Podcast
Important Information:
Supply Chain Resilience
Important Information:
Nicolas Morales discusses his research on supply chains and the factors that can help them resist and recover from economic shocks, like the COVID-19 pandemic. Morales is an economist at the Federal Reserve Bank of Richmond.
Transcript
Tim Sablik: Hello, I'm Tim Sablik, a senior economics writer at the Richmond Fed. My guest today is Nicolas Morales, an economist in the Richmond Fed's Research department.
Nicolas, welcome back to the show.
Nicolas Morales: Hi, Tim. Thanks for the invitation.
Sablik: Today we're going to be discussing supply chains, the topic that we all became much more familiar with during the COVID-19 pandemic. I'm sure many of our listeners may have experienced long waits for appliances, cars or even grocery store items during the last few years. Some products and industries seem to have been disrupted more than others.
You co-authored a paper last year in which you tried to identify what makes supply chains resilient, or more adaptable to sudden economic shocks. Was this something that researchers were looking at before the pandemic?
Morales: Yes, definitely. In the past decade in the literature of international economics, there has been very active progress on our understanding on how supply chains work.
On one hand, it's been very well documented that the fragmentation of production — having different firms specializing in producing specific inputs and then trading them with each other, specialization which forms our supply chain as these firms depend on each other — has been shown to generate efficiency. Firms can just focus on producing one thing very well.
But on the other hand, it's been also documented that supply chains have the potential to amplify economic shocks. Before COVID-19, the type of shocks that were mostly studied were natural disasters. A very famous example is the earthquake that happened in Japan. A few papers in the literature quantified the direct impact of the earthquake — that's firms that were actually hit by the earthquake. They show that the direct impact got amplified through a supply chain. [The papers] looked at the consumers and the suppliers of those firms that were not hit by the earthquake and they were also affected, up to the second and third degree and even more. So, the idea was that supply chains helped contribute to this amplification of a negative shock.
When COVID-19 lockdowns were being proposed throughout the world, this was a new type of shock. People were worried that supply chains could amplify the economic consequences that these lockdowns were already having in the local economy.
A pandemic brought the need to understand this concept of supply chain resilience, which in the paper we define as the ability of a supply chain to resist an economic shock and how fast they can recover from that negative economic shock. In this paper, with my co-authors, we wanted to use this unexpected shock that was the COVID lockdowns to study specifically which characteristics of supply chains make them more resilient to shocks, which is a question where we did not have much empirical evidence on.
Sablik: What sort of data did you need to study this question?
Morales: I think a key piece of data we need in general to study supply chains is a measure of how much firms buy and sell from each other. Once we have that, we can construct the firm-to-firm connections and measure how important these connections are by building the supply chain empirically.
Unfortunately, this type of data is very rare [and] will not be available in most countries. For example, in the U.S. firms do not record anywhere specifically what are the internal transactions within the country. So, it's hard to construct the internal supply chain network for the U.S. For other countries, this data is starting to become more and more available. Some notable countries that have been used in other research have been Belgium, Chile, Costa Rica, Turkey and, in this case, we're using India.
We need a system where firms need to pay some sort of tax when they sell goods to another firm. Then, the transaction from each firm to another firm gets recorded. If you can get access to his tax data, you can use it to construct the supply chains and study important questions on supply chain dynamics.
Sablik: Right.
As you mentioned, you and your co-authors use data from India gathered from the start of 2018 to the end of 2020. Why did you choose India?
Morales: We chose India mainly for two big reasons. The first one is the one you already mentioned: that is data. India has a special tax system that was imposed in 2018. If firms need to ship goods from one destination to another, they need to pay a value added tax for the amounts that was being shipped. They need to pay this tax if the transaction is above something like $700.
What we did was we obtained data for one Indian state of all these firm-to-firm transactions, as long as at least the buyer or the supplier for that transaction was located in our state. Once we have that and we could construct the full network of buyers and suppliers that were located in that state, we can understand how the network changed over time, particularly in the years when there was big lockdown policies imposed.
A second reason why we picked India is that India is, in some sense, a sort of ideal case study when trying to use COVID lockdowns as an economic shock. India imposed lockdowns of different stringencies across space. We use that in the paper as a differential shock that affects different types of farms.
Sablik: Yeah, maybe you could tell us a bit more about the pandemic experience in India for our listeners who are not familiar with it.
Morales: If we jump back to March 2020, which probably many of us remember very well, COVID was spreading throughout the world. Many countries were imposing different lockdowns with different types of stringencies.
India was a particular case because COVID was actually not a very big deal back in March 2020. There were not a big number of cases. There were not a lot of deaths compared to what the size of India is. Still, there was a fear that, because of the high population density of India, COVID could spread very fast. What the federal government did was they imposed a federal lockdown policy, but different districts had different stringencies of the lockdown based on this initial number of cases that was actually fairly low.
