Julian Sanchez header image 2

photos by Lara Shipley

Privacy and Tiered Pricing

May 29th, 2010 · 11 Comments

Here’s Kevin Drum on trading personal information for discounts—at the supermarket and the newsstand:

Today, overall supermarket prices are still the same as they’ve always been, they’re just tiered differently: those with cards pay less and those without cards pay more. So on average, consumers haven’t benefited. What’s more, competition is generally fierce in the supermarket biz, which means that overall profit margins are also the same as they’ve always been. So supermarkets haven’t benefited.

So who has benefited? Well, as near as I can tell, the answer is: marketing firms. Loyalty cards generate mountains of purchasing data that allow third parties to target advertising more effectively. This is great news for marketing companies and their clients. Whether it’s great news for the rest of us is a little harder to determine.

But this might be the news model of the future. Basically, you’ll be able to get access to the Times two ways: either by paying for a subscription or by registering with your Facebook/Twitter/LinkedIn ID and agreeing to give the Times access to your online life. This is roughly the same trade that we’ve made in the supermarket biz: pay more and maintain your privacy, or pay less in return for giving it up.

So, I’m going to admit up front that I’m not at all familiar with the empirical literature on loyalty cards or their effects in various sectors. But two things occur to me offhand. First, the spread of grocery loyalty cards has coincided with the rise of high-end mass market options like Whole Foods and Wegmans. Depending on how that price comparison is calculated, static average real prices could very plausibly reflect a variety of trends pushing price in opposite directions, so one can hardly infer that the cards haven’t reduced prices on net without controlling for a bunch of other factors. But I’m willing to assume Kevin would have thought of this as well, and knows of some studies that have done this.

The more important point, though is that price discrimination can make consumers better off on net even if the average price is constant. I assume, after all, that Kevin thinks a move from flat to progressive taxation (bracketing incentive effects for the moment) is welfare improving because of the declining marginal utility of money, even if tax revenues are constant. So a world where the least price elastic or most privacy sensitive consumers pay a little more and the others get a discount could certainly be a net benefit to consumers, even if it’s not of uniform benefit to every consumer. It is, of course, perfectly legitimate to ask whether there are discomfiting unforseen externalities accompanying that sort of shift, but that’s a separate question.

Tags: Economics · Privacy and Surveillance



11 responses so far ↓

  • 1 Joseph // May 29, 2010 at 5:31 pm

    I think your response is a bit off point by failing to address directly the negative externalities. Drum outlined a dichotomy between privacy on the one hand and purchasing products on the other. I think the general point here is that in the not to distant future, most people can have one or the other but not both. Drum is troubled by a situation, where only the rich have the option of any privacy.

    Your response on its face is sensible enough. “Look,”you can argue,”its true that some privacy has been sacrificed but look at benefit X. This benefit on net outweighs for society outweighs the cost to privacy centric individuals.”

    And then you add a semi-caveat about the possibility of unforseen externalities. But, of course, the negative externalities is exactly the point and one that I think underpins both Drum and mine’s concern. More to the point, these aren’t unforeseen, they are well known.

    First, consider incentives. The more information advertisers have, the more easily they can target their ads, and the more easily businesses can track who is purchasing their products and why. This is clearly invaluable information to businesses. The overwhelming incentive is to shred consumers privacy to retrieve this information. Rather than a continumm, its a binary conflict with three sides: 1) businesses who want your information 2) consumers who don’t care about privacy 3) and consumers that do care about their privacy but don’t have the financial means to protect it. In addition, this incentive is further sharpened by the business model for most social media sites. To survive, they have to have access to that information and then leverage that into ad revenue. Since facebook is a private company, we don’t know their financials. But I would not be surprised if their profit margins were actually razor thin. For Facebook, shredding your privacy is a matter of survival. Seen in that light, facebook’s constant torturous privacy kabuki is understandable. So in a nutshell: it seems that a situation will eventually occcur where people are defined by digital identities that anyone can access. Do you disagree? If so, what countervailing force is going to push back? Facebook hasn’t changed its business practices at all and the only real threat might be some sort of government regulation…but I imagine that’s not something that a libertarian would be a fan of.

