The Story of the Missing Carrots

About two months ago, a friend of mine called me saying that she was not able to get carrots and was missing them. We had a quick discussion on where they come from and why something as commonplace as carrots would not be available at the local “Sabzi Mandi” or vegetable market. But as many of us are finding out, the global pandemic is causing us to re-examine our definition of what commonplace is. I began to realise that the infallible supply chain in our globalised world is suddenly showing signs of fragility when I tried to buy a mask about two months ago. The deliveries were delayed because the brand explained they sourced raw materials from nine separate locations. The highly infectious nature of the virus has meant that consignments needed for the production process were held up due to lockdowns thereby causing delays in the final delivery. Prior to the pandemic the ease with which global supply chains functioned, we would never have had cause to question the wisdom of procuring raw materials based on the famous economist David Ricardo’s theory of comparative advantage.

This blog is to establish the importance of fine-tuning supply chains which is a fairly complex exercise in the best of times. With the onset of the digital age, consumers are spoiled for choice and a well defined supply-chain strategy will help enhance customer experience and retention. However with the world reeling under the impact of the pandemic, supply chain optimisation has assumed a significance like never before. Not just at a micro level but even at a macro level where even governments have to get involved in ensuring the smooth functioning of supply chains in certain key areas. I have collated information from a few research papers to be able to provide a rough framework of how this could be done. This is going to be especially important as we have to grapple with getting essential commodities to the most needy people. Some of the questions that both businesses and governments have to contend with are, how will the vaccine when it is finally developed, be sent out to every part of the world? Will we have to source most of our raw-materials locally, given the challenges in transportation and unpredictable lockdowns? Changing consumption patterns because of economic uncertainty means that many are concerned with consuming just the bare essentials. Firms will have to work on their supply chains taking into consideration the new reality of procuring inputs and supplying output to rapidly changing consumption patterns.

When I talk about a supply chain, I mean the channels for procurement of raw materials for final production and subsequent distribution to customers. In reality, this is far more complex because the raw materials may in turn need raw materials and the final product may be raw material for another product and so on. Later in the blog we will look at a relatively simple example of a supply chain and my hope is that you would be able to extrapolate this information to more complex situations. The question that we would hope to answer through analyses such as these is whether we could simplify these supply chains to eliminate some of these linkages and in these turbulent times, these systems would be able to work with minimum support.

A paper which documents the impact of natural disasters on supply chains has studied the impact of the earthquake in Japan (2011) on the supply chains of firms in the affected areas. The findings were that the disaster impacted the growth of both suppliers and downstream customers of the firm based in the impacted area. The impact was felt by not just the direct suppliers and customers but those who were indirectly linked as well though the impact was greater for those firms who were closer in their linkage. The paper also indicated that the impact on the downstream customers tended to be more than that on the upstream suppliers and the consequences of the supply chain disruptions were experienced tangibly at both micro and macro levels.

So how exactly do we fine-tune or in mathematical lingo optimise our supply chains? To be able to achieve the objective of optimising the supply chain, we would have to be able to identify firms that are essential to keep the production process going. This is not really going to go into all details that may have to be employed but the general approach that a problem of this sort entails based on evidence from academic research. This piece is inspired largely by the work of Prof Vasco Carvalho and Dr Matthew Elliot along with several others on a few papers. The latest paper in which they are studying the impact of the Covid-19 pandemic is in the works right now. However their previous published work does provide a framework to analyse and recommend simplification of supply networks to minimise the risk of disruption especially during emergencies. This is by no means a simple problem to solve but it is does start at a firm level and then can be built up. The nature of the inputs have to be considered, i.e. is it easy to substitute or hard to substitute. Are firms locked into agreements to procure from a particular vendor or is it a fairly expensive and cumbersome process to switch providers for raw materials? For the analysis to be usable it does have to consider some of these complexities.

