Market-shaping approaches to accelerate COVID-19 response: a role for option-based guarantees?

post by DerekF, Davidmanheim · 2020-04-27T22:43:26.034Z · LW · GW · 2 comments

Contents

  Summary
  The problem
  Potential solutions
      1. Prizes
      2. Public-private partnerships
      3. Direct purchase orders
      4. Option-based guarantees
  Recommendation
  Annex 1: Potential variations on standard put options
    Declining payout
    Priced contracts
    Early-ending bonus
  Annex 2: Key questions about option-based guarantees
      Q1: Isn't this a giveaway to corporations? 
      Q2: Isn't this wasteful? 
      Q3: Won't there be fraud or unnecessarily high costs?
      Q4: Won’t safety and quality suffer?
      Q5: How do you ensure the final product is affordable?
None
2 comments

This is a policy brief directed at decision makers in the UK government, with a view to accelerating production of tests, drugs and vaccines for COVID-19; but it could be adapted for a wide variety of countries, products and crises. Critical feedback is very welcome.

Thanks to Sam Hilton, Tim Colbourn and anonymous others for input on previous drafts.

Summary

Effectively tackling COVID-19 will require rapidly scaling up the production of diagnostic tests, pharmaceutical treatments and vaccines. In each case, preparations for large-scale manufacturing, such as building factories, are typically delayed until the product is proven safe and effective. This makes sense from a commercial perspective, but incurs great costs in terms of lives lost and damage to the economy.

There are several potential solutions, but the most promising appears to be “option-based guarantees”. In essence, the government commits to paying a proportion of the manufacturer’s preparation costs should the product turn out not to be viable. (If the product is viable, it can be sold as normal.) This reduces the risk to the company while maintaining an incentive to produce a high-quality product quickly and at scale.

The problem

The UK, like most of the world, faces an urgent need for increased COVID-19 testing capacity. As more people become infected and recover, large-scale screening for past exposure (antibody tests) will be necessary, but the shortage is especially acute for tests of active infection (“swab” tests). The government recognises this need and has set a target of 100,000 tests per day – a combined figure for antibody and swab tests – by the end of April. While this increase is welcome, safely bringing the country out of lockdown could require far more widespread swab testing, potentially as many as 10 million tests a day. Post-lockdown policy is still being developed so this may become the strategy within a few weeks. The UK should prepare now by deciding how market forces can be leveraged to rapidly scale up testing.

A long-term solution to the crisis will involve effective pharmaceutical treatments or vaccines (ideally both). Promising candidates have been identified, but most will take at least several months to complete Phase 3 studies – perhaps less if “human challenge trials” are permitted. Since many products will not prove viable, companies have little incentive to invest in production facilities before the product achieves regulatory approval. Thus, scaling up production is likely to add a few more months to the overall timeline, costing thousands of lives and billions of pounds of lost GDP in the UK alone – far more than the cost of preparing to manufacture products that do not end up reaching the market.

Potential solutions

There are several ways to address this problem.

1. Prizes

The government could offer financial rewards for solutions to supply shortages. For example, companies could compete to offer the best idea for rapidly scaling up vaccines, and a contract to produce them could be part of the prize. By only paying out for the best solution (and perhaps not any, if certain criteria are unmet), this can be a fairly cheap option that incentivises innovation. However, there is necessarily a substantial delay between announcing and awarding the prize; and because the “losers” get nothing, there may not be adequate financial incentive to participate.

2. Public-private partnerships

PPPs can be an effective means of achieving social objectives, such as building infrastructure, by sharing the risk between government and private companies. However, they generally take a long time to negotiate and implement. Unless this process can be greatly accelerated, they are unlikely to be sufficient to ensure the most urgent needs are met in the current situation.

3. Direct purchase orders

The government could pre-order tests, pharmaceuticals and vaccines directly, well before efficacy or safety is established. This would legally guarantee that producers have a market and that the company will supply the product, thereby reducing risk to both parties.

However, this requires the government to first identify suppliers and producers, negotiate prices, and make orders. The process for governmental purchasing is complex, and purchasing something from a new vendor, or purchasing products not shown to be safe, will potentially be a violation of the UK’s public procurement policy. It is also likely to be wasteful as many final products will be unused, and it gives little incentive for producers to improve quality, speed, or cost-effectiveness through innovation.

4. Option-based guarantees

A new approach is for governments to enter into agreements with companies using “put” options. A “put” (as in “put up for sale”) gives the holder the right, but not the obligation, to sell an asset at a specified price, by (or on) a specified date, to the provider of the put. So the government could supply a put option giving companies the right to sell certain items (e.g. vaccine factories, drugs, or diagnostic tests) to the government for a certain percentage of the cost of making them. Because there is no requirement to exercise the put, companies could sell viable products as normal, and would only use the option if their product turns out to be non-viable.

