Go Figure – An extract from The Undercover Economist
On a sunny day in London you can purchase a cappuccino and sip away as the capsules on the Eye, the capital’s landmark Ferris wheel, rotate high above you, occasionally passing between you and the sun… one of life’s simple pleasures. Everywhere you look around the Eye you can see vendors with scarce resources, trying to exploit that scarcity. There is only one coffee bar in the immediate area, for instance. There is also a lone souvenir shop doing brisk business. But the most obvious example is the London Eye itself. It towers over the majority of London’s most famous buildings and is the world’s largest observation wheel. The scarcity power is clearly considerable, but it is not unlimited: the Eye may be unique, but it is also optional. People can always choose not to go on it.
Further along the river, the Millennium Dome is similarly unique, ‘the largest fabric structure in the world’, boasts the local authority. Yet the Dome has proved a commercial disaster because uniqueness alone wasn’t enough to persuade people to pay enough to cover the vast costs of its construction. Business with scarcity power cannot force us to pay unlimited prices for their products, but they can choose from a variety of strategies to make us pay more. It’s time for the Undercover Economist to get to work and find out more.
The only coffee provider beside the London Eye wields plenty of scarcity power over the customer. It€’s not innate, but is reflected glory from the amazing setting. As we know, because customers will pay high prices for coffee in attractive locations, the coffee bar’s rent will be high. Their landlords have rented out some of this scarcity value to a coffee bar, just like the owners of Manhattan’s skyscrapers, or railway stations from Waterloo to Shinjuku. Scarcity is for rent – at the right price.
But how should the bar’s managers exploit the scarcity they are renting from the London Eye? They could simply raise the price of a cappuccino from £1.75 to £3. Some people would pay it, but many would not. Alternatively, they could cut prices and sell much more coffee. They could cover wages and ingredients by charging as little as 60p a cup. But unless they were able to increase their sales dozens of times over, they’d not make enough to cover their rent. That’s the dilemma: higher margins per cup, but fewer cups; or lower margins on more cups.
It would be nice to side-step that dilemma, by charging 60p to people who are not willing to pay more and £3 to people who are willing to pay a lot to enjoy the coffee and the view. That way they would have the high margins whenever they could get them, and still sell coffee at a small profit to the skinflints. How to do it, though? Have a price list saying, ‘Cappuccino, £3, unless you’re only willing to pay 60p’?
It does have a certain something, but I doubt it would catch on with the coffee-buying public of London’s South Bank. However, for years, the previous incumbent coffee bar, Costa Coffee, appeared to have achieved this. Costa, like most other coffee bars these days, offers ‘Fair Trade’ coffee – theirs comes from a leading fair trade brand called Cafedirect. Cafedirect promises to offer good prices to coffee farmers in poor countries. Fair trade coffee associations make a promise to the producer, not the consumer. If you buy fair trade coffee, you are guaranteed that the producer will receive a good price. But there is no guarantee that you will receive a good price. For several years, customers who wished to support third-world farmers – and such customers are apparently not uncommon in London – were charged an extra 10p. They may have believed that the 10p went to the struggling coffee farmer. Almost none of it did.
Cafedirect paid farmers a premium of between 40p and 55p per pound of coffee, and that premium was reflected in the price they charged to Costa. That relatively small premium can nearly double the income of a farmer in Guatemala, where the average income is less that $2,000 a year. But since the typical cappuccino is made with just a quarter-ounce of coffee beans, the premium paid to the farmer should translate into a cost increase of less than a penny a cup.
Of the extra money that Costa charged, more than 90 per cent did not reach the farmer. Cafedirect did not benefit, so unless using the fair trade coffee somehow increased Costa’s costs hugely, the money was being added to profits. The truth is that fair trade coffee wholesalers could pay two, three or sometimes four times the market price for coffee in the developing world without adding anything noticeable to the production cost of a cappuccino. Because coffee beans make up such a small proportion of that cost customers might have concluded that the extra 10p was to cover the cost of the fair trade coffee, but they would have been wrong. A certain Undercover Economist made some inquiries and found that Costa worked out that the whole business gave the wrong impression, and at the end of 2004 began to offer fair trade coffee on request, without a price premium.
