Sunday, 18 August 2019

Always be a new customer - price discrimination and 'loyalty taxes'

Price discrimination is one of the most pervasive pricing practices that firms use. It involves the seller the seller selling the same product to different consumers for different prices, and where those differences in prices don't reflect differences in cost. Specifically, the seller should charge a higher price to consumers who have less elastic demand (consumers who are less price sensitive, and those who are less likely to go elsewhere), and a lower price to consumers who have more elastic demand (those who are more price sensitive).

Some of the examples seem really perplexing, until you think about them in terms of the price elasticity of demand. For instance, consider the example of Delta Airlines charging a higher price to its frequent fliers than to those who are not frequent fliers (one of my favourite examples to use in my ECONS101 class). That seems to make no sense, until you realise that frequent fliers are less likely to go elsewhere, because they want to keep accumulating air miles. That means that the frequent fliers have less elastic demand than other consumers do, and a price-discriminating airline should be charging them a higher price.

Given how common price discrimination is, I was interested to read this article in The Conversation last month, by Allan Feis (University of Melbourne), about what he refers to as 'loyalty taxes':
A “loyalty tax” occurs when discounts are offered to new customers while longer-term customers pay more. Often this involves increasing premiums at the first and subsequent renewals...
Our research last year showed, on average, customers renewing their insurance policy paid 27% more than new customers. Our most recent data indicates the gap has risen to 34%. This translates to hundreds of dollars for the average home and contents insurance policy.
Loyalty taxes appear to be widespread in Australia. The Australian Competition and Consumer Commission concluded from different pricing inquiries that loyal customers of both banks and energy providers end up paying more. It also demonstrated the price difference for insurance in northern Australian – with one insurer on average charging renewing customers 15-20% more than new customers.
In Britain, regulators have calculated that customers are, by their fifth renewal, paying about 70% more than a new customer. The Competition and Markets Authority estimates the total cost of loyalty taxes in five British markets – mortgage, savings, home insurance, mobile phone contracts and broadband – to be about £4 billion (about A$7 billion) a year.
Translating this British estimate to the equivalent sectors in Australia (taking into account differences in population and GDP), the cost to consumers could be as high as A$3.6 billion, or at least $140 a year per person. This estimate does not include the energy sector, where evidence suggests the practice of charging longstanding customers more is rife.
Loyal customers are, by definition, less likely to go elsewhere. Common sense suggests that companies should look after these customers, especially if the cost of acquiring new customers exceeds the cost of keeping existing customers (apparently, five times as much). However, maybe that heuristic is breaking down, especially if loyal customers are profitable in the short term, because their less elastic demand means that firms can charge them a higher price until they leave.

In my ECONS101 class, we also talk about customer lock-in, where customers find it difficult or costly to change provider once they have started buying from one. The cost might be monetary (such as a contract termination fee), or it could simply be the time and effort required to find a new seller. One way a firm can increase the number of locked-in consumers is to offer a low price initially. Once the consumers are locked in, it makes sense for the firm to raise its price to those consumers. This is referred to as multi-period pricing.

'Loyalty taxes', as described by Feis, cover both of these situations (price discrimination, and multi-period pricing). It is difficult for consumers to avoid the loyalty tax if it arises from multi-period pricing (especially if they are locked in by a contract termination fee), but less costly to avoid price discrimination, if firms are simply charging loyal and long-term customers a higher price.

The take-away message from this for consumers should be: make sure that you regularly change suppliers, for electricity, broadband, insurance, etc. If you're not locked in by a contract, you should be checking for alternatives on a regular basis (and there are government-provided services available to help, like whatsmynumber in the case of electricity providers in New Zealand). If the firms are only giving better deals to new customers, you should be aiming always to be a new customer.

