Received: 17 July 2018 Revised: 19 March 2019 Accepted: 19 March 2019

DOI: 10.1111/rsp3.12198

ORIGINAL ARTICLE

Global and regional effects of the US tariffs on

iron, steel and aluminium: A SMART combination
of models with a focus on Spain

Carlos Llano1 | Julián Pérez2 | Federico Steinberg3 |

Geoffrey J.D. Hewings4

1
Departamento de Análisis Económico: Teoría
Económica e Historia Económica & L.R.Klein Abstract
Institute (Centro Stone) & CEPREDE, Facultad On 8 March, US President Donald Trump approved new
de CC.EE y EE. Módulo E‐1, Universidad
Autónoma de Madrid, Spain import tariffs on “iron and steel” and “aluminium” of 25%
2
Applied Economics Department & L.R.Klein and 10%, respectively. The aim of this paper is to estimate
Institute (Centro Stone) & CEPREDE, Facultad
the impact that this decision might have on the Spanish
de CC.EE y EE. Módulo E‐XIV, Universidad
Autónoma de Madrid, Spain regions. Our empirical strategy combines three powerful
3
Departamento de Análisis Económico: Teoría models to address the trade effect both over the global
Económica e Historia Económica & Real
Instituto Elcano, Facultad de CC.EE y EE.
economy, and more specifically, in each Spanish region.
Módulo E‐XIV, Universidad Autónoma de First, by means of the SMART simulation model, we esti-
Madrid, Spain
4
mate the trade effect for 131 countries; then, we compute
Regional Economics Applications Laboratory,
University of Illinois at Urbana‐Champaign, IL, the intersectoral effects of such shock using the World
USA Input–Output Database (WIOD); finally, we estimate the
Correspondence
interregional and intersectoral effects within Spain, using
Carlos Llano, Departamento de Análisis
Económico: Teoría Económica e Historia an interregional input–output table. This last step combines
Económica & L.R.Klein Institute (Centro Stone)
both, the immediate product shock over the metal sector in
& CEPREDE. Facultad de CC.EE y EE. Módulo
E‐1. Universidad Autónoma de Madrid. Spain, and the global effect computed by means of the
Campus Cantoblanco. 28049 Madrid, Spain.
WIOD. The results obtained shows how the US tariffs might
Email: [email protected]
Funding information induce a total job loss of 185,000 employees worldwide,
Secretaría de Estado de Investigación, while in Spain, it may cost 3,500 jobs.
Desarrollo e Innovación, Grant/Award Num-
ber: ECO2016‐79650‐P
KEYWORDS

inter‐regional input–output models, inter‐regional trade, Spain, trade

policy, US

JEL CLASSIFICATION
F13; F14; R15

-------------------------------------------------------------------------------------------------------
© 2019 The Author(s). Regional Science Policy and Practice © 2019 RSAI

Reg Sci Policy Pract. 2019;11:525–547. wileyonlinelibrary.com/journal/rsp3 525

 17577802, 2019, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/rsp3.12198 by Benemerita Universidad Autonoma De Puebla, Wiley Online Library on [12/02/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
526 LLANO ET AL.

I concur in the Secretary's finding that steel articles are being imported into the United States in such
quantities and under such circumstances as to threaten to impair the national security of the United
States (…) I have decided to adjust the imports of steel articles by imposing a 25 percent ad valorem
tariff on steel articles, as defined below, imported from all countries except Canada and Mexico. In
my judgment, this tariff is necessary and appropriate (…) This relief will help our domestic steel industry
to revive idled facilities, open closed mills, preserve necessary skills by hiring new steel workers, and
maintain or increase production, which will reduce our Nation's need to rely on foreign producers for
steel and ensure that domestic producers can continue to supply all the steel necessary for critical
industries and national defense. Donald Trump (2018)

1 | I N T R O D U CT I O N

On 8 March 2018, US President Donald Trump approved the establishment of new import tariffs on steel and alu-
minium of 25% and 10%, respectively. As indicated in the explanatory memorandum of the approved norm, Trump,
based on the report of the Secretary of State for Commerce (USTR, 2017), and considering the previous experiences
of the US commercial policy, justified the measure on the basis of national security; the tariffs were activated on 23
March. Canada, Mexico the EU, as well as other countries such as Argentina, Australia, Brazil and South Korea, were
temporally exempted. It is not clear if this exemption was part of the original intention of the Trump administration, or
if it was decided after intense diplomatic negotiations, or a cost–benefit evaluation of the potential countermeasures
announced by the EU. Indeed, the US administration said the suspensions could be renewed or revoked, depending on
satisfactory alternative measures to address the threatened impairment to US national security. Finally, on 1 June, all
exemptions were revoked except the ones for South Korea, Argentina, Brazil and Australia.1 The EU announced retal-
iatory measures and took the US decision to the World Trade Organization (WTO). China also confirmed the appli-
cation of retaliation measures against US products. Nonetheless, the trade confrontation between the US and China
extended well beyond steel and aluminium. On 6 July, the Trump Administration announced additional measures
against Chinese products, in response to intellectual‐property abuses by China. China's retaliation started a tit‐for‐
tat dynamic that could constitute the beginning of a trade war.
In this context, the aim of this paper is to estimate the impact that this protectionist policy might have on Spain.
More specifically, we want to evaluate its potential effect on the exports, income and employment in the Spanish
economy, with a focus on each of the Spanish regions (autonomous communities at the NUTS 2 level). However,
in order address this question adequately, we first consider the potential effects from a wider perspective, computing
the direct trade effects of the new US tariffs on the Spanish products affected, as well as the additional trade effects
induced in each Spanish sector by the trade effects generated in each country in the world. We are therefore provid-
ing an overall global assessment of the impact of the tariffs, as well as their specific impact on Spanish regions.
The US has a long history of protectionism in the steel sector (Read, 2005). Read (2005) explores the political
economy affecting the previous tariffs imposed by the Bush administration in 2001. He describes the main forces
determining the policy, as well as the potential effects of the measure, revising some key impact evaluation reports,
both ex ante (Francois & Baughman, 2001; US International Trade Commission, 2001) and ex post (Francois &
Baughman, 2003; US International Trade Commission, 2003a, 2003b). All these analyses illustrate how difficult it
is to predict the final effect, given the number of inter‐sectoral reactions, retaliations and competitive effects.