But why had it been profitable to charge a higher mark-up on fair trade coffee than on normal coffee? Because fair trade coffee allowed Costa to find customers who were willing to pay a bit more if given a reason to do so. By ordering a fair trade cappuccino, you sent two messages to Costa. The first was: ‘I think that fair trade coffee is a product that should be supported.’ The second was: ‘I don’t really mind paying a bit extra.’ Socially concerned citizens tend to be less careful with their cash in coffee bars, while unconcerned citizens tend to keep their eyes on the price. Perhaps another price list saying, ‘Cappuccino for the concerned £1.85. Cappuccino for the unconcerned £1.75’?
In fact, any well-run business would seek to charge each customer the maximum price he’d be willing to pay – and they do. There are three common strategies for finding customers who are cavalier about price. Let’s cover two for now and leave the best until last.
The first is what economists call ‘first-degree price discrimination’, but we could call it the ‘unique target’ strategy: to evaluate each customer as an individual and charge according to how much he or she is willing to pay. This is the strategy of the used-car salesman or the estate agent. It usually takes skill and a lot of effort: hardly surprising, then, that it is most often seen for items that have a high value relative to the retailer’s time – cars and houses, of course, but also souvenirs in African street stalls, where the impoverished merchant will find it worth bargaining for some time to gain an extra pound.
Now, however, companies are trying to automate the process of evaluating individual customers to reduce the time it takes to do so. For instance, supermarkets accumulate evidence of what you’re willing to pay by giving you ‘discount cards’, which are needed to take advantage of sale prices. In return for getting a lower price on certain items, you allow the stores to keep records of what you buy and then in turn offer you vouchers for discounts on products. It doesn’t work perfectly, because supermarkets can only send ‘money off’ vouchers, not ‘money on’ vouchers. ‘Money on’ vouchers have never been a success.
The second approach, the ‘group target’ strategy, is to offer different prices to members of distinct groups. Who could complain about reduced bus fares for children and the elderly? Surely it must be reasonable for coffee shops to offer a discount to people who work nearby, and for tourist attractions to let locals in for a lower rate? It often seems reasonable because people in groups who pay more are usually people who can afford more, and that’s because people who can afford more are usually people who care less about the price. But we shouldn’t forget that this is a convenient coincidence. Companies trying to increase their profits and get the maximum value out of their scarcity are interested in who is willing to pay more, rather than who can afford to pay more.
For instance, when Disney World in Florida offer admission discounts of more than 50 per cent to local people, they’re not making a statement about the grinding poverty of the Sunshine State. They simply know that for a reduced price, locals are more likely to come regularly. But tourists will probably come once, and once only, whether it is cheap or expensive.
This example gets to the heart of things, and tells us what we really mean by ‘price sensitivity’ or ‘being lavish’ or ‘being cavalier about prices’. The important concept is this: when I raise the price, how much do my sales fall? And when I cut the price, how much do my sales rise? Economists tend to call this ‘own-price elasticity’. I prefer ‘price sensitivity’.
Tourists visiting Florida are less price-sensitive than locals, which means that if Disney World raises its prices, locals are more likely to skip a day at the park. By the same token, if the admission price falls, locals may make repeat visits in a way that tourists probably won’t. Being rich is sometimes connected with being insensitive to prices, but not always. Business-class air travel is expensive, because companies are willing to pay and airlines have the scarcity power to take advantage of that fact.
The same is true of discounts at coffee bars for local workers. The AMT bar in Waterloo station will knock 10 per cent off the cost of your coffee if you work locally. This isn’t because the local workers are poor: they include top Whitehall mandarins and the extravagantly remunerated employees of the gigantic oil company, Shell. The discount reflects the fact that local workers are price-sensitive despite being rich. Commuters who pass through Waterloo in a hurry see only one or two coffee bars and are willing to pay high prices for convenience. Local workers pop out of the office at 11am for coffee and could walk in any direction. They can buy from several cafes, all equally convenient, all of which they will have had a chance to sample. They are bound to be more price-sensitive, even if they are rich.
The ‘individual target’ strategy is difficult, partly because it requires a lot of information and partly because it tends to be very unpopular. Despite the difficulties, however, it’s so profitable that companies always explore new ways to do it. The ‘group target’ strategy of discounts for students or locals is less effective, but easier to put into action, and usually it’s socially acceptable. Either will deliver more profits than simply treating all customers as a homogenous mass.
