The contractionary effects of protectionist trade policy in a dollarized economy

Grijalva, Diego F.; Uribe-Terán, Carlos; Gache, Ivan Working Paper The contractionary effects of protectionist trade policy in a dollarized economy IDB Working Paper Series, No. IDB-WP-1480 Provided in Cooperation with: Inter-American Development Bank (IDB), Washington, DC Suggested Citation: Grijalva, Diego F.; Uribe-Terán, Carlos; Gache, Ivan (2024) : The contractionary effects of protectionist trade policy in a dollarized economy, IDB Working Paper Series, No. IDBWP-1480, Inter-American Development Bank (IDB), Washington, DC, https://doi.org/10.18235/0005501 This Version is available at: https://hdl.handle.net/10419/289910 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by-nc-nd/3.0/igo/legalcode The Contractionary Effects of Protectionist Trade Policy in a Dollarized Economy Diego F. Grijalva Carlos Uribe-Terán Ivan Gachet IDB WORKING PAPER SERIES Nº IDB-WP-1480 January 2024 Department of Research and Chief Economist Inter-American Development Bank January 2024 The Contractionary Effects of Protectionist Trade Policy in a Dollarized Economy Diego F. Grijalva* Carlos Uribe-Terán* Ivan Gachet** * Universidad San Francisco de Quito (USFQ) ** World Bank Cataloging-in-Publication data provided by the Inter-American Development Bank Felipe Herrera Library Grijalva, Diego F. The contractionary effects of protectionist trade policy in a dollarized economy / Diego F. Grijalva, Carlos Uribe-Terán, Ivan Gachet. p. cm. — (IDB Working Paper Series ; 1480) Includes bibliographical references. 1. Commercial policy-Ecuador. 2. Protectionism-Ecuador. 3. Tariff-Ecuador. 4. Dollarization-Ecuador. I. Carlos Uribe-Terán. II. Gachet, Iván. III. Inter-American Development Bank. Department of Research and Chief Economist. IV. Title. V. Series. IDB-WP-1480 Copyright © Inter-American Development Bank. This work is licensed under a Creative Commons IGO 3.0 AttributionNonCommercial-NoDerivatives (CC-IGO BY-NC-ND 3.0 IGO) license (http://creativecommons.org/licenses/by-nc-nd/3.0/igo/ legalcode) and may be reproduced with attribution to the IDB and for any non-commercial purpose, as provided below. No derivative work is allowed. Any dispute related to the use of the works of the IDB that cannot be settled amicably shall be submitted to arbitration pursuant to the UNCITRAL rules. The use of the IDB's name for any purpose other than for attribution, and the use of IDB's logo shall be subject to a separate written license agreement between the IDB and the user and is not authorized as part of this CC-IGO license. Following a peer review process, and with previous written consent by the Inter-American Development Bank (IDB), a revised version of this work may also be reproduced in any academic journal, including those indexed by the American Economic Association's EconLit, provided that the IDB is credited and that the author(s) receive no income from the publication. Therefore, the restriction to receive income from such publication shall only extend to the publication's author(s). With regard to such restriction, in case of any inconsistency between the Creative Commons IGO 3.0 Attribution-NonCommercial-NoDerivatives license and these statements, the latter shall prevail. Note that link provided above includes additional terms and conditions of the license. The opinions expressed in this publication are those of the authors and do not necessarily reflect the views of the Inter-American Development Bank, its Board of Directors, or the countries they represent. http://www.iadb.org 2024 Abstract This study analyzes the firm-level impacts of temporary safeguard import tariffs implemented in Ecuador from 2015 to 2017. Employing a difference-in-differences methodology, we explore the policy’s effects on a unique dataset combining firmand product-level data. We focus on the direct effects on importing firms and indirect effects through the value chain. The analysis shows, that while the safeguards significantly reduced imports, they also resulted in short-run negative scale effects on firms. These include reduced sales, employment, labor costs, and material costs, without positive impacts on local firms in import-competing industries. Overall, our findings suggest a contractionary effect of protectionist policies, particularly in a dollarized economy, highlighting the complex implications of trade measures on firm performance and economic sectors. JEL classifications: F13, F14, F16, O24, O54 Keywords: Trade policy, Protectionism, Input-output linkages, Emerging markets, Latin America Acknowledgments: We thank Pablo Astudillo and Servicio de Rentas Internas (SRI) for providing comprehensive yearly transaction-level data for all registered firms in Ecuador. We are also grateful to the Latin America and the Caribbean Research Network of the Inter-American Development Bank Group for their financial support for the project “The Political Economy of Trade Policy in Latin America and the Caribbean.” Thanks to Paul E. Carrillo, Kasper Vrojlik, Priyaranjan Jha, and an anonymous reviewer at the Inter-American Development Bank for helpful comments on an earlier version of this paper. We also appreciate discussions with participants at the USFQ Brownbag seminar, the LACEA-LAMES annual meeting, and the IEA Annual Congress. A special mention to Sara Brborich for her excellent work in creating the final dataset. The opinions in this paper are ours and do not represent those of our institutions. Replication files are available from the authors upon request. 1. Introduction Trade tariffs throughout the world have remained stable or declined in recent decades, yet the use of temporary trade barriers (TTBs)—antidumping measures, countervailing duties, and safeguards—has ballooned (Bown,2011;Grübler and Reiter,2021). Several articles have analyzed the firm-level effects of recent cases of countries using TTBs, but have focused mainly on antidumping (see e.g. Jabbour et al.,2019;Konings and Vandenbussche, 2013;Vandenbussche and Zanardi,2010), which tends to affect a relatively small number of firms. In contrast, we look at the firm-level effects of temporary safeguards that affected about 80% of the universe of importing firms in Ecuador, a dollarized emerging economy. Following a decade-long oil boom, at the end of 2014, Ecuador’s oil price fell to less than half of its 2011-2013 levels, reducing government revenue and threatening a large trade deficit. The Ecuadorian government responded by implementing a broad set of temporary safeguard import tariffs to limit imports. The safeguards came into effect on March 11, 2015 and affected approximately one-third of all imports, including intermediate inputs (capital and raw materials) and final goods (consumption). The Ecuadorian implementation of safeguards provides an ideal case study because it used tariffs (as opposed to non-tariff barriers), and because of its broad-based, temporary nature. Since it affected around one-third of imports (close to 3,000 HTS 10-digit subheadings), the policy generated heterogeneous exposure across firms and industries. Also, since the safeguards covered final and intermediate goods, they had an effect on firms’ activity both as producers and as consumers of goods and services. Finally, the policy was largely unexpected, as it started at most five months after the initial decline in oil prices. The implementation of the Ecuadorian safeguards also allows us to neatly isolate their effect, without several of the usual confounding factors. First, because the policy was implemented unilaterally under the provision of Article XVIII of the WTO, there was no reciprocation by Ecuador’s trading partners. Second, the safeguards were initially designed to last for 15 months, and therefore firms considered them temporary.1Third, since Ecuador is a dollarized economy, there were no effects on the nominal exchange rate and the Central Bank did not respond. We analyze the influence of safeguard import tariffs on the performance of Ecuadorian firms by taking advantage of their varying exposure to the policy. Given that firms and industries have unique import profiles, they were affected differently by the introduction of safeguards. This diversity allows us to evaluate the causal impact of the policy by applying a difference-in-difference approach. Although import surcharges were implemented 1The safeguards were extended in April 2016 following a strong earthquake that affected the Ecuadorian Coast. They were fully phased out in June 2017. 