The role of import protection in industrial policy has long been a subject of contentious debate. From Alexander Hamilton (1791) to Donald Trump, US leaders have argued that American manufacturing should be protected from foreign competition. Today economists tend to be sceptical of such ideas, pointing to the advantages trade liberalisation brings by reducing input costs and exposing domestic firms to the rigors of international competition.
Yet economic theory also formalises conditions under which import protection may support industrial development. For example, Paul Krugman (1984) develops a model where import protection is export promoting at the industry level. This arises because production is subject to economies of scale at the industry level, meaning that average production costs decline as output increases. Consequently, if import competition reduces domestic output, it also increases production costs, making domestic firms less competitive not only at home, but also abroad. Hence, import liberalisation may lead to export destruction and, conversely, import protection can be complementary to export success.
Does the Krugman mechanism operate in the real world? Evidence from US-China trade
In recent work (Breinlich et al. 2022), we examine the empirical relevance of Krugman’s argument by studying the ‘permanent normalisation [of US] trade relations’ (PNTR) with China in 2001. PNTR granted China ‘most favoured nation’ (MFN) status on a permanent basis, removing the threat of higher tariffs on Chinese imports, which had been a significant factor in discouraging Chinese exporters from entering the US market. Indeed, previous research by Handley and Limão (2013 and 2017) has demonstrated that PNTR can account for over one-third of the growth of Chinese exports to the US in the period 2000-2005. Thus, PNTR was, de facto, a substantial liberalisation of US imports from China.
According to the mechanism proposed by Krugman, PNTR should decrease US exports, as US producers lose domestic market share to Chinese import competition, reducing their overall output and increasing their costs. To investigate this possibility, we study how US export growth following liberalisation varied across goods industries with different exposure to PNTR. We measure exposure using the NTR gap, defined as the tariff increase Chinese imports faced if the US revoked China’s MFN status. Pierce and Schott (2016) find that PNTR led to lower employment growth in US industries with higher NTR gaps.
As Figure 1 shows, US export growth declined after PNTR in industries with higher NTR gaps. We confirm this relationship using a more sophisticated regression analysis that controls for other factors potentially influencing US exports to third markets, such as demand growth in destination countries and global technology shocks. These results support the existence of the Krugman mechanism and imply that US production features industry-level economies of scale.
Figure 1 PNTR and US export growth
Notes: Change in US export growth post-PNTR is defined as the annualized change in log total exports between 2000 and 2007 minus the annualised change between 1995 and 2000. Each dot represents one industry. The solid line shows the fitted relationship from a linear regression.
We also find that PNTR had a positive impact on US exports by reducing the costs of intermediate inputs used by US exporters. This export promotion effect offset the export destruction effect due to greater Chinese import competition. Which of the two mechanisms dominates depends on their relative importance for a given industry. Our estimates show that the net effect ranges between negative 24% (for cigarette manufacturing) and positive 38% (ice manufacturing) and is negative for 41% of industries. However, this does not mean that PNTR was export destroying for 41% of industries because our regression analysis cannot account for the general equilibrium effects of PNTR. Most notably, PNTR is likely to have increased income levels in China, generating more demand for US exports.
Accounting for general equilibrium effects
We account for such general equilibrium effects using a state-of-the-art quantitative trade model with scale economies that builds upon Kucheryavyy et al. (2020). The model has 12 economies, including the US and China, and 24 sectors, including 15 goods sectors. We use our empirical findings to calibrate the strength of scale economies in the model by matching the simulated effect of the NTR gap on US exports to our regression estimates.
Using the model, we find that the overall effect of PNTR on US exports was positive because increased demand, when combined with the input cost effect mentioned previously, was sufficient to offset the negative import competition effect. Specifically, we find that total US exports (relative to GDP) increased by 3.6% and exports rose in 13 out of 15 goods sectors due to the PNTR-induced liberalization.