The cleverest and most common way to persuade turkeys to vote for Christmas is the ‘self-incrimination’ strategy. To get customers to give themselves away, companies have to sell products that are at least slightly different from each other. So they offer products in different quantities (a large cappuccino instead of a small one, or an offer of three for the price of two) or with different features (with whipped cream or white chocolate or fair trade ingredients) or even in different locations, because a sandwich in a station kiosk is not the same product as a physically identical sandwich in an out-of-town superstore.
It’s reasonable to ask how common this tactic really is. Because the products are different, you never quite know whether the company is using a price-targeting trick or merely passing on added costs. It could be that it really does cost 10p more to put fair trade coffee in a cappuccino; maybe cans of whipped cream are expensive to refrigerate and troublesome to clean and the staff hate using them; perhaps large cups of coffee take longer to drink, and so the charge is for table-space not coffee – in which case, charging a higher price is not a strategy to get me to incriminate myself, but simply the coffee shop passing its costs through to me. But I think it’s safe to say that companies are always alert for ways to squeeze the maximum advantage out of whatever scarcity power they have, and price-targeting is the most common way to do that. If it looks like price-targeting, it probably is.
Supermarkets have turned price targeting into an art, developing a vast array of strategies to that end. Above the main concourse of Liverpool Street station, there’s a Marks and Spencer ‘Simply Food’ store, catering for busy commuters on the way in and out of London. Knowing what we do about scarcity value, we shouldn’t be surprised to find that this shop isn’t cheap – even compared with another branch of M&S merely 500 metres or so away, at Moorgate.
I picked up five products at random in the Liverpool Street store and managed to locate four of them in the Moorgate store. Every single one was about 15 per cent cheaper there. Big salads were down from £3.50 to £3.00, sandwiches from £2.20 to £1.90. But even when such discrepancies come to light, few City workers would be willing to stray that distance to save 30p. A bold and effective piece of price-targeting.
Other supermarkets are more circumspect about their pricing policy. Going undercover once again, I made a comparison between the smallish Sainsbury’s supermarket in Tottenham Court Road, and the large store in Dalston, one of east London’s less prosperous neighbourhoods. It was harder to find examples of identical products selling for different prices, although by no means impossible. Does this mean that Sainsbury’s doesn’t price-target as much as M&S? Not at all. They simply go about the whole process with more finesse.
When researching Sainsbury’s, my approach was the same as with M&S: walk into the shop and see what caught my eye. As you probably know, what catches our eye as we walk into the supermarket is no coincidence; it’s the result of careful planning designed to throw attractive but profitable products in the path of customers. What constitutes an attractive product depends on who those customers are. In Tottenham Court Road the obvious goods were all quite expensive: Tropicana orange juice at £1.95 a litre, Tropicana ‘Smoothies’ at £1.99 for 100ml, Vittel mineral water at 80p for 750ml, and so on. It wasn’t that these products were more expensive in Tottenham Court Road than in Dalston (only the Vittel was), it was just that in Dalston cheaper substitutes sprang into view far more readily.
For instance, I couldn’t find inexpensive orange juice in the Tottenham Court Road store, but in Dalston, Sainsbury’s own brand of fresh chilled juice was sitting next to the Tropicana at about half the price, and the concentrated juice was almost six times cheaper than the Tropicana. Brand-name pasta was the same price in both shops, but only in Dalston was it sitting next to Sainsbury’s pasta, which again was almost six times cheaper. The effect was to target the whole Tottenham Court Road store at shoppers who are indifferent to prices, but to aim the Dalston stores at shoppers with a keener eye for a bargain – while of course giving any price-blind Dalston shoppers plenty of opportunity to show their true colours.
Another very common pricing strategy is sale pricing. We’re all so used to seeing a store-wide sale with hundreds of items reduced in price that we don’t pause and ask ourselves why on earth shops do this. When you think hard about it, it becomes quite a puzzling way of setting prices. The effect of a sale is to lower the average price a shop charges. But why knock 30 per cent off many of your prices twice a year, when you could knock 5 per cent off year-round? Varying prices is a lot of hassle for shops because they need to change their labels and their advertising, so why does it make sense for them to go to the trouble of mixing things up?
One explanation is that sales are an effective form of self-targeting. If some customers shop around for a good deal and some customers do not, it’s best for stores to have either high prices to prise cash from the loyal (or lazy) customers, or low prices to win business from the bargain-hunters. Middle-of-the-road prices are no good: not high enough to exploit loyal customers, not low enough to attract the bargain-hunters. But that’s not the end of the story, because if prices were stable then surely even the most price-insensitive customers would learn where to get particular goods cheaply. So rather than stick to either high or low prices, shops jump between the two extremes.
