Supply shocks have a significant impact on the economy
Attinasi et al 21 (Maria Grazia Attinasi; an Economist in the External Developments Division of the European Central Bank; Micro Balatti; a PhD economist working in DG International at the European Central Bank; Michele Mancini; Senior Economist at the Bank of Italy; Luca Metelli; Economist at the European Central Bank; “Supply chain disruptions and the effects on the global economy”; European Central Bank; August 2021; https://www.ecb.europa.eu/pub/economic-bulletin/focus/2022/html/ecb.ebbox202108_01~e8ceebe51f.en.html) [DTD]
Supply chain disruptions are putting a drag on activity and trade at the global level. The most relevant elements are i) difficulties in the logistics and transportation sector, ii) semiconductor shortages, iii) pandemic-related restrictions on economic activity, and iv) labour shortages. Global shipping of merchandise goods has been severely disrupted owing to container misplacement and congestion on the back of not only the rapid recovery in the global economy, the rotation of consumption demand from services to goods, and the associated high import volumes, but also port closures because of localised and asynchronous outbreaks of COVID-19.[ 2 ] As a result, shipping costs, especially from the main Asian ports to the United States and Europe, have skyrocketed since the end of 2020. Semiconductor shortages started to materialise in the second half of 2020 and are especially pronounced in the automotive sector. During the great lockdown, car producers reduced their chip orders, while demand for chips used in other electronic equipment rose significantly (mostly on account of the work from home instruction). Producers were surprised by the sharp increase in new car orders in the second half of 2020, and with little spare capacity left in the semiconductor industry, chip production was unable to keep up with the high demand – possibly also as a result of underinvestment in the years prior to the pandemic.[ 3 ] Labour shortages appear to be less widespread and more concentrated in certain economies, such as the United States and the United Kingdom. In both countries, indicators of labour market tightness are already above their pre-crisis levels, in contrast to the slow recovery after the global financial crisis. Declines in both matching efficiency and labour force participation partly reflect increases in unemployment benefits, early retirements and the need to care for children and other family members during the pandemic, as well as a reluctance to work in contact-intensive sectors.[ 4 ] Finally, the impact of the aforementioned factors in terms of clogging up supply chains might be exacerbated by the “bullwhip-effect”, a standard amplification channel phenomenon whereby firms build up their inventories because they are expecting robust demand amid a shortage of key inputs in the production process, such as raw materials and intermediates. The lengthening of suppliers’ delivery times across advanced economies since the end of 2020 is the most evident manifestation of widespread strains in global production networks. One of the indicators most commonly used as a proxy for such strains is the global Purchasing Managers Index suppliers’ delivery times (hereinafter referred to as the “PMI SDT”), which quantifies developments in the time required for the delivery of inputs to firms. One key advantage of the PMI SDT is that it is able to capture capacity constraints of a different nature (e.g. intermediate goods shortages, transportation delays or labour supply shortages), making it an all-encompassing indicator of strains in global production networks.[ 5 ] This indicator suggests that suppliers’ delivery times have lengthened massively in recent months (Chart A, panel a) and that the lengthening is proving to be more protracted than during the initial COVID-19 shock. The chart also suggests that there is a significant amount of heterogeneity between advanced economies and emerging economies, with economies like the United States, the euro area and the United Kingdom being much more affected than key emerging economies. Finally, while the increase in the PMI SDT is common to most sectors, it is particularly pronounced for certain types of product, such as technology equipment and machinery (Chart A, panel b), suggesting that the shortage of intermediate products is more severe in those sectors. Suppliers’ delivery times reflect strains in production networks and display some procyclicality vis-à-vis output fluctuations. The PMI SDT tends to co-move closely with the global PMI manufacturing output, which is a proxy for the business cycle, suggesting that as output increases, delivery times tend to lengthen. In order to purge movements in the PMI SDT from the normal lengthening associated with cyclical fluctuations, we use a monthly bivariate vector autoregression (VAR) model for the global (excluding euro area) PMI manufacturing output and the global PMI SDT, in which shocks stemming from the recovery in demand and supply chain disruptions are identified using sign restrictions.[ 6 ] More specifically, we assume that disruptions to supply chains lengthen delivery times and reduce output, while the rise in demand induced by the economic recovery increases both delivery times and output. This approach enables us to recover the structural shocks underlying movements in the PMI SDT, and in particular the supply-side shock, which we take as our measure of supply chain shocks. Our empirical analysis suggests that supply chain shocks account for around one-third of the strains in global production networks. The historical decomposition shows that, even though demand factors played a primary role in driving the overall level of the PMI SDT, supply chain disruptions accounted for one-third of the lengthening in delivery times over the last six months, and their contribution has been growing (Chart B). By contrast, the greater contribution of demand factors is not surprising given the procyclicality of delivery times in periods of economic recovery and the unprecedented economic recovery that has followed the initial COVID-19 shock.[ 7 ] Supply chain disruptions have a negative impact on global industrial production and trade, and a positive impact on inflation. Our analysis aims to quantify the impact of the aforementioned supply chain shock on activity, trade and prices, and, in turn, the headwinds it creates for the economic recovery. To achieve this, we estimate a companion VAR, with five endogenous variables (exports, imports and industrial production, together with the inflation rates for the consumer price index and the producer price index). The estimated supply chain shock is plugged into the model as an exogenous variable. In order to quantify the headwinds for activity, trade and prices, we then generate a counterfactual scenario by running a conditional forecasting exercise for the period from November 2020 to September 2021, which assumes that there are no supply chain disruptions (i.e. the supply chain shock is set at zero throughout). For that period, we find that world trade would have been around 2.7% higher cumulatively in the absence of supply chain shocks, while global industrial production would have been around 1.4% higher (Chart C, panel a). The effects are greater on trade than on industrial production because the weakness in the logistics sector disproportionately affected trade. Moreover, the shift towards domestic suppliers and domestic goods might have mitigated the repercussions on industrial production. Our findings also suggest that supply chain disruptions have a significant – and increasing over time – effect on prices, which is much more prominent in the producer price index than in the consumer price index (Chart C, panel b).[ 8 ] This could be attributed to the fact that producers are more directly exposed to supply chain disruptions than consumers. Moreover, rising producer prices are passed on to consumers only partially and/or with a lag. Finally, it is worth noting that the aforementioned aggregate results mask significant heterogeneity across countries given that not all countries are affected by supply bottlenecks to the same degree. For instance, we find that the effects are greater in the United States, where trade and industrial production stand at 4.3% and 2.0% below the disruption-free counterfactual scenario respectively.