2 between 2015 and 2017, we estimate their immediate (2015), short-run (2016 and 2017), and medium-run (2018-2021) effects, all relative to a 2014 baseline. Using firm-level variation in exposure, we assess their direct impact on imports and estimate the elasticity of imports with respect to tariffs (direct exposure). Subsequently, we examine how these protective measures have affected various performance indicators of importing firms, including total factor productivity (TFP), sales and the likelihood of firm exit. We also look at the channels through which safeguards might have affected these performance measures, specifically employment, labor costs and material costs. Likewise, using industry variation at the ISIC 4-digit level, we evaluate the safeguards’ impact on the same performance indicators of importing-exporting, local import-competing, and local non-import-competing firms. We estimate the effects of industry-level protection generated by the safeguards (output exposure) and, using ISIC 4-digit total requirements input-output tables, we also estimate the impact of safeguards on downstream industries through their value chain exposure (input exposure). Figure 1provides a simple overview of the methodology considering the import dynamics. It shows the monthly percentage variation of the imports value with respect to the average of January 2014–February 2015, separated by whether the HTS subheadings were affected by the safeguards. The figure also shows the percentage variation of total imports. Before the implementation of the safeguards, each group of imports behaved very similarly, but this was not the case during and after the implementation of the safeguards: imports of affected products fell significantly more relative to the January 2014–February 2015 average. Our results point to a clear shortand medium-term trade-off between reducing a trade deficit by increasing the cost of imports and the firm-level costs of using safeguards to achieve this goal. In 2017, the last year the policy was in place, an additional 1% of direct exposure led to a firm-level decrease of 2.6% in the rate of import growth. Imports elasticity reached -0.15 in this year, a relatively low value consistent with the temporary nature of the safeguards. This reduction in imports among exposed importing firms was associated with large negative performance results. The mechanism is a reduction in the firms’ scale (Head and Ries,1999): In 2017, an additional direct exposure of 1% was associated with a reduction of 0.86% in sales growth, 0.40% in employment growth, and 0.99% in material costs growth. These effects persisted until 2021, beyond the implementation of the policy. Along with these results, we find that exposure to the safeguards was associated with a higher probability of firms exiting the market. By 2021, a 1% increase in exposure resulted in a 1.37% higher probability of exit. 3 Start of safeguards End of safeguards -80 -60 -40 -20 0 20 40 Percentage variation 2012m1 2013m1 2014m1 2015m1 2016m1 2017m1 2018m1 2019m1 2020m1 2021m1 2022m1 Non-affected Affected All imports Figure 1: Import Evolution of Affected and Non-affected HTS 10-digit Subheadings This figure displays the import dynamics of affected and non-affected HTS 10-digit subheadings before (2012-2014), during (2015-2017), and after (2018-2021) the period of implementation of the safeguard import tariffs in Ecuador. It presents percentage changes compared to the average import levels between January 2014 and February 2015. Data for total imports come from the Central Bank of Ecuador’s Commerce dataset. Monthly high-frequency variation is smoothed out using a threeperiod moving average. We also analyze the effects of safeguard import tariffs on other types of firms using industry-level variation in output and input exposure (Corden,1966,1971). Importingexporting firms and local import-competing firms were not affected. The latter result is important because it shows that the firms that could have benefited from import protection did not obtain any advantage. Local firms operating in non-import-competing industries were negatively affected by the policy through their value chain exposure. These firms experienced a temporary decline in productivity growth between 2016 and 2018. In the last year, a 1% increase in input exposure led to a 0.71% decrease in the growth rate of TFP. An additional 1% in input exposure was also associated with a persistent decline in the growth rate of sales and material costs, reaching 6.67% and 7.61% in 2021, respectively. The growth of labor costs among these firms also decreased between 2018 and 2020. In the last year, a 1% increase in input exposure implied a reduction of 13.63% in the growth rate of labor costs. 4 Finally, we show that the safeguards did not have an effect on the share of new firms by industry. Together with an increased probability of exit among importing firms, this result provides evidence of a net negative effect of safeguard import tariffs on firm creation. The paper contributes to the recent literature that finds that protectionist trade policy has contractionary effects even in the case of a fixed exchange rate (Barattieri et al.,2021). We provide empirical microeconomic evidence of these negative effects in the context of a dollarized emerging economy, which displays some elements of a fixed exchange rate, except for the expansionary policy needed from the Central Bank to sustain the fixed exchange rate. To the best of our knowledge, this is the first paper to provide a systematic evaluation of a broad-based short-term protectionist policy in a dollarized economy. More generally, our paper contributes to the literature that analyzes the effects of trade policy on firms’ performance. However, in contrast to the common focus of the literature on the long-run consequences of permanent tariff reductions, we focus on the shortand medium-run effects of a type of temporary trade barriers (TTBs), safeguard import tariffs. This is important because the effects of increasing tariffs are not symmetric with those of falling tariffs (Furceri et al.,2021). The analysis provides evidence on a policy tool that is increasingly used by developing countries facing balance of payments problems, but that has been scantly analyzed.2 The paper also contributes to the literature that looks at the effects of trade policy through value chains. Specifically, we provide empirical evidence on the shortand mediumrun effects of safeguards through vertical production linkages. Finally, it also contributes to the still-scant literature on trade policy effects at the firm level in Latin America in general, and Ecuador in particular. Relation to the Literature This paper is related to the literature that discusses the effects of trade policy under different exchange rate regimes (Auray et al.,2022;Barattieri et al.,2021), focusing on the context of dollarization. It is also related to the large set of studies that looks at the effects of trade policy on firms’ performance, particularly productivity. Caliendo and Parro (2022) and Goldberg and Pavcnik (2016) review the literature on the effects of trade policy in general and De Loecker and Goldberg (2014); Harrison and Rodríguez-Clare (2010) and Melitz and 2According to WTO Stats (available at https://stats.wto.org/) since 1996 43 countries have had safeguards in force at some point. In any given year, around 10 countries had at least one measure in force, and in 2015— the peak year and the year that we analyze for Ecuador—17 countries had a total of 43 safeguard measures in force. Since 2001 Ecuador had a total of 9 years with safeguard measures in force. 