1
By the time when our simulation was conducted, it was not clear whether Canada and Mexico will be affected or not by the new tariffs. Initially, both were
exempt. Then, the exemptions were eliminated, and Canada joined the EU announcing retaliation measures. However, given the fact that NAFTA agreement
was successfully renegotiated in the summer of 2018 (and it is now called USMCA), it is still unclear to what extent and for how long the tariffs will hold.
That is why we decided to exclude Canada and Mexico from the list of countries affected by the new tariffs, something that fits with the main exempted
countries in the previous tariff rise implemented by president Bush in 2001 (Read, 2005).
 17577802, 2019, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/rsp3.12198 by Benemerita Universidad Autonoma De Puebla, Wiley Online Library on [12/02/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
LLANO ET AL. 527

Also, from the methodological viewpoint, the discussion about the optimal procedures for an ex ante evaluation of
the potential effects of a new tariff is broad and inconclusive. Interesting reviews have been recently provided
(Piermartini & Teh, 2005; WTO‐UNCTAD, 2012), with the aim of orienting the analysis conducted both by
researchers and policy makers. In this regard, the WTO‐UNCTAD (2012) describes the virtues and limitations of par-
tial equilibrium (PE) approaches and computable general equilibrium frameworks (CGE). Regarding this point, the
WTO‐UNCTAD (2012) notes that the trade‐off underlying the choice between both approaches, highlighting the
advantages of the CGE on the accountability of inter‐market linkages, which is usually applicable for long‐term
analysis at the aggregate level, whereas a PE model is usually more suitable for short‐term and more disaggregated
policy experiments. WTO‐UNCTAD (2012) also reviews a list of powerful models, such as SMART, GSIM, TRIST or
ATPSM.
After considering all possible options, our empirical strategy combines different PE models able to compute a first
impact of the effect that the new US tariffs will have, both globally and within each Spanish region. This strategy
allows us to understand the causes and direction of the effects considered, although it lacks the inter‐sectoral com-
petitive effects covered in a CGE framework. However, the strategy places emphasis on the ease of use, replicability,
and transparency of the tools used, which integrated three powerful models able to address inter‐country‐inter‐
regional effects at the product level. The estimation process can be summarized in three steps:

• First, we simulate the effect that the US tariffs might generate worldwide at the country level, using the SMART
simulation model developed by the World Bank and UNCTAD.
• Then, the trade effects computed at the national level in the first step are plugged into the newest World Input–
Output Database (WIOD). The trade effect enters directly in the corresponding sector for the 43 countries that
have an individual IO table within the WIOD, while the others are aggregated in the “Rest of the World.” Thus, by
means of the WIOD, it is possible to compute the inter‐sectoral effects that the original trade effects generate in
the main economies in the World, excluding the US and Spain.
• Finally, all the previous effects are regionalized within Spain using a unique inter‐regional input–output table
(SIRIO), whose structure is very similar to the WIOD, but accounting for the inter‐sectoral relationships between
the 18 Spanish regions, and of those with the rest of the world. In this regard, two different effects are consid-
ered: first, the immediate trade effect of the US tariffs over the Spanish metal sector computed by the SMART
model in the first step (product effect); secondly, the global effect, quantified as the sectoral‐regional effects within
Spain generated by the initial impact of the US tariffs over the world (Spain excluded), as they are captured by
the WIOD.

According to the initial results obtained by SMART, the total world exports of “Iron and steel” and “aluminium” to
US would fall by more than US$8.4 trillion. The six countries (assumed to be) exempted from the new tariffs benefit
by an increase of their exports of almost US$2.0 trillion, with Canada the main winner. The rest of the countries
reduce their exports by more than US$10.0 trillion, with Russia the most hit.
Using the WIOD system, the trade effect computed by SMART would provoke a direct net job loss of about
72,500 employees in basic metal industries around the world; while the total effect, including direct and indirect out-
put impacts, would increase the total job losses to 185,000 employees, about 0.007% of world total employment.
In Spain, the new tariffs may produce a product effect over the Spanish export's worth of €150 million. The so
called global trade effect, computed by combining the SMART with the WIOD, would add an additional decrease of
€163 million. Then, once that the inter‐sectoral and inter‐regional linkages of our SIRIO model have been computed,
the final effect over the Spanish economy rises to €616 million, with a net job loss of up to 3,200 employees (0.02%
of national totals). Moreover, the induced effect through the reduction in household consumption, will add other €16
million of GDP and 255 additional jobs lost. In conclusion, the new tariff will generate a €673 million decrease in the
Spanish output and the loss of 3,500 jobs. The Basque Country will be the one that suffers the most, followed by
Catalonia and Madrid.
 17577802, 2019, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/rsp3.12198 by Benemerita Universidad Autonoma De Puebla, Wiley Online Library on [12/02/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
528 LLANO ET AL.

