The argument is not entirely wrong. There is some empirical evidence that private providers tend to treat healthier patients, and that this can have adverse effects on nearby NHS providers. But in its generalised form, it represents sloppy thinking. It is not always a problem if private providers concentrate on easier cases, and whenever it is a problem, it is either caused by a faulty payment structure or by improperly specified contracts. It cannot be used as a general argument against allowing private providers into healthcare.
Let’s go back to the basics. What exactly is cherry-picking? Let’s start by illustrating what it is not.
When the Labour government first allowed the establishment of Independent Sector Treatment Centres (ISTCs) in the early 2000s, they had a very specific division of labour between the NHS and the ISTCs in mind. They wanted ISTCs to concentrate on relatively simple high-volume procedures, taking some pressure off local NHS providers so that they could concentrate properly on the difficult cases. In healthcare, the relationship between the private and the public sector was meant to resemble that of a senior professional and their intern: ‘You can do the routine stuff, but leave the serious business with me’.
Whatever one makes of this split of responsibilities, it is not an example of cherry-picking. If ISTCs ended up with more simple cases, it was because that was the idea. The private-public division of labour was meant to work that way.
You might object that ISTCs are a bad example because they do not directly compete with the NHS, so let’s have a look at a scenario in which public and private providers do compete for the same patients, and where cherry-picking could be a problem. Suppose there are 100 patients suffering from the same condition. For half of them (the easy ones), the treatment cost is 40 thaler, and for the other half (the difficult ones), it is 60 thaler. There is one private and one public hospital, both of which offer the treatment. Will the private hospital cherry-pick the easy patients?
It depends. If they are paid on the basis of average treatment cost (say, the reimbursement formula is average cost plus 10%, i.e. 55 thaler), and if the hospital can select patients on a case-by-case basis, the answer is yes. They will then end up with a surplus of 750 thaler, while leaving the public hospital with a loss of 250 thaler.
Yet if the hospitals are paid on the basis of individual cost, they should be indifferent between the two patient groups. Moreover, in a pure fee-for-service payment system, we could even get ‘cherry neglect’. Patients in generally good health would then be the least profitable ones. They will not require any extras, and they are unlikely to be back anytime soon. If hospitals were pizzerias, uncomplicated patients would be the equivalent of a customer who shows up once a year to order a Pizza Margarita and a glass of house red. Those are not the customers who a get a free Limoncello after their course, and they are not greeted with a ‘Ciao, amico!’ either. A pure fee-for-service system is generally undesirable because it encourages oversupply, but fixed fees with appropriate risk-adjustments are enough to make cherry-picking economically unattractive.
Alternatively, cherry picking can also be avoided through block contracts, where providers commit to taking care of the whole fruit basket, not just the cherries, or lose the whole contract. Healthcare commissioning contracts would then be more like employment contracts, where you get paid for a package of services, some of which are more enjoyable than others. Discrimination can, of course, occur in more subtle ways, so this may not be enough. But a coupling of financial incentives with well-specified contractual obligations can rule out cherry-picking. It works in almost every other developed country, so why would it not work here?
Most objections to private healthcare are not really about practical downsides, of course. In the minds of NHS purists, the entrance of private providers would be a corrupting influence that debases an institution they perceive as noble and elevated. They object to it for the same reason that churchgoers would object to the idea of leasing out a part of their church to Burger King, so that they could sell Double Whoppers with fries and Coke during mass.
But this blog is about economics, not spirituality. From an economic perspective, there is no good reason to shut private for-profit providers out of the health sector. Cherry-picking is not an issue in mixed systems elsewhere, and it need not be an issue here.
 This assumes, of course, that hospitals can identify which patient belongs to which group, that the public hospital will accept any patient, and that the private hospital has a way of attracting the easy patients.