The sunk cost effect in customer loyalty schemes.

The sunk cost effect in customer loyalty schemes.

Gary was bored at work sat at his desk. He was wasting time before a meeting by browsing some of the popular shopping websites looking for a new book to read for his commute to work. Like many of us, Gary has several membership subscriptions for websites that he shops on a lot. The subscriptions that he buys for with a small annual cost guarantees Gary quick delivery, priority ordering and discounts on a large range of products. Since he was bored at work Gary thought that he’d take advantage of his subscription and order another book – afterall he pays for the service so why not make the most of it?

Almost all large companies have subscription membership schemes that offer their customers exclusive benefits. Mobile phone (cell phone), internet, utility (e.g., gas and electricity), insurance, and travel providers all offer their main services alongside additional schemes with small membership charges for exclusive offers. These schemes are everywhere. Subscription-based membership schemes are very profitable for most companies, and rely on customers opting for easy, lazy decision-making over well thought through decision-making. The historical cost of purchasing the subscription for a small fee is an important aspect of this easy decision-making.

Post 7 - image 1.png

Traditional economic theories explain that when making decisions historical costs should be irrelevant, they should not factor into present decisions (Kramer, 2017). One of the most prominent cognitive biases were historical costs are an important influencer of decision-making is known at the sunk cost fallacy. Kelly (2004) summarized the sunk cost fallacy into two claims: i) individuals give weight to sunk costs in their decision-making (i.e., subscription costs), and ii) it is irrational for them to do so.

Two popular examples of subscription-based membership schemes that benefit from the sunk cost fallacy are the Bahn Card 50 and Amazon Prime. The Bahn Card 50 is the original customer loyalty scheme for Deutsche Bahn, the main German railway operator. The Bahn Card 50 was introduced in 1992, as of 2014 it had 5 million subscribers. The fee for the card is EUR 255 per year which gives the owner of the card a 50% discount on train fares. Like all of the other subscription-based membership schemes the owner of the Bahn Card 50 regards the subscription cost as a sunk cost (historical cost), in doing so they try to make the most of the scheme (Tacke & Firmer, 1992).

Post 7 - image 2.png

Amazon’s subscription-based membership scheme is called Amazon Prime. Prime was introduced 2004. For a fee of $99 (as of November 2014) subscribers get faster delivery and priority to access to products on Amazon’s shopping website. Prime has become an important part of Amazon’s business model. After the introduction of Prime Amazon recorded a 150% increase in purchases by Prime members in the year immediately after joining the scheme. Prime is responsible for 20% of Amazon’ overall sales in the United States. For a company that makes billion in profits every year exploiting the sunk cost fallacy has proven to be very profitable.

The sunk cost fallacy is just one of many cognitive biases that influence decision-making. As we have seen above for people like Gary, and of course most of us, the sunk cost bias can affect the way in which we make decisions through subscription-based membership schemes. It is easier to think that since we have already paid a membership fee that gives us a discount we should take advantage of this and use it. If you sit back and think about the situation not exploiting the discount actually saves us more money than the discount would because we are not spending anything more. Big companies such as Amazon have learnt to use these cognitive biases, thereby making huge profits from our quick, easy and effortless decision-making.

The gambler’s fallacy, framing, anchoring and hindsight bias in judicial decision-making.

The gambler’s fallacy, framing, anchoring and hindsight bias in judicial decision-making.

Months before William had moved into a new house with his wife, Sarah. They were excited about their move to their new home. William and Sarah unlocked the door to their new house and waited for the furniture removal staff to arrive at the end of their driveway. As the removal van drove up the driveway the van struck the side of William and Sarah’s car, writing the car off. Several months later William and Sarah headed to court to claim for the damages of their car against the removal company. Over many years their lawyer had learnt to be careful in the way that she words her cases in court. Their lawyer asked the judges “How much would you award the plaintiff in compensatory damages?” rather than “We are claiming ‘x’ amount from the defendant.” William and Sarah’s lawyer had learnt that framing a case in such away could exploit the framing bias.

The judicial system is one of the most important systems in all countries with a court of law. Judges play a significant role in society deciding on the outcome of many cases and setting precedencies for future cases. The public expect judges to decide on the outcome of each case fairly, without systematic errors or bias. However, like any instance where a person must make a judgement or decision judges are subject to systematic biases.

One of the most important aspects of law for many people is immigration law. Asylum judges must make decisions that can determine the fate of the individual in court. The United States offers asylum to foreign national who can (i) prove that they have a well-founded fear of persecution in their own countries, and (ii) that their race, religion, nationality, political opinions, or membership in a particular social group is one central reason for the threatened persecution. In 2016 a study by Harvard academics investigated the use of heuristics and biases in the asylum courts of the United States (Chen et al., 2016). Chen and colleagues accessed data through a Freedom of Information Act request of 699 decisions that were made by 357 judges in 45 courts from 1985 to 2013. All cases in the US are handled on a first-in-first-out basis with no quotas as to how many individuals are granted, or not granted asylum. On analysis of the data Chen et al found that judges were subject to the gambler’s fallacy. Judges were more likely to grant (or deny) asylum after denying (or granting) asylum to a previous applicant. The judges presiding over the fate of the asylum applications believes in a representation of randomness, as such, they were more likely to alternate between granting and denying asylum.

Post 6 - image 1.png

There are many other heuristics and biases that play an important role in the judicial decisions of judges in the US these include framing (as demonstrated by William and Sarah’s lawyer), anchoring, and the hindsight bias. In a joint study with Cornell Law School researchers at a major conference issued questionnaires to 167 federal magistrates. The questionnaires contained examples of cases that a court judge would preside over. From the responses to the questionnaires Guthrie et al (2002) found that judges were subject to anchoring in personal injury claims, framing in copyright action cases, and the hindsight bias in cases of medical negligence.

Post 6 - image 2.png

A retrospective study by Rachlinski et al (2000) examined the effect of the hindsight bias on decisions in real court cases. The hindsight bias is a mental shortcut that states that we have an inclination, after an event has occurred, to see that the event was predictable. The researchers found that judges failed to appreciate the problems associated with judging an event from hindsight. One example (Chase v. Pevear) from the Supreme Judicial Court of Massachusetts decided that the trustees of two high risk investments should have known the outcome of the investments with a nearly omniscient ability. In the case of First Alabama Bank v. Martin the Alabama Supreme Court held that another group of high risk, high yield equities were speculative, as shown by the fact that trustees lost money by selling ‘at the bottom of the market.’ The courts in both cases assumed that the investors should have known that the price of the equities were recovering, and hence the trustees should not have sold them. One court even held a trustee liable for failing to predict the stock market crash of 1929.

The public expects our judges to make well balanced judgements and decisions without falling for cognitive biases. As humans tasked with make important decisions the judicial system is not isolated from any of the shortcuts that we use to make decisions. For William and Sarah seeking their compensation for the damage to their car the experience that their lawyer had with the framing bias came in useful. In many of other cases, such as with the example of the gambler’s fallacy in the asylum courts these cognitive biases and mental shortcuts can become problematic. We have seen how the framing bias, anchoring heuristic and hindsight bias are all important when making decisions, even at the highest levels of decision-making in society.