This is a hypothetical example of a relatively simple supply chain which is based on the Leontief Production Function which many of us probably studied in a basic Microeconomics course. It consists of 3 raw materials which are supplied to produce intermediate goods A, B & C which in turn are supplied to manufacture goods E, F & G. The final products H & I are manufactured using E, F, I & D which are then sold. The production process entail different proportions of the inputs. The crux of the simplification process lies in identifying the firms that are essential and then ensure that adequate quantities of the raw and intermediate inputs are available to ensure that the production is smoothly executed. Put quite simply, the problem at hand is how we can eliminate some of these firm to simplify the supply chain. In the real world, this would be far more complex, involving many other inputs and outputs and the raw materials can be sourced from domestic firms as well as those which may be located in other parts of the world which, add to the complexity of the problem. This firm level optimisation problem could then be aggregated across many such production units and analysed to provide policy recommendations at a macro level. The government may not be able to optimise across all sectors of the economy but may be able to focus on some key sectors in terms of the value to the economy and/ or covering the basics. At the moment, a vast amount of research has shown that there has been a shift in consumption patterns due to uncertainty in the labour markets across the globe. A large proportion of consumers are sticking to the essentials with the consumption of many non-essentials being shelved for the time being. Hence I would expect that the government has a focus on the sectors that will have a direct impact on the ability of the country to fight the pandemic and also ensuring that basic needs to the country are being met at the very minimum.

From the methodology point of view this is a computationally hard or a NP Complete problem which would mean that we may have to be satisfied with an approximate solution. This may also entail omitting certain portions of the network but the evidence suggests that the may be a very small fraction of the supply chain. The analysis has to be set up taking into account, nuances of the specific contexts such as ensuring certain key firms are part of the final solution in certain cases, probably trying to work with firms that are closest to the end-customer etc. Academics working on this issue have employed linear programming which have proved to be an efficient and scalable solution to the problem. As part of the analysis, it is important to be able to determine the production capacities for each of the firms involved and to ensure that that excluding certain firms will not compromise meeting the final demand and ensuring customer satisfaction. It is possible to assign manufacturing to the firms on the basis of historical data. In the absence of data on the output manufactured by firms that are part of the supply chain, it is possible to employ a hierarchical clustering algorithm to identify firms that are similar in nature across the different levels in the supply chain. Like with most analyses, it is important to understand that we may have to analyse the problem several time to be able to choose the solution that is best and provide the most actionable results for the problem at hand.

Academic research has shown much progress has been made to allow for this analysis to be carried out for the economy as a whole. The problem is analysed using a standard general equilibrium model and does generate hypotheses that may be tested and validated. There are numerous challenges in carrying out this analysis and the framework as it currently stands could be considered a version that is not strictly an aggregation of the firm level data.

Given the times that we are living it (its surreal - never would have imagined living in these times) this is a highly relevant analysis because all of us have probably struggled to get some essentials since March of this year. This is probably very crucial for firms as they struggle to stay afloat with all the talk of recession and plummeting growth rates. Not being able to satisfy demand because of sourcing challenges and other supply chain disruptions would be a travesty. I would strongly recommend revisiting your supply chain if you are faced with problems of this sort. At a macro level, governments across the world are grappling with some very serious supply chain questions such as, how do we ensure that we have an adequate supply of PPE for the people who are working on the frontline, similarly we have to ensure adequate supplies of ventilators etc. These could potentially be treated as sector level supply chain questions and analysed as such. Also given that we are faced with challenges of climate change and having to reduce our carbon footprint, there may be credible reasons to consider optimisation of supply chains. There are some sectors such as agriculture, food and drink supply that are notoriously difficult to manage in terms of supply chain and falling in the essentials category which, would need greater attention. So the case for ensuring carrots and other vegetables are available at the “Sabzi Mandi” remains stronger than ever.


Supply Chain Bottlenecks in a Pandemic by Vasco Carvalho, Matt Elliot & John Spray (

Carvalho, Vasco M. (2014), “From Micro to Macro Via Production Networks”, Cambridge INET Working Paper Series No: 2014/18, Cambridge Working Paper in Economics: 1467

Carvalho, Vasco M., Makoto Nirei, Yukiko U. Saito, Alireza Tahbaz Salehi (2020 - Updated), “Supply Chain Disruptions: Evidence from Great East Japan Earthquakes”, Cambridge INET Working Paper Series No. 2016/25, Cambridge Working Paper Economics: 1670

McKinsey & Company (October 2020), Consumer sentiment and behaviour to reflect the uncertainty of the COVID-19 crisis (