The structure of the rest of this paper is as follow: we first contextualize Trump's protectionism and its possible
impact on the multilateral trading system. Second, we review the economic rationale behind our analysis, revisiting
the trade literature on the effects of an increase in tariffs. Next, we describe the empirical strategy, divided in each
of the three models used. Finally, we analyse the results obtained at the country and regional level, highlighting the
inter‐sectoral‐inter‐regional effects produced in Spain with a final section providing some concluding remarks.

2 | T R U M P ' S TR A D E W A R S I N C O N T E X T

The protectionist spirit Trump promised during his electoral campaign finally moved into action in mid‐2018. The
Administration promoted reasons of national security to establish steel and aluminium tariffs, which the WTO allows
under its Article XXI, but only in exceptional circumstances that, in principle, do not correspond to the current sce-
nario (Bown & Keynes, 2018). The announcement was accompanied by an “incendiary tweet” in which Trump claimed
that “trade wars are good and easy to win,” ignoring both the codes of good diplomacy and the lessons of economic
history. With this decision, President Trump weakened still further the confidence of his allies and continues to
undermine the rules‐based multilateral trading system. As Posen (2018, p. 28) argues “U.S. President Donald Trump
has rejected the idea that the world's economies all benefit when they play by the rules … if the United States con-
tinues its retreat from economic leadership, it will impose serious pain on the rest of the world—and on itself.”
So far, the Trump Administration imposed tariffs on imported washing machines and solar panels in January 2018.
However, the steel and aluminium tariffs are different. On the one hand, they are accompanied by bellicose rhetoric
that undermines the confidence of allies. On the other hand, the decision makes clear that the US has no interest in
resolving trade issues through multilateral dialogue and that, from now on, it will adopt a more openly aggressive uni-
lateralism. A G20 initiative already underway to deal with the global excess steel capacity caused by China has been
derailed by these actions. Furthermore, in contrast with the tariffs on washing machines and solar panels (which were
temporary and declining over time), the invocation of the national security clause puts the WTO in a delicate situa-
tion. If the organization allows the national security justification and authorizes the tariffs, other countries could use
the same excuse for closing their markets and the world could enter a protectionist spiral like that of the 1930s
(Kindleberger, 1973). However, if the WTO does not authorize them, it will open the door for the US, the world's
largest economy, to leave the organization, dealing it a potential death blow. In any case, as the WTO decision will
take time; the protectionist spiral could accelerate while these deliberations proceed.
In fact, the EU has already announced its reaction. The EU opted to stand up to President Trump and to defend
the multilateral trade order, even at the cost of economic damage that now will be larger as a result of the likely tariff
escalation. The EU probably has for some time been preparing this position in the eventuality that the US Adminis-
tration should begin to explicitly undermine the multilateral system (Dadush, 2018).
The EU announced three actions: (i) to take the US to the dispute resolution body of the WTO; (ii) to impose com-
pensatory tariffs on specific US products (Harley‐Davidson Motorcycles, bourbon and Levi's jeans have been men-
tioned, but the list will be much longer, including 100 products worth €2.8 billion); and (iii) to set temporary tariffs
on steel and aluminium entering the EU, anticipating that some of the steel that no longer goes to the US will come
to Europe, with the resulting negative impact on domestic production. All of these measures are, in principle, compat-
ible with WTO regulations. In the June 2018 Trump‐Junker summit, the EU and the US agreed to pause tariff esca-
lation and to launch new negotiations to liberalize trade in manufactured goods, agriculture and energy. However, as
these negotiations proceed, the US has announced that it will study establishing new tariffs on European (and other
foreign) automobiles, also on a national security basis.
Events will continue to unfold slowly and their economic impacts will take time to materialize. Since tariff esca-
lation is a real possibility, it is useful to be ready to anticipate the potential effects of such external shocks that, in
the case of Spain, as in other EU countries, promise to compromise recovery from almost a decade of very high
and persistent unemployment rates in the less resilience regions.
 17577802, 2019, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/rsp3.12198 by Benemerita Universidad Autonoma De Puebla, Wiley Online Library on [12/02/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
LLANO ET AL. 529

3 | E C O N O M I C R A T I O N A L E BE H I N D OU R I M P A C T A N A L Y S I S

In this section we revisit the literature on import tariffs as described in Laird and Yeats (1986), Laird (1997), Jammes
and Olarreaga (2005), Piermartini and Teh (2005), Feenstra (2015) and WTO‐UNCTAD (2012). The trade literature
explains how the establishment (elimination) of a tariff over the imports of a given country can produce trade, reve-
nue and consumer effects, both within the importing and the exporting country (Feenstra, 2015). The results depend
to a large extent on the market structure, the penetration of the products in the importing country and the size of the
country applying the tariff.
As said before, our empirical strategy is three‐fold, and has been summarized in Figure 1.
In the first step, we use the SMART simulation model, which allows us to cover all these effects for a sudden
increase of the two tariffs considered, using the existing values of trade between the US and almost all countries
in the world. Further details are given in the next section.
As described by Laird and Yeats (1986), the trade loss (creation) effect captures the trade reduction (expansion)
induced by an increase in the tariffs that leads to the displacement of efficient (inefficient) producers that were deliv-
ering their less‐expensive products in the importing country (the US in our case). Such imports will be substituted by
less‐efficient domestic products. SMART assumes that there is full transmission of price changes when tariff distor-
tions are introduced (eliminated). The exact expressions for this effect are included in Laird and Yeats (1986) and
others (Jammes & Olarreaga, 2005; Khorana, Kimbugwe, & Perdikis, 2009; Sadni Jallab, Lahsen Abdelmalki, &
Sandretto , 2007), which considers the import demand and export supply functions for any given pair of countries
for the product subject to the tariff.
Thus, trade loss effect (creation) depends on the current level of imports of US in the two products under consid-
eration (HS72; HS76), their import demand elasticity and the relative tariff change.
Then, SMART also computes the trade diversion effect, which can be defined as the substitution of goods coming
from one set of foreign suppliers for goods from another set of foreign suppliers. This results from the changes in the
relative import prices (including the tariff) of goods from the different sets of foreign suppliers because of changes in
the differential in the rates. In contrast, the trade diversion effect can expand or contract trade globally. Again, follow-
ing Laird and Yeats (1986) and others (Jammes & Olarreaga, 2005, Khorana et al., 2009, Sadni Jallab et al., 2007), the
diversion effects depends on the elasticity of substitution, which can be expressed as the percentage change in rel-
ative shares of imports from two different countries due to a 1% change in the relative prices of the same product

