May 23, 2021
People in policy debates are not supposed to question the desirability of patent monopolies as a mechanism for financing the development of new drugs and vaccines. After all, why ask a question that could jeopardize the profits of some of the world’s largest corporations? But, since I live out in Southern Utah, far away from the great centers of policy debate, I thought I would ask the question in reference to vaccines against Covid.
To be specific, suppose that instead of funneling money into drug companies to subsidize the patent monopoly financed system, we instead use this money, and added more to it, for the purpose of fully prefunding the development of vaccines. The condition of accepting funding is that all the work would be fully open-source.
This means that all the findings would be posted on the web, so that researchers around the world could build on them. It also means that any patents would be in the public domain so that any manufacturers, anywhere in the world, could produce the vaccines developed through this system, if they had the necessary expertise. The requirement for openness would also apply to the results of clinical trials, so it would be possible for researchers to know which vaccines were most effective for specific demographic groups and against which variants of the virus.
This system would require some sort of international agreement on sharing research costs. While agreements can take time to work out, getting the numbers exactly right should not have been an impediment. In the context of a worldwide pandemic costing millions of lives and trillions of dollars of lost output, it is not a big deal if the United States or China pays 25 percent too much or too little. The simplest course would be to agree to rough numbers with a commitment to re-examine the issue down the road.
The Open Source Alternative
Having raised the idea of open source research in various contexts over the last year or so, I realize that many people are not familiar with the concept. The basic idea is that the research is fully open, and posted on the web as soon as practical, so that any researcher in the world can read it and build on it.
I have had people ask me why anyone would do research if they had to immediately share the results, rather than secure a patent. The answer is that they are getting paid for it. This is in fact the reason the vast majority of researchers are doing their work now. For the typical scientist at Moderna, Pfizer, or Merck it doesn’t matter whether these companies are making their money through ownership of a patent monopoly or through a government contract. They get a paycheck every month either way. (I recognize that many scientists may actually be concerned about controlling a pandemic and saving lives, but I am not expecting that motivation to be a replacement for money in a paycheck.)
If the government is paying for research, it can impose the condition that everything must be open-source. If a company doesn’t like the deal, it doesn’t have to take it. That’s pretty simple.
This is a big deal with the current crop of vaccines being used, especially the mRNA vaccines. In addition to having already paid for much of the basic research developing mRNA technology through NIH, the government also paid for the cost of developing and testing Moderna’s vaccine.
Many people have pointed out that Moderna had invested in developing a platform that allowed them to be able to quickly turnaround and design a vaccine against the coronavirus. There is some truth to this. Arguably the government would have had to make some additional payment to Moderna, and likely other drug companies to get them to make all the information about their vaccines, and their production process, fully public.
It would take some serious analysis to get a good estimate of a fair payment, but to put a rough cap on what we would be looking at, in the case of Moderna, it would almost certainly be less than $2 billion, and likely considerably lower with other manufacturers working with less novel processes. (The Oxford-Astra Zeneca vaccine was developed almost entirely with public money, so there should not have been a need for any substantial payments to make it fully open-source.)
As far as the $2 billion figure for buying full rights to open-source Moderna’s technology, it had a market capitalization of around $9 billion before the pandemic. Paying the company $2 billion for the technology directly related to producing the coronavirus vaccine would mean giving them almost 25 percent of the value of the company for this one specific line of work. (Alternatively, we can envision paying $9 billion to buy the company, and then selling off the components that are not directly related to producing a coronavirus vaccine.)
If Moderna was reluctant to sell its rights for $2 billion (or some comparable sum) the government could attempt to go around the company and simply pay its top researchers and engineers large sums (e.g. $1 million a month) to share their knowledge with the world. The government would also have to agree to cover all legal fees and settlements that could result from any subsequent suits by Moderna.
It is worth noting the logic of open-source, which would actually make $2 billion sound very generous. Suppose that other manufacturers but not Moderna, agreed to a buy out of rights and do future work as open-source. This would mean first, that Moderna would need to come up with the billion dollars that the federal government paid to cover its development costs and Phase 3 clinical trials.
Even more importantly, Moderna would know that as soon as its vaccine had been through the testing and approval process, it would be competing with other vaccines, which may be every bit as effective, and selling as cheap generics. In addition, it would be having a very difficult time getting the necessary inputs, since the federal government would be using the Defense Production Act to steer all the inputs in short supply to produce the vaccines developed by companies that agreed to open-source their work.
Also, the government could throw a large monkey wrench into Moderna’s plans by demanding harsh terms for the use of a NIH patent that is essential to the vaccine. The government’s rights to this patent could guarantee that Moderna would not be making big profits by refusing to come to terms.
Many people have argued that these sorts of terms would destroy incentives for drug companies to develop vaccines in the future. This claim is difficult to understand. Presumably drug companies undertake investment with an understanding of both risks of failure and expected payoffs. No one is proposing a situation where the drug companies would not be making a profit from the development of the vaccines for Covid. The issue is simply whether they would get a super-bonanza from having developed a vaccine against a pandemic. This path takes away the super-bonanza.
