When the Value of Information is Negative

It's 2023, and data scientists develop a DNA test that perfectly predicts each person's future disease and thus each person's exact lifetime medical expenses.

• In one country, insurers price all "insurance policies" to fully reflect these future costs. Everyone pays exactly what they will get back.

• In a second country, insurers are required to provide insurance at standardized rate which ignores the DNA test and everyone is required to buy it.

Question 1: Ceteris paribus, which country has better insurance? Which has higher overall welfare? If you haven't taken the DNA test yet, which would you rather live in?

Question 2: Would any of your answers change if the test was imperfect and only predicted some of your medical expenses?

France Protects Booksellers from Amazon through Legislation

According to an article on FT.com today:

France’s parliament has passed a law preventing internet booksellers from offering free delivery to customers, in an attempt to protect the country’s struggling bookshops from the growing dominance of US online retailer Amazon.

The report highlights a risk that Amazon faces - its cost leadership in France is seen as a threat against which local bookshops need to be protected through legislation. Rather than be seen as an isolated incident, this perhaps is an indication that Amazon has reached a size where its action will attract law makers’ attention.

Is Mass Media a Myth in the Information Economy?

Hal Varian, Chief Economist at Google, recently gave an interesting speech where he discussed the economics of the newspaper industry. Varian proposes that tablets give newspapers a way to reclaim some lost audience.

Jeff Jarvis, author of the book entitled 'What Would Google Do?' takes an opposing viewpoint. He contends that that mass media is a myth and that newspapers should personalize content to stay relevant.

How Big Data is Disrupting the Management Consulting Industry

In a article entitled Consulting on the Cusp of Disruption, HBS professors Clay Christensen, Dina Wang, and Derek van Bever, discuss how multiple forces are disrupting the management consulting industry led by the heavyweights: McKinsey, Bain and BCG.

A chief factor at work is technology, particularly big data. Some pertinent snippets from the article on this topic:

As BTG’s CEO, Jody Miller, puts it, “Democratization and access to data are taking out a huge chunk of value and differentiation from traditional consulting firms.”

... 

Scores of start-ups and some incumbents are also exploring the possibility of using predictive technology and big data analytics to deliver value far faster than any traditional consulting team ever could. One example is Narrative Science, which uses artificial intelligence algorithms to run analytics and extract key insights that are then delivered to clients in easy-to-read form. Similar big data firms are growing explosively, fueled by private equity and venture capital eager to jump into the high-demand, high-margin market for such productized professional services.

Only a limited number of consulting jobs can currently be productized, but that will change as consultants develop new intellectual property. New IP leads to new tool kits and frameworks, which in turn lead to further automation and technology products. We expect that as artificial intelligence and big data capabilities improve, the pace of productization will increase.

---

The steady invasion of hard analytics and technology (big data) is a certainty in consulting, as it has been in so many other industries. It will continue to affect the activities of consultants and the value that they add. Average costing and pricing analysis have been automated and increasingly insourced; now Salesforce.com and others are automating customer relationship analysis. What’s next? 

Consider the disruption that technology has already introduced. The big data company BeyondCore can automatically evaluate vast amounts of data, identify statistically relevant insights, and present them through an animated briefing, rendering the junior analyst role obsolete. And the marketing intelligence company Motista employs predictive models and software to deliver insights into customer emotion and motivation at a small fraction of the price of a top consulting firm. These start-ups, though they lack the brand and reputation of the incumbents, are already making inroads with Fortune 500 companies—and as partners to the incumbents. 

Income Redistribution in the Information Economy

In her TED talk, Chrystia Freeland discusses how globalization and the technology revolution are economic drivers of income redistribution towards the top end.