FIGURE 1 Scheme describing the empirical strategy

Source: Own elaboration based on SMART simulation.
 17577802, 2019, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/rsp3.12198 by Benemerita Universidad Autonoma De Puebla, Wiley Online Library on [12/02/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
530 LLANO ET AL.

TABLE 1 Spanish exports to the US of the products potentially affected by the new tariffs. 2017

Taric‐2 digit items Tons Thousands of € % of US consumption.

72 iron and steel 275,680 201 0.3%

76 aluminium and articles thereof 36,401 91 0.7%
Total 332,081 292 0.5%

Source: Own elaboration using data from: Spanish exports (DataComex). US demand: Worldsteel.org and U.S. Geological Sur-
vey, Mineral Commodity Summaries.

from them. Trade diversion then depends on the current level of imports from the US and ROW, the percentage of
increase of tariffs facing US imports with those remaining unchanged (the six countries exempted) and the elasticity
of substitution of the imports from the two sources. The higher the value of the elasticity of substitution, the greater
will be the trade diversion effects.
Adding the trade creation and diversion derives the total effect on trade, which is the one that we will use for com-
puting the global trade effect and the product trade effect entering the Spanish economy in the third step.2
Departing from the trade effect estimated by SMART, we compute in step 2 a country level analysis, plugging the
SMART trade effect vector (without Spain) into the inter‐country input–output model defined by the WIOD data-
base. By doing so, we obtain an impact vector on the Spanish economy, which is one of the 44 countries included
expressly within the WIOD. The rest of the analysis will be the standard one in any input–output impact analysis,
where the driver of shocks will be channelled through the intermediate demand at the country level (WIOD) and
at the regional level within Spain (SIRIO). More technical details are covered in subsection 5.4.
Thus, the final effect computed for each sector and region within Spain, considers two sources of effects (product
and global trade effects), which lead to total, direct, indirect and induced components, in the way they have been tra-
ditionally considered in the input–output literature.

4 | DESCRIPTIVE ANALYSIS

We start this section by showing, in Table 1, the main data relating to total Spanish exports to US in 2017. The vol-
ume of the total exports from Spain to the US in 2017 reached almost 300 million tons, with a value of €332 million.
Using the estimated US demand for these three products, Table 1 also shows how the Spanish exports in 2017 rep-
resented around 0.5% of apparent US consumption of these products, with aluminium having the highest relative
presence, with about 0.7% of this US demand. These data illustrate that although the intensity of Spanish exports
of these products to US is relatively high, they account for a very small share in the whole US demand.
Now, the focus of attention will be on the importance of the US market for each of the Spanish regions. Consid-
ering the aggregate Spanish exports of these products, the US market accounts for around 4% of the Spanish total.
As shown in Figure 2, within this global amount, some regions such as the Basque Country, La Rioja, or the
Comunidad Valenciana, accumulate shares of their total exports that are over 6%.
Moreover, it is interesting to analyse the concentration of total exports of these products by regions, knowing the
regional relevance of the potentially affected products. This share will be very much by the productive specialization
of each region. As reported in the two panels of Figure 3, for “iron and steel products,” the Basque Country, Galicia
and Cantabria, absorb 75% of the total Spanish exports to the US; while in “aluminium,” the origin would be more
territorially distributed, with more prominent contributions from the Basque Country and Madrid.

2
SMART also delivers the price, revenue and welfare effects from a variation in the tariffs in a given country or set of products. In our case, we focus just on
the trade effects generated by the tariffs in the ROW, avoiding the discussion about these effects.
 17577802, 2019, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/rsp3.12198 by Benemerita Universidad Autonoma De Puebla, Wiley Online Library on [12/02/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
LLANO ET AL. 531

FIGURE 2 US relevance on the Spanish exports of HS 72 and 76 chapters by region. Percentage over each region
export total to the US of these products
Source: Own elaboration using data from DataComex.

5 | E M P I R I C A L ST R A T E G Y

Given this overview, we now proceed to estimate the impact of the new tariffs over the Spanish economy describing
in detail the three steps mentioned before:

1. Computing the global trade effect of the tariffs using the SMART‐WITS simulation model.

2. Obtaining the inter‐sectoral effects by country using the WIOD.

3. Obtaining the inter‐sectoral and inter‐regional effects using the Spanish inter‐regional input–output tables
(SIRIO).

5.1 | Simulating the global effect of the tariffs using theSMART‐WITS model

The empirical strategy starts by computing the global trade effect of the new tariffs using a partial equilibrium model
—SMART—developed by the WTO and the UNCTAD, included in the World Integrated Trade Solution (WITS, 2019).
This framework facilitates the development of the simulation, using the most recent data on inter‐national flows by
country and products, as well as the updated Trade Analysis Information System ( TRAINS) dataset about tariffs appli-
cable to each specific delivery (2017). The theoretical and practical underpinnings of the model are described by
Jammes and Olarreaga (2005) and Laird and Yeats (1986). The tool makes possible the determination of the range
of products subject to the tariff, computing the total trade effects in the exporting countries affected by the tariffs
(total, creation and deviation effects). It also computes the welfare effects in the importing economy that is applying
the tariff variation (the US in our case). The model has been used in several analysis such as: Khorana et al. (2009)
Sadni Jallab et al. (2007), Gómez‐Abella, Pereira‐Villa, and Gaitán‐Guerrero (2013), Gaalya Micah (2015),
Makochekanwa (2014), Tian and Yu (2014), Kabir and Salim (2011), Munemo (2013), Pereira Villa, Gómez Abella,
and Omar Herrera (2012). In the latter case, for example, an analysis has been conducted integrating results using
the SMART partial equilibrium model with another widely used general equilibrium alternative (GTAP), revealing
how one approach complements the other.
 17577802, 2019, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/rsp3.12198 by Benemerita Universidad Autonoma De Puebla, Wiley Online Library on [12/02/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
532 LLANO ET AL.

FIGURE 3 Spatial concentration of the Spanish exports to US of HS 72&76 chapters. Percentage over national
total exports to US of the three potentially affected products
Source: Own elaboration using data from DataComex.

Following this literature, two main scenarios have been applied using SMART: the imposition of a 25% new tariff
rate to all products included in the rubric HS‐72 “iron and Steel” and a 10% new tariff rate to the HS76 “aluminium.”
All countries are considered as potential economies affected by the tariff, except for Argentina, Australia, Brazil, Can-
ada, México and South Korea.
Regarding the technical aspects of the simulation:

• Import demand elasticity values used by default in SMART are the same for all the countries affected by the new
tariff but may vary by product. Given that the product‐mix within the HS72 and HS76 sectors is different across
countries, our simulation will have different import demand elasticities by country at that level of aggregation.
 17577802, 2019, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/rsp3.12198 by Benemerita Universidad Autonoma De Puebla, Wiley Online Library on [12/02/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
LLANO ET AL. 533

• By default, the export supply elasticity in SMART is set as 99 for infinite elasticity for all products and partners, a
value that is supported by trade theory for small countries. Given the large size of the US and the ROW, we set a
demand elasticity below infinity (75).
• Regarding the substitution elasticity, we assume the value set by default in SMART (1.5), which is slightly higher
than the one we previously obtained (around 1), using a time series analysis, in a previous version of this paper
(available under request).

The results obtained are reasonable considering previous impact analysis for very similar cases (Read, 2005), but
are subject to the corresponding criticism of any PE simulation.

5.2 | Obtaining the inter‐sectoral effects by country using the WIOD

The output generated by the SMART simulation is a T HS72;HS76

c→US column vector with dimension (131 × 1) containing the
changes in total exports of each of the c = 131 countries affected by the new tariffs. The effect on Spain is removed
at this step, so it can be entered directly into the SRIO model in a further stage. The vector of this trade effects over
the HS72 and HS76 for the rest of the countries are included as an export shock in the WIOD.3
WIOD offers tables from 2000 to 2014. Our analysis uses the latest one, 2014, which includes detailed IO
relationships for 28 EU countries, plus 15 additional large economies and a rest of the world (ROW) aggregated
for the remaining countries. The 2014 WIOD considers 56 sectors. Since the employment data is only available
for 35 sectors, we aggregate the original WIOD to this number of sectors (the details are reported in the
Appendix).
Employment data for the 43 countries included in the original table are obtained from the WIOD Database, while
the data for the ROW aggregated was prepared by Arto et al. (2018).

6 | OBTAINING THE INTER‐SECTORAL AND INTER‐REGIONAL EFFECTS

WITHIN SPA I N U SI NG S I RIO

In this final step, using an interregional input–output table for Spain (SIRIO), we estimate the inter‐regional inter‐
sectoral effects induced by the trade effects computed in the previous two sections. The origin of the SIRIO dates
to the late 1990s when the Institute L.R. Klein, addressed a project to create Inter‐regional input–output tables of
the Spanish economy, funded by the Ministry of Economy and Finance. The first table was built for 1995, which
was then updated for 2001, 2004, 2007 and 2010. The methodological details can be found in Pérez, Dones, and
Llano (2009); García and Pérez (2005) and Llano (2004a, 2004b, 2009). Hence, a full set of interregional trade data
integrated within an interregional system has been produced by the C‐Intereg Project (www.c‐intereg.es) for each
branch of activity (Llano, 2004a; Llano et al. 2017; Pérez, Dones, & Llano, 2009; Pérez, Llano, & García, 2009). The
most recent SIRIO table is used for 2010, and includes 21 activity branches for the 18 Spanish regions.
We now proceed to obtain the inter‐sectoral inter‐regional effects within the Spanish economy, considering two
possible shocks:

1. First, we compute the regional spillovers caused in the Spanish economy by the total trade effect in the “basic
metals sector,” obtained from the SMART simulation. Note that this shock was removed from the vector plugged

3
Full references and databases can be found in the WIOD project website (www.wiod.org/home) The methodological details can be found in Dietzenbacher
et al. (2013), Timmer, Dietzenbacher, Los, Stehrer, and de Vries (2015) and Timmer et al. (2016).
 17577802, 2019, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/rsp3.12198 by Benemerita Universidad Autonoma De Puebla, Wiley Online Library on [12/02/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
534 LLANO ET AL.

into the WIOD in the second step. Thus, we can now compute the total output, employment and GVA reduction
generated by this product trade effect on the Spanish metal sector, and the sectoral and regional backward
effects as computed from the input–output model.
2. Thereafter, we compute the additional shock affecting the entire Spanish economy generated the inter‐sectoral
effects obtained from the WIOD in the second step, providing a vector of sectoral effects on the Spanish econ-
omy. We label this additional shock as the global trade effect, since the shock is coming from all the 131 countries
(US and Spain excluded) considered in the first step, once they have been plugged into a WIOD system with 44
countries and 35 branches.