    The other point is this: what are the consequences to poeple with no privacy? What will third parties do with this information? Well for starters, I predict for job interviews, corporations will delvelop robust data mining software that will scrub the internet and then find out every single word, picture, or video of you. Then, it will be analzyed. If you have done something, even once, that could be considered unacceptable according to corporate policy…well then we won’t be offering you that job. Many businesses are already doing this with peoples’ credit scores. I see a future where people’s digital identities become grounds for judging their worth for employment and social status–in effect a non-normative or “bad digital identity” will be as bad as having a criminal record. Dara Lind alluded to this when writing about facebook (http://yglesias.thinkprogress.org/archives/2010/05/mark-zuckerbergs-silver-spoon-vanguardism.php).

    Am I wrong? Am I overreacting? I sure hope so.

  • 2 Dr Duck // May 29, 2010 at 5:32 pm

    So if it weren’t for the tradeoff of personal data, loyalty cards would be an unqualified win, right?

    So join (or form) an informal card swapping club. Just rotate cards every week or so among several households. This will make the ‘personal’ data useless to the marketers, but still give you the lower prices.

  • 3 Julian Sanchez // May 29, 2010 at 7:37 pm

    I’m hardly insensitive to those concerns; I kept my point narrow because, as your comment demonstrates, it would have taken a much longer post to get into those issues.

  • 4 Robert Jackson // May 30, 2010 at 1:44 am

    Joseph, I think your point is spot on. I, too, believe that your digital life, being quite public, will be the basis of future decisions by others about you. This will related to employment and many other areas of life as well.

    I think we need to do all we can to protect our identities. After all, it is not just Facebook who holds our data. If we use Gmail or Hotmail (or the like) then our messages are likely on someone else’s servers (out of our control). It only takes on bug to expose all that data. Once it is exposed, it is hard to control it.

    The good news is that there are some things we can do to control our privacy and they don’t need to cost a lot. Indeed, some don’t cost anything.

    Take TrueCrypt. It allows you to encrypt your hard drive / files. If your hard drive is ever stolen you will be glad you did this. Oh, yes, it is free.

    Take TrulyMail. It allows you to encrypt your email to others. If someone access your email stored on someone’s server, no worries – it is all unreadable to them. Yes, this is free, too.

    Maintaining one’s privacy does not always cost money. It does require thinking things through, though.

  • 5 The Economics of Supermarket Rewards Cards «  Modeled Behavior // May 30, 2010 at 9:44 am

    […] Julian Sanchez […]

  • 6 RickRussellTX // May 30, 2010 at 2:18 pm

    Guys, I was there on the front lines when Lotus Marketplace came out. I sent my angry letter to Mitch Kapoor. As a young idealist, I railed against the dangers of caller ID on our primitive pre-web forums like “USENET” and “bulletin boards” (you could probably find me in the Google news database)…

    And it all came down to no real threat. We feared Lotus would join smart IT concepts with targeted marketing, and it never really happened. Even today, it’s not happening; if anything, the data is showing companies how useless most advertising is, with a consequent contraction in advertising spending. We feared the caller-ID assessed phone number would become a universal ID and that people would be branded as good or bad on their home phone, but the recycling of phone numbers and proliferation of mobile devices has rendered that problem irrelevant.

    Can you imagine what would happen to a pizza place TODAY that refused to take orders from a mobile phone?

    Joseph claims that there are no countervailing incentives that would rein in data gathering and mining. But there are costs that, in many cases, would overwhelm the benefit to the miner. The information to be gleaned from purchase information (shopper’s cards) or voluntary information placed on Facebook is of limited value. It has errors that are difficult to correct and break the ability to link across services. Even if it were trivial to obtain or store this information, the analysis has some cost (e.g. the interview screener has to look at a lot of frat party photos to catch you doing the Captain Morgan stance with a full bottle at your lips).

