Question 1 (easy): Mobile phone companies often offer plans with a fixed regular subscription fee plus a per-minute usage charge (a two-part tariff). Why isn’t gasoline sold in this way?

Question 2 (slightly harder): Why don’t hairdressers offer plans with, say, an annual membership fee plus a per-haircut charge?

I’m returning from Japan to New Zealand in a few months. I don’t think we’ll be back to Japan within 12 months, so I wanted to buy a one-way ticket. I knew it’s going to be more than half the price of a return ticket, but I expected it to be a little cheaper. So you can imagine my surprise when I found that a one-way direct ticket costs three times the price of a return ticket. I am not kidding. A return economy-class ticket costs about 100,000 yen and a one-way is 300,000.

I’m not complaining too much, because I think 100,000 yen (about 940 US dollars) is a reasonable price for a 12 hour flight on a good airline (and a very good price for return). But it’s still kind of puzzling. Obviously, with this pricing structure, no one will buy one-way tickets, as you can just buy the cheaper return ticket and cancel the return portion (with no refund of course). So why sell one-way tickets at all if they’re more expensive than return tickets? Are one-way travellers so costly or troublesome for the airline that they want to discourage them all together? What is going on here …?

To make it even more confusing, in the other direction, from New Zealand to Japan, you can buy a one-way ticket for about 25% less than a return ticket. However the price of one-way from New Zealand to Japan is higher than a return ticket originating in Japan. I’d be really interested to know the factors that are driving these price differences.

Tyler Cowen discusses one of the topics in behavioural economist Dan Ariely’s new book “Predictably Irrational” — an experiment to measure the value of giving something away for free rather than selling it. In the experiment, two types of chocolate were sold, a Lindt Truffle (high quality) and a Hershey’s Kiss (lower quality). Initially the prices were set at 15c for the truffle and 1c for the kiss, and 73% of people bought truffles. Then both prices were reduced by 1c, so 14c for the truffle and the kisses were free. In this case 69% of people chose the kiss.

On his blog, Dan talks about some of the possible explanations that were ruled out in their study, such as transaction costs. In the original paper (PDF), “affect” is cited as the most likely cause. I don’t quite understand the psychology jargon, but I think it means that people experience a psychological boost (extra utility) by getting something for free which they don’t get even if they have to pay a tiny amount such as 1c. Thus giving kisses away for free increased demand by much more than the simple 1c price cut would imply.

This has important implications for business models in the content industry. As I said the other day, you can still make money by giving content away for free if you can charge for complementary, customised or improved goods or services. Once the demand boost of going from a positive price to a zero price is taken into account, these strategies become even more attractive.

Lightspeed Venture Partners has an interesting blog post about the value of “digital gifts” sold on Facebook. For those who don’t know, Facebook members can send virtual gifts to each other. These are basically just a small graphic of something that is displayed on your profile, such as a “voodoo doll” or “tissues”. Facebook creates scarcity of these objects by limiting the quantity that are available, and to send someone a gift costs US$1 (though some discounts are available). These gifts give a whole new meaning to the phrase it’s the thought that counts.

Lightspeed provides some detailed data on sales of digital gifts, and estimate that Facebook is earning revenues of $15 million per year from them. Best sellers seem to be Christmas-related items like “Santa bear” and “Mistletoe” which sold about 10,000 items per week around Christmas. A few items were given away for free, and the quotas of these vanished very quickly. However, most items sold quite slowly and I think it’s because the price is too high.

Lightspeed’s data can be used to investigate this a little bit further. There are two types of gifts in the dataset, those that were free and those that were $1. There are four free gifts and 143 non-free with valid sales data. Using this data, I have done the most horrendous piece of econometrics ever to see the light of day:

The estimated weekly demand per gift is Q = 150,780 - 149,933P. Based on this rather dodgy regression, Facebook would maximise revenues (and hence profits since the marginal cost of a gift is essentially zero) per gift at a price of about 50 cents rather than $1. This would increase average revenues per gift from $864 to $37,908, and based on Lightspeed’s $15 million total estimate, gift revenues would increase to $672 million. Not bad!

Note to Facebook: You owe me 657 million dollars …

Ok, I admit the demand is probably not linear. But I still think a price cut will increase profits.

This video compares prices of identical Apple products in Australia and the US. The Australian prices are all significantly higher, when converted to US dollars:

In Apple’s defence, you could say that the US dollar is historically very low at the moment, and you can’t expect Apple to update its Australian prices every day in response to exchange rate changes to keep them in line with US prices. On the other hand, I’m pretty sure similar differences hold if you use, say, the average exchange rate over the past 12 months, and surely 12 months is a long enough time for Apple to update its prices. In the past I’ve observed similar things with New Zealand Apple prices too, while Japanese prices seem pretty much in line with the US.

It’s hard to see an obvious reason for persistently higher Australian prices. The video suggests that shipping costs could be some part of it, but at least some Apple products are made in China, and China is roughly the same distance from the US as it is from Australia. Local retailing costs can’t explain it either, as I don’t believe that wages or retail space are significantly more expensive in Australia than the US. And finally, I don’t believe that Australians have significantly higher willingness to pay for Apple products on average compared to Americans. So does anybody have any ideas about what’s really going on?

Related: Joshua Gans wonders why iTunes movie rentals will not be available in Australia.