Note that in step 2, the 131‐country vector have been aggregated to fit into the WIOD 43 + ROW spatial disag-
gregation; while in step 3, the Spanish aggregated effect obtained with the WIOD has been regionally distributed
using the regional structure of exports for the specific product (iron‐steel and aluminium) in 2017. For the so called
global trade effect, the regional output structure in 2010 is used to translate the 35 sector effects estimated for the
Spanish economy through the WIOD, into the 21 sectors detailed for each region in the SIRIO.

7 | T E C H N I C A L N O T A T I O N F O R E S T I M A T I N G T H E SH O C K S I N TH E W I O D
AND T HE SIRIO

Both, the WIOD and the SIRIO, are presented in an interregional input–output framework (IRIO) originally defined by
Isard (1951), where the spatial and sectoral origin and destination of the final and intermediate demand are captured.
Following Miller and Blair (2009), an IRIO model with N regions and n sectors is denoted as in Equation (2):

2 3 2 11 3 2 3 2 13
x1 A A12 ⋯ AN1 x1 f
6 2 7 6 21 7 6 7 6 7
6x 7 6A A22 ⋯ AN2 7 6 x2 7 6 f2 7
6 7¼6 7x6 7 þ 6 7 (2)
6 ⋮ 7 6 7 6 7 6 7
4 5 4 ⋮ ⋮ ⋱ ⋮ 5 4 ⋮ 5 4 ⋮ 5
xN AN1 AN2 ⋯ A NN
xN fN

For A we have a block matrix containing the intraregional input coefficients in the on‐diagonal elements, and the
interregional input coefficients in the off‐diagonal ones. The A matrix contains elements ars
ij , which accounts for the

input requirements produced by a sector i in region r needed to produce an additional unit of sector j in region s.
These coefficients can be computed row‐wise based on a similar Z matrix, containing the intermediate inputs pro-
duced by a sector i in region r consumed by sector j in region s (zrs
ij ).

The solution of the model is obtained following the Equation (3), where the intra and interregional coefficient
matrix (A), the corresponding Leontief inverse matrix (I‐A)−1, the output vector (x), and the final demand vector (f),
now contain sectoral and regional (country in the case of the WIOD) data organized by blocks:

23 02 3 2 11 31−1 2 3
x1 I ⋯ 0 A ⋯ A1N f1
−16 7 B6 7 6 7C 6 7
x ¼ ðI−AÞ f 4 ⋮ 5 ¼ @4 ⋮ ⋱ ⋮ 5−4 ⋮ ⋱ ⋮ 5A ×4 ⋮ 5 (3)
N
xN 0 ⋯ I AN1 ⋯ A NN
f

The estimation of direct and indirect effects induced by the increases of final demand was calculated applying
Equation (3). By means of Equation (4) to the impact vectors Δf, a Δx vector of total output effect is obtained:

Δx ¼ ðI−AÞ−1 Δf (4)
 17577802, 2019, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/rsp3.12198 by Benemerita Universidad Autonoma De Puebla, Wiley Online Library on [12/02/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
LLANO ET AL. 535

Then, the indirect effects Δo are obtained as follows, subtracting direct from total effects:

Δo ¼ Δx − Δf (5)

This procedure allows to distinguish between direct and indirect output effects captured by the region that orig-
inally suffered the export reduction and each of the remaining regions in the country. Further indirect effects are
obtained by making the final demand interregional spillover effects endogenous. Similarly, Equation (6) provides
the mechanism to compute direct and indirect effects in terms of employment Δe:4

Δe ¼ Δb eðI−AÞ−1 Δf
ex ¼ b (6)

8 | RESULTS

Drawing on the initial results obtained from SMART, the total world exports of Iron/Steel and Aluminium to the US
would fall by more than US$8,400 billion, which almost 80% is iron and steel and the remaining 20% is aluminium.
Looking at the country specific results, the six countries exempted from the new tariffs, would benefit by an
increase of their exports of almost US$1,954 billion of, with Canada the main winner with nearly 50% of the total.
In contrast, as shown in Figure 4, other countries would reduce their exports collectively by more than US$10,386
billion of, with the Russian Federation the most damaged country, followed by Japan, Germany Turkey and China.
Using the employment coefficients from the WIOD, these output reductions induced by the reduction of exports
to USA would provoke a direct net job loss of about 72,500 employees in basic metal industries around the world;
the total effect, including direct and indirect impacts, would increase the total job losses to more than 185,000
employees, about 0.007% of world total employment.
As Figure 5 reports, the total impacts in terms of total countries' employment, ranges from a positive effect of
0.03% in Canada, to negative effects below −0.05% in The Netherlands, Sweden, Portugal and Luxembourg.
Focusing on Spain, the results derived from SMART (product effect) show that the new tariffs may produce a
direct impact on Spanish exports worth of €150 million, of which 86% would be concentrated in the steel industry.

FIGURE 4 Main impacts on exports of Iron/Steel and Aluminium to USA as computed by SMART simulation tool
(billions of US$)
Source: Own elaboration based on SMART simulation.
 17577802, 2019, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/rsp3.12198 by Benemerita Universidad Autonoma De Puebla, Wiley Online Library on [12/02/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
536 LLANO ET AL.

