More ammo: Improving resilience against extreme surges in demand

by Ajay Shah.

The Javelin anti-tank guided missile is important for the defence of Ukraine. Under normal times, the production capacity seems to be about 3600 a year. The Ukrainians seem to be using 500 per day, or roughly one missile per kilometre of battlefront per day. The peak load is about 50 times bigger than normal times.

Vershinin, 2022 estimates the Russian army is using 7,176 artillery rounds a day, and argues that these numbers are challenging for the modern Western military manufacturing capacity. He estimates that present US annual artillery production would last for about two weeks of combat in Ukraine. In more recent times there are estimates about Russian use of as much as 60,000 rounds/day.

It may appear that with precision guided weapons, a smaller number of weapons will be required to get the job done. However, precise information about targets is lacking, and the military is reduced to shooting at numerous low probability targets. There are more pathways to target acquisition owing to drones, low earth satellites, night vision, etc., and therefore there are more opportunities to use ammo per unit time. The Ukrainian Armed Forces (UAF) innovated with their new `GIS Art for Artillery' system, where rumoured gains on the delays in the kill loop run from 20 minutes to 30 seconds. As a consequence, modern wars are facing production constraints. As an example, in the small air war in Libya in 2011, the UK and France quickly ran out of precision guided munitions (PGMs).

Such problems with the peak-to-base ratio are not unique to ammo. Consider medical oxygen. The peak load in the delta wave was much bigger than normal times. Alongside this, the bulk of the oxygen production is in the economically advanced peninsula, the biggest demand was in the Hindi heartland, and transporting oxygen is difficult as the refrigerated trucks cannot go at over 25 kph.

Or consider surgical masks and personal protective equipment (PPE). The peak demand during the pandemic perhaps went up by 50 to 100 times when compared with normal times.

Or consider medical education. With students returning from Ukraine, there was a small surge in demand for medical education in India. In a healthy economy, there should be a supply response. In a well functioning society, the resource allocation is not fixed.

Or consider electricity. Electricity demand peaks in the evening, windmills are unreliable, the sun shines in the day and can be obscured by clouds. There is substantial intra-day variation of demand (that is quite predictable), but supply is unpredictable and has a different natural intra-day variation. The puzzle of the energy system lies in dealing with the peak-to-trough ratio.

How should we think about such problems? How does the price system respond to these challenges? Is there market failure? What, if anything, is the role for the state in improving things?

Surge capacity as an option

The right but not the obligation to buy is an option. When the buyer has the right, but not the obligation, to buy 2$\times$ more or 20$\times$ more from the seller, at a preset price, this is an option.

As we know from the field of options, options are always valuable (i.e. they come with a non-zero cost). Being there with excess capacity is not free, for the seller. And, the value of the option goes up when there is more volatility. While financial options loom large in the imagination, the world is full of real options.

Surge capacity in the price system

Prices move, from moment to moment, till supply equals demand. When faced with a shortage, prices go up so as to ration out many prospective buyers. And, equally, those high prices tickle firms into producing more. Vast amounts of patience and intelligence are put in, by buyers and sellers, in order to reduce demand (e.g. by finding substitutes) and increase supply (e.g. by producing in innovative ways). Every surge in prices contains the seeds of its own demise, as buyers establish alternatives, and through the supply surge that follows. Covid vaccines were always going to be a short hot market, and production is now being shut down.

When demand surges or supply drops, the price system sends out signals for firms to produce more through high prices. This tends to be accompanied by a lot of hand-wringing about shortages and high prices. If you think "something should be done", to increase output, you should be happy at what is unleashed by the price system, as there is no force more powerful than high prices, in encouraging buyers to buy less and sellers to produce more.

The market economy is not a bureaucracy; it thinks in all sorts of creative ways. If the price of oxygen is high enough, steel factories will stop making steel and sell oxygen into the public market. If prices go sufficiently high, oxygen cylinders from the Indian peninsula, and from abroad, would be airlifted into the Hindi heartland. The sources of increased supply will always surprise us.

But with the best of effort, mobilising enhanced output is hard and takes time. There is a cost to reallocating the resources of the economy, in order to shift from making widget $x$ to widget $y$. The price system finds this reallocation at the lowest cost to society, at the lowest disruption to society, without harming the incentives for sound behaviour and long-term growth.

Many times, a disruption on the output side is also a disruption of the inputs of the firm. When vast increases in output are required, the inputs (whether physical raw materials or the precise human capital) also become costlier. Both supply and demand curves change in many a surge. Such a combination of factors exacerbates the price rise.

Resilience

The price system will sort things out, in the sense of finding the price at which demand equals supply. It is interesting to go one step further and ask: How big was the supply response, of masks produced per day at its peak divided by masks produced per day before the pandemic? A more resilient economy is one where the price system induces a bigger output surge in a shorter time while requiring a smaller rise in the price.

