This article originally appeared at techfreedom.org
Yesterday, the FTC reiterated its age-old formula: there are benefits, there are risks, and here are some recommendations on what we regard as best practices. The report summarizes the workshop the agency held in October 2014, “Big Data: A Tool for Inclusion or Exclusion?”
Commissioner Ohlhausen issued a separate statement, saying the report gave “undue credence to hypothetical harms” and failed to “consider the powerful forces of economics and free-market competition,” which might avoid some of the hypothetical harms in the report.
“The FTC is essentially saying, ‘there are clear benefits to Big Data and there may also be risks, but we have no idea how large they are,’” said Berin Szoka. “That’s not surprising, given that not a single economist participated in the FTC’s Big Data workshop. The report repeats a litany of ‘mights,’ ‘concerns’ and ‘worries’ but few concrete examples of harm from Big Data analysis — and no actual analysis. Thus, it does little to advance understanding of how to address real Big Data harms without inadvertently chilling forms of ‘discrimination’ that actually help underserved and minority populations.”
“Most notably,” continued Szoka, “the report makes much of a single news piece suggesting that Staples charged higher prices online to customers who lived farther away from a Staples store — which was cherry-picked precisely because it’s so hard to find examples where price discrimination results in higher prices for poor consumers. The report does not mention the obvious response: if consumers are shopping online anyway, comparison shopping is easy. So why would we think this would be an effective strategy for profit-maximizing firms?”
“The FTC can do a lot better than this,” concluded Szoka. “The agency has an entire Bureau of Economics, which the Bureau of Consumer Protection stubbornly refuses to involve in its work — presumably out of the misguided notion that economic analysis is somehow anti-consumer. That’s dead wrong. As with previous FTC reports since 2009, this one’s ‘recommendations’ will have essentially regulatory effect. Moreover, the report announces that the FTC will bring Section 5 enforcement actions against Big Data companies that have ‘reason to know’ that their customers will use their analysis tools ‘for discriminatory purposes.’ That sounds uncontroversial, but all Big Data involves ‘discrimination’; the real issue is harmful discrimination, and that’s not going to be easy for Big Data platforms to assess. This kind of vague intermediary liability will likely deter Big Data innovations that could actually help consumers — like more flexible credit scoring.”
The Technology Liberation Front is the tech policy blog dedicated to keeping politicians' hands off the 'net and everything else related to technology.