FIGURE 5 Total effects in labour markets computed by plugging SMART output into the WIOD
Note: None positive impact has been considered for USA, neither any direct negative impact are included in Spain, so
in both cases only spillovers effects are computed.
Source: Own elaboration.

The Basque Country would be the most affected Spanish region. As Figure 6 shows, the initial impact of tariffs would
translate into a reduction of €67.2 million in exports from the Basque Country, in Galicia (–€22.1 million), in Cantabria
(–€14.6 million) and in Madrid (–€12.5 million). The least affected regions would be Murcia (–€0.11 million), La Rioja
(–€0.13 million), Canary Islands (–€0.17 million) and Extremadura (–€0.18 million); the Balearics Islands and Ceuta
&Melilla with almost no effects.
In addition, the so called global trade effect, computed by combining the SMART with the WIOD, would add an
additional shock to the Spanish output of €163 million, whose regional and sectoral distribution are show in
Figure 7, panels (a) and (b) respectively. In Figure 7 panel (a), the most affected regions would be Catalonia, with –
€30.7 million, followed by the Basque Country, Madrid and Andalusia with losses of more than €20 million each. Sec-
tor wise, panel (b) shows how the ‘manufacture of basic metals’ (–€56.2 millions) and ‘mining, quarrying and energetic
industries’ (–€34.0 millions) are affected the most.
Then, once that the inter‐sectoral and inter‐regional linkages of our SIRIO model are considered, the €150 million
initial reduction in exports could lead to a total reduction of national output of about €331 million, while the initial

FIGURE 6 Estimated direct ‘product effect’ of the tariffs on the Spanish regions (Millions of €)
Source: Own elaboration based on the SIRIO model for the product effect.
 17577802, 2019, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/rsp3.12198 by Benemerita Universidad Autonoma De Puebla, Wiley Online Library on [12/02/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
LLANO ET AL. 537

FIGURE 7 Estimated global trade effect in the Spanish economy (Millions of €)

Source: Own elaboration based on the SIRIO model when computing the global trade effect.

impact of €163 million derived from the global trade effect increases to €306 million. The reduction in national out-
put would mean a decrease in total GDP of more than €205 million and a net job loss about 3,200 employees,
approximately 0.02% of the national total.

TABLE 2 Regional distribution of total effects by source (millions of €)

Product effect Global trade effect Induced

Direct Indirect Direct Indirect Effect Total %

Basque Country −67 −56 −23 −17 −5 −169 25.1%

Catalonia −10 −24 −31 −29 −7 −101 15.0%
Community of Madrid −12 −27 −21 −23 −8 −90 13.4%
Andalusia −2 −6 −20 −18 −3 −49 7.3%
Galicia −22 −12 −8 −5 −2 −49 7.2%
Cantabria −15 −13 −3 −3 −1 −35 5.2%
Asturias −10 −12 −6 −5 −1 −34 5.1%
Valencian Community −1 −6 −12 −11 −2 −32 4.7%
Aragon −2 −7 −9 −8 −1 −26 3.9%
Castile and León −1 −6 −7 −5 −2 −21 3.2%
Navarre −4 −5 −4 −4 −1 −19 2.8%
Castilla‐La Mancha −2 −3 −6 −4 −1 −16 2.4%
Region of Murcia 0 −1 −3 −3 −1 −8 1.3%
Canary Islands 0 −1 −3 −3 0 −7 1.1%
Balearic Islands 0 −1 −2 −3 −1 −7 1.1%
La Rioja 0 −2 −1 −1 −1 −5 0.8%
Extremadura 0 −1 −2 −1 0 −4 0.6%
Ceuta & Melilla 0 0 0 0 0 0 0.0%
Total −150 −181 −163 −143 −36 −673 100.0%
% 22.3% 26.9% 24.3% 21.2% 5.3% 100%

Source: Own elaboration.

 17577802, 2019, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/rsp3.12198 by Benemerita Universidad Autonoma De Puebla, Wiley Online Library on [12/02/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
538 LLANO ET AL.

Additionally, the employment reduction induces an additional adjustment by the income and the household's con-
sumption, which leads to and additional output reduction, which can also be estimated in the SRIO model. This
income effect, named induced impact, worth of €16 million of GDP and 255 additional jobs lost.
It is possible to split the total impact into five different sources (direct product impact, indirect product impact,
direct global trade impact, indirect global trade impact and induced impact), whose regional distribution in terms of
total output is shown in Table 1.
As it is shown in Table 2, the main source of impacts is the indirect impact of product effects that accounts for the
26.9% of the €673 million reduction of Spanish output, followed by direct impact of global trade effects 24.3%, direct
impact of product effects 22.3%, and indirect impact of global trade effects 21.3%. The spatial distribution of the
shock is uneven, with the Basque Country, which represents the 6.1% of total GDP, absorbing more than one quarter
of the total impact. In the same way, but with less intensity, regions like Cantabria (1.1% total GDP), Asturias (2.0%),
Galicia (5.2%) and Navarre (1.7%), show relative impacts below their GDP shares. In contrast, other large communities
as Catalonia (19.2%), Madrid (18.9%), Andalusia (13.4%) or Valencian Community (9.3%) experience a total output
shocks that are relatively smaller than their weights in the national GDP.
In fact, looking at the results shown in Figure 8, the total impacts in terms of GDP related to regional totals range
from 0.070% in Cantabria, or 0.058% in the Basque Country, the two regions where the metallurgic industry is part of
the core of the entire regional economic structure, to less than 0.01% in Ceuta & Melilla, Canary Islands, or
Extremadura.
This reduction in production and income would also affect the labour market, which could lose more than 640
direct jobs due to the ‘Product effect’, and another 807 because of global trade effect. These amounts would be
increased by the indirect effects (intermediate demand linkages) 967 and 844 for each effect respectively, plus the
induced effect (final demand linkages), that accounts for another 255 jobs. Adding all the effects together, the total