5 Variable Importer Importerexporter Local, import competing Local, non-import competing Imports 2.38M (8.12M) 2.32M (8.92M) 101.14K (782.80K) 73.08K (777.88K) Exports 61.77K (1.07M) 11.51M (29.93M) 78.87K (1.55M) 20.01K (739.70K) TFP 4.63 (12.03) 24.72 (44.56) 12.91 (20.46) 1.63 (4.38) Sales 8.13M (41.37M) 19.46M (60.50M) 2.68M (9.44M) 3.90M (24.33M) Wages 757.60K (2.93M) 2.09M (4.65M) 429.88K (1.38M) 265.91K (1.38M) Employment 63 (295) 263 (626) 50 (142) 30 (191) Materials 5.00M (28.25M) 11.94M (28.57M) 1.48M (5.68M) 3.05M (19.30M) Table 2: Descriptive Statistics for Full Sample Period (2012-2021) by Firm’s Trade Status This table presents the mean and standard deviation of imports, sales, TFP, wages, employment, and cost of materials by firm’s trade status for the entire panel of firms used for the estimation. Standard deviations are presented in parentheses. Mis millions, Kis thousands. temporary work or consulting activities directly related to the production process. Labor costs also include costs related to regulations such as mandatory social security contributions made by the employer on behalf of the employee, and the payment of the 15% share of profits that the firm is required to pay to employees on an annual basis. Employment only considers workers hired under formal contracts. Importer-exporter firms are much larger than the rest of the firms in the economy in terms of sales, number of workers, or purchase of materials. They are also the most productive, followed by local firms in import-competing industries (see Table 2). This result highlights the relevance of foreign competition. Importer-exporter firms face competition in international markets, while local firms in import-competing industries face competition from imported goods. 3. Firms’ Exposure to Trade Policy Trade policy affects economic activity through different channels. Therefore, firms’ exposure to trade policy is highly heterogeneous. To see why, we propose a broad classification of firms according to their engagement with international markets: firms that are 12 importers, those that are importer-exporters, local firms that sell their products in importcompeting markets, and local firms that sell their products in non-import-competing markets. In our classification, firms are designated as importers if they maintain an importsto-sales ratio of at least 0.05 for three consecutive years.5We posit that these firms use their imported goods either for direct sales to final consumers or to sell as intermediate goods to other firms. Consequently, safeguard import tariffs have an immediate impact on these firms’ business operations. We classify these firms as directly exposed, indicating that safeguard import tariffs significantly influence their output. Firms are classified as importer-exporter if they maintain importand export-to-sales ratios of no less than 0.05 over a span of three years. We postulate that these firms directly import materials as inputs for their production, purchase additional inputs locally, and subsequently distribute their products domestically and internationally. These firms can be affected through two channels. First, they are directly exposed via the output purchased internationally for use in their manufacturing processes. Second, they face an indirect exposure due to their dependence on locally sourced inputs from other domestic firms, which may themselves be directly affected by the policy. This affects the firm through the exposure of its value chain to the safeguard import tariffs, something that we call input exposure. It should be noted that, for both groups of firms (importers and importer-exporters), protectionist trade policies are expected to exert adverse effects, primarily through increased production costs (refer to Konings and Vandenbussche,2013, for an analysis on exporters). Assessing the impact of protectionist trade policies on local firms presents a complex challenge. For firms operating in markets that compete with imports, safeguard import tariffs can lead to two potential indirect effects. The first is output protection, stemming from tariffs imposed on competing foreign goods. This effect is indirect because, while the firm itself is not subject to tariffs, the market for its products is influenced by the policy. Consequently, trade protection measures are likely to confer a competitive advantage on these firms relative to foreign suppliers. This group of firms could also experience an increase in production costs if they rely on inputs from import-dependent firms affected by safeguard import tariffs. In such scenarios, a decline in firm performance is anticipated due to their input exposure. This concept aligns with the notion of effective protection, as developed by Corden (1966,1971), which 5The rationale for using a three-year period to define firms’ trade status is associated with the establishment of our baseline sample for the pre-policy period in the empirical analysis. 13 highlights the dual impact of protectionist trade policies. Amiti and Konings (2007), in their empirical analysis of the Indonesian context, introduced the terms output exposure and input exposure to describe these distinct but interrelated effects. Firms operating in markets that do not directly compete with imports are not immediately subject to the direct effects of protectionist policies. Indirectly, however, these firms can be affected through their supply chains. Specifically, they may face substantial increases in production costs due to their suppliers’ susceptibility to protectionist measures (input exposure). Consequently, we anticipate observing detrimental impacts of such policies on these firms’ performance, primarily driven by the increased operational costs associated with their input procurement. 4. Identification Strategy Based on the previous discussion, our empirical strategy exploits firms’ heterogeneous exposure to safeguard import tariffs and the different channels through which trade policy can affect economic activity. In this section, we discuss the design of the measures of exposure and their distributions across firms and years, followed by the details of our identification strategy. 4.1. Measures of Exposure In Section 3, we delineated three measures of exposure to capture the impacts of the safeguard import tariffs. For clarity, consider t= 0 as the immediate pre-policy period. The policy is implemented in period t= 1 and persists during t= [1,˜ T], where ˜ T > 1. We observe firm activities for a total of T > ˜ Tperiods. Denote τi,t as the safeguard import tariff levied on good iin period t. According to the policy framework, it is evident that τi,t = 0 ∀iat t= 0,τi,t ≥0for t= [1,˜ T], and then it reverts to τi,t = 0 ∀iat t > ˜ T. Extending the work of previous studies (Corden,1966,1971;Amiti and Konings,2007) that evaluated output and input exposure at the industry level, we introduce a measure of direct exposure at the firm level. This measure serves as a treatment indicator defined only for importers and importer-exporters. Let Mi,j,0be the value of imports of product i by firm jimmediately preceding the implementation of the policy. The cumulative direct exposure of firm jup to period tis expressed as: ed j,t =1 Mj,0 t X s=1 Ij,0 X i=1 τi,sMi,j,0,(1) 14 where Ij,0represents the total number of imported varieties by firm jin the pre-policy period, and Mj,0is the total value of imports by firm jin the same period. This exposure metric reflects the potential effective rate that a firm would face under the current safeguard import tariff regime τi,t, considering its import structure prior to the change in policy. Assuming no anticipation effects, direct exposure is deemed exogenous. The range of values for ed j,t spans from zero to the highest tariff rate established in the policy. Incorporating the framework of Amiti and Konings (2007), our analysis also includes measures of output and input exposure, calculated at the industry level. Output exposure captures the level of protection that an industry receives against international products that directly compete with its output. Following (Corden,1966,1971), output exposure constitutes the initial aspect of what is termed effective protection. Let Mi,k,0represent the total value of imports of product iby industry k. The accumulated output exposure for industry kup to period tis formulated as: eo k,t =1 Mk,0 t X s=1 Ik,0 X i=1 τi,sMi,k,0,(2) where Ik,0denotes the spectrum of products within industry kbefore the start of the policy. The concept of effective protection extends to a second layer, reflecting the potential rise in production costs for firms that source inputs from those affected by the policy. The input exposure for industry kis defined as the weighted average of output exposure endured by its supplier industries. Formally: ex k,t =X m ωk,m,0·eo k,t,where ωk,m,0=yk,m,0 yk,0 .