With their “Commercial Privacy Bill of Rights Act of 2011,” John Kerry and John McCain proposed this week to limit the sale and sharing of our sensitive information between companies. It’s a quaint idea, even a good one, but overestimates government’s ability to reverse the way marketers have worked for almost a century.
While data collection online or through our cell phone activities may be new, data collection is as old as commercial record keeping. Direct marketers of the junk-mail era used public records such as driver’s licenses, home deeds, mortgages and school registrations to learn about us. They purchased more information from magazines about their subscribers, department stores about their credit card holders and, yes, obstetricians about their expecting patients.
Market research firms housed huge libraries of index cards on which data about each of our households was categorized according to income level, demographics and stage of life – all so that direct mail advertisers wouldn’t waste stamps on unlikely prospects.
Their industry computerized long before we consumers went online ourselves. By the 1970s our profiles had migrated from index cards to databases, where statistical modeling created much more complex portraits of our behaviors. Armed with a technique called “factor analysis,” statisticians learned to predict one set of behaviors from another, seemingly unrelated set. If dog owners between 30 and 45 who drove two-door cars tended to respond favorably to advertising appeals for decaffeinated coffee, that’s all that mattered.
All the internet really does is give those same market research firms more data to work with. What did we click on? What web page did we visit and for how long? Where did we come there from and where did we go next? This information helps create a behavioral model about us. It doesn’t matter how sensitive the data; the imprint is just as real.
Sure, it’s often quite wrong, as Wall Street Journal reporter Jeffrey Zaslow joked in his 2002 column, “Oh No, My Tivo Thinks I’m Gay.” In that article, the digital video recorder’s efforts to predict its owner’s tastes based on statistical modeling may have gone awry, but the fact that companies are employing these techniques at all should alarm us all the same. If they’re using them to please their consumer users, they are also using them to please their commercial clients.
Kerry and McCain might do well to research how their own presidential campaigns use data-mining and factor analysis in communicating with potential voters. While the Republicans may once have had more experience in data collection and analysis, by John Kerry’s 2004 campaign the Democrats had caught up.
In swing states, the campaign used factor analysis of voters’ consumer profiles to identify precious undecided voters to target with campaigns. One group in Ohio went so far as to knock on doors and show short advertisements to voters on handheld displays, with content tailored to the particular household’s modeled profile. Republicans, meanwhile, used church membership files to determine which voters would respond favorably to incendiary attacks on Democrats’ godlessness and Bible hatred.
Would the senators’ “Commercial Privacy Bill of Rights” limit these campaign tactics as well? Not in its current, revised form, which is limited to “sensitive” data like health records and financial data. But that limitation should soon be meaningless. If the statistical modeling programs have enough data to crunch, they should be able to infer our sensitive data from the less sensitive sort.
We have to welcome any effort to limit transmission and sale of our activities from one company to another, especially when most of those companies’ only intention is to figure out how much cash they can extract from us. But we mustn’t allow the implementation of any privacy act to fool us into complacency.
The sheer volume of data we create as we move through our digital environment is almost immeasurable. This means there’s more than enough to measure, whatever restrictions we might like to place on them now.
They have our numbers. They always have, and they always will.