Consumers Capture Economic Value Through Big Data

In a review of the book Big Data: A Revolution That Will Transform How We Live, Work, and Think, Loyd E. Eskildson summarizes the story of Farecast:

Oren Etzioni, frustrated to learn that many passengers booking a flight after he had, were able to pay less - contrary to conventional wisdom. He then 'scraped' information from a travel website from a 41-day period to forecast whether a price was a good deal or not, founding Farecast to offer this new ability. Etzioni next went on to improve the system by digesting data from a travel stie that covered most American commercial routes for a year - nearly 200 billion flight-price records. Before expanding to hotel rooms, concert tickets and used cars, Microsoft snapped up his firm ($110 million) and incorporated it into it Bing.

This is an excellent example of how big data enables consumers to capture economic value in the information economy.

Etzioni went on to co-found Decide, which in the same vein, uses big data to help consumers decide whether they should buy a product now or wait for the price to drop. A sample recommendation:

Today, it was announced that Decide was purchased by eBay for an undisclosed amount.

Will Big Data Create A Personalized Pricing Nirvana for Retailers?

In a fascinating blog post, Adam Ozimek makes the case that we will see much more individualized pricing (which economists call "first degree price discrimination") as more data mining becomes available.  

For instance, in a new working paper, Ben Shiller is able to use big data to massively improve his ability to predict demand for Netflix subscriptions by any given individual:

Adding the full set of variables ... including web-browsing histories and variables derived from them, substantially improves prediction – predicted probabilities range from close to zero to 91%….I find that web browsing behavior substantially raises the amount by which person-specific pricing raises variable profits.

As more and more data become available, it's easy to imagine retailers using these data to offer personalized prices for each consumer, especially for information goods where margins are large and pricing flexibility is greatest. "You want to watch Elysium tonight? Special price just for you!"  

This can substantially increase profits, and it is also likely to make products and services available at lower prices to some people who couldn't previously afford them, increasingly overall economic efficiency. At the same time, total consumers' surplus is often falls when differential pricing is introduced, as consumers with high valuations make larger payments to sellers for goods and services they would have purchased anyway.

However, that's not the end of the story.  Remember that big data can work for consumers, too.  Thanks to Google and other tools, consumers also have more and better search engines and recommendation services available.  That can intensify competition among sellers, which tends to increase consumers' surplus.

Who will come out ahead? Ultimately, it's an arms race, with those best able to use data and technology outcompeting other buyers and sellers.  In the long run, history suggests that the invisible hand tends to favor consumers as each new wave of technology is invented and diffuses.

The State of Performance-driven Management in America

I teach at a School of Management so you won’t be surprised to learn that I think good management can make a huge difference in the performance of companies, and ultimately the economy.  But you may be surprised that there is very little economic research on the effects of management.  Sure, there’s lots of speculation and countless management books and articles, but a recent review of the economic literature by Chad Syverson concluded: “No potential driving factor of productivity has seen a higher ratio of speculation to empirical study [than management practices].”  The biggest problem has been simply a lack of a comprehensive, reliable data set of management practices.

To address this gap, I recently helped formulate the U.S. Census Bureau’s survey of management and organizational practices at more than 30,000 manufacturing plants across the country--the first large-scale survey of management in America. Along with Nick Bloom, Lucia Foster, Ron Jarmin, Itay Saporta and John Van Reenen, we examined three types of practices-- performance monitoring; setting targets, and offering incentives—which we called “Structured Management.”

Analysis of the data reveals several striking results about the relationship between performance goals and improved business. Specifically, setting business goals and monitoring results are among the practices that actually yield better business productivity and growth, according to this comprehensive survey of U.S. management conducted in 2011. The survey was funded by the National Science Foundation and had administrative support from the National Bureau of Economic Research and the MIT Center for Digital Business. It was a joint, academic-U.S. census bureau collaboration.

My fellow researchers and I set out to determine whether, and what type of management practices influence bottom-line business results such as productivity, output and growth. Based on the responses, we found a tight link between Structured Management and performance outcomes such as growth, expenditures and innovation as indicated by R&D and patent intensity.