Saturday, 17 August 2019

Junk food discounts at supermarkets

In The Conversation yesterday, Adrian Cameron (Deakin University, and no relation of mine) and others wrote about junk food discounts at supermarkets:
Half-price chips, “two for one” chocolates, “buy one get one free” soft drinks: Australian supermarkets make it very easy for us to fill our trolleys with junk food...
We looked at supermarket specials over a year to see how healthy they were. The results of our research, published today, show junk foods are discounted, on average, twice as often as healthy foods...
The way supermarkets choose what products are on special each week is complex.
Food manufacturers pay large premiums to have their products featured in supermarket catalogues, at end-of-aisle displays or near the checkout. The arrangements between food manufacturers and supermarkets are often governed by contracts that specify the way products are to be promoted.
Food manufacturers and supermarkets know unhealthy food is often bought on impulse, making price discounts a great way to entice customers to make those impulse choices.
This was quite timely, because last week I covered pricing strategy in my ECONS101 class, and the week before that we covered elasticity. The combination of elasticity and pricing strategy, along with transaction utility from behavioural economics, do a good job of explaining what supermarkets are doing, and why.

Consumer demand for junk food is likely to be relatively price elastic. Most junk food items are relatively inexpensive, so they take up only a small proportion of our income, and that is associated with relatively more elastic demand. They also have many substitutes (there are lots of items to choose from), so our demand for any particular item is also likely to be relatively more elastic. Finally, they tend to be luxury items (in contrast with necessities), which also have relatively more elastic demand.

When demand is elastic, a change in price has a bigger effect (in percentage terms) on the quantity that we purchase. So, a 10 percent price discount on an item with elastic demand will lead to an increase of more than 10 percent in the quantity purchased. That increases revenue for the seller. [*]

This also explains why they would discount junk food items but not fruit or vegetables. Fruit and vegetables are necessity items, not luxuries - they have price elasticities of demand that are less than one (as noted here). So, fruit and vegetable sales do not respond much to a decrease in price, so discounting them would decrease revenue for the seller. Discounting fruit and vegetables is a sure-fire way for a supermarket to destroy their profitability.

However, elasticity by itself doesn't explain discounting, because if it was the only explanation, then the seller would better off to keep the price low permanently. A complementary explanation is transaction utility (as I discussed in this post earlier this year). When we buy an item, we get utility (satisfaction or happiness) from receiving the item (which we call consumption utility), plus we get utility from the transaction itself (transaction utility). If we feel like we are getting a good deal, that makes us happier about our purchase. It doesn't make us any more satisfied with the item itself, but it increases our transaction utility. Higher total utility (consumption utility plus transaction utility) makes us more likely to buy the item. By offering discounts on different items every time, they avoid giving consumers the perception that the price is lower, so each time a discount cycles back to an item, there has been time enough for consumer perceptions about the 'usual' price to reset.

So, if an item has relatively elastic demand (which is true for junk food, but not for fruit and vegetables) and the seller can make us feel good by offering a discount, then it can make sense for them to do so.

All of this is somewhat related to another practice of supermarkets, which is loss leading. That is where a seller sells some products at a loss in order to increase sales of other products. However, it seems unlikely that discounting junk food is an example of loss leading. As the quote above notes, junk food is an impulse purchase. In contrast, the ideal loss-leading product is one that has elastic demand and will therefore bring a lot of customers into the store. Nobody chooses their supermarket based on a discount for their favourite chocolate bar (I think?).

Anyway, none of this behaviour by supermarkets should be a surprise to us. It only takes a little bit of knowledge about consumer behaviour and price elasticity to explain why supermarkets discount junk food and not healthy food. The article finishes with:
Imagine what it would be like to shop at a supermarket where healthier food was on special more often, and with bigger discounts. Where customers were enticed by discounted fruit and vegetables instead of half price chips, chocolate and soft drinks.
You'll have to use your imagination. No such store exists, and if it did, you'd better get in fast because it's not going to last long before it fails.


[*] Economists' usual assumption is that firms are trying to maximise profits, not revenue. For simplicity, I'm ignoring that assumption here. For a supermarket, with high fixed costs and the power to negotiate steep quantity discounts from suppliers, there probably isn't too much difference between maximising revenue and maximising profits.