(3) Here, ωk,m,0, the weight in the equation, reflects the proportion of products that industry kprocures from industry m(yk,m,0) in relation to its total acquisitions (yk,0) in the pre-policy period. These weights are set before the policy’s implementation to avoid endogeneity and ensure the validity of the treatment variable. To get a sense of the structure of the measures of exposure, Figure 3presents the cumulative annual distributions of firms and industries corresponding to direct, output, and input exposures. Each distribution is conditional on the relevant group of firms. Direct exposure is conditional on importers; output exposure is conditional on importer-exporters and local firms in import-competing industries; and input exposure is conditional on importerexporters and local firms in importand non-import-competing industries. The distribution of direct exposure is at the firm level, while the distributions of output and input exposures are at the industry level (ISIC 4-digits). 15 0 .2 .4 .6 .8 1 Cumulative distribution 0 .1 .2 .3 .4 .5 Firm-level direct exposure 2015 2017 2019 2021 (a) Direct 0 .2 .4 .6 .8 1 Cumulative distribution 0 .1 .2 .3 .4 .5 Industry-level output exposure 2015 2017 2019 2021 (b) Output 0 .2 .4 .6 .8 1 Cumulative distribution 0 .05 .1 .15 .2 .25 Industry-level input exposure 2015 2017 2019 2021 (c) Input Figure 3: Distribution of Measures of Exposure This figure shows the distributions of the three measures of exposure: direct, output, and input. Each distribution is conditional on the relevant group of firms. Direct exposure is conditional on importers and importer-exporters; output exposure is conditional on importer-exporters and local firms in import-competing industries; and input exposure is conditional on importer-exporters and local firms in importand non-import-competing industries. The distribution of direct exposure is at the firm level, while the distributions of output and input exposures are at the industry level (ISIC 4-digits). A substantial proportion of firms and industries are not exposed to the safeguard import tariffs. Approximately 20% of firms have zero direct exposure, while around 17% of industries show no output exposure. Input exposure, however, presents a different pattern. Due to the inter-industrial linkages within value chains, almost all industries experience some level of exposure to safeguard import tariffs. Despite this, the magnitude of input exposure is generally lower compared to direct and output exposures. While direct and output exposures can reach up to 45% (aligned with the policy’s maximum tariff rate) in the initial year of implementation, input exposure seldom exceeds about 20%. The dynamic nature of the policy and the cumulative design of the exposure indices mean that the exposure distributions evolve over time. Since imports are kept constant at baseline levels, exposure to the policy gradually diminishes due to the phasing-out of safeguard import tariffs. This trend is evident in the evolution of our three exposure measures over time. By 2017, the overall reduction in exposure is modest, yet some stochastic dominance is observable in Figure 3. By 2019, even though the policy is no longer active, the accumulated nature of our measures still allows us to track the exposure of firms and industries. As expected, a marked decrease in exposure levels is apparent, becoming more pronounced by 2021, which marks the last year of our analysis. 16 4.2. Empirical Specifications Our specifications are based on the linear difference-in-difference models previously used by Machin et al. (2003); Draca et al. (2011) and Harasztosi and Lindner (2019) in the labor market literature.6We use different specifications for each type of firm in line with the exposure that they face. To determine the effects of safeguard import tariffs on importers, we use our measure of direct exposure ed j,t as a treatment variable, where jdenotes firms and tdenotes years. Our regression can be written as zj,t −zj,0=αt+βted j,t +γtXj,0+εj,t,(4) where the left-hand-side is the log-variation in outcome z,αtare time-specific fixed effects, and γtmeasures the effects of firm-specific characteristics. Our parameter of interest is βt, which quantifies the effect of direct exposure to the import safeguards on each specific outcome. We are technically computing a weighted average of the average treatment effect (ATE). However, as pointed out by Callaway et al. (2021), there are several caveats that must be taken into account when using the two-way fixed-effects estimator to summarize the effect of a continuous treatment variable. One major concern is the possibility of bias arising even if the classical parallel trends assumption is satisfied. However, in our setting, we are confident that we can invoke the strong parallel trends assumption established by Callaway et al. (2021) due to the way we constructed our measures of exposure. Specifically, any bias resulting from some units receiving a dose different from the one specified in the original treatment design should not be problematic. Consequently, we interpret βt(and the relevant coefficients in the subsequent models) as a weighted average of the ATE effects across expected tariffs. If the firm-level impact of the policy is profound enough, it will affect importing firms’ decisions on the extensive margin of their activities, i.e., some firms may shut down their operations due to the increased production costs. To evaluate this effect, we use a modified version of (4), where the outcome variable is a binary variable, dj,t, that takes the value of 1 if a firm shuts down in period twhile it was open in period 0. Mathematically, dj,t = Φ(θt+ξted j,t +νtXj,0+ϵj,t),(5) 6Handley et al. (2020) use a similar strategy to estimate the effect of import tariffs on exports in the United States through the lens of value chain linkages. 