Figure 2 decomposes the impact of PNTR on sectoral exports into the real market potential effect from changes in import competition, the input cost effect and the foreign demand effect. The real market potential effect is negative in almost all sectors, consistent with Krugman’s hypothesis. This effect leads to total sectoral export growth being negatively correlated with the NTR gap, meaning that PNTR shifted US comparative advantage away from sectors with higher NTR gaps. However, in most sectors the positive input cost effect outweighs export destruction either on its own or when taken together with the foreign demand effect. Interestingly, we find that the existence of scale economies magnifies the input cost effect, highlighting an important interaction between scale economies and input-output linkages.
Figure 2 Decomposition of export changes due to PNTR
Notes: Decomposition of simulated sectoral changes in exports due to PNTR into a real market potential (or import competition) effect, an input cost effect and a foreign demand effect. Sectors ordered with NTR gap increasing from left to right. Goods sectors only. Textiles and Leather not shown.
Our calibrated model also shows that lowering import barriers in a single sector will generally reduce that sector’s exports. This is because lower import barriers reduce exports through the import competition channel. At the same time, the input cost effect is typically weak because the liberalised sector uses inputs from other sectors that have not been liberalised. And any export-promoting increases in foreign demand will be shared equally across liberalising and non-liberalising sectors. Consequently, the combined input cost and foreign demand effects will generally not be sufficient to offset the negative import competition effect in the liberalising sector. This of course also implies that increasing trade barriers in a single sector can potentially increase that sector’s exports, as Krugman hypothesized. However, it is important to note that this argument does not apply to overall exports nor to an across-the-board change in import barriers, as our study of PNTR illustrates.
Welfare effects for the US and China
We also use our model to study the consequences of PNTR for aggregate US welfare (measured as changes in real income). Our results show that the presence of scale economies lowered US welfare gains from PNTR because liberalisation shifted resources out of goods sectors whose outputs are used as inputs by other sectors. This effect partially offsets the gains from trade operating through cheaper imports and higher foreign demand. Nevertheless, overall welfare gains are positive if rather small (a 0.08% increase in real income). China gained substantially more from PNTR than the US, with real income increasing by 0.47%, reflecting the fact that China’s economy was small compared to the US at the time of PNTR.
We find evidence for Paul Krugman’s idea that import protection might act as export promotion and, conversely, that import liberalisation might reduce exports. This is consistent with historical evidence from the Napoleonic Wars by Juhász (2015, 2018). But we also show that the mechanism highlighted by Krugman is only one part of the full story. While scale economies mean that import liberalization indeed reduces exports through a decline in domestic output, liberalization also raises exports by allowing firms to import cheaper intermediate inputs and by increasing foreign demand. Taking all general equilibrium effects into account, we find that the permanent normalization of US trade relations with China boosted overall US exports, even though export growth was lower in more exposed sectors.
Breinlich, H, E Leromain, D Novy, and T Sampson (2022), “Import liberalization as export destruction? Evidence from the United States”, CEPR Discussion Paper 17031.
Hamilton, A (1791), “Report on the subject of manufactures”, Submitted to Congress on December 5.
Handley, K and N Limão (2013), “Does policy uncertainty reduce economic activity? Insights and evidence from large trade reforms”, VoxEU.org, 23 November.
Handley, K and N Limão (2017), “Policy uncertainty, trade, and welfare: Theory and evidence for China and the United States,” American Economic Review 107(9): 2731-2783.
Juhász, R (2015), “Temporary protection and technology adoption: Evidence from the napoleonic blockade”, VoxEU.org, 15 January.
Juhász, R (2018), “Temporary protection and technology adoption: Evidence from the napoleonic blockade”, American Economic Review 108(11): 3339-3376.
Krugman, P (1984), “Import protection and export promotion: International competition in the presence of oligopoly and economies of scale”, in H Kierzkowski (ed.), Monopolistic competition and international trade, Clarendon Press, Oxford, pp. 180-193.
Kucheryavyy, K, G Lyn, and A Rodríguez-Clare (2020), Grounded by gravity: A well-behaved trade model with industry-level economies of scale, Mimeo, University of California, Berkeley.
Pierce, J and P Schott (2016), “The surprisingly swift decline of US manufacturing employment”, American Economic Review 106(7): 1632-1662.