It’s hard to avoid advertisements for Amazon Prime. If you’re not already a member, Amazon will ask why not whenever you open the website or browse Amazon videos on your television. Joining is as easy as pushing a button. Amazon, the inventor of the one-click purchase, makes buying smooth and easy. Nothing should get between a customer reaching for their wallet and the company charging their card.
If you want to cancel your subscription to Amazon Prime, however, the process is very different. First, you have to select “Account & Lists” from the drop-down menu at the top right of the home page. That choice opens a long list of options that might include Your Account and Your Orders as well as Your Pets and Your Garage. Your Prime Membership will be somewhere in that list.
Find and select the Prime Membership option, and you’ll have more searching to do. Most of the page is taken up by a table showing the benefits of membership. Amazon will also show you how much you’re paying and how you’re paying. Links will offer options to change your payment method and print a receipt. Buried somewhere in the list of links is an option to “End Membership and Benefits”.
Clicking that link doesn’t end your subscription however. It takes you to a page that again provides a reminder of membership benefits and no fewer than four different buttons. One button lets you use your benefits; one lets you keep them (it’s a cancel button); and a third lets you procrastinate the cancellation for three days. Only the fourth button, marked “I Do Not Want My Benefits” takes you further on the path towards cancellation.
Even pressing that button though doesn’t do the trick. Amazon will give you two more screens and two more buttons to press before you can actually complete the cancellation. The company does let its subscribers stop paying but it makes the process as difficult and time-consuming as possible, with plenty of opportunities to stop, re-think, or just give up.
The process is an example of what Cass Sunstein has called “sludge.” Writing in an academic journal about public policy in 2019, Sunstein defined sludge as “excessive frictions” that make a process harder to complete. “Much sludge consists of dreary or duplicative paperwork, understood to include time spent online,” he writes. “Some of it involves waiting time, in person or online.”
In the public sector, it’s often “an indirect, covert and inadequately scrutinized method of achieving particular policy goals by limiting access to various programs and benefits.” Sludge can make applying for welfare benefits, a visa or an immigration card harder than it needs to be. It can cause students to give up on applications for financial aid and patients to leave their medical insurance registration forms unfilled. It imposes particularly heavy obligations on people short of time, and in the public sector, it falls largely on the poor and the needy who draw the most from government services.
Sludge is a kind of negative version of “nudge,” Sunstein and co-author Richard Thaler’s idea that governments and companies can push people in a positive direction without the need to regulate or require. Automatically enrolling new employees in a pension plan, for example, is a kind of nudge. It assumes that the employees wants a pension, takes the necessary action for them, but still gives the employee the freedom to withdraw from the plan if they want. Nudge, explains Thaler, in a 2018 article in Science, improves what he and Sunstein call the “choice architecture” to enable people to make wiser choices without restricting any options. Driving navigation app don’t just offer a range of different routes to a destination. They also point out what they believes will be the fastest route: they nudge the driver towards the best choice.
Sludge makes the best choice—or the only desired choice—particularly difficult without making it impossible. Amazon can’t ban Prime subscribers from cancelling their subscriptions. But it can give subscribers every opportunity to change their minds. Phone companies do something similar when they make it difficult for customers to leave for a rival service. Agree to couple a new phone with a subscription, for example, and it becomes much harder to accept an offer from a company with a cheaper subscription. You either have to pay the balance on the phone or continue paying for it on an inflated basis even though you’re no longer the company’s customer.
One of the most common—and clearest—examples of sludge is the mail-in rebate. A banner in a retail stores will tell customers that their $1,000 computer is sold with a mail-in rebate from the manufacturer. When they get home, all they need to do is send a form and proof of purchase to an address, and they’ll receive a check for $200 or so, significantly cutting the price of the computer. The offer usually comes with a time limit. One study found that that people believed there was an 80 percent chance that they would claim the rebate within the 30 days allowed. The same study found that the actual redemption rate was only 31 percent.
The combination of the need to print a form, complete it, add a proof of purchase, mail it, and above all overcome the inertia of doing nothing was enough to ensure that most people don’t bother to claim money that a company is willing to give them. The study’s authors tried three different interventions to raise the redemption rate. But neither warning participants that other people tended not to claim their rebates nor issuing two reminders changed the participants’ confidence that they would apply for their rebate—or raise the redemption rate. Only when the authors simplified the process and made the form easier to mail back did the redemption rate rise to 54 percent.