One common situation is for two supermarkets to be competing for the same customers. As we’ve discussed, it’s hard for one to be systematically more expensive than the other without losing a lot of business, so they will charge similar prices on average, but both will also mix up their prices. That way, both can distinguish the bargain hunters from those in need of specific products, such as people shopping to pick up ingredients for a recipe they are making for a dinner party. Bargain-hunters will pick up whatever is on sale and make something of it. The dinner-party shoppers came to the supermarket to buy specific products and will be less sensitive to prices. The price-targeting strategy only works because the supermarkets always vary the patterns of their special offers, and because it is too much trouble to go to both stores. If shoppers could reliably predict what was to be discounted, they could choose recipes ahead of time, and even choose the appropriate supermarket to pick up the ingredients wherever they’re least expensive.
In fact, it is just as accurate, and more illuminating, to turn the ‘sale’ on its head and view prices as premiums on the sale price rather than discounts on the regular price. The random pattern of sales is also a random pattern of price increases – companies find it more profitable to increase prices (above the sale price) by a larger amount on an unpredictable basis than by a small amount in a predictable way. Customers find it troublesome to avoid unpredictable price increases – and may not even notice them for lower-value goods – but easy to avoid predictable ones.
Try to spot other odd mix-ups next time you’re in the supermarket. Have you noticed that supermarkets often charge 10 times as much for fresh chilli peppers in a packet as for loose fresh chillies? That’s because the typical customer buys such small quantities that he doesn’t think to check whether they cost 4p or 40p. Randomly tripling the price of a vegetable is a favourite trick: customers who notice the mark-up just buy a different vegetable that week; customers who don’t have self-targeted a whopping price rise. I once spotted a particularly inspired trick while on a search for crisps. My favourite brand was available on the top shelf in salt and pepper flavour and on the bottom shelf, just a few feet away, in other flavours, all the same size. The top shelf crisps cost 25 per cent more, and customers who reached for the top shelf demonstrated that they hadn’t made a price-comparison between two near-identical products in near-identical locations. They were more interested in snacking.
Perhaps you are a company director rubbing your hands with glee as you read this, planning to deploy a range of clever price-targeting strategies in your own business. Before you get too excited, you’ll need to deal with the leaks in your price-targeting system. There are two potentially catastrophic leaks or great holes in an otherwise brilliant marketing scheme. If you don’t deal with them, your plans will be in ruins.
The first problem is that supposedly price-insensitive customers may not play the self-targeting game. It’s not hard to persuade price-sensitive customers to steer clear of an expensive product, but sometimes it is more difficult to prevent the price-insensitive customers from buying the cheaper one. This is not a problem in the case of small price differences; we have already seen that you can get some customers to pay a modest mark-up in absolute terms, but the mark-up can be huge in relative terms.
Some of the most extreme examples come from the transport industry: travelling first class by rail or air is much more expensive than buying a standard ticket, but since the fundamental effect is to get people from A to B, it may be hard to wring much money out of the wealthier passengers. In order to price-target effectively, companies may have to exaggerate the differences between the best service and the worst. There is no reason why standard-class railway carriages shouldn’t have tables, for instance, except that potential first-class customers might decide to buy a cheaper ticket when they see how comfortable standard class has become. So the standard-class passengers have to do without.
The 19th-century French economist Emile Dupuit pointed to the early railways as an example: ‘It is not because of the few thousand francs which would have to be spent to put a roof over the third-class carriage or to upholster the third-class seats that some company or other has open carriages with wooden benches… What the company is trying to do is prevent the passengers who can pay the second-class fare from travelling third class; it hits the poor, not because it wants to hurt them, but to frighten the rich… And it is again for the same reason that the companies, having proved almost cruel to the third-class passengers and mean to the second-class ones, become lavish in dealing with first-class customers. Having refused the poor what is necessary, they give the rich what is superfluous.’
The shoddy quality of most airport departure lounges across the world is surely part of the same phenomenon. If the free departure lounges became comfortable, then airlines would no longer be able to sell business-class tickets on the strength of their ‘executive’ lounges. And it would also explain why flight attendants sometimes physically restrain passengers from the cheap seats from stepping off the plane before the passengers from first and business class. This is a ‘service’ aimed not at economy-class passengers but at those looking on in pity and disgust from the front of the plane. The message is clear: keep paying for your expensive seats, or next time you might be on the wrong side of the flight attendant.