Read more:

Tuesday, 13 August 2019

African Swine Fever, China and demand for New Zealand beef

In my ECONS102 class this week, we've been discussing international trade. Having spent a lot of time going through the various ways that governments can intervene in markets with international trade, and the deleterious effects of those interventions, doesn't leave a lot of time for interesting questions about market dynamics in markets with international trade. So, I thought I would take a moment here to do so, motivated by this New Zealand Herald article from last month:
African Swine Fever has played a part in China overtaking the United States as the biggest market for New Zealand beef, the Meat Industry Association (MIA) said...
Demand for New Zealand red meat in China had been growing before the disease arrived in Asia a year ago, but Ritchie said its onset had altered the world supply/demand dynamic.
"We have been lucky in the sense that there has been that buying demand for protein in recent years," he said.
"Right now, with the impact of African Swine Fever in China, they are talking about potentially a 25 to 30 per cent cut in production, so that has to be extraordinarily significant given pork's position as the biggest meat and with China representing about 50 per cent of the world's pork market," he told the Herald.
"Clearly, the alternative proteins have been caught up in that, and demand there has such a huge impact on the world's meat markets," he said.
China has seen a large decrease in the supply of pork, due to African Swine Fever. This raises the price of pork. Beef is a substitute for pork, and when the price of pork increases, some consumers will switch to buying beef. This increases the demand for beef. Now, China is an importer of beef, and how that increase in demand affects the market for beef is shown in the diagram below. If China was not able to trade for beef, the market would operate in equilibrium, where the price of beef is P0 and the quantity of beef traded is Q0. The increase in demand (from D0 to D1) would raise the price of beef from P0 to P1, and increase the quantity of beef traded from Q0 to Q1.

However, China is able to trade for beef on the world market. In other words, they can buy beef from the market by paying the world price for beef, which in the diagram is PW. The world price PW is lower than the Chinese domestic price P0, because China has a comparative disadvantage in beef production - they can produce and sell beef, but only at a higher cost than other countries (those other countries have a comparative advantage in beef production). In other words, the rest of the world is willing to supply China with beef at the price PW. We represent this with the kinked (red) supply curve S+imports (the supply of beef to the China market, once we account for imports). Now, Chinese consumers only have to pay the lower price PW instead of P0, so they will buy more (QD). However, Chinese beef suppliers have to compete with the lower world price PW, so they will sell less (QS). The difference between QD and QS is the quantity of beef imports.

When demand increases from D0 to D1, that doesn't affect the Chinese beef suppliers any more. They are already supplying as much as they wanted to at the low price PW. The Chinese consumers will increase the quantity that they purchase though, to QD1. The difference between QD1 and QS is a larger quantity of beef imports.

Now consider how that will affect New Zealand, as a beef exporting country. This is shown in the diagram below. New Zealand has a comparative advantage in beef production, so the world price PW is above the New Zealand domestic price that would obtain if there was no trade (P2). In other words, New Zealand can produce and sell beef at a lower cost than other countries. The rest of the world is willing to demand New Zealand beef at the price PW. We represent this with the kinked (red) demand curve D+exports (the demand of beef from New Zealand, once we account for exports). At the higher world price of PW, New Zealand beef suppliers are willing to supply more beef (QS3), and New Zealand beef consumers are willing to purchase less beef (QD3), than they would at equilibrium. The difference between QS3 and QD3 is the quantity of New Zealand beef exports.

When Chinese demand for beef from the world market increases, this pushes up the world price of beef (the Chinese economy and population are large enough that an increase in demand from China is enough to shift world market prices - this would not be the same for New Zealand in most markets!). We won't go back to our earlier diagram on the Chinese market and make this change, but in the New Zealand market, the world price increases from PW to PW1. The D+exports curve moves up to D+exports1. Now, New Zealand consumers have to compete with a higher world price, so they reduce their beef purchases to QD4. New Zealand beef suppliers increase their production to QS4, to take advantage of the greater profit opportunities from the higher world price. New Zealand exports of beef increase to the difference between QS4 and QD4.