17 where Φ(·)represents the cumulative distribution function of the Normal distribution. Here, ξtcaptures the impact of the safeguard import tariffs on firms’ exposure, which may result in them shutting down. As we discussed in Section 3, identifying the effect on the rest of firms in the economy is more complex. In the case of importer-exporter firms, they suffer direct exposure to the policy because of their importing activity, but they also face input exposure, because they might still buy inputs from firms in industries that were affected by the policy. The model we estimate to measure these effects can be written as: zj,k,t −zj,k,0=αt+βted j,t +θtex k,t +γtXj,0+εj,k,t,(6) where θtmeasures the effect of input exposure, which is measured at the industry level. The exposure of local firms in import-competing industries occurs through two channels. First, these firms have local sales that might be competing with imports. Thus, the first channel is the protection triggered by output exposure. The second channel is, again, through the exposure of these firms’ supplier chain, which is captured by input exposure. The model for this group becomes the following: zj,k,t −zj,k,0=αt+βteo k,t +θtex k,t +γtXj,0+εj,k,t,(7) where both measures of exposure are measured at the industry level. Finally, local firms in non-import competing industries are mainly affected by the exposure faced by their supply chain. The model in this case simplifies to: zj,k,t −zj,k,0=αt+θtex k,t +γtXj,0+εj,k,t,(8) where, again, exposure is measured at the industry level. Because of the differentiated effects that the policy design can have on firms and industries depending on their import structure, we include a wide set of firm and industry characteristics set at baseline (2014) as controls for all our specifications. In particular, we include industry-level fixed effects, dummy variables that identify firms’ size according to their sales, Herfindahl indexes for sales and imports (computed at the ISIC 3-digit level), and firms’ capital stock. Additionally, we consider the degree of import penetration at the industry level (ISIC 3-digits), province-level fixed effects (defined according to the firms’ fiscal identification number), a dummy that identifies big corporations, and one that identifies firms that belong to business groups. The last two dummy variables are key to 18 controlling for variation in the policy that could be explained by the degree of lobbying power that large firms might have in the design of the policy (Bombardini,2008). 5. Results We present the results of our analysis in three sections. First, we examine the impact of safeguard import tariffs on importers. We start with an assessment of the direct effects on imports, then explore the implications on firm performance, and conclude with an analysis of the potential channels of influence. Additionally, this section incorporates a placebo test to assess the robustness of our identification strategy. We further expand our baseline analysis by extending the study period to evaluate the long-term effects induced by the policy. Subsequently, we turn our attention to non-importing firms. Mirroring our approach with importers, we study the performance effects on non-importing firms and delve into the underlying mechanisms driving these results. We conclude the results section by examining how safeguard import tariffs may have influenced the emergence of new firms in industries with higher exposure levels. 5.1. Effects on Importers In this section, we focus on the effect of safeguard import tariffs on importers. We see these results as the main effects of the policy because importing firms are the ones facing the safeguards directly. In this case, the treatment variable corresponds to the measure of direct exposure defined in equation (1). 5.1.1. Effects on Imports The primary objective of the safeguard import tariffs is to protect the balance of payments by reducing imports. We estimate equation (4) using firm-level total imports as the outcome variable. The results can be found in Panel A of Table 3, which shows the effect of the safeguards on cumulative total imports for the three years that the policy was in place. Safeguard import tariffs were highly effective in limiting imports throughout their implementation. To guide the interpretation, recall from Figure 1that the period of safeguard implementation was characterized by a general decline in imports, with a deeper contraction on affected products. Our results show that this effect found at the product-level filters its way up to the firm level: an additional 1% increase in firm-level direct exposure is associated with a 1.4% reduction in the growth rate of imports during 2015. 19 Changes in outcome between tand 2014 2015 2016 2017 Panel A: Change in firm-level imports Expected tariff −1.429 ∗∗∗ −3.010 ∗∗∗ −2.649 ∗∗∗ (0.257) (0.553) (0.571) Mean ∆z−0.081 −0.307 −0.144 Observations 2891 2826 2795 Panel B: Imports elasticity Expected tariff −0.088 ∗∗∗ −0.188 ∗∗∗ −0.147 ∗∗∗ (0.021) (0.045) (0.036) Observations 2891 2826 2795 Table 3: Effects of Safeguards on Imports and Imports Elasticity This table shows, in Panel A, the relationship between firm-level output exposure to safeguards and the change in imports. In Panel B, the table presents the estimate of imports elasticity to changes in safeguard import tariffs. The estimates in panel A correspond to equation (4). To obtain the estimates of imports elasticity, we estimate (4) separately for the change in imports and the change in paid safeguards. Then, we use the seemingly unrelated regression framework to estimate the non-linear combination of the βtcoefficients of both regressions. In all cases, standard errors are clustered at the industry level (ISIC 3-digits). The estimations for each year include controls for economic sector (ISIC 1-digit), firm size (CAN classification), whether or not the firm belongs to a business group (identified by SRI), Herfindahl indexes for sales and imports, the logarithm of the firm-level capital stock, province-level fixed effects, and the firm-level share of imports and exports to the European Union. The variation in the outcome variable is winsorized at the bottom and top 0.5%. To get a better sense of the magnitude of the effect, note that the average firm-level import contraction between 2014 and 2015 was 8%. Therefore, a 1% additional exposure to safeguard import tariffs implies a further reduction of 0.1 percentage points in this growth rate. This effect more than doubles during the second year of implementation and persists during the third year, when a 1% additional exposure is associated with a 2.6% decrease in imports growth. All these results are significant at the 1% level. Recall that our results provide information on the cumulative changes between the baseline year (2014) and each of the years reported. The larger effect in 2016 is consistent with an additional full year of safeguard import tariffs, while the lower coefficient in 2017 is consistent with the fact that firms were only partially exposed to safeguards in 2017, as they were phased out by June of that year. Next, we estimate equation (4) for effective paid tariffs and imports as outcomes in a set of seemingly unrelated regressions, following Harasztosi and Lindner (2019). The ratio between the βtcoefficients for imports and effectively paid tariffs represents the import elasticity with respect to ad valorem safeguard import tariffs. We calculate two types of elasticity: impact elasticity, which captures firms’ responses to the policy during its first 20 year of implementation, and short-term elasticity, which accumulates the effects for 2016 and 2017. These parameters are relevant for policy design since they enable the parameterization of models capable of estimating trade policy effects ex ante. We present these estimations in Panel B of Table 3. The elasticity of imports with respect to the safeguard rate is −0.088 on impact. This elasticity reaches its lowest point in 2016 (−0.188) and then increases to −0.147 in 2017. The elasticity on impact represents 47% of the strongest response of imports to the increase in tariffs. Our findings reveal a range of elasticities that is lower than recent estimates in the literature. Specifically, Boehm et al. (2023) report an impact elasticity of −0.26 and a short-run elasticity of −0.76. A key factor driving these discrepancies is likely the distinct policy designs considered in each analysis. Ours is based on the escalation in tariff rates induced by safeguard import tariffs, which due to policy restrictions must be temporary. Accordingly, we expect firms’ response to be smaller, as they expect the surcharges to be short-lived. In contrast, Boehm et al. (2023) utilize exogenous variation resulting from changes in most favored nation tariffs, which tend to influence trade over more prolonged time periods. 