In the supermarkets, we see the same trick: products that seem to be packaged for the express purpose of conveying awful quality. Supermarkets will often produce an own-brand ‘value’ range, displaying crude designs that don’t vary whether the product is lemonade or bread or baked beans. It wouldn’t cost much to hire a good designer and print more attractive logos. But that would defeat the object: the packaging is carefully designed to put off customers who are willing to pay more. Even customers who would be willing to pay five times as much for a bottle of lemonade will buy the bargain product unless the supermarket makes some effort to discourage them. So, like the lack of tables in standard-class railway carriages and the uncomfortable seats in airport lounges, the ugly packaging of ‘value’ products is designed to make sure that snooty customers self-target price increases on themselves.
Consider a hypothetical organisation, TrainCorp, a passenger train company. TrainCorp owns a train that always travels full. Some of the seats go at a discount of £50 to leisure travellers who booked in advance, to senior citizens, to students or to families. The other tickets cost the full price of £100 and are bought by commuters and other business travellers. This is a fairly standard group-targeting strategy: by giving away a few low-price tickets, TrainCorp restricts supply and acquires the ability to demand high prices by offering tickets to only the buyers with the highest willingness to pay. (It might be profitable for TrainCorp just to fence off some of the seats and restrict supply that way, but it’s even better for them to fill the spare seats if they can.)
We know at once – if we are economists – that this is inefficient. In other words, we can think of something that would make at least one person better off without making anyone else worse off.
That something is to find a commuter who was willing to pay a little less than £100, say £95, and who decided to travel by car instead, and offer him a seat for £90. Where does the seat come from, since the train is full? Well, you take a student who is in no great hurry and was willing to pay a little more than £50, say £55, for the seat and politely throw him off the train. But you refund the price of his ticket, plus an extra £10 for his trouble.
Where do we stand now? The commuter was willing to pay £95 but only paid £90. He’s better off by £5. The student was willing to pay £55 for a £50 ticket, so if he’™d been allowed to ride, he’d have been only £5 better off. But he has just been given £10, so the student is also happy. And what about TrainCorp? Well, TrainCorp just transformed a £50 ticket into a £90 ticket and made a more profitable sale. Even after paying £10 compensation to the student, the company is £30 ahead. Now everyone’s a winner; or they would be if TrainCorp adopted this system instead of its group price-targeting strategy. But of course, that’s not what happens, because if TrainCorp tried it, commuters who were willing to pay £100 would hang around for the £90 tickets, and students who weren’t willing to pay £50 would buy tickets anyway and wait to be paid to get off. The whole affair would turn out badly for TrainCorp, who is the one who gets to set the prices.
In case your head is spinning a little, here’s the quick-and-dirty summary: the group price-targeting strategy is inefficient because it takes seats away from customers who are willing to pay more, and gives them to customers who are willing to pay less. Yet airlines and railways still use it, because the alternative of individual price-targeting isn’t feasible.
OK, so sometimes price-targeting is less efficient than a uniform price; sometimes it’s more efficient than a uniform price. But we can say more than that. Whenever price-targeting fails to expand the number of sales and merely moves products from people who value them more, like commuters, to people who value them less, like students, as in the case of TrainCorp, it will definitely be less efficient than a uniform price. Whenever price-targeting opens up a new market without affecting the old market, it will definitely be more efficient than a uniform price.
And there’s a middle position. A lot of group price-targeting does a bit of both: it opens up some new markets but also wastefully moves products away from high-value users to low-value users. For example, my book, The Undercover Economist, is published in hardcover at a high price, and the paperback edition emerges later, at a lower price. The aim is to target a higher price at people impatient to hear what I have to say and at libraries. One good result is that the publisher will be able to sell paperbacks more cheaply, because some costs will be offset by the hardcover sales, and so the book will reach more people. One bad result is that the early version is much more expensive than it would be if there was only a single paperback edition, and some buyers will be put off. That’s what life is like in a world of scarcity: when companies with scarcity power try to exploit it, the situation will almost always be inefficient, and – equivalently – we economists will almost always be able to think of something better.
Published in FT Magazine, 22 October 2005