Figure 1: Better Performance is Associated With More Structured Management
 

While the survey did not focus exclusively on digital technologies, the conclusions may partly reflect the increasing adoption of information technologies, like Enterprise Resource Planning (ERP) systems, which make data collection and processing much cheaper, easier and more effective. Structured Management scores for data use have improved the most, according to the data. Presumably this reflects the growing use of IT in modern firms.

The study also highlights the important rise of data-driven decisionmaking, which the MIT CDB has championed for several years. Most of the rise in structured management practices, for example, has come among businesses that have implemented data-driven performance monitoring.

Figure 2: Average Management Scores Increased between 2005 and 2010,Especially for Data Driven Performance Monitoring

It was also interesting to note that adoption of structured management practices has increased between 2005 and 2010, particularly for those practices involving data collection and analysis.  This is consistent with my earlier research with Lorin Hitt and Heekyung Kim on Data-Driven Decisionmaking.

Among other key findings:

-- There is a substantial dispersion of management practices across the establishments. Eighteen percent have adopted at least 75% of these more structured management practices, while 27% adopted less than 50% of these.

--There is a positive correlation between structured management practices and location, firm size, establishment-level measures of worker education, and export status.

Going forward we will continue to analyze the data and explore causality. Additionally, we may do another survey in 2015 to establish longer-term data and perhaps will focus on the retail or health-care sector.

Let me know what other ideas you think we should explore.

Tags: u.s._census_bureau, performance_monitoring, productivity, u.s., management, manufacturing, data_analysis, data-driven_decisionmaking, erik_structured_management, erik_brynjolfsson, u.s._economy


This post first appeared on the MIT CDB Community site.

Hollowing Out

Tom Edsall does a nice job summarizing the increasing hollowing out of the job market in his New York Times column today.  The employment/population ratio has fallen drastically since 1999 even as Real GDP hit an all time high this month.  Edsall quotes Andy McAfee and me arguing that technological progress is part of the story for both these trends. Fewer people are working in America today than in the late 1990s, even though overall income is higher.

James Hamilton and Amar Bhide are quoted by Edsall as being skeptical that the restructuring of the economy contributed to job losses.  While Andy and I agree that the Great Recession is undoubtedly the biggest driver of the job losses since 2007, we also see a longer-term forces at work.  In fact, employment growth was sluggish well before 2007.

Digital technologies have advanced very rapidly in recent years.  This can and does create enormous wealth. That’s the good news.  But there’s no economic law that everyone will share in this wealth.  In fact, as David Autor has noted, in recent years, the demand for middle skill jobs, involving routine cognitive and/or physical skills has plummeted. This is reflected both in wages and in employment. At the same time, those in the top 1% have seen their incomes soar.  The median family actually has less income today than 15 years ago, even though the nation is producing more goods and services than ever before.

Of course, creative destruction has always been important to the US economy.  90% of Americans worked in agriculture in 1800.  By 2000, it was less than 2%. This switch did not result in mass unemployment.  Instead, new industries, from autos to computers, were created to employ people more productively.  However, this time around, job destruction is happening faster than job creation, at least for certain types of workers. Demand for jobs involving routine work is rapidly falling since those jobs are the easiest to automate.  As entrepreneurs discover and invent new ways to employ the laid off workers, the economy should come to a new equilibrium and re-employ those who lost their old jobs.  However, even though the overall economic pie will likely grow, the new equilibrium may involve lower wages for many types of workers, and many may choose to drop out of the labor force entirely, as they have over the past decade.

The first step to addressing the challenges of this great restructuring is correctly diagnosing it.  It won’t do to assume that, just because things worked out in the past, everything will ultimately work out this time as well.

Study the Digital Economy at MIT

If you'd like to study the Digital Economy with me and my colleagues at the MIT Sloan School, you may want to apply to our PhD Program.

We're accepting PhD applications thru January 15, 2012

Please spread the word to your smart, creative, ambitious friends and family members.