5.1.2. Effects on Performance In this section, we discuss the effects of safeguard import tariffs on the performance of importing firms. To gauge this impact, we track annual changes in total factor productivity (TFP), sales, and the number of firms during the policy implementation period (2015 to 2017), using 2014 as a reference point. The results are summarized in Table 4, offering a comprehensive view of how import surcharges influenced the operational metrics of importing firms during the specified time frame. Contrary to previous research (e.g. Amiti and Konings,2007;De Loecker et al.,2016; Topalova and Khandelwal,2011), we find that increased trade protection is not associated with a reduction in the productivity growth of directly exposed firms. Although the coefficients are consistently negative, the observed decrease in productivity is never statistically significant (Panel A). Regarding sales, our data reveal a clear relationship between greater direct exposure and reduced sales growth. Specifically, a 1% rise in direct exposure is associated with a 0.27% decrease in sales growth in the first year of policy implementation. This negative trend becomes more pronounced over the next two years, with sales growth declining by 0.55% in 2016 and 0.86% in 2017. These findings are statistically significant at the 1% level (Panel B). The large decline in sales is the first sign of the negative scale effect associated with increased trade protection (Head and Ries,1999). 21 These outcomes are consistent with protectionist trade policies being generally contractionary for directly affected firms. For importing firms, which rely heavily on imports for their operations, import surcharges directly impact their cost structure, leading to the observed scale reduction in line with Head and Ries (1999). Although the results for 2020-2021 are consistent with those of previous years, it is important to note that these years are affected by the impact of the Covid-19 pandemic. Still, there is no a priori reason to expect that firms that were more exposed to the safeguards based on their 2014 import structure were also more affected by the pandemic, providing support for our analysis. 5.2. Effects on Other Firms In this section, we estimate the impact of the safeguard import tariffs on the rest of firms, i.e., those that also export in addition to importing and those that do not trade with the rest of the world. We categorize these firms into three groups based on their international trade involvement: importer-exporters, local firms in import-competing industries, and local firms in non-import-competing industries. We assess the effects on five variables: Total factor productivity (TFP), sales, employment, labor costs, and material costs. In addition, for importer-exporters, we examine the impact on exports. Importer-exporters could be influenced by import surcharges in two ways: direct exposure due to their imports and indirect exposure through local supply chains. However, we find no significant impact on this group through either channel (see Figure 7in the Appendix). The absence of notable effects on importer-exporters might be attributed to their larger size and international market presence, potentially allowing them to absorb the additional costs induced by the policy. However, it is important to note that the number of observations within this group of firms is rather limited, representing just 7% of our sample firms (Table 1). This small proportion and outcome variability could impact the precision of the estimation. For local firms in import-competing industries, we apply the concept of effective protection. While these firms might benefit from trade protection, they could also face higher input costs due to value chain exposure. We observe no significant effects of the safeguard tariffs on these firms, and our estimates show minimal variation. This finding is important because, as Jabbour et al. (2019) suggest, the group of firms most likely to benefit from short-term protectionism comprises import-competing firms. The fact that we do not find significant effects strongly supports the idea that the benefits of protectionism are minimal or non-existent. The results are detailed in Figure 8in the Appendix. 28 -3 -2 -1 0 1 2 3 Effect 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 Point estimate 95% CI (a) Total factor productivity -12 -10 -8 -6 -4 -2 0 Effect 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 Point estimate 95% CI (b) Sales -12 -10 -8 -6 -4 -2 0 2 Effect 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 Point estimate 95% CI (c) Employment -20 -15 -10 -5 0 5 Effect 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 Point estimate 95% CI (d) Labor costs -14 -12 -10 -8 -6 -4 -2 0 Effect 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 Point estimate 95% CI (e) Materials Figure 5: Dynamic Effects of Safeguards on Local Firms in Non-import-competing Industries This figure shows the dynamic relationship between industry-level input exposure to safeguards and the change in the outcomes of local firms that operate in non-import-competing industries. The outcomes include firm-level TFP (Panel a), sales (Panel b), employment (Panel c), labor costs (Panel d) and material costs (Panel e). In all cases, standard errors are clustered at the industry level (ISIC 3-digits) and all the control variables from Table 3are included. The estimations for each year include controls for economic sector (ISIC 1-digit), firm size (CAN classification), whether or not the firm belongs to a business group (identified by SRI), Herfindahl indexes for sales and imports, the logarithm of the firm-level capital stock, province-level fixed effects, and the firm-level share of imports and exports to the European Union. The variations in the outcome variables are winsorized at the bottom and top 0.5%. Lastly, we analyze the impact on local firms in non-import-competing industries. Most of these firms (93%) belong to wholesale & retail industries. They do not import directly but likely purchase imported goods from local importers for local resale without foreign competition. Therefore, these firms do not face either direct or output exposure, but they may be affected through their value chain. The results are shown in Figure 5. Safeguard import tariffs consistently harm economic activity among non-import-competing firms. From the second year of the policy, total factor productivity (TFP) begins to decrease. By 2016, a 1% increase in input exposure leads to a 0.45% drop in TFP growth, significant at the 5% level. This negative trend continues through 2017 (-0.48%) and 2018 (-0.71%), even a year after the policy ended, intensifying to -0.95% in 2019, although with less significance due to variability among firms. 29 Sales are even more adversely affected. Although not significant in 2015 and 2016, by 2017, a 1% increase in input exposure causes a 2.49% decrease in sales growth. This negative impact on sales persists and worsens, reaching a 6.67% decrease by 2021. We find no significant effects on formal employment, with the exceptions of 2015 and 2021. In the first year of policy implementation, a 1% increase in input exposure corresponded to a 1.14% decrease in formal employment growth. Although this negative effect persisted in subsequent years, it was not statistically significant. Regarding labor costs, no significant effect was detected during the period of active safeguard import tariffs. However, a notable change occurred between 2018 and 2021, with labor costs decreasing significantly, especially in 2020, the year of the pandemic, where they fell by 13.63% in response to a 1% increase in input exposure. Material costs also decrease from 2016 onward. A 1% increase in input exposure results in a 2.50% reduction in the growth of material cost in 2016, falling to -8.07% by 2020. These effects are significant at the 1% level from 2016 to 2021. These findings reflect the broader impact of