Tim Harford draws our attention to the limitations of conventional economics. He says that gift vouchers are "awful currencies" and that economists can only scratch their heads at why people prefer to give and receive them rather than cash.
Strictly speaking, he's right. Vouchers are clearly inferior to cash in that they often expire before being used and can only be spent in a very limited number of shops.
But despite this, vouchers are hugely popular. In Scroogenomics - the standard work on the economics of Christmas - Joel Waldfogel points out that they are often near the top of lists of what people want for Christmas.
There are two reasons for this.
One is that vouchers carry social meanings which cash gifts don't. As Tim says, giving cash is "grubby and unimaginative." At best, it signals less thoughtfulness and effort than vouchers. At worst, it's simply crass; whilst cash gifts are acceptable from, say, grandparents to teenagers, they are frowned upon between folk of similar ages. Tim says economists are perplexed by this. But the fault is wholly theirs. Interactions between people often carry meanings beyond mere exchange. To pretend that gifts are simply optimizing acts of rational asocial individuals is the sort of autism that discredits orthodox economics.
Secondly, vouchers make sense once we depart from the paradigm of complete rationality and recognize behavioural economics. Vouchers appeal to mental accounting. A gift of cash tempts us to spend the money on everyday items, whereas a voucher invites us to get something we really want. Of course, the voucher is as fungible as cash in the sense that it frees up money for us to spend on dull things. But not all of us see it that way.
In this sense, the anomic rational maximizing paradigm of economics makes a massive predictive error. It predicts that a market shouldn't exist when in fact it does. This is not the only such error that paradigm makes. It predicts that the market for active fund managers should at least be small, but it's not - perhaps because the optimism bias and overconfidence cause people to over-estimate the potential for fund managers to add value.
On the other hand, though, the paradigm also predicts the existence of some markets which don't in fact exist - and not because they are illegal. For example, there's less (pdf) international trade than the Heckscher-Ohlin model predicts, and fewer important state-contingent markets than there should be. These absences might be explained by behavioural factors such as the home bias, path dependency (if markets haven't existed yesterday, they'll not exist today), or optimism bias; we under-estimate risks and hence the need for insurance against them.
My point here is simple. It's a cliche that mainstream economics makes serious predictive errors. What's not so appreciated, though, is that one set of such errors is that it fails to predict which markets should exist, and which shouldn't.