    In addition, the utility is limited. Companies that set their screening filters too tight will lose great candidates, and suffer competitive failures when their army of boring-as-milk-soup employees can’t generate any insanely great ideas. Stores that target their marketing too tightly will find that consumers are bored with repetitive advertising that keeps pushing the same products. So I bought Honey Nut Cheerios once, when I start seeing commercials for it at every break, I’m just going close Hulu and go read some Julian Sanchez.

    So, yes, data utility may go up and data cost may go down, and inasmuch as that keeps happening, people will push to gather more data. We’ve seen a handful of abuses (teachers fired over pictures of a college party), and we’ll probably see more, and we can (and should) visciously criticize organizations that abuse the data.

    But this really comes down to free association and free speech — any regulated limitation on data gathering can be applied across the board, or worse enforced selectively by authorities against people they don’t like (e.g. activists tracking bad corporate or government behavior). Given the choice, I’ll take my chance with data gathering and not give the government another excuse to put people in jail.

  • 7 Seth // May 30, 2010 at 5:56 pm

    Dr. Duck is right — so many people provide false information that these databases must be riddled with errors. I know one guy who signed up as “Mickey Mouse” — no questions were asked. And many stores give you the card and ask you to fill out the information form later — which of course is insanely unlikely to happen.

    The data is still valuable even if the identities are false, though — it gives companies the ability to build shopper profiles, which are probably more valuable than direct marketing databases anyway. It’s not that important to Marketer X that they be able to sell directly to you — they just want to know what people like you like to buy. Put another way, aggregate information, read the right way, is good enough to let you build techniques to sell to most individuals.

  • 8 Noah Yetter // Jun 1, 2010 at 2:22 pm

    My loyalty cards are registered under an address and phone number that are no longer mine. Advertise THIS, fools!

    It doesn’t matter anyway because advertising doesn’t work. They’re welcome to all the information they could ever want about what I buy, good luck trying to make any money off of it.

  • 9 mds // Jun 3, 2010 at 1:00 pm

    “And it all came down to no real threat.”

    Is it XKCD imitating life, or the other way around? I’m having trouble telling anymore.

    And with the usual disclaimer about the plural of anecdote, as someone who grocery shopped on Wegman’s original home turf, I can contribute an observation. The other regional supermarket, Tops, didn’t originally have a shopper’s card. Some of us preferentially shopped there on that basis. Then they introduced one, and in fact the “discount” prices were the old default prices. It was a way to mark everything up, with the rubes thinking they were saving money by signing up for the card. If they had wanted to say, “discounts without coupon clipping,” fine. Instead, they dropped store coupons and indulged in a fine imitation of Darth Vader altering the bargain. So I guess there’s an irrational component in my quest for establishments without loyalty cards: I view them as a signifier of more general dickish behavior. (Still shopped at Wegmans sometimes, though … I just didn’t get 20 cents off of four boxes of raisins.)

  • 10 Eli // Jun 3, 2010 at 9:58 pm

    On privacy and loyalty cards, credit card data, etc. – I can’t imagine this stuff is very useful until you can get at broad trends. I mean, the fact that I, specifically, bought a slice, filled up my tank, rented Top Gun, then 2 gallons of milk and some toiletries must be beyond even the most craven identity hound’s purview. Although I wouldn’t put it past some slick future algorithm.

    I tend to say: bring it on. I want my next flavor of slurpee to reflect my preference in store-brand boxers.

  • 11 lemmy caution // Jun 7, 2010 at 6:17 pm

    867-5309 will work in every supermarket in the US. You may have to pronounce it “eight, sixty seven, fifty-three, zero, nine” to avoid being laughed at.