safeguard import tariffs on the production scale of local non-import-competing firms. Our results also support existing literature on protectionism’s negative impact on productivity, especially through input exposure in downstream industries. Furthermore, our analysis reveals that value chain impacts are more substantial than direct industry protection. This is in line with previous findings that input tariffs have more pronounced effects than output tariffs. Overall, our results show a general negative scale effect, with sales, labor, and material costs declining across various firm groups, particularly noticeable through input exposure among local non-import-competing firms. Contrary to some studies, we find no negative impact on exporters’ sales or other performance metrics. The null effect on importer-exporters and the negative impact on local firms align with the context of the tariff implementation, which was a response to falling oil prices and the resulting aggregate demand shock. Local firms faced reduced demand, with exposed firms hit harder, whereas exporters were less affected. Contrary to expectations, we find no positive impact on local import-competing firms, though they fare better than non-import-competing firms by not experiencing negative impacts from value chain exposure in labor or material costs. 30 5.3. New Firms We conclude this section showing results on the effect of safeguard import tariffs on firm entry. Since we cannot see firms that were willing but did not enter the market, we focus on the change in the share of new firms in each industry. First, we define new firms as those open between 2015 and 2021 but not present in 2014. Unlike previous sections, our benchmark is composed of all firms open in 2014 as opposed to only those included in the balanced panel. We then calculate their share at each point in time as a percentage of all active firms in their industry and trade status category. Second, we adjust our estimation approach. The outcome variable is the share of new firms by industry at the ISIC 4-digit level. We consider only output and input exposure, since direct exposure is defined at the firm level. Other controls are industry averages set at the 2014 baseline. Third, due to limited data on importer-exporters, which causes problems with the computation of standard errors (bootstrapped, with 1,000 repetitions), we include importerexporters in the same category as importing firms. This results in three trade status categories instead of four. With these adjustments, the results of our estimations are presented in Figure 6. Protective import tariffs can have varying effects on the share of new firms, depending on their participation in international trade. For importer-exporters, output exposure can act as a barrier to entry, imposing higher tariffs during safeguard periods. In contrast, output exposure might offer a protective advantage for new firms in import-competing industries. Input exposure, however, likely poses an entry barrier for all firms due to increased input costs. Our estimates for importer-exporters (Panels a and c in Figure 6) indicate no significant impact of safeguard tariffs on the share of new firms. Output exposure shows negligible, non-significant, negative effects, while input exposure’s effect is virtually zero. For local firms in import-competing industries (Panels b and d, Figure 6), the expected trends are observed, although none of the estimates is statistically significant. Output exposure, offering protection, yields positive effects, whereas the raised production costs from input exposure lead to negative point estimates. Local firms in non-importcompeting industries see larger (in absolute value) yet non-significant changes due to input cost increases from value chain exposure. In summary, safeguard import tariffs do not significantly affect the share of new firms in more exposed industries, neither as barriers nor as protective measures. Coupled with our findings of a positive exit probability among importers (Figure 4d), we observe a net negative effect of the policy on firm-level decisions at the extensive margin. 31 Panel A: Output Exposure -.01 -.008-.006 -.004 -.002 0 .002 .004 Effect 2014 2015 2016 2017 2018 2019 2020 2021 Point estimate 95% CI (a) Importer-exporters -.004-.002 0 .002 .004 .006 .008 .01 Effect 2014 2015 2016 2017 2018 2019 2020 2021 Point estimate 95% CI (b) Local, import competing Panel B: Input Exposure -.03 -.02 -.01 0 .01 .02 .03 Effect 2014 2015 2016 2017 2018 2019 2020 2021 Point estimate 95% CI (c) Importer-exporters -.025-.02-.015-.01 -.005 0 .005 .01 .015 Effect 2014 2015 2016 2017 2018 2019 2020 2021 Point estimate 95% CI (d) Local, import competing -.25 -.2 -.15 -.1 -.05 0 .05 .1 .15 Effect 2014 2015 2016 2017 2018 2019 2020 2021 Point estimate 95% CI (e) Local, non-import comp. Figure 6: Effects of Safeguard Import Tariffs on the Industry Share of New Firms This figure shows the dynamic relationship between industry-level output and input exposure to safeguards and the change in the share of new firms in each industry by trade status. The group of importer-exporters includes importing and importing-exporting firms due to the small number of observations in the second group. In all cases, standard errors are bootstrapped (1,000 repetitions) and all previous controls are included at the industry level (ISIC 4 digits). The estimations for each year include controls for economic sector (ISIC 1digit), firm size (CAN classification), whether or not the firm belongs to a business group (identified by SRI), Herfindahl indexes for sales and imports, the logarithm of the firm-level capital stock, province-level fixed effects, and the firm-level share of imports and exports to the European Union, all aggregated at the industry level. The variations in the outcome variables are winsorized at the bottom and top 0.5%. 6. Conclusion In this paper, we analyzed the firm-level effects of safeguard import tariffs implemented by the Ecuadorian Government in response to a balance of payments crisis sparked by a sharp drop in oil prices at the end of 2014. The policy is particularly instructive due to its use of tariffs rather than non-tariff barriers, its temporary and largely unforeseen nature, and its impact on a wide range of goods, both final and intermediate. We employed a quantitative approach, taking advantage of the heterogeneous prepolicy exposure of firms and industries. We analyzed direct exposure for importing firms, and output and input exposure at the industry level, in line with the literature on effective protection. Using this framework, we conducted a difference-in-difference analysis at both firm and industry levels. 32 For importing firms directly affected by the safeguard tariffs, we studied the effects on imports and calculated the tariff elasticity of imports, a crucial metric for policy evaluation. Our findings indicate that despite lower short-term elasticities compared to other studies, the tariffs significantly curbed imports. We also assessed the impact of the policy on performance metrics such as productivity (TFP), sales, and exit probability of affected firms. The results showed significant, persistent, negative impacts on sales and an increasing likelihood of firm exit post-policy, though no immediate effects were observed during the policy period. Analyzing the underlying mechanisms, we observed significant, sustained negative impacts on employment and material costs growth rates, with no notable effects on labor costs, possibly due to the specificities of the Ecuadorian labor market. We extended our analysis to other firm categories, including importer-exporters, local firms in import-competing, and non-import-competing industries. The most adversely affected firms were the local non-import-competing firms, experiencing negative, lasting effects on productivity, sales, labor, and material costs. For other firms, the policy did not have significant effects on these parameters. Our research offers crucial insights for policymakers. Specifically, it sheds light on how trade policy impacts production scale. We found that protective trade measures often result in a significant decrease in production scale and a corresponding drop in input demand. These measures tend to have a recessionary effect, even in economies without active exchange rate policy. This highlights the need for developing countries to focus on enhancing competitiveness. Policies aimed at facilitating access to international markets and encouraging growth by exposing local businesses to foreign competition are essential. 