So, we can see how Chinese demand for beef translates into impacts on the New Zealand economy. New Zealand beef exporters will be better off, and beef exports increase, but New Zealand beef consumers can expect to see higher prices. I wonder - are we already seeing higher beef prices at New Zealand stores?

Monday, 12 August 2019

Book review: The Case Against Education

I just finished reading Bryan Caplan's The Case Against Education. The subtitle is "Why the education system is a waste of time and money". You might wonder why I would read such a book, given that I work in the higher education sector. Isn't it a book that argues against the very thing that I do? Indeed, I think my in-laws raised their eyebrows on seeing the book on our coffee table.

Caplan writes well, and is contrarian by nature, so I thought this might be an eye-opening read. The book covers both high school and university-level education (as well as graduate school, but there is less data available at that level). Essentially, Caplan argues that there are two private benefits to education that lead to a wage premium for those with more education: (1) an increase in the student's skills and knowledge (or human capital); and (2) a signal to employers that the student is worth employing (because they are intelligent, conscientious, and conformist).

None of this is particularly new, even the idea of signalling, and I have blogged about it before in exactly this context. What is new about Caplan's argument is that he breaks down how much of the education premium relates to human capital rather than signalling. He extensively reviews the literature (not just in economics, but also in educational psychology and sociology), and he contributes his own analysis based on US General Social Survey data. From this, he arrives at shares of 20% human capital and 80% signalling. That suggests that most of the benefit of education is in its ability to sort students into those who are more (or less) worthy of employment, and much less benefit derives from the skills and knowledge they are supposed to be being taught. As Caplan notes:
Most of what schools teach has no value in the labor market. Students fail to learn most of what they're taught. Adults forget most of what they learn.
And that is the inconvenient truth in the whole education sector (that might have made a good alternative title for the book, if it wasn't already taken!). The education system is not teaching skills and knowledge that students are going to make use of in the labour market. Caplan takes particular aim at clearly non-vocational subjects like the arts, music, history, social studies, civics, and physical education. The end result is that many (less able) students would probably be better off not going to university, and doing some vocational education instead (Caplan is rather more bullish about the value of vocational education).

Caplan isn't done there though. If 80% of the private gains to education are from signalling, then there is a strong case against public funding of education. While society gains from increases in students' skills and knowledge, society gains virtually nothing from signalling, since that is simply a way of sorting good and bad future employees. So Caplan argues that the social gains from education are much lower than the private gains, and therefore the costs to the public of funding education could outweigh the benefits (and certainly in the case of low-ability students). Here I think Caplan over-plays his hand, but he does a good job of making his case, even if I may not agree with it entirely (or at least, I haven't yet been able to bring myself to agree with it entirely). He even addresses social justice, which I was expecting him to have left alone (as many economists in the same position would have):
Yes, awarding a full scholarship to one poor youth makes that individual better off by helping send a fine signal to the labor market. Awarding full scholarships to all poor youths, however, changes what educational signals mean - and leads more affluent competitors to pursue further education to keep their edge. The result, as we've seen, is credential inflation. As education rises, workers - including the poor - need more education to get the same job. Where's the social justice in that?
Many readers will disagree with the points that Caplan raises, but it would do good for more people to be engaged with these ideas. There is a growing Assurance of Learning Industrial Complex, driven by accrediting agencies such as those that accredit business schools, engineering schools, and so on. If education is mostly signalling, then the majority of assurance of learning is little more than an educational equivalent of the mechanical Turk.

Caplan's libertarian values will also not appeal to many readers, who might be appalled by his willingness to engage with the idea of promoting child labour. However, he does base the policy prescription on his data and analysis - if vocational skills are mostly learned on-the-job and not in school, then if the goal of education is to provide children with vocational skills, then it would be more effective to have them working rather than at school learning history or physical education.

Overall, this was an interesting read, and my ECONS102 students can expect to see me pick up on a few of Caplan's less-outspoken ideas when we get to the economics of education later this semester. Recommended for teachers (and especially economics teachers)!

Read more: