Sunday, 30 June 2013

Charging visitors to the UK, adverse selection, and moral hazard

A few months ago the Deputy Prime Minister Nick Clegg was expounding on the benefits of a security bond for visitors to the UK. The basic idea was that visitors to the UK would have to deposit some money with the government and they would get the money back when they left. I was hoping the plan would be quietly dropped, and it seemed to have been. But, unfortunately, the story is back and now looks worryingly likely to happen. Visitors from a select list of countries, including India and Nigeria, will be expected to deposit £3,000 with the UK government if they want to visit the UK. The government's policy on immigration has been a disaster for many years, but the idea of a security bond seems to be taking up a notch the level of stupidity. 
       To put this plan for a bond in context we need a bit of background. A large proportion of the UK electorate is anti-immigration. The current coalition government's answer is to bow to the electorate and promise a tough line on immigration. The message has gone out that Britain is closed to foreigners. Foreign students, for example, are no longer allowed to work in the UK for one year after their graduation. And universities find it ever tougher to hire foreign professors. The government's approach has catastrophically failed. It has not appeased the electorate, who have started voting for the UK Independence Party. And, more importantly, it has annoyed foreigners. Given that immigration is a good thing, this is not something to celebrate. Indeed, at the same time the government was closing the door to foreigners, the Prime Minister and others were trying to sell Britain's advantages in emerging markets, like India. Which doesn't make much sense! If the government would just get on with arguing the case for immigration rather than being trapped by public opinion we might see progress. Instead, we have the latest idea to annoy foreigners. 
       The logic behind the bond is simple enough: We have people who visit the UK and then illegally stay in the country longer than allowed. The bond is an incentive for people to leave when there supposed to. Unfortunately, this logic is fundamentally flawed. What we have here is a game of incomplete information. Suppose that there are two types of foreign visitors. Genuine visitors who want to visit friends and family. And illegal visitors who want to stay in the country indefinitely. The government wants to let in the genuine and block the illegal. But, it cannot tell which is which. That leads to the problems of adverse selection and moral hazard. 
        Adverse selection. Who do you think is going to be more willing to deposit £3000 to visit the UK, the genuine or the illegal? My money would be on the illegal. £3000 is a small price to pay to live in the UK and we know that many illegal immigrants pay a lot more than that to enter the country. Someone visiting friends, by contrast, may not have £3000 to hand. The government is, thus, likely to hand pick the very opposite of what it wants. This is a classic example of adverse selection: the least desirable are more likely to be willing to pay the bond.
      Moral hazard. Let's focus now on the people who paid the bond and have been let into the country. Will he £3000 bond change their behavior. My guess is that it will. We know that a financial payment can crowd out more pure incentives.* So, a genuine may decide to stay in the UK illegally. In other words, someone who may have returned home after they stay, because, say, they felt guilty at staying illegally, may now decide to stay. The bond removes their guilt. Again, this is the exact opposite of what the government wants to happen! 
         The security bond is a bad idea, even if the goal is to reduce immigration. So, hopefully, there is still time for the policy to be quietly dropped.

* Frey, B. S., & Oberholzer-Gee, F. (1997). The cost of price incentives: An empirical analysis of motivation crowding-out. The American economic review, 87(4), 746-755.

Saturday, 22 June 2013

Waiting times in A&E

The UK's National Health Service seems to have been constantly in the news in recent months for the wrong reasons. One issue has been waiting times at Accident and Emergency Departments. The government's target is to treat 95% of patients at A&E departments within 4 hours. Whether or not the target is met has become a general indicator of pressure within the NHS. Recently the government missed the target. But, how useful are such targets?
      Let us look at a hypothetical A&E department at 6pm on the 2nd July 2013. The waiting room is full of people. To be efficient we need to work out the benefit of treating each patient and compare that to the cost. The benefits and costs are shown in the diagram below. To illustrate how the benefit side works we can pick out two of the patients: David has had a heart attack and needs treatment or he will die, while Brian has sprained his ankle playing football. The benefit of treating David far exceeds that of treating Brian. On the cost side we recognize the capacity constraint in the department. The department can treat Q patients at any one time without much cost, but cannot realistically treat more than that. Who should the hospital treat? Clearly they should treat David and leave Brian waiting.
      Fast forward one hour. David is now in intensive care and Brian is still waiting. A new set of patients has arrived. Amongst these new patients is John who was in car accident and has serious injuries. Again, the benefit of treating John far exceeds that of treating Brian and so Brian is going to be left waiting a bit longer.
      Its now 10pm and Brian has been waiting four hours. The benefit and cost trade-off, however, has not changed. David and John have been treated but many new patients have arrived in more need of treatment than Brian. For example, Sarah has just arrived after a fall that caused serious head injuries. Who should the hospital treat? On any ethical and moral grounds they should treat Sarah. There are, however, 'strategic' or 'management' incentives to treat Brian in order to meet the target. Essentially, the benefit of treating Brian becomes artificially higher due to the target.
         One would hope that decisions are made on medical grounds. Unfortunately, however, we have evidence that sometimes they are not. Serious problems, for example, were uncovered at Stafford Hospital. An inquiry into the poor standard of care found a management culture that put targets ahead of patient care. Receptionists were left deciding who to treat in A&E. Brian might get treated ahead of Sarah. 
       Our hypothetical example illustrates the way that targets can distort incentives.They provide an incentive for the department to treat Brian ahead of more needy patients. And things work the other way too: why did Brian turn up at A&E with a sprained ankle? The promise of being treated in under four hours may have been a deciding factor. Which is why increasing the capacity of A&E is not necessarily the answer to the problem.
        This is not to say that targets are not a good thing. Performance measures and indicators are clearly essential to motivate performance. It is, however, necessary to be careful how targets are used. Setting a target that can be artificially met is not good - it is easy for a hospital to meet the waiting time target by sacrificing patient care. Setting a target that makes no sense is also not good - On neither efficiency or equity grounds do we want a hospital treating Brian ahead of Sarah. Undue focus on one target is also not a good thing - a mix of targets avoids distorting incentives in one particular way. Such nuances are, though, difficult to sell to a public audience that likes simple black and white tests of whether the NHS is performing up to standard.

Saturday, 8 June 2013

Why would you read an investment newsletter?

The latest issue of the Hargreaves Lansdown Investment Times arrived in the post last week. As always it was full of advice on which investment funds are good bets for your money. Adherents to the efficient market hypothesis would suggest that such investment newsletters are basically a waste of time. But, I always enjoying reading through my copy of the Investment Times. So, why can investment newsletters be useful?
       To answer that question let's start by explaining why investment newsletters are supposed to be useless. The efficient market hypothesis says that stock, commodity, bond, fund prices etc. should always reflect all the information available at that time. If, therefore, a freely available, published newsletter claims 'here's a great opportunity to invest' it shouldn't remain a great opportunity by the time you get the newsletter! The person writing the newsletter, for one, has an incentive to act on the advice. By the time you get the newsletter, prices should have adjusted, and the great opportunity will have disappeared.
      There's no denying that this argument has some truth in it. It is, for instance, unlikely a fund manager can outperform high profile indices, like the FTSE100, by 'good stock picking'. In such highly traded markets, prices will adjust immediately to new information and so 'good stock picking' equates to a heavy dose of  'good luck'. You would be better to invest in a tracker fund than pay extra for the services of a manager. An investment newsletter's claim to have found the best stock picker is likely to reflect past performance rather than future potential. They can tell you who was lucky, but are not so good at telling you who will be lucky!
     But, investment newsletters are not a complete waste of time. The 'textbook' benefit of a newsletter is that it can advise on the risk, return trade-off. Some investments are more risky than others and an investor would be well advised to take this into account. Someone close to retirement, for example, should have their money in relatively safe assets while someone far from retirement can play a more risky strategy. A good investment newsletter can explain this, and advise which investments are relatively safer or riskier.
      . To appreciate further benefits of a newsletter, we need to look at the limitations of the efficient market hypothesis. We know that in practice asset prices can be well above or below fundamental value. The issue then becomes one of time and patience. Suppose, for example, that shares in the FTSE100 companies are under-priced. Prices need not adjust immediately because speculators fear a further drop in prices. Indeed, for speculators eager to make money here and now, it is not particularly useful to know that companies are under-valued. For someone willing to take a longer term view, however, under-priced shares area a good opportunity to make money. Prices might go lower in the short run, but over time the price should readjust. Investment newsletters can advise on such opportunities. But, note, that you don't need a good stock picker to realize such gains. It is enough to invest in a tracker fund.
      The efficient market hypothesis also becomes less relevant when we come to markets with a smaller volume of information and investors. This puts attention on small companies and markets in developing countries. The lack of information creates the possibility for asymmetric information, and this can be exploited. Consider, for example, a small start-up company. Who knows whether this company will be a success or not? A good fund manager might be better than others at answering this question. Asymmetric information comes from the manager's efforts to study the firm, meet the CEO, ask the right questions etc. An investment newsletter can advise on who the good managers are. Such investments are likely to be quite risky, and so we can think if this as advice on the best risky investments.
     So, investment newsletters are not a complete waste of time. But, the advice needs to be treated with care. Claims on how to outperform highly traded markets are almost certainly overblown. Advice on which markets to invest in and who to trust with your money to in less traded markets is potentially more valuable.

