A Freakonomics Proposal to Help the British National Health Service

In the first chapter of our new book, Think Like a Freak, we recount an ill-fated interaction that Dubner and I had with David Cameron shortly before he was elected Prime Minister of the U.K. (In a nutshell, we joked with Cameron about applying the same principles he espoused for health care to automobiles; it turns out you don’t joke with Prime Ministers!)

That story has riled up some people, including an economics blogger named Noah Smith, who rails on us and defends the NHS.

I should start by saying I have nothing in particular against the NHS, and I also would be the last one to ever defend the U.S. system.   Anyone who has ever heard me talk about Obamacare knows I am no fan of it, and I never have been.

But it doesn’t take a whole lot of smarts or a whole lot of blind faith in markets to recognize that when you don’t charge people for things (including health care), they will consume too much of it. I guarantee you that if Americans had to pay out of their own pockets the crazy prices that hospitals charge for services, a much smaller share of U.S. GDP would go to health care. And, of course, the same would be true in the U.K.

Smith ends his critique by writing:

But I don’t think Levitt has a model. What he has is a simple message (‘all markets are the same’), and a strong prior belief in that message.

Smith could not have known, based on what’s in Think Like a Freak, that we actually do have a model for the NHS. And, indeed, I proposed the model to Cameron’s team after he left the meeting.

If nothing else, the model is admirably simple.

On January 1 of each year, the British government would mail a check for 1,000 pounds to every British resident. They can do whatever they want with that money, but if they are being prudent, they might want to set it aside to cover out-of-pocket health care costs. In my system, individuals are now required to pay out-of-pocket for 100 percent of their health care costs up to 2,000 pounds, and 50 percent of the costs between 2,000 pounds and 8,000 pounds. The government pays for all expenses over 8,000 pounds in a year.

From a citizen’s perspective, the best-case scenario is that they use no health care, so they end up 1,000 pounds to the positive. Well over half of U.K. residents will end up spending less than 1,000 pounds on health care in a given year. The worst case for an individual is that he/she ends up consuming more than 8,000 pounds of health care, so that he/she ends up 4,000 pounds in the red (he/she spends 5,000 pounds on health care, but this is offset by the 1,000 gift at the beginning of the year).

If it turns out that consumers are sensitive to prices (i.e., that the most basic principle of economics holds, and demand curves slope downwards), total spending on health care will decrease. In simulations we’ve run at The Greatest Good, we estimate that total health care costs might decline by roughly 15 percent. That is a decrease in spending of nearly 20 billion pounds. This decrease comes because (a) competition will likely lead to increased efficiency; and (b) consumers will cut out the low-value healthcare services they are currently using only because the services come for free.

Everyone remains protected against catastrophic illness.

Like any government program, there are winners and losers. The majority of Brits will be better off in the scenario I laid out, but those who need to spend a lot on health care in a particular year will be worse off. That is because the system I propose provides only partial insurance – which retains incentives for consumers to make prudent choices. The healthcare system would then mimic the rest of life. When my TV breaks, I have to buy a new one. I’m worse off than the guy whose TV did not break. When my roof needs to be replaced, it’s expensive, and I’m worse off than if the roof didn’t need replacement. There’s nothing immoral about this; it is just the way the world usually works.

There are, no doubt, many improvements that could be made to this simple proposal. For instance, maybe the cash payment to the elderly at the beginning of the year should be larger than that to those who are younger. Maybe the cash payment is bigger to those who have chronic illnesses, etc.

I have no idea whether this sort of plan could be politically viable, but I have done some informal polling of the British electorate. Every time I take a cab in London, I ask my driver whether he would be in favor of my proposal. Probably the cabbies are just being polite, but roughly 75 percent of them say they would prefer my plan to the current system.

Perhaps, then, it is time for another audience with the Prime Minister …


I don't see how your model accounts for preventative care. It would seem to dis-incentivize the low-cost screening, which invariably leads some percentage to the more costly (by orders of magnitude) catastrophic medical bills, which we all pay for anyway.


"But it doesn’t take a whole lot of smarts or a whole lot of blind faith in markets to recognize that when you don’t charge people for things (including health care), they will consume too much of it. "

I'm not sure that's strictly true. If I leave a wheelbarrow of apples on my front lawn with a sign saying 'please help yourselves' people would generally take only as much as they had a need for; rather than take more than they need because they are there.

