The Economics Of Queuing, Booking & Paying

At the weekend, my wife and I went to a pub restaurant that operated on a first-come, first-served table system. When we arrived, there were no tables available, so we were added to a waiting list. After having a drink, we were seated within 15 minutes. The dining experience was wonderful, and the food was fabulous. I got talking to the manager about their non-booking system, and how they'd made a success of it, maximising turnover in the process. From an economist's perspective, there are pros and cons to both types of system (booking and non-booking) - but to make the latter work, you typically have to offer a top-notch customer experience and have an excellent reputation.

Just as this restaurant made queuing a pleasant experience for its patrons, I predict that with continually advancing technology we will have to queue a lot less than we do now. There'll be far less queuing in shops, in bars, on roads, etc because automated bots will be bringing us our goods, serving our drinks and driving our cars. But until that day comes, let's have a further chat about queues, booking and paying.

We all know what it is like to decide which checkout line to go to in a busy supermarket. The human motivation of all shoppers is to get out of the store as quickly as possible. To do this, one must first do a quick scan at the number of people and the number of goods in each shopping trolley in each checkout line, to get a sense of how long each person in the queue may take. During this time, we'll be on the lookout for potential delays, such as old ladies with vouchers or chequebooks, items that have hard to read barcodes, items like fruit and veg that may need a manual entry from the cashier, single people packing their own bags vs. couples with one of them doing the packing, that sort of thing.

This is the basis of complex systems theory: individual agents trying to maximise their own utility, whereby in just a few seconds the mind executes some rapid computations to ascertain which of a number of possibilities is the optimal one. Because of this, in busy supermarkets, most checkout lines most of the time will appear to involve roughly the same perceived waiting time (and usually the same actual waiting time too).

Queues frustrate many people, but we use queues as a way to deal with short-term fluctuations in demand. Queues are usually a problem of supply meeting demand without any additional costs. But the best way to understand queues is that they are a constraint on the supplier's ability to provide a good or service at the price or speed the consumer (and often the provider) desires. Additionally, it's usually to do with number of personnel, skills of personnel, amount of space, etc - but whenever you have to wait in line, there is a constraint occurring somewhere. 

With the qualities of the free market, you are all but guaranteed (through price theory) to facilitate the most rational, incentive-driven allocation of resources possible. In theory, if the price is set right according to supply and demand, there should be virtually no queuing. If prices are too low, demand exceeds supply, and queues are expected to form. As the price rises, we lose the consumers who are not willing to pay more, so the queue diminishes. We reach equilibrium when the price is high enough to ensure the quantity demanded equals the quantity supplied - which is the point at which we'd expect no queue. In other words, if prices are set correctly, demand will fall until the queue reaches zero.

I don't queue very often because I rarely care enough about any consumable good to wait 20 or 30 minutes in line for it. Because there are many people like me, queues engender lots of opportunity costs for providers and suppliers. Imagine a queue at a Building Society in which one customer arrives every two minutes, and one customer every two minutes is dealt with by a member of staff. All it takes is a hold up somewhere in the Building Society (a customer with a complex problem, one of the team on a lunch break, someone off sick, or an influx of people joining the queue), and you could have a queue of ten people. That means anyone joining the queue has to wait for at least twenty minutes to be served. While a Building Society may not lose much custom this way, a food stall surrounded by lots of competition probably would. Queues allocate resources efficiently, but not optimally, because they do not distinguish between Jack, who wants a good or service really badly, and Jill, who doesn’t care much about the good or service but joined the queue simply because she saw there was only one person waiting in line.

Waiting in line is an example of a sub-optimal event, which has been improved by technology that improves sale experiences for consumers. For example, being able to buy a cinema seat online in advance is a far more useful way of allocating the scarce resources of a popular movie than queuing outside in the hope that the cinema won't sell out of tickets. Improved technology that enables consumers to pay according to how much they value something is superior to waiting in line, where there is no way of telling exactly how much someone values something. To that end, popular restaurants that operate under a table booking system should charge for booking a table at peak times as well as for eating the meal. It's obvious to everyone that the laws of supply and demand factor in to the dining out experience too. A 6pm booking on a Tuesday night at a restaurant that has been open for 10 years is bound to be in much lower demand than a 7:30pm booking on a Saturday night at a popular restaurant that has only been open a few weeks.

That is why it's so easy to distort the true signals of value. A couple that phones up and books a table at random, or a few friends who walk past and grab a table on a whim, may not value their table as much as people that would have paid an extra surcharge to eat in there. Consequently, charging for table bookings at high demand restaurants increases the chances that the people who most value a dining experience have that experience, while at the same time leaving room for less-discerning people to choose other restaurants. Moreover, if non-price sensitive people pay more at peak times, price-sensitive people should find cheaper meals of the same quality at non-peak times. 

So why, then, don't such restaurants charge for booking a table? It could be for the same reason that hugely popular concert tickets don't sell for more. But it's probably also the case that popular individual restaurants that adopted this policy would unilaterally place themselves at a disadvantage against other popular restaurants that chose not to charge a booking fee. In all likelihood, this is why reservations do not have the kind of prices that would allocate diners with restaurants more optimally, and create extra societal value in doing so.