Saturday, 1 June 2013

Last minute deals and price discrimination

I was flicking through a magazine yesterday when I saw an advert for the travel company Great Railway Journeys. The advert caught my eye because it guaranteed that 'you'll never pay more than last-minute bookers. If we reduce a holiday price for any reason, we'll give the same saving to anyone who has already booked'. To someone brought up on the microeconomics textbook this can sound a bit weird. The textbook tells us that price discrimination - charging different people a different price for the same good - is one of the main ways a company can increase profit. So, why would a company guarantee that it will not discriminate?
     Last minute deals are an example of second-degree price discrimination. This is where a company knows there are different types of buyer but cannot tell them apart. By offering a menu of packages the company can potentially get customers to reveal their type and charge them accordingly. To illustrate: Holiday makers may differ in their willingness to pay for comfort. Hotels with standard rooms, executive rooms, junior suites, and presidential suites are offering a menu of packages for holiday makers to choose amongst. And anyone staying in a presidential suite can expect to pay a lot more than someone staying in a standard room. The hotel can profit from discriminating in this way.
   The main hope of last minute deals is to discriminate between customers willing to pay a lot for a holiday - who will book early - and those willing to pay less for the holiday - who will book at the last minute. Many travel companies clearly use this strategy. But, there is a basic flaw - a time-inconsistency problem. If customers expect the price to drop at the last minute then why book early? Customers have an incentive to delay purchase and pick up the bargains. Unfortunately, this is the exact opposite of what the company wants: it creates uncertainty about the level of demand and means the company has failed to discriminate between customers.
     To understand why last minute deals can fail to work, its worth mentioning that most examples of second-degree price discrimination also come by the name of hurdle pricing. The idea being that to get a cheap price the customer has to overcome some hurdle. This will only work if the hurdle is high enough. For a traveler who values their comfort and has money to burn, staying in a standard room is a big hurdle; to get a cheap price, they have to sacrifice the comfort of the presidential suite for the cramped standard room. Booking late, however, is not much of a hurdle, particularly when there are lots of companies with holidays on offer. To get a cheap price, the customer merely has to wait a bit. And given people's present bias propensity to delay, waiting a bit is not a big hurdle!
    The time-inconsistency problem means that last minute deals can be more of a curse than opportunity for profit making companies. So, an ability to price discriminate is not always advantageous. Which explains why a company like Great Rail Journeys would want to 'tie their hands' and rule out last minute bargains. The guarantee 'if we reduce a holiday price for any reason, we'll give the same saving to anyone who has already booked' does this in a seemingly credible way. With a guarantee like that they have no incentive to drop the price. Moreover it should reassure customers that the list prices are not too high.