In the paper, we referred to these lockdowns in three levels of stringencies. A red area was an area [which] had the harshest lockdown — public transportation was suspended, retail businesses were mostly closed, transportation of goods was very limited. It was very hard to go to work. In green areas, there were still some restrictions, but they were much milder. For example, public transportation operated at 50 percent capacity and business activity was still operating to some extent. Orange areas were somewhere in the middle between red and green. These lockdowns were imposed at the end of March and they started getting lifted around May. India did not really experience a large wave of COVID infections until the Delta wave that came later in 2021.
What we do in this paper is that we use these different lockdown policies across space to quantify the supply chain disruptions or to quantify supply chain resilience — that is a term we use in the paper.
For example, we compare two very similar firms that are located in the same region. So, they are exposed to the same lockdown themselves. They produce the same type of goods [and] they're in the same industry, so they are very similar. The only difference is that, say, firm one before the pandemic [was] buying from a lot of suppliers from red areas which later we're going to hit with harsh lockdowns, while firm two before the pandemic was buying mostly from suppliers in green areas that later were not going to be hit by harsh lockdowns. What we do is we can compare firm one and two and then the difference on how these firms, how disrupted their production is after the lockdowns. We can quantify this economic shock.
What we find is that suppliers in harsh lockdown areas are more likely to stop purchasing from their old suppliers. They find it harder to find new suppliers — it's not that easy to rebound after the economic shock. Overall, suppliers in harsh lockdown areas decreased their overall production and their sales to their consumers. They experienced a higher supply chain disruption and they're less resilient after the shock.
Sablik: Yeah, thanks very much for that overview.
Digging a little bit more into your results, when it comes to gauging supply chain resilience, what specific characteristics were you looking at?
Morales: We started by looking at two sets of characteristics.
The first characteristic we looked into is how differentiated the products transacted were, the complexity of the products the firms were purchasing. For example, imagine a firm that is buying computers. Computers require a lot of different inputs to be produced themselves. At the same time, there's fewer suppliers in the market available to sell computers, so it's a more rare input to source. We consider that as a complex differentiated input.
On the other hand, think about firms buying a wooden table or a wooden chair. Those type of inputs are less complex — first, because they require fewer inputs to be produced themselves, you probably need much fewer inputs than a computer; and second, presumably there's also more suppliers available to sell that good in the market. So that's the first characteristic — we look at just different measures of complexity of the products.
The second set of characteristics is more related to the suppliers who were transacting before the shock happened. For example, we looked at whether the suppliers that you were buying goods from were big, were well connected, and how dependent firms were on their suppliers — most of your purchases are concentrated in one single supplier.
Sablik: Did you have any hypotheses about which of these characteristics would matter most for supply chain resilience when you were going into this study?
Morales: Well, there's not much empirical evidence on which characteristics matter, so for a lot of them we actually didn't have a strong prior. But before the study began, we thought that firms buying more complex products would have been the ones that were the most in trouble after a shock. If they faced a disruption, they might find it harder to find alternatives and to rebound after the shock. That was sort of what we're thinking when we're getting into the project.
Sablik: And what did you actually find?
Morales: Well, surprisingly, you actually find the exact opposite. We found that firms that were buying more complex inputs seem to resist the lockdown shock of their suppliers much better. They were less likely to separate from their suppliers and they exhibited a much lower input drop than firms that bought less complex inputs.
Part of the explanation is that if you were a firm that depended on these complex inputs [and] you knew these inputs were rare, you were likely more prepared when the lockdowns hit to withstand the supply chain disruption. If you know it's more risky if you happen to lose your suppliers, you'd likely invest more in maintaining the relationship with your suppliers, in finding alternatives [and] in making the relationship stronger.
On the supplier characteristics, we found things that are more in line [with] what we expected. We found that if firms before the lockdowns were buying from larger and better connected suppliers, they faced fewer supply chain disruptions after the shock. Likely, their suppliers were more resilient and more robust to the shock, so that made the impact on the firm much milder than if you were buying from less connected and smaller suppliers.
Sablik: What are some of the general lessons that you think we can take away from this study of Indian firms?
Morales: In general, these results help us identify which firms face the largest disruption when an unexpected shock hits. In this case, the unexpected shock is the COVID lockdowns to our suppliers. The firms that were most at risk of facing supply chain disruptions and reducing production were small firms, firms that bought less complex inputs, and firms that were connected to smaller suppliers and less central suppliers in the network.
Just thinking of policy, if we want to minimize economic disruptions of future economic shocks and improve resilience of the overall network, then focusing on these type of firms will be a first step to do that.
Sablik: Well, Nicolas, thank you so much for joining me today to talk about your research.
Morales: Thanks for having me.
Sablik: Listeners can find links to the papers we discussed today on the show page. And if you enjoyed this episode, please consider leaving us a rating and review on your favorite podcast app.