So, if we think about the incentive structure going forward, if governments just handed out patent monopolies, including on research that they had paid for, then drug companies could incorporate in their planning that there is maybe a one in fifty or one in a hundred chance that there will be a pandemic in the relevant time-frame for their investment. If we pursue the route of forcing open-source development, then the drug companies will not incorporate the low-probability prospect of a pandemic super-bonanza in their planning. We should be able to live with that.
The Impact on Production
It would be difficult to determine how much vaccine production could have increased if we had gone the route of open-source research and international cooperation, but it is likely that the gains would have been substantial. First and foremost, sharing of technology would have allowed for it to be replicated widely. If all the information needed to produce the Pfizer, AstraZeneca, or Sinopharm vaccines were posted on the web, then any company anywhere in the world with expertise in production of vaccines or related items could start converting or building a factory.
BioNTech, Pfizer’s German partner, bought a Novartis plant in the middle of September of 2020. It converted the plant to producing its mRNA vaccines. It now expects to be able to produce 250 million doses in the first half of 2021, and have an annual production rate of 750 million a year. This indicates a very rapid rate of turnaround. The Novartis plant was already a sophisticated manufacturing facility, so this surely saved some time, but there are other sophisticated facilities around the world which are not now being used to produce Covid vaccines.
If there had been a clear international commitment to producing vaccines, even with the risk that some might prove ineffective or unneeded, this sort of conversion could have happened much sooner and with other facilities simultaneously. It would be an enormous risk for an individual company to spend hundreds of millions of dollars to convert a factory to produce a vaccine that may ultimately not be used. However, this risk is trivial for a government spending trillions of dollars to sustain its population through the pandemic. In fact, the purchase and conversion of the Novartis plant actually was funded by the German government.
Also, while the mRNA vaccines have proved to be the most effective in stopping the spread, there are several other vaccines that have been approved by various governmental regulatory agencies. These other vaccines use different technologies, which should mean more facilities could be used and they would not be competing for inputs. Even if their effectiveness rate is somewhat lower, vaccines that are 70 to 90 percent effective will still enormously slow the spread. Also, these other vaccines have advantages over the mRNA vaccines in that they do not have to be frozen, and, the case of the Johnson and Johnson vaccine, only require one shot.
If we had gone this route of full international governmental commitment, we also could have acted to remove bottlenecks in production. As the industry has repeatedly pointed out, many of the inputs for vaccines are in short supply. According to Thomas Cueni, the director general of the International Federation of Pharmaceutical Manufacturers and Associations, there is a major shortage of items like syringes and vials that is impeding the effort to vaccinate the world. (This assertion can be found at 21.10 here.) While even the best coordinated efforts may not be able to remove all bottlenecks, surely we could have produced an adequate supply of syringes and vials one year into the pandemic.
The other side of this sort of international commitment and open-sourcing is that we almost certainly would have developed more innovations in the production process. Pfizer reported in February that its engineers had figured out a way to alter their production process to cut the time needed almost in half. They also discovered that it was not necessary to super-freeze the vaccine at minus 94 degrees Fahrenheit. It can instead be stored in a normal freezer for up to two weeks, which hugely eases the distribution process.
If Pfizer’s production technology had been open-sourced early in 2020, it seems hard to imagine that no engineer in the world could have discovered the route to improve efficiency before February of 2021. There presumably are other ways to make its technology more efficient, which its engineers have not yet discovered. There would be a similar story with the other vaccine manufacturers. Also, if Pfizer had open-sourced its technology, it would be surprising if someone somewhere had not discovered it wasn’t necessary to super-freeze the vaccine before March of this year. In fact, the company did not even realize that its vials contained enough material for six shots rather than five (a 20 percent increase), for more than a full month after they were being widely distributed.
The Benefits in Lives and Money
There clearly were, and undoubtedly still are, huge potential gains from open-sourcing the technologies used to produce the Covid vaccines and key inputs, as well as internationalizing the production process. If this had been done when the pandemic was first recognized, we could have hugely increased the pace of vaccine output and gotten the pandemic under control more quickly. Given the possibilities for converting and building more facilities, as well as increasing the efficiency of the existing facilities, we can envision that we would have two or three times the number of vaccines as we do today. This would have had an enormous impact on lives and the economy.
To get some idea of how a doubling of available vaccines might have impacted the course of the pandemic, it is worth remembering that the worst wave in Europe and the United States was hitting in January and February, just as vaccines were starting to be distributed. By January 15th, 3.2 percent of the population in the United States, or 10.5 million people had received at least one shot. This was a period in which the number of infections was averaging more than 230,000 a day and deaths were over 3,400 a day.
If we had twice the available vaccines, and had a competent delivery system (this was still during the Trump presidency), we would have had 21 million people vaccinated at that point. This would have covered all of our health care workers and the bulk of our older and most vulnerable population. Of course, many of these people would have only had one shot, which means they don’t have the full benefit of the vaccine, but the number of people getting seriously ill and dying would have been sharply reduced.