For example, suppose a manufacturer wishes to produce 100 million of a new type of test, but is delaying production because the product is currently being evaluated. They could approach the government, which could agree to provide a put option for, say, 90% of costs, capped at the company's initial project cost estimate. If the test is found viable, the company would not exercise the option, the government would pay nothing, and the company would be able to sell the tests normally to the NHS and others. If found non-viable, however, the company would have an incentive to stop production and exercise their option. At that point, a financial audit of costs would take place, and the government would accept delivery of any items purchased, built, and/or produced in exchange for 90% of costs. A further independent evaluation might be useful to resolve disputes about reasonableness of costs.

This approach has a number of advantages.

  1. Commercial companies can continue to use traditional, non-governmental methods for financing and constructing a product without any government supervision.
  2. Companies will be willing to take a larger risk in manufacturing not-yet-proven technologies, because the costs (to the company) of failure are reduced. This incentivises starting production earlier.
  3. Both haste and high quality are still incentivised through normal market mechanisms: being the first and/or the best product on the market will increase sales and therefore profit.
  4. Compared to some alternatives, it is relatively cheap.
  5. It could potentially be implemented quickly – a very important consideration in the current circumstances, especially for diagnostics.

Recommendation

Overall, guarantees based on put options seem to show the most promise. When the viability of the final product is uncertain, they provide a relatively quick, low-cost way of incentivising the rapid production of new technologies. However, options 1–3 are also worth exploring further, and the optimal approach (or combination of approaches) may vary among products, time periods, and companies. For example:

Over the coming weeks and months, making the right choices could save thousands of lives in the UK and millions around the world, while enabling economies and communities to reopen.

Annex 1: Potential variations on standard put options

Declining payout

The payout for the put options could be declining over time, so that the payment is, say, 95% at the outset, and declines by 1% per month. This will incentivise companies to exit as soon as possible if they think the project will fail.

Priced contracts

The government could decide to charge for the contracts, to dissuade unqualified or undercapitalised companies from taking huge immediate risks with small probabilities of paying off.

Early-ending bonus

Alternatively or additionally, there could be payments for ending the contract early. In this case, the government might refund a portion of the initial payment or pay some fixed amount if the company decides to end the option without exercising it. This would create another incentive for companies to move quickly, and reduce uncertainty and risk on the government’s side.

Annex 2: Key questions about option-based guarantees

Q1: Isn't this a giveaway to corporations?

A: Yes, but in a sense it is a minimal giveaway. It does not subsidise companies to undertake projects that they expect cannot succeed, but does allow them to move forward schedules for production. Under the circumstances, it seems worthwhile.

Q2: Isn't this wasteful?

A: Yes, it is nearly certain that some items produced will not be viable, so the government will pay for unused products. However, the companies have an incentive not to spend more than needed, since they recover only part of the costs; and most importantly, the successful products will be available far faster. (The guaranteed percentage of costs can be adjusted to reach the desired trade-off between avoiding waste and hastening development of the needed product.)

The government may also be able to reduce the costs of the program by reselling some items: for example, a plant designed to manufacture an ineffective vaccine could eventually be adapted to produce an approved vaccine. There have been intermittent shortages of other vaccines, pharmaceuticals and tests, so excess capacity may not be entirely wasted.

Q3: Won't there be fraud or unnecessarily high costs?

A: There is a risk that some companies will try to take advantage of the programme. The put options, however, will pay less than the cost, so there is a reduced risk that companies will attempt to participate if they do not think the product has a reasonable chance of success. (Again, the right balance can be struck by adjusting the guaranteed percentage.)

For subcontracting, the structure of the payout means that the company, not the government, takes on the risk that costs will be considered excessive, so they have incentives to ensure they are not overpaying.

Q4: Won’t safety and quality suffer?

A: With less “skin in the game”, there may be a reduced incentive for companies to be successful. But if a product is viable, they will not want to exercise the put option, and they will have the same incentive to ensure quality and safety as they would otherwise.

Q5: How do you ensure the final product is affordable?

A: Put options do not ensure the final test, drug or vaccine is available at a reasonable price, but nor do they preclude price controls. This is an important but entirely separate issue that applies equally to products developed through other means. It is worth noting that, while the product must be cheap enough to roll out at massive scale, the price must also be high enough to reimburse the costs of constructing the production facilities.

2 comments

Comments sorted by top scores.

comment by Davidmanheim · 2021-12-26T16:34:00.851Z · LW(p) · GW(p)

This was a promising and practical policy idea, of a type that I think is generally under-provided by the rationalist community. Specifically. it attempts to actually consider how to solve a problem, instead of just diagnosing or analyzing it. Unfortunately, it took far too long to get attention paid, and the window for its usefulness has passed.

comment by Davidmanheim · 2020-12-14T08:52:13.947Z · LW(p) · GW(p)

See the original twitter thread for a slightly earlier version of this idea, a video presentation to Foresight on the topic, as well as their extensive report that discussed that presentation, and the new paper for a more polished version.