33 References Ackerberg, D. A., K. Caves, and G. Frazer (2015): “Identification Properties of Recent Production Function Estimators,” Econometrica, 83, 2411–2451. Amiti, M. and J. Konings (2007): “Trade Liberalization, Intermediate Inputs, and Productivity: Evidence from Indonesia,” American Economic Review, 97, 1611–1638. Amiti, M., S. J. Redding, and D. E. Weinstein (2019): “The Impact of the 2018 Tariffs on Prices and Welfare,” Journal of Economic Perspectives, 33, 187–210. Auray, S., M. B. Devereux, and A. Eyquem (2022): “Self-enforcing trade policy and exchange rate adjustment,” Journal of International Economics, 134, 103552. Autor, D. H., D. Dorn, G. H. Hanson, and J. Song (2014): “Trade Adjustment: WorkerLevel Evidence,” The Quarterly Journal of Economics, 129, 1799–1860. Barattieri, A. and M. Cacciatore (2023): “Self-Harming Trade Policy? Protectionism and Production Networks,” American Economic Journal: Macroeconomics, 15, 97–128. Barattieri, A., M. Cacciatore, and F. Ghironi (2021): “Protectionism and the business cycle,” Journal of International Economics, 129, 103417. Bas, M. (2012): “Input-trade liberalization and firm export decisions: Evidence from Argentina,” Journal of Development Economics, 97, 481–493. Boehm, C. E., A. A. Levchenko, and N. Pandalai-Nayar (2023): “The Long and Short (Run) of Trade Elasticities,” American Economic Review, 113, 861–905. Bombardini, M. (2008): “Firm heterogeneity and lobby participation,” Journal of International Economics, 75, 329 – 348. Bown, C. P. (2011): “Taking Stock of Antidumping, Safeguards and Countervailing Duties, 1990–2009,” The World Economy, 34, 1955–1998. Bown, C. P., P. Conconi, A. Erbahar, and L. Trimarchi (2021): “Trade Protection Along Supply Chains,” . Broz, J. L., M. J. Duru, and J. A. Frieden (2016): “Policy responses to balance-of-payments crises: the role of elections,” Open Economies Review, 27, 207–227. 34 Caliendo, L. and F. Parro (2022): “Chapter 4 - Trade policy,” in Handbook of International Economics: International Trade, Volume 5, ed. by G. Gopinath, E. Helpman, and K. Rogoff, Elsevier, vol. 5 of Handbook of International Economics, 219–295. Callaway, B., A. Goodman-Bacon, and P. H. C. Sant’Anna (2021): “Difference-inDifferences with a Continuous Treatment,” . Cavallo, A., G. Gopinath, B. Neiman, and J. Tang (2021): “Tariff Pass-Through at the Border and at the Store: Evidence from US Trade Policy,” American Economic Review: Insights, 3, 19–34. Corden, W. M. (1966): “The structure of a tariff system and the effective protective rate,” Journal of Political Economy, 74, 221–237. ——— (1971): The Theory of Protection, Clarendon Press. De Loecker, J. and P. K. Goldberg (2014): “Firm Performance in a Global Market,” Annual Review of Economics, 6, 201–227. De Loecker, J., P. K. Goldberg, A. K. Khandelwal, and N. Pavcnik (2016): “Prices, Markups, and Trade Reform,” Econometrica, 84, 445–510. Dix-Carneiro, R. and B. K. Kovak (2019): “Margins of labor market adjustment to trade,” Journal of International Economics, 117, 125–142. Draca, M., S. Machin, and J. Van Reenen (2011): “Minimum Wages and Firm Profitability,” American Economic Journal: Applied Economics, 3, 129–51. Décamps, J.-P., S. Gryglewicz, E. Morellec, and S. Villeneuve (2016): “Corporate Policies with Permanent and Transitory Shocks,” The Review of Financial Studies, 30, 162–210. Fajgelbaum, P. D., P. K. Goldberg, P. J. Kennedy, and A. K. Khandelwal (2020): “The Return to Protectionism,” The Quarterly Journal of Economics, 135, 1–55. Fernandes, A. M. (2007): “Trade policy, trade volumes and plant-level productivity in Colombian manufacturing industries,” Journal of International Economics, 71, 52–71. Flaaen, A. and J. Pierce (2019): “Disentangling the Effects of the 2018-2019 Tariffs on a Globally Connected U.S. Manufacturing Sector,” Finance and Economics Discussion Series 2019-086, Board of Governors of the Federal Reserve System, Washington. Furceri, D., S. A. Hannan, J. D. Ostry, and A. K. Rose (2021): “The Macroeconomy After Tariffs,” The World Bank Economic Review, 36, 361–381. 35 Goldberg, P. K., A. K. Khandelwal, N. Pavcnik, and P. Topalova (2010): “Imported intermediate inputs and domestic product growth: Evidence from India,” The Quarterly journal of economics, 125, 1727–1767. Goldberg, P. K. and N. Pavcnik (2016): “The effects of trade policy,” in Handbook of commercial policy, Elsevier, vol. 1, 161–206. Grijalva, D. F., I. Gachet, P. Lucio-Paredes, and C. Uribe-Terán (2022): “The Political Economy of Trade Policy in Ecuador: Dollarization, Oil, Personalism, and Ideas,” in Political Economy of Trade Policy in Latin America, ed. by J. Cornick, J. Frieden, M. Mesquita Moreira, and E. Stein, 215–253. Grübler, J. and O. Reiter (2021): “Characterising non-tariff trade policy,” Economic Analysis and Policy, 71, 138–163. Handley, K., F. Kamal, and R. Monarch (2020): “Rising Import Tariffs, Falling Export Growth: When Modern Supply Chains Meet Old-Style Protectionism,” Working Paper 26611, National Bureau of Economic Research. Harasztosi, P. and A. Lindner (2019): “Who Pays for the Minimum Wage?” American Economic Review, 109, 2693–2727. Harrison, A. and A. Rodríguez-Clare (2010): “Chapter 63 - Trade, Foreign Investment, and Industrial Policy for Developing Countries*,” in Handbooks in Economics, ed. by D. Rodrik and M. Rosenzweig, Elsevier, vol. 5 of Handbook of Development Economics, 4039–4214. Head, K. and J. Ries (1999): “Rationalization effects of tariff reductions,” Journal of International Economics, 47, 295–320. Jabbour, L., Z. Tao, E. Vanino, and Y. Zhang (2019): “The good, the bad and the ugly: Chinese imports, European Union anti-dumping measures and firm performance,” Journal of International Economics, 117, 1–20. Konings, J. and H. Vandenbussche (2013): “Antidumping protection hurts exporters: firmlevel evidence,” Review of World Economics, 149, 295–320. Levinsohn, J. and A. Petrin (2003): “Estimating production functions using inputs to control for unobservables,” The Review of Economic Studies, 70, 317–341. 36 Machin, S., A. Manning, and L. Rahman (2003): “Where the minimum wage bites hard: Introduction of minimum wages to a low wage sector,” Journal of the European Economic Association, 1, 154–180. Melitz, M. J. and S. J. Redding (2014): “Chapter 1 - Heterogeneous Firms and Trade,” in Handbook of International Economics, ed. by G. Gopinath, E. Helpman, and K. Rogoff, Elsevier, vol. 4 of Handbook of International Economics, 1–54. Olley, G. and A. Pakes (1996): “The dynamics of productivity in the telecommunications equipment industry,” Econometrica, 64, 1263–1297. Pavcnik, N. (2002): “Trade Liberalization, Exit, and Productivity Improvements: Evidence from Chilean Plants,” The Review of Economic Studies, 69, 245–276. Topalova, P. and A. Khandelwal (2011): “Trade Liberalization and Firm Productivity: The Case of India,” The Review of Economics and Statistics, 93, 995–1009. Vandenbussche, H. and M. Zanardi (2010): “The chilling trade effects of antidumping proliferation,” European Economic Review, 54, 760–777. Wong, S. A. (2007): “Market-Discipline Effects of Trade Liberalization: Micro-Level Evidence from Ecuador, 1997-2003,” Applied Econometrics and International Development, 7. ——— (2009): “Productivity and trade openness in Ecuador’s manufacturing industries,” Journal of Business Research, 62, 868–875. 37