By March 1 we had vaccinated 15.2 percent of the population or 50.2 million people. If we had twice the number of vaccines available, we would have vaccinated 30.4 percent of the population, or 100.4 million people. At this point we would have enough vaccines to fully cover older people and others at risk. We were still seeing almost 70,000 cases a day at that point and 2,000 deaths. With this higher level of vaccination, the point actually reached at the start of April, the number of infections and deaths would be correspondingly lower. By the start of April, the number of infections was averaging just over 60,000 a day and the number of deaths had dropped below 1,000 a day.
If we maintained a doubling of our actual rate, by the end of March, almost 60 percent of the population, more than 195 million people, would have been vaccinated. (On May 22, we are at 48.6 percent.) At this point, increasing vaccination is a question of persuading vaccine resistant people to get the shot, it is not an issue of supply. While it would be necessary to maintain a stockpile, additional production could be given to other countries. We also could envision that the economy would have been largely reopened by the beginning of last month.
There is a similar story in Europe, which also saw its infections peak in January and February. In mid-January, France was seeing 17,000 cases a day, and almost 400 deaths, Italy averaged 16,000 cases and almost 500 deaths, and the UK over 50,000 cases and almost 1,100 deaths. The share of the population that was vaccinated as of January 15th in France, Italy, and the UK was 0.6 percent, 1.9 percent, and 5.3 percent, respectively. Doubling these rates would still only allow for a small share of highly vulnerable people to be vaccinated in France and Italy. In the UK, a 10.6 percent share would largely cover its health care workers and older population.
By March 1, these percentages had risen to 4.6 percent for France, 5.1 percent for Italy, and 30.2 percent for the UK. Doubling these figures would largely cover health care workers and the older population in France and Italy, and would imply that virtually anyone who wanted a shot in the UK had one.
The hardest hit countries in the developing world, India and Brazil, could have benefitted hugely from a more rapid rollout of vaccines. India hit its peak infection rates of just under 400,000 a day around May 10th. Its reported deaths averaged over 4,000 a day. (These figures are almost certainly gross understatements, as testing is limited and many Covid deaths are going unreported.)
On May 1, 9.2 percent of its population was vaccinated. If it had doubled this figure, 18.4 percent of the population would have been vaccinated, a share that would be large enough to substantially slow the spread and protect the most vulnerable segments of the population. In addition, with countries like the U.S. and UK having largely saturated demand by late March, vaccines could have been re-allocated, at essentially zero risk, to help India cope with the pandemic.
Brazil saw a peak of infections and deaths somewhat earlier. Its infections peaked at more than 75,000 a day in mid-March, with deaths averaging over 3,000 a day in early April. (There is also serious under-reporting in Brazil.) Its vaccination rate was 4.5 percent on March 16th. A doubling to 9.0 percent would have allowed the country to cover its health care workers and most vulnerable population. As with India, Brazil could have benefitted from vaccines being reallocated from the U.S., U.K. and other countries that had satiated demand by the end of the month.
It would be interesting to see a more careful calculation on the reduction in infections and deaths that would have resulted from a doubling of the production rate of vaccines, but it clearly would have been substantial. This is without even considering questions of reallocations between countries, like should health care workers and older people in developing countries be vaccinated before young healthy people in the United States and Europe.
There also is the over-riding question of the risk of the development of a vaccine resistant strain. While it may not be easy to get a good estimate of the probability that a vaccine resistant strain will develop in the months and years ahead, the costs are clearly catastrophic. Even if the existing vaccines can be adjusted to make them effective against a new strain, we will see many months of infections, deaths, and shutdowns, until the new vaccine can be tested and produced and distributed in sufficient quantity to protect the bulk of the population. Countries that have largely succeeded in vaccinating their populations should be willing to pay a large price to avoid this risk.
Big Gains from Overcoming Inertia
Anyone who has spent any time in Washington policy debates knows that inertia is an incredibly powerful force. Really bad policies can survive for decades just because it’s much easier to leave things as they are than to change them.
This is clearly true for our system of patent monopoly financing of prescription drugs and vaccines. The pandemic is a full out emergency that should have led us to do everything possible to stem its spread as quickly as possible, coordinating as much as possible with other countries (even Russia and China) in a genuine international effort.
Unfortunately, that was certainly not the agenda of the Trump administration. The result was many unnecessary infections and deaths, and enormous economic waste. While we are some distance down the road now, we can still gain from taking the path of openness and cooperation.
 This would also be the case with manufacturers of inputs for vaccines, who would have both patents and industrial secrets connected with their production process. The government would have to be buy these rights in order to make them fully open-source.
 My own view is that we should move away from patent monopoly financing for the development of new drugs and vaccines and switch to direct open-source funding (see Rigged, chapter 5 [it’s free]), but that is beside the point.