FIGURE 8 Total output effect measured in relative terms to the regional GDP
Source: Own elaboration.
 17577802, 2019, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/rsp3.12198 by Benemerita Universidad Autonoma De Puebla, Wiley Online Library on [12/02/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
LLANO ET AL. 539

FIGURE 9 Total impact on employment by regions using the SIRIO model

Source: Own elaboration.

impact would be more than 3,500 jobs in Spain. By region, the Basque Country will be the one that suffers the most,
with around 318 direct jobs losses (product + global effects), and 359 indirect and induced ones. Catalonia would be
ranked second, with 219 direct and 373 indirect plus induced full‐time equivalent job losses, while the Madrid region
will lose 172 and 371 jobs respectively (Figure 9).
Table 3 illustrates how the initial effects allocated into the exporting activity in the affected sector (metallurgy and
metallic products) in each region, will spread to other sectors, causing job losses in the whole Spanish economy. The

TABLE 3 Ranking of the 20 main indirect + induced effects in terms of job losses (%)

Source: Own elaboration.

 17577802, 2019, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/rsp3.12198 by Benemerita Universidad Autonoma De Puebla, Wiley Online Library on [12/02/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
540 LLANO ET AL.

key here will be the regional and sectoral linkages within the intermediate (indirect) and final demand (induced)
included in the SIRIO model. Interestingly, although the first largest [indirect + induced] impacts takes place in the
Manufacture of basic metals itself in the Basque Country (6.9% of overall indirect and induced effects in terms of
job losses), the following largest effects occur in other sectors. For example, large effects are obtained in the other
professional and personal services in Madrid (5.0%), Catalonia (3.6%) and Basque Country (2.9%) or wholesale and retail
trade activities in Catalonia (3.8%) or Madrid (3.7%). Also, big indirect + induced effects are obtained in transportation
services and other manufacturing activities.
It is important to view these results in the light of the emerging literature on the resilience of the EU regions with
respect to the Great Depression (see, for example, Crescenzi, Luca, & Milio, 2016; Cuadrado‐Roura, Martin, &
Rodríguez‐Pose, 2016 and Fratesi & Rodríguez‐Pose, 2016), and more particularly in Spain (Cuadrado‐Roura &
Maroto, 2016). According to these authors, the Spanish regions that have recovered better from the Great Depres-
sion are those that had previously specialized in more dynamic industries and sectors. Thus, a small group of resilient
regions, with a more diversified sectoral structure and which specialized in manufacturing activities, energy and some
market services, were able to better exploit their higher productivity potentials during the period after the crisis. In
contrast, the less‐resilient regions that were specialized in construction, extracting activities and primary industries,
are the ones that suffered the most from the impact of the Crisis and are now facing the more significant problems
in the recovery. All this has manifested itself in a process of widening the regional productivity gap.
In this context, Table 4 shows how the largest total effects accumulate in Basque Country (19.3%), Catalonia
(16.9%) and Madrid (15.5%), whose unemployment rates in 2018 are 10.0%, 11.5% and 12.2% respectively, the low-
est (!) in the Spanish context. These three rich regions, which have been described as the most resilient Spanish
regions, accumulate more than 50% of the overall effect, given their capacity to attract spillovers (in this case, nega-
tive) from direct effects generated in the rest of the country. However, if we focus on the list of the next nine regions
in the ranking, which accumulates around 42% of the total effect, the situation is quite different. For example, Galicia,
which accounts for 8% of the total effects of the US tariffs, has an unemployment rate of 13.3%, the Valencian Com-
munity, with 6.5% of the total effects, has currently a 15.6% unemployment rate. More dramatically, Andalusia, which
might account for the 7.1% of the job losses due to US tariffs, has a 23% unemployment rate.

TABLE 4 Total effects of US tariffs vs unemployment rates

 17577802, 2019, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/rsp3.12198 by Benemerita Universidad Autonoma De Puebla, Wiley Online Library on [12/02/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
LLANO ET AL. 541

9 | C O N CL U S I O N S

The aim of this paper is to offer a first estimate of the potential effect that the recent import tariffs on iron, steel and
aluminium imposed by the Trump administration may have in Spain. This protectionist measure was justified based
upon national security reasons. Our estimates end up suggesting a non‐negligible effect on some of the most indus-
trialized regions in Spain, such as the Basque Country, Catalonia or Madrid, in terms of exports, GDP and job losses.
The effects obtained across regions should be evaluated along with the high and persistent unemployment rates reg-
istered in some of the Spanish regions during and after the 2009–2014 double dip recession. Moreover, if tariff esca-
lation continues and affect other products such as automobiles, and the US‐China trade war reduces economic
growth globally, the impact on the Spanish economy could be more significant.
Our empirical strategy combines three different powerful tools. First, we depart form the SMART partial equilib-
rium model developed by the World Bank and UNCTAD. Next, we plug the trade effects on the two sectors affected
into the WIOD, to obtain a global trade effect over every sector in the Spanish economy. Finally, the product effect
and the global trade effect are introduced in a unique input–output table for the Spanish Economy (SIRIO), obtaining
the total (direct, indirect and induced effects) impacts in each region.
In an economy of global value chains, collateral effects of any unilateral action might be important. Indirect effects
of the protectionist measures could come from the spillovers generated in third countries such as China. Given the
high level of interdependence of the global economy, it is important to develop methodologies able to offer quick
estimates of potential spillovers and feedback loops affecting a given sector