“Here, then, is concrete evidence of the potentially significant effect of sludge reduction in increasing people’s access to valuable benefits,” Sunstein concludes.
Auditing Your Sludge
To reduce that sludge, companies should conduct regular sludge audits, recommends Sunstein. The audits could save banks, insurance companies, hospitals, universities and publishers a great deal of money by reducing sludge, he suggests, and greatly improve the experience of the people who interact with them. “They might even be able to change people’s lives.”
The audits could be formal or informal, involve quantification or be more qualitative. Having completed the audit, just three reforms would go a long way to removing the sludge that makes reaching goals difficult for customers and users.
Institutions, says Sunstein, should look back to see whether current requirements are justified, and remove any that seem outmoded, pointless, or costly. They should choose the least burdensome method of achieving their goals—in effect, increasing their own efficiency to improve the efficiency of their customers. And institutions should make sure that the benefits of administrative burdens justify their costs. Such sludge audits would immediately show both public and private institutions that an existing level of sludge is not in their interests. Children will find it easier to obtain free school meals. Students will have better access to financial aid. Car showrooms will find it easier to bring buyers from the test drive to the dotted line.
But that assumes that sludge is not in the interest of the company or the public body. Sludge can help government administrations reduce the cost of welfare programs without the political cost of cutting them themselves. Sludge that makes voting difficult for some demographics (in the name of voter fraud reduction) can help political parties increase their chances of winning re-election. Companies want the extra business that a mail-in rebate offers but without the cost of actually delivering the rebate. Sludge might be inefficient and prevent people from doing what they want but they can give businesses and organizations what they want.
“From the moral point of view, intentional imposition of sludge is especially objectionable,” warns Sunstein. He compares it to a form of ‘phishing,’ “as, for example, through self-conscious efforts to exploit inertia or present bias so as to lead people to buy unduly costly products or plans.”
A utilities firm that fails to reduce its rates for long-standing customers even as new customers pay less, while deliberately making plan changes difficult and time-consuming, is clearly exploiting the power of sludge. It’s a move that that can be checked with regulation. But sludge isn’t always negative.
Sunstein himself points to “Are-You-Sure-You-Want-To” questions such as those that slow the purchase of guns or the cancellation of anti-virus software as examples of sludge that can produce good results.
But it’s possible for companies to go further. A computer retailer might expect that only about 30 percent of its customers would take up the mail-in rebate. In fact, it could set the value of the rebate based on that take-up rate. If, at the point of sale, store assistants helped customers fill out the form and mailed it for them, raising the redemption rate above 90 percent, the company might find that the offer is no longer worth offering. It would either have to reduce the value of the rebate or remove it altogether. The sludge ensures that the rebate goes to those who are most motivated to receive it.
That makes for an important difference between sludge in commercial and public spaces. Sludge that stops the indigent from claiming welfare or students from receiving financial aid makes life hard for those who have it hardest. Sludge that stops people from obtaining something that they don’t really want, while still making it accessible to those who want it most, ensures that benefits go where they’re most valued. A wealthy family that buys a computer with a $200 mail-in rebate is less likely to put in the effort to redeem their funds than a poor family that places a higher value on that $200. If those wealthy families were to redeem their rebates, the value of those rebates might only be $50. The sludge redistributes benefits, sending them where they most needed.
That makes sludge a tool that companies can use morally but one they have to use carefully. Businesses have to ask themselves what the sludge is stopping and where the benefits are flowing. Sludge shouldn’t stop someone from downgrading a subscription plan to a smaller, cheaper plan that suits them better. But it can make people think twice before they cancel a plan that the company knows they need, even if they don’t realize they need it at the time. If an anti-virus software firm, for example, knows that a significant portion of its former customers are infected within six months of cancelling their subscription, it could inform customers of that rate between the cancel button and cancel confirmation.
When the sludge benefits the customer, it’s much easier to justify.
Stopping customers from cancelling their Amazon Prime subscription though? That only benefits the company and it’s the kind of sludge that’s much harder to justify.