Alternatively, we can focus on quantities and wonder, under what conditions can very large surges be achieved? We can identify a few sources of resilience.

Complexity in the production process

In a country where many things are produced, and in a country with deep pools of skilled people, there will be more headroom for adaptation. If there is a civilian aircraft industry, it can more easily retool to make military craft. If the world's biggest vaccine manufacturer is in our backyard, it can license a good vaccine from abroad and mass produce it. Tractor factories can become tank factories. For a contrast, a country like Saudi Arabia or Russia has simple structures of production, and the price system has inferior raw material to work with.

A successful software tool
is one that was used to do something
undreamed of by its author.
-- S. C. Johnson

The most important ingredient is the human capital: the managers, the chemists who know multiple routes to get to a given molecule, the creative people who can hack a machine or a software system to do things that were `undreamed of by its author'. Resilience comes from deep pools of these individuals, who are sparked into self-interested action by the price system. It is equally about the raw STEM knowledge, and about the creative thinking of the business folk who see profit opportunities, who imagine new kinds of deals, who innovate. These pools of capacity lie in the private sector. Even when a government-controlled system has the creative people, it does not have the incentives for them to think, take risks, innovate, and solve problems.

Of particular importance are adjacent products and dual-use technologies. A factory that makes vaccines can be the right starting point to rapidly get a factory to make Covid vaccines. Cylindrical engineering products made using special alloys, for civilian applications, can be rapidly retooled to make ammo. The lowest costs for augmenting supply come from the presence of these neighbours to the desired product.

In normal times, the optimal structure of production tends to become monolithic. The market tends to collapse into a small set of firms and techniques of production. Monolithic methods of production are inherently risky. When crises come along, we see the value of more diversified and more eclectic methods of production. Price surges, in a crisis, create profit opportunities for obscure strategies for production, and obscure producers. These occasional bouts of profiteering serve to keep these obscure firms, these option sellers, alive.

Private sector confidence

The private sector will stand ready with option-like capabilities, it will be alert, it will move mountains to produce when there is a price surge, all in search of one outcome: high profit rates in those brief extreme moments. A society that views supernormal profits as unjust, and tries to expropriate these private firms, is a society where private firms will layer risk premia on top of their ordinary market-based responses. In other words, we would require an even bigger price surge to elicit the supply side response when the probability of expropriation of the firms goes up.

Globalisation

High domestic prices incite imports; the productive capacity of the whole world is brought to bear upon the shortage within one country. Covid vaccine manufacturing in India was about an Indian facility that licensed a British vaccine design, and used numerous imported materials. A deep engagement with globalisation also increases resilience by fostering higher human capital of the elite. An inward oriented economy, with barriers to cross-border activities in the laws and in the minds, is likely to be less resilient.

Storage

If more ammo, oxygen or PPE are held in storage, this creates greater resilience. There is no free lunch; this storage has costs in terms of the opportunity cost of capital, the cost of storage and depreciation. Someone has to pay for this.

Production capacity that has an upside

Consider a factory that makes ammo. If the private person has a contract where there are the assembly lines and staff running at 1 shift, but are ready to jump up to 3 shifts, then there is headroom for a 3$\times$ increase in output. Sometimes assembly lines can be designed in a way where additional workers can be added and the line then runs faster. This can potentially create space for another 2$\times$ increase in output.

In the case of oxygen, firms in the field of industrial gases can have additional equipment on standby, through which medical oxygen output can go up on demand.

As with storage, there is no free lunch. The private firm would have to have contracts with skilled workers in order to be able to surge the production on demand, and design a production system with this kind of headroom. As with the `disaster recovery systems' in the world of software, the principal should randomly trigger these provisions every once in a while, and verify that each agent is indeed able to surge output as promised under contract.

Capabilities in government contracting

When there was a sudden requirement to fly students from Warsaw to India, the best pathway lies in the government rapidly running an auction, where global airlines compete to deliver the lowest price. Surge capacity for the state lies in the combination of (a) A capable and innovative private sector and (b) A state that is able to enter into contracts with private persons.

Is there a role for public policy here?

Fighting wars is a service that is produced by the government. The strategic planners in the field come up with a requirements document such as `We need to be able to sustain a war for 3 months where we are using 100 tubes a day'. Establishing this level of surge capacity is required as part of production of the public good of defence.

In the case of health, what is required is a careful counting of deaths owing to Covid-19, and assessing the number of deaths which are attributable to the shortage of medical oxygen. A careful analysis is then required, where the statistical value of a life is compared against the costs to society of higher surge capacity for oxygen. If a certain enhanced surge capacity for oxygen is able to save lives, while spending less than the statistical value of a life, there is market failure, and then there is a case for public policy to think about state action.

The fact that there is a surge in oxygen demand does not necessarily imply that there is market failure. We can envision private hospitals propositioning health insurance companies and to individuals, saying that they have established the following kinds of surge capacity for oxygen. This is not unlike the work that private hospitals do, in order to assure themselves of electricity in the event of a disaster. We should skeptically evaluate whether we want a government to do something.