Health care isn't like candy; you don't consumer healthcare for the sake of it, you consume it out of necessity. I don't see a run on the police station or fire station because they don't charge you for calling them out when you are in distress.

Dubner is not the first person to suggest using incentives to reduce the demand on the NHS; Tim Harford did something similar in his Undercover Economist book. However there is limited evidence that this would actually reduce demand on the service that wouldn't result in a detrimental affect on the nation's health. The proposal appears to be based around a limited subset of conditions that the NHS deals with - namely illnesses and accidents that otherwise well-people get and then can be treated within a defined period.

The unintended consequence of this is that those on lower incomes are incentivised not to seek medical health when they need it so that they can use the money they save for other basic necessities such as food, clothes and housing.

The real cost savings in the NHS are changing the Victorian model of having a hospital per town into a system of having large superhospital with clear specialities and resources that are fully utilized. But no MP is going to support closing the local infirmary...

Above all, the NHS is a statement about our nation, which is why it played so well when it was highlighted in the Olympic Opening Ceremony (oh yer, forgot the US Networks cut the bits they didn't agree with). It is a clear statement that everyone, regardless of their ability to pay, should be afforded the dignity of a healthy life; and that when they are struck down by accident or illness, we shall all play our part in supporting them back to health. If that means paying a few more pounds on my tax bill to live in a country that thinks that's important, I'm happy to do that.


Stephen Werner

Has a comparison been made between Medicaid patients and poor working class who have insurance through their employers and have a co-pay.
I'm pretty sure that Meddicaid patients don't have a co- pay.
Has a study been made were 500 Medicaid patients who don't work, 500 Medicaid patients that do work, 500 patients who are low income patients but have medical insurance through their job, with a co-pay, and 500 stay at home patients, and how often does each group seek medical services?

Mike Toreno

I just wish there were some data - I wish the UK had actually started the proposed NHS, or that some other country had instituted a national healthcare system. Until that happens, we have nothing but speculation and thought experiments, and Levitt's idea is as good as any other.


Cab drivers are not usually thought to be a reliable source in most public policy.
The UK does not over consume health and the problem with the NHS is not frivolous demand, in any case most cost is consumed at the beginning and end of life


"The worst case for an individual is that he/she ends up consuming more than 8,000 pounds of health care, so that he/she ends up 4,000 pounds in the red (he/she spends 5,000 pounds on health care, but this is offset by the 1,000 gift at the beginning of the year)."

This is pretty bad for an elderly person with a low fixed income. It is also especially bad for a child whose parents decide not to give him adequate care for his illness. It's also bad for low income individuals with chronic illnesses who are 4000 dollars in the red every year for the rest of their lives. And I still doubt it would save very much money. Healthcare spending follows the 80/20 rule, if not the 90/10 or 95/5 rule. You don't know anything about healthcare, so I suggest you stop opining on it.

Twm Owen

I guess the author doesn't own a car? If he did he would recognise that while regular servicing can seem like a costly annual or perhaps twice yearly expense it will usually save the owner from a bigger bill by preventing costly major failure. It will also help maintain the vehicle in good working order for a longer period.
Free health care can help the earlier detection and treatment of diseases/health problems - thus saving money longer term and leading to a healthier population.


We don't aim to look after the majority of Brits. We look after all of them.

Money has different influences on consumption of services in each system. I delivered babies in the NHS and in Belgium, and saw different attitudes in patients and providers. (No, Brussels insurance people, I don't want homeopathy. No, not even if it's free. No really, nor a thalasso spa treatment once a year. Really, no.)

But efficiency is a secondary concern. The majority of Brits will always be ok, like the majority of Belgians and the majority of Americans. The reason Brits are so passionate about the NHS is that we look after everyone. Every single person in Britain has 100% full health cover. The homeless and the Queen.

An accountant might not call that efficient, but a demographer might. And the homeless dude with cancer.

We're proud of that with good reason.


What Levitt doesn't get is that the NHS isn't about the money, it's about what's RIGHT.


Hey guys. You raise a really interesting point about our health system, although in your system what would be the net positive for the British people? Efficiency savings? There are studies showing that the NHS is remarkably efficient - http://www.theguardian.com/society/2011/aug/07/nhs-among-most-efficient-health-services

Personally, I'd argue that the system we have now carries a far stronger incentive - living in a society in which nobody ever had to worry about a medical bill. Knowing no matter what happens you'll be cared for. That peace of mind is so much more valuable to the people than an abstract number like £20 billion.

