. An economic consultancy, it focuses on auction design and strategy and charges higher fees than established players such as McKinsey for its services. “Auctions can be gamed,” its CEO explains.
“This does not mean to break rules, but rather to understand them and use this knowledge to the benefit of the client,” says Battilana. “It can mean gaining a strategic advantages over other bidders or setting the rules for the auction as profitably as possible for the auctioneer.”
And the payoff for such strategic advantages can be huge. In an auction of mobile phone frequencies in Switzerland in 2012, Swisscom paid $394 million to acquire 42 percent of available frequency bands, while its competitor Sunrise spent $528 million to acquire only 40 percent. Advised by an analyst of Auctionomics, Swisscom paid $134 million less than its competitor and still acquired more of the spectrum. In a similar auction in 2006 in the U.S., economic consulting helped two firms save $1.2 billion.
You might assume that in an environment dominated by big data, the theoretical side becomes less relevant, as the data “can speak for itself.” And it is true that technological progress in data analytics means economists can directly test their hypotheses through randomized trials using previously unimaginable amounts of information. But the problem is that analyzing such data volumes costs up to thousands of dollars and requires hundreds of machines working simultaneously. By providing the right direction, economists can save time and a lot of money.
“If you ask the right questions, you have a greater ability to let the data speak, and so you can be much less reliant on assumptions. You still need a strong conceptual framework to understand what’s coming out,” explains Susan Athey, a John Bates Clark medalist and an economic adviser to Microsoft.
So the opposite is really the case: Academic theory is going hand in hand with practical influence. Hal Varian was an academic expert in information economics before joining Google in 2002, where he was put in charge of refining AdWords, a product that generates over $40 billion in annual revenues. Athey initially won worldwide recognition for her academic work on timber auctions. She then found herself tasked with turning Bing into a serious competitor to Google and teaches, among classes, Economics of the Internet at Stanford.
Making economics useful, and generating value and improving what they teach, tech economists are an interesting response to the many criticisms leveled at their field. With academia and business becoming ever closer, the number of tech economists is likely to grow.