The price of information

By Eileen Schaps Stein

Does Google deserve some of the blame for Mitt Romney's loss in November? Just before the election, some enterprising Web surfers found that a Google image search for "completely wrong" returned screens full of photos of the Republican presidential candidate.

It wasn't a prank perpetrated by tech-savvy Obama supporters. The search results were actually a freak consequence of a series of online articles that quoted Romney uttering the phrase "completely wrong" - and featured images of his mug.

Of course, we can't pin Mitt's electoral misfortune on the world's most popular search engine. But the discovery did demonstrate Google's chief shortcoming: It's not capable of providing interpretation or analysis of the information it delivers. And given that research today, no matter how detailed or arcane, often begins and ends on Google's home page, this dearth of insight can be disastrous.

More than any other company, Google embodies the information age. After all, it has set out to do no less than organize all the information the world has to offer.

The search engine and its peers strive to do so with algorithms that calculate the relevance of different pages on the Web. But tech wizards and others consistently game those algorithms, marginalizing potentially higher-value and more credible information.

Take, for instance, the notion that childhood vaccines may be linked to disorders such as autism and diabetes. A 2011 report by the Institute of Medicine reviewed more than 1,000 studies and found no connection between vaccines and autism or diabetes. However, a Google search of the terms returns several pages on the first screen that connect vaccination with all sorts of health problems.

Whom do people believe? According to a recent NPR-Thompson Reuters Health Poll, 21 percent of Americans think there's a connection between autism and vaccines. For a chunk of the population, scientific evidence apparently has nothing on the junk science Google offers.

The search engine has proven popular - and loosened old media's grip on information - partly because it's free. And news and research organizations of all kinds are struggling to figure out how to produce information in a world where consumers have come to expect Google to deliver it for free.

High-quality content costs money. As New Yorker editor David Remnick put it, "I cannot give you everything on the Internet for free and make you think the New Yorker is something that comes out of the faucet."

Whether it's delivered by Google or someone else, free information is often worthless without rigorous analysis to accompany it. Take the world of commodities, in which information is public enough that there's no such thing as insider trading in these markets. The U.S. Department of Agriculture publishes weekly reports on the state of the country's crops, for example, and anyone can wander into a field to see how the corn is doing.

And yet America is not filled with commodities-trading millionaires. That's because while Google may be able to give you a USDA report, it can't tell you what it means. Such analysis takes time, expertise, and money.

More people are beginning to realize that when information is free, you get what you pay for. Consequently, information bearing a price tag may be on the cusp of a renaissance.

For example, although the New York Times has encountered difficulties in implementing an online "paywall," more than 500,000 people now cough up a total of about $100 million to read the digital version of the paper. Digital Times subscriptions may outstrip print within two years. And the Wall Street Journal has successfully charged users to read it online for years.

Paywalls don't mesh with Google's preference for free information. Consequently, the slices of reality behind them - particularly those without a brand name on the order of the New York Times' - don't have much of a presence in Google's survey of the Internet.

But reliable information will always be worth paying for. And while free information from the likes of Google can be useful, it can also be "completely wrong."