Also, I'd advise against using London cabbies as a sample of UK opinion. They're notoriously far to the right of your average brit, not really a representative sample.

Appreciate the insight!


>I have no idea whether this sort of plan could be politically viable

No, it's not. Attempting to dismantle the NHS would destroy the politician's career. And you don't really seem to understand how the NHS works anyway. There are no low-value healthcare services that people only use because they're free; and people can't just demand to use something, unless they want to pay for it privately (which they already have the choice to do).

And the taxi drivers in London are almost certainly agreeing with you to butter you up because you're in London, the tourist capital of the world, and they think you're likely to tip them money if they Yes Man you. Try leaving London and speaking to real people. Especially in the North.


I like this concept. I think it has two problems: 1) people know very little about prices of health care (is your "cold" going to cost 100 dollars or 10,000 dollars?) and 2) it IS immoral to let the sick suffer (if I can't replace my TV because it broke, oh well. But if I break your leg, I will have to pay whatever it costs to get fixed).

How about we start with figuring out how much everything costs?


I don't think that people are rational consumers when it comes to health care. Most people do not have the education, time, or motivation to evaluate medical treatment based on scientific evidence in combination with the costs. One example is the people who refuse to vaccinate their kids. Another example is the people who wait too long to go to doctor and their condition becomes more expensive to treat. A thought experiment:

Let's say I have a stomach ache, and it hurts for a few days, but I've had stomach aches before and they always get better. I don't really want to spend $100 out of my free $1000 for a doctor to tell me I need rest. How do I evaluate the probability that it will get better instead of worse, and if that chance is worth $100 to me? If my appendix bursts the next day, it will probably cost way more than the $8000 cap.

I highly recommend this playlist (http://www.youtube.com/playlist?list=PLkfBg8ML-gIngk82SUbTp6Og_KkYfJ6oF). It goes over the health care systems of different countries. The last one they covered was England.



I have had health insurance in 6 different countries, including the UK. I would take the NHS system anytime for providing the greatest good in the simplest way to the greatest number of people.


How do you incentivise people with power to look after vulnerable people?

- Free at the point of delivery
- Payment direct from salary contributions with each time providing very little opportunity for using their disproportionate power to opt out from self-interest

Historically, it took a world war to make the majority of people act as a unified population and support the 'deserving' vulnerable. We won't get another one of those (I sincerely hope). If the NHS is destroyed by consistent budget cuts and moves towards market systems, we get the worst of all worlds.

Discuss all you like, be free, open and creative thinkers.... but please represent those in the NHS struggling through nightshifts at appalling salaries, risking physical harm from late night drunks and stopping work early on benefits from bad backs who are motivated by nothing but service for all people. All of my nurse and doctor friends could earn more and work less in other sectors (private practice or other careers) but have chosen to do this. Their incentives are neither money nor thanks. Just good work and a belief in the values NHS.




Hi Steven,

fascinating insight into the world of economics. Can I raise a couple of points? I've tried to keep these to questions on your argument rather than on the fairness of your argument.

1. The post compares having good health to having access to a television set. As an economist are you proposing that the value to a wider society of having healthy citizens to the value to a wider society of citizens having access to television? If so are you thinking of any particular channels or shows to extend the comparison - let's say, losing access to Britain/America's Got Talent might be equivalent to losing access to one of the less-useful toes in the middle of the foot in terms of damage to wider society?

2. Did the simulations at The Greatest Good factor in the cost of education to ensure that all citizens understood the cost and likelihood of health care being required and for society to intervene where a citizen was unable to make a choice but had spare funds, say because they were an orphaned child or an elderly retired academic with poor reasoning faculties?

3. Could you release the data and methodology behind the simulations?

4. Similarly could you release the data and methodology behind the evidence presented from the research performed with taxi drivers? I'd be delighted to understand how the towns, cities and taxis were selected.

5. What are your views on the higher cost per citizen in both the current and prior US models as compared to the UK model?

6. Your worst case example includes the citizen spending money. Would you accept that in the worst case the death rate will rise with associated effects? How did you factor this into your model?

I've got some other questions but happy to wait for these to be answered first