How does the age of ever-bigger data help consumers, or threaten annoyance or harm? The Federal Trade Commission raised those questions as politely as possible last week while again turning a spotlight on the data-broker industry, part of its ongoing effort to get a grasp on something that's about as slippery as seawater.

The FTC studied practices at nine little-known companies such as Corelogic, Intelius, and Rapleaf that amass, buy, and sell data on nearly every U.S. consumer. Its conclusion? A call on Congress to "consider" guaranteeing consumers more rights, including the right to see what's collected about them and to dispute mistakes.

There is nothing inherently good or bad about data, of course. But don't be fooled by the FTC's measured tone as it pushes for greater "transparency and accountability" from data brokers. Though the agency can't have much hope for quick congressional action - who does, about anything? - its report shows why privacy advocates and others have serious concerns about this loosely overseen, largely invisible industry.

How do consumers benefit from data collection and analysis? One way is through antifraud software that instantly identifies aberrant patterns.

I got a taste of that speed recently when my daughter's credit cards were lifted from her purse as it lay by her feet at a Philadelphia theater. Before the movie was over, American Express had e-mailed her to question a transaction it had blocked as suspicious. It stopped a theft before it occurred.

The FTC also allows that consumers may benefit from data- driven marketing, including the personally targeted ads that seem to follow you website to website. If you want to better grasp how such marketing methods work, the FTC's report (at is a great primer.

But marketing based on detailed consumer profiles has its critics, such as University of Pennsylvania media scholar Joseph Turow. His most recent book, 2012's The Daily You: How the New Advertising Industry Is Defining Your Identity and Your Worth, looked at how data collection and analysis affect what companies know about consumers - or at least think they know.

Among Rapleaf's pitches, for instance, is that it can predict consumers' behaviors - such as whether you'll make your mortgage payments on time - based on data about how your friends behave.

It's easy to see the harm of such predictions if they're based on erroneous data - the old computer-science acronym is GIGO, for "garbage in, garbage out." But Turow also warns of more subtle harms, such as social or price discrimination, that occur even with good data.

Data-driven discrimination takes many forms, Turow says. He saw one example last year as he and other passengers rushed to reschedule after United Airlines canceled their flight.

"On the monitor to my left was a warning that said the amount of time it will take to deal with you will be based on your status with the airline," Turow recalled. "I had high status because I travel a lot. But I realized that a lot of people standing there might not make their connections."

The worst harm from erroneous data? Jim Francis, a Philadelphia lawyer who specializes in data cases, says clients have had wages wrongly garnished - or couldn't get jobs in the first place - because of GIGO data.

One was James Hines Jr., who applied three years ago for a job as a physical-therapy assistant with a Virginia health-care company, CareSouth. He was turned down, CareSouth told him, because of his criminal record.

There was just one problem: Hines didn't actually have that criminal record - a particularly ugly one belonging to a sex offender with a similar name.

The errant practices behind that were bad enough. Worse was the fact that the source of the bad data, Corelogic subsidiary National Background Data L.L.C., claims to not be covered by the Fair Credit Reporting Act's key protections, because it's a data broker and didn't sell the data directly to CareSouth.

Transparency, accountability, and a simple path for consumers or job-seekers to see what's amassed on them and challenge anything wrong - yes, those would be a good start. It's time to bring data brokers out of the shadows.