[media-credit name=”CafeCredit” link=”http://www.cafecredit.com” align=”alignnone” width=”640″]
[/media-credit]
This is a slightly edited version of this post from Harold Feld’s personal blog on WetMachine.com.
The Federal Communications Commission has an ongoing proceeding to apply Section 222 (47 U.S.C. 222) to broadband. For those unfamiliar with the statute, Section 222 prohibits a provider of a “telecommunications service” from either disclosing information collected from a customer without a customer’s consent, or from using the information for something other than providing the telecom service. While most of us think this generally means advertising, it means a heck of a lot more than that — as illustrated by this tidbit from Cable One.
Briefly, according to Cable One CEO Thomas Might, Cable One used predictive analytics to determine that customers with low FICO credit scores are “hollow value” customers (i.e. customers who often pay late or don’t pay, and who don’t buy additional services regularly). These customers either provide little profit to the company, or actually cost the company money (due to the expense of maintaining customers while their bill is disputed, and any costs associated with cancellation and bill collection). While Might explained they did not “turn people away,” they used FICO scores to determine what quality of customer service to provide. Cable One technicians aren’t going to “spend 15 minutes setting up an iPhone app” for someone with a low FICO score, according to Might — regardless of whether the customer in question actually pays her bills on time.
Worse, the use of FICO credit scores for this purpose disproportionately impacts communities of color, and vulnerable populations such as those struggling after a prolonged period of unemployment. These same populations are the ones least likely to have a home subscription to broadband. But without strong privacy rules in place, even those who can afford a home broadband subscription may not get the customer service they deserve because their provider considers them a “hollow customer” based on their FICO score. This is exactly the kind of discrimination that civil rights organizations have warned about in their 2014 “Civil Rights Principles For The Era of Big Data,” and on which the FCC has sought comments for in its Notice of Proposed Rulemaking.
Cable One violating subscriber privacy hurts the subscriber regardless of who else can get the same information.
The Cable One example illustrates why so many of the arguments against the FCC adopting its proposed privacy rules for broadband are simply wrong. For example, some argue that because Google and other edge providers and data brokers can also track broadband subscribers online, it does no good to regulate broadband provider privacy practices. I suspect the customers receiving poor customer service from Cable One on account of their low FICO credit scores would disagree. When a customer gives Cable One personal information (e.g., name, address) to provide service, and Cable One uses that information without the customer’s consent to (a) obtain a FICO credit score; in order to, (b) determine whether or not to give that customer decent customer service — that customer suffers a real life injury regardless of whether anyone else can get access to her FICO credit score.
Cable One shows how broadband providers can hurt you more than Google.
Even more ridiculous, as the Cable One situation clearly illustrates, is the argument that somehow preventing broadband providers from abusing their customers in this fashion is a “favor to Google.” Like the argument ad hominem, the argument ad Google has surface emotional appeal. But as this ACLU blog post shows, like the argument ad hominem, the argument ad Google is a logical fallacy that leads to bad policy.
While broadband providers want to make this an argument about competition with other advertisers, Cable One shows us how broadband providers can — and do — use information obtained from customers for much darker, harmful purposes. Google can certainly obtain enough information about you to get your FICO credit score. That’s certainly bad and a privacy violation. In theory, however, I can use things like encryption and flushing cookies to protect myself from Google — or even avoid Google entirely. In the case of my broadband provider, I have no choice. I must turn over information usable in obtaining a FICO credit score to my broadband provider so my provider can actually come out to my house, activate my connection, and bill me on a regular basis.
Worse still, given the state of broadband competition, I may have little or no choice for a high-speed connection. If my cable broadband provider decides I’m unworthy of good customer service because of my FICO score, but I must have a high-speed connection so my kids can do their homework or so I can make what living I can selling my handmade knickknacks on Etsy.com, then I take what I can get.
What does Google have to do with preventing any of that? How does preventing a broadband provider from abusing the information I must provide to get service — without even the paltry protection of clear disclosure so I know what to expect — do “favors for Google?” Do people really want to argue that what Google and Facebook and other edge providers do justifies letting broadband providers abuse customers in new and exciting ways? Do people really want to make the argument that it is okay for broadband providers to use FICO credit scores to decide whether to provide quality customer service based on a metric we know discriminates against people of color, those with limited income, and other vulnerable populations “because Google?”
Without rules, we can expect these and other discriminatory practices to continue and grow.
Cable One CEO Thomas Might told analysts at the JP Morgan Conference that since they started “qualifying” customers with FICO credit scores, bad debt declined by 70%. That’s very good for the company. Mind you, because the targeting of customers with low FICO credit scores is predictive, this invariably meant that some potentially “good” customers got flushed out with the “hollow value” customers. But companies don’t worry about fairness to customers. They worry about profit maximization. The loss of marginal income from those customers with low FICO credit scores who would have consistently paid on time has more than made up for in the savings from the “hollow value” customers with low FICO scores that Cable One pushed off its network and cut cost of service.
Given the profitability of Cable One’s “qualifying” scheme, it seems very likely to spread through the industry (assuming it hasn’t already, since we have no way of knowing unless another company CEO starts to brag about using customer data to screw them over). Why shouldn’t other broadband providers use customer information customers have no choice about providing to determine whether to “waste time” providing them with real customer service? It maximizes profit, and it’s not against the law. Don’t broadband operators have a responsibility to their shareholders to use customer data to discriminate like this? Sure, it may hurt people of color, the unemployed, or the millions of people who have low FICO scores by mistake. But broadband providers are profit-maximizing businesses — not charities or social justice warriors. As long as it’s not against the law, aren’t broadband providers obligated to their shareholders to maximize revenue and discriminate against potentially “hollow value” customers?
Mind you, if the FCC actually adopts the proposed broadband privacy rules, using customer information to obtain a FICO score without disclosure and consent would violate the law. Cable operators like Cable One would not be able to “qualify” customers and would need to give everyone the same level of customer service.
But hey, it’s Google. So whatcha gonna do?
