According to the OECD the world spends about 1.15 trillion dollars a year on research and development. The US spends the most at 0.46 trillion, with China second at 0.41 (but rapidly catching up and likely to exceed the US soon). That is about 2.3% of worldwide GDP. That includes all sources, public and private.

From any perspective, that is a lot of money. It is also a good thing – the world is investing a significant amount of its activity and resources in the future. I don’t know what the optimal percentage for such investment is, and I am sure someone could make a reasonable argument that it should be higher. What I want to discuss here, however, is not how much we invest but how we invest in research. In broad brushstrokes – what should be our research priorities?

If you think about it, where we invest in research essentially determines the path forward that our civilization takes in terms of science and technology. So it’s worth thinking about how that quadrillion plus dollars gets spent.  Right now we have a decentralized R&D infrastructure, with many different facets. There is a lot of “bottom up” research, meaning that individual researchers, companies, labs, and other institutions are determining for themselves what to research based upon their own priorities. There is also some “top down” research in which large funders, mostly governments, determine research priorities through their granting process. There are strengths and weaknesses to a diffuse system like this, but I think overall it’s pretty good. Essentially free-market forces are at work with some nudges here and there.

Private researchers are largely going to prioritize R&D that makes them and their investors money. That’s fine, as this also produces incentives to make things faster, better, cheaper. All governments can do in this situation is look for perverse incentives and mitigate them through regulations. For example, I strongly believe we should not allow industry to simply externalize the costs of their profit-making enterprises. That amounts to a hidden subsidy, and also creates a perverse incentive to externalize costs (for example, by dumping into the environment).

Governments can also give direct incentives to companies for researching things that society may value, but are not very profitable. The first thing that comes to mind is the orphan drug law, which gives pharmaceutical companies incentives to invest in research in rare diseases that would not otherwise be profitable.

And of course governments, but also large charitable organizations, can influence the direction and priority of research through direct funding. This is a way for society to collectively determine research priorities, through our elected officials or by directly contributing to a charity dedicated to a goal you support.

Given all of this, the conversation I would like to see addresses the question of whether or not we are collectively prioritizing cost-effectiveness and sustainability in our research priorities. First let me talk about my own field of medicine.

Rising health care costs are a significant problem in developed countries. According to government statistics:

U.S. health care spending grew 4.3 percent in 2016, reaching $3.3 trillion or $10,348 per person.  As a share of the nation’s Gross Domestic Product, health spending accounted for 17.9 percent.

There are several factors increasing health care costs, including an aging population and increases in obesity, but the largest factor is advancing medical technology. We do more now than we did previously, because we can do more. So essentially the direction of medical research is not sustainable because it is driving up the cost of health care to unsustainable levels. Many countries deal with this problem by simply rationing care. This is a solution that is unlikely to be politically acceptable anytime soon in the US, and also isn’t the optimal solution. Just refusing to deliver high-tech expensive care may save money, but sucks if you are the person being deprived of treatment.

Of course there is room for improvement in efficiency and eliminating waste. But even if we eliminated 100% of waste in medical care, that would only delay the problem, not solve it. We also have to think carefully about how much money we spend in ultimately futile end-of-life care.

Despite these important but tangential issues, the bottom line is that we cannot really afford to give the full care that we have the technology to deliver. Despite this we continue to research even more expensive high-tech interventions that provide marginal improvements in outcome, or may even just be equivalent to existing treatments.

This is where the question of sustainable research comes in. It seems to me that we need to prioritize much more than we currently do R&D into treatments that may be no better than existing treatments, but have the sole virtue of being cheaper. Of course we still need to do research into treatments that improve outcomes and give us more treatment options, but only prioritizing these factors has lead to an explosion of health care costs that is not sustainable. Identifying the most expensive interventions, and then researching other options that are reasonably equivalent in outcome but substantially less expensive could ultimately have more of a benefit on overall health outcomes than making an expensive but incremental advance.

The simple way to do this is to prioritize cost effectiveness in NIH funding. Get the word out to researchers that there is money to fund their lab if they are trying to developer cheaper, rather than just better, medicine.

The harder nut to crack will be private research, such as the pharmaceutical industry. They love high-tech very expensive treatments, because that’s where the profit is. Often the only check on them is the insurance industry, who doesn’t want to pay for expensive but marginal treatments. But if the expensive treatment is the only option, they may have no choice. So we need to get creative and think of ways to alter the incentives for pharmaceutical companies to research inexpensive treatments. Perhaps, for example, the length of their patent could depend on the cost of their drug. The cheaper it is, the longer they get to keep their patent, while expensive drugs expire quickly. Perhaps tax breaks could be used to reduce the risk of upfront R&D costs for cheaper products on the back end.

Regardless of the ultimate details, we need to move our healthcare infrastructure through our research priorities in the direction of less costly interventions. Right now we are moving in the direction of more and more expensive interventions for incremental improvements.

Another example of sustainable research priorities is developing technology to do things more sustainably, even if they are a lateral move in terms of all other factors. For example, plastic pollution is a huge and growing problem. Part of the problem is that plastics are not infinitely recyclable. Their quality breaks down with each cycle, and they eventually end up in a land fill or just in the environment, where they last an incredibly long time.

It would therefore be a huge advantage to develop a plastic substitute that either breaks down quickly and completely, or can be infinitely recycled. One recent study does show a significant advance in this latter direction. Scientists have developed a polymer that can be easily broken down into its original form, and used an unlimited number of times.

But this type of polymer is not likely to be cheaper or better in any other respect, and in fact may have some inferior qualities. So the question is – how much do we prioritize the feature of being recyclable? That gets us back to externalizing cost – who pays the bill for plastic waste? In a very real way, we all do. If we attach a real cost to those contributing to plastic waste, then a more ecofriendly plastic might become cost-effective.

The ultimate question of how we prioritize our research is fairly complex, because we cannot reliably predict how research programs will pan out. By definition, we research what we don’t know, so there will always be an element of the unknown in all this. There is also evidence that just letting researchers follow their interest, without setting specific practical goals for their research, does result in significant and practical advances. We need to fund a certain amount of basic science for this research.

But R&D specifically to develop practical results are directed at certain priorities. We need to think carefully about what those priorities should be, looking at not only the big picture but the long term. That is ultimately what I am talking about – adequately taking a long term perspective in our research priorities. If we look at only short term outcome, we will not necessarily move in the direction we want long term. We may research our way into a corner we will find it difficult to get out of.