Effects of Tariff Shocks across Time, Space, and Sectors : The Role of Production Networks

Luo, Shuya and Cardi, Olivier and Motta, Giorgio (2026) Effects of Tariff Shocks across Time, Space, and Sectors : The Role of Production Networks. PhD thesis, Lancaster University.

Text (2026ShuyaLuoPhD)
2026ShuyaLuoPhD.pdf - Published Version
Available under License Creative Commons Attribution-NonCommercial-NoDerivs.
Download (3MB)

Abstract

This study examines how protectionist trade policy shocks affect sectoral value-added and employment by focusing on shock transmission through production networks and dependence on imported intermediates. The research proceeds from a country-specific setting to progressively broader evidence, examining China and later exploring a sectoral panel design from a cross-country network perspective. Therefore, mechanisms observed in a special case are identified more systematically and then generalized. All empirical analyses implement a two-step identification strategy that (i) extracts exogenous tariff innovations from sectoral VARs and (ii) traces dynamic adjustments via Jordà-style local projections. Annual tariff measures from WTO TAO are matched to sectoral outcomes from the GGDC/UNU-WIDER Economic Transformation Database (ETD) for 1990-2018; the cross-country component additionally links WIOD/GGDC input–output tables to measure upstream exposure. The identification strategy in this analysis is applied to matched WTO–ETD series with sectors classified by tradability by focusing on China. Estimates from the traded activities show continuous decline in value added after tariff shocks. However, employment adjustments are transient and cushioned in the near term. Activities which are non-traded display material negative spillovers with broad contractions in both value added and employment as higher upstream input costs pass through. These dynamics establish the baseline patterns that motivate disaggregation and robustness in the subsequent panel design. Building on that baseline, a sector-level panel framework replaces aggregate time-series averages with disaggregated dynamics for 41 developing and emerging economies (1990-2018), combining panel VARs with local projections and incorporating sector and year fixed effects. This design quantifies heterogeneity across industries and the strength of spillovers into non-traded activities, possibly implied by input–output linkages. The estimates indicate that tariff shocks depress value added in traded activities while near-term employment buffering is concentrated in politically sensitive segments. A compact comparative extension places China’s responses alongside those of other economies within the same data infrastructure which shows similar adverse effects elsewhere and a somewhat attenuated sensitivity for China. The cross-country chapter reinforces the spillover mechanism developed in the earlier analysis. It constructs a harmonized sectoral panel across countries by incorporating input–output exposure measures and applying a two-step identification strategy to trace adjustment over a six-year horizon. The findings reveal that the protected traded activities tend to register gains in value added without corresponding durable increases in employment. On the other hand, downstream non-traded activities face immediate and persistent declines in both value added and employment. This implies that the effects of tariffs on sectoral performance operate primarily through production-network linkages rather than within isolated sectoral boundaries. Further, the extent of dependence on imports conditions the outcome. As a more dependent country has global value chains experience sharper and longer-lasting contractions, whereas those with lower dependence display comparatively greater resilience.