Consider the field of masks and other personal protective equipment (PPE) at the early stages of the pandemic. When demand went up by 50 to 100 times, prices skyrocketed. Some policy makers were red in the face and barged into the economy, with export bans, with efforts to supplant the managers of private firms and organise production. But the right response was to do precisely nothing. High prices created near-magical responses by the private sector; there was a surge of import and production, and competition drove down prices.

State intervention that harms surge capacity

When the price system gets going, solving the mismatch between supply and demand through high prices, we often get many calls for state intervention into the working of the economy with tools like price limits and ordering private firms to operate in certain ways. It is ironic that the very feature that incites more production and reduces the demand -- high prices -- is what irritates some people.

Firms will earn supernormal profits in a surge. These supernormal profits are the fair return for (a) The hard work to modify production capacity in a short time; (b) The alertness and risk taking when faced with an incipient surge; and (c) The long years of holding option-like capabilities which are not earning high returns in normal times. When a society begrudges these supernormal profits, and uses state power to expropriate firms, the response of firms is to be less alert, take less risk, do less hard work in modifying production capacity and hold less real options, all of which worsens the problem faced by society in responding to the surge.

To commandeer resources, to order private firms, without proper compensation, is expropriation. During the second wave, many private firms were forced to stop their production in order to transfer oxygen to medical applications. If they were not compensated for their lost production, this constitutes expropriation.

Price controls hamper the very process of healing. High prices kick off the modification of the resource allocation in order to produce more and consume less. When policy makers use state coercion to force transactions to take place at artificially low prices, this reduces both responses. The one thing worse than a price that moves rapidly by a lot is one that does not.

There are always eclectic and opportunistic firms that jump into the fray and reap huge profits when a certain situation presents itself. These firms might even earn nothing at normal times, and just provide options to society. When the state interferes with the 'profiteering', their viability is adversely impacted.

State intervention that gets surge capacity at an excessive cost

One path to having the requisite amount of peak ammunition is to build a large number of public sector factories, which are idle in normal times, where the full cost of a factory is paid and the workers do nothing. While this does get the job done, it is an inefficient path; it does not harness the cleverness of private firms to get the same surge capacity at a lower price.

When there is a shortage in the country, it is tempting to ban exports. This appears to augment supply in the country, and bring down prices, in the short term. But it harms the trust of all firms to produce in India and thus harms India's long-term growth.

The Indian state attacked firms who were importing oxygen concentrators at the time of their peak demand [example]. This amplified the required rate of return for doing this important work.

The most damaging state interventions are those that directly control the resource allocation (e.g. forcing factories to close down so as to grab their oxygen), and in violating the rule of law with outright threats to coerce private persons. When the state becomes such a bull in the china shop, it tends to disrupt the complexity and sophistication of the resource allocation of the market economy. This encourages private people to produce less in India.

The discussion here, of unwise state intervention, is related to the problem of supply chain resilience when faced with Chinese exports of APIs to the Indian drugs industry. There also, it is possible to do clumsy things. Bambawale et. al. 2021 show how to do this better, how to go with the grain of the price system.

Going with the grain of the price system for surge capacity in ammunition

In the field of defence, strategic thinking should ideally generate a requirements document such as `We need to be able to sustain a war for 3 months where we are using 100 tubes a day'. Alongside this, there may be a peacetime requirement of 5 tubes a day, i.e. a peak-to-base ratio of $20\times$. This problem would get handed off to defence economics.

The best way for defence economics to solve this problem is to undertake the following kind of contract:

1. To ask for multiple private vendors who add up to a peak capacity of 100 tubes/day while actually running every day in peacetime at one-twentieth this rate;
2. The private firms would find the cost-minimising paths for obtaining this flexibility in production, and they would do this better than a PSU or a government department;
3. Each private firm would be subjected to random fire drills, where they are asked to suddenly up their production by $20\times$ for a period of $n$ days with $n < 90$.

In this procedure, we have fixed the surge capacity and are procuring on the price. Alternatively, the procurement can fix the price of the tube, and ask for bids which promise the highest surge capacity.

Through this, the energy and intelligence of the private sector would be brought to bear on the problem of obtaining surge capacity for the public goods of defence. It is better to have multiple private vendors, rather than one, so as to avoid single points of failure/attack, and to set off the spiral of quality where private firms compete with each other to deliver bigger surge capacity at a lower price.

This requires complexity in government contracting. Government contracting is a critical homeostatic capability that is required by all states, which works poorly in India. This is an important field for research.

Once contracts are in place, state actors must work within the rule of law: they must not not coerce private persons to behave in ways which were not contracted. Once the Indian state has behaved correctly for a few generations, the private sector will become more comfortable, and will require reduced safety factors in their pricing.

I thank Akshay Jaitly, Amrita Agarwal and Pranay Kotasthane for useful conversations.