Understanding how foreign shocks affect economies has become a policy priority. Using domestic firm-to-firm transaction data matched with customs records, this column shows that Italy’s exposure to external risks, such as US tariffs and a sudden stop of Chinese supplies, is far larger than standard trade statistics suggest. Indirect exposure transmitted through domestic supply chains exceeds direct export and import exposure in most Italian regions. A small number of large traders and wholesalers act as key transmission hubs. Widely used input–output tables substantially understate these risks. Micro-level supply-chain data are essential to detect local vulnerabilities and design mitigation policies.

Following escalating geopolitical tensions and resurgent protectionist tendencies, many governments have introduced new trade barriers in recent years (Aiyar et al. 2023, Attinasi and Mancini 2025). Tariffs, export controls, and related restrictions have become increasingly widespread, marking a shift from efficiency-driven globalisation toward a more fragmented and politically shaped world economy (Baldwin and Ruta 2025). Such interventions can have substantial local consequences, especially in regions highly exposed to foreign trade — whether directly or indirectly through domestic supply-chain linkages (Caliendo et al. 2019). Understanding how local economies are embedded into global value chains is therefore essential for designing policies that enhance resilience in an increasingly volatile global environment.

In a new work (Borin et al. 2025), we examine how international trade shocks may transmit through domestic supply chains, shaping local economic vulnerabilities. To do so, we match domestic firm-to-firm transaction records with customs-based foreign trade data and balance-sheet information for the universe of Italian firms in 2022. ¹ Using these data, we measure the exposure of Italian local labour markets (LLMs) to two major external risks: (i) direct exposure through transactions with foreign firms and (ii) indirect exposure propagated via domestic supplier-buyer networks. We focus on exports to the US, which are subject to the recent increase in US import tariffs (Conteduca et al. 2025a), and on imports from China, in light of rising tensions between Western and Eastern geopolitical blocs (Conteduca et al. 2025b). Methodologically, we build on Dhyne et al. (2021) by quantifying firms’ direct and indirect exposure to foreign trade — on both the export and import sides — and extend their framework to the sectoral and geographical dimensions of trade‐shock exposure. Such a micro-level approach to measuring indirect exposure to foreign shocks is also close in spirit to macro-level studies based on Inter-Country Input–Output tables (Baldwin et al. 2023, Pauwels and Imbs 2022).

Domestic supply chains amplify and spread exposure to foreign shocks

Our analysis highlights two key aspects of the exposure to foreign shocks.

First, domestic supply chains amplify the exposure to foreign shocks. Indirect exposure exceeds direct exposure for nearly 80% of local labour markets in the case of exports to the US and about 90% for imports from China. As shown in Figure 1, 3.4% of total revenues are linked with exports to the US at the aggregate level. More than 50% of such exposure, however, is due to Italian firms’ indirect connection to the US market through other national firms. Similarly, we find that 3.6% of firms’ costs relate to imports from China, with about 60% due to the indirect connection to the Chinese market. Therefore, international trade data — which capture only direct links — substantially underestimate the potential disruptions coming from foreign shocks.

Second, domestic supply chains spread the exposure to foreign shocks across regions. Although direct foreign transactions are concentrated in a relatively small set of local labour markets, domestic firm-to-firm linkages can transmit external shocks to other regions. Consequently, many regions that appear insulated from foreign shocks can nonetheless be affected indirectly through domestic supply chains. As shown in Figure 1, the distribution of exposure across LLMs becomes wider once indirect effects are accounted for, especially in the case of imports from China.

Figure 1 Distribution of revenue and cost exposures across Italian local labour markets

Note: Kernel densities of the average exposures to US exports (left panel) and imports from China (right panel) across Italian local labour markets (LLMs) in 2022. The green solid line represents the direct exposure, the purple solid line the indirect exposure, and the pink solid line the total exposure. Vertical lines correspond to nationwide averages. Averages are obtained by aggregating firm-level exposures using total revenues or total costs as weights.

Domestic networks reshape the geography of foreign exposure

Italian firms are most likely to trade with partners in distant local labour markets, followed by those within their own local labour market, and least with firms in neighbouring areas. This organisation of domestic production networks implies that geographically localised shocks can travel substantial distances, shaping exposure patterns across regions well beyond their point of origin. As shown in Figure 2, for exports to the US, about 70% of indirect exposure stems from non-neighbouring local labour markets, with the share particularly high among local labour markets with limited direct exposure. By contrast, only 19% of indirect exposure arises within the same local labour market — consistent with the structure of integrated local production districts — while neighbouring local labour markets contribute just 11%.

Figure 2 Decomposition of exposure by proximity of partner local labour markets across total exposure quintiles

Note: Decomposition of exposure to US exports (left panel) and Chinese imports (right panel) across the quintiles of the distribution of local labour markets’ total exposure by type of partner local labour market in 2022. Same LLM denotes the contribution to indirect exposure from firms located in the same LLM. Neighbouring LLM captures the contribution from firms located in contiguous LLMs, whereas Not neighbouring LLM refers to the contribution from firms in non-contiguous LLMs.

The geography of local vulnerabilities differs markedly when comparing exports to the US with imports from China. As shown in Figure 3, on the export side, many local labour markets in the Centre-North exhibit exposure levels above the median in both direct and indirect terms, whereas in the Centre-South, local labour markets with below-median exposure dominate. On the import side, the pattern is quite different. Direct exposure remains concentrated in the Centre-North, while high indirect exposure is more evenly distributed across the country, with numerous districts in the Centre-South exceeding the median.

Figure 3 Direct versus indirect exposure by local labour market

Note: Distribution of the combination of direct and indirect export exposures to the US (left panel) and import exposures to China (right panel) by local labour market in 2022. We classify exposure in each local labour market as high or low based on whether its exposure value is above or below the national median.

Top traders and wholesalers: Key channels for foreign exposure

Which firms act as the main conduits of foreign shocks? We find that top trading firms — primarily those exporting to the US — and wholesalers — mainly handling imports from China — are major sources of indirect exposure.

Top trading firms are crucial to generate aggregate exposure. Excluding the top 100 exporters to the US reduces national exposure by more than 40%, highlighting a high degree of concentration (Figure 4, Panel a). Instead, excluding the top 100 importers from China reduces national exposure by about 25%. For the US, even excluding the top 10 firms has a substantial impact, as total exposure is lowered by almost 20%.

Trading, logistics, and distribution sectors often account for a significant share of direct export and import activities, serving as intermediaries that connect domestic producers and consumers with foreign markets (Bernard et al. 2007). We find that wholesale firms serve as critical nodes within the Italian production network. Excluding wholesale firms directly trading with the US and China would reduce the exposure to exports to the US by roughly 10% and the exposure to imports from China by more than 30% (Figure 4, Panel b). 

Figure 4 Direct versus indirect exposure by local labour market

Note: In panels (a) and (b), we reallocate exports to the US and imports from China of the top-k exporters to the US and importers from China to other countries and compute the resulting drop in exposure. In panel (b), we remove the direct trade connections of wholesale firms with the US and China and compute the resulting drop in exposure. Exposures are aggregated at the national level using revenues and costs in 2022.

Measures based on Inter-Country Input-Output (ICIO) tables understate indirect foreign exposure

Finally, we find that indirect exposure measured from firm-to-firm transaction data is substantially higher than that derived from Inter-Country Input-Output (ICIO) tables. Even compared with the regionally disaggregated FIGARO-REG database (Lopez Alvarez et al. 2025), our firm-level estimates are on average nearly three times larger at the regional level. This discrepancy arises for several reasons. ICIO tables assume all firms in a sector and region use the same input mix, ignoring the central role of large, trade-intensive firms with extensive buyer-supplier networks (Bems and Kikkawa 2021, Bernard et al. 2022). Additionally, capital-goods purchases recorded at the firm level are attributed to final domestic demand in ICIO frameworks, further underestimating inter-firm linkages. Together, these factors explain why microdata reveal substantially higher indirect exposures.

Conclusions and way forward

Our findings show that firm-to-firm transaction data reveal exposure channels that remain invisible in traditional trade statistics and are understated even in state-of-the-art inter-country input–output frameworks. With trade tensions and geopolitical disruptions increasingly common, aggregate statistics alone are no longer sufficient to detect the vulnerabilities hidden within domestic production networks. Incorporating microdata that capture the structure of supply chains into policy analysis substantially improves the ability to detect which firms — and which regions — are most vulnerable to foreign shocks. In a parallel study (Benecchi et al. forthcoming), we demonstrate that these exposure measures are not just descriptive: they meaningfully predict firms’ vulnerability to tariff-related tensions. Firms with higher exposure, both direct and indirect, were significantly more likely to report sales losses from the 2025 tariff shock.

Investing in the collection and sharing of these data is therefore a strategic policy choice. In this sense, we echo Pichler et al.’s (2023) call for an ‘alliance to map global supply networks’: a coordinated effort that would markedly strengthen the capacity of governments and institutions to monitor, predict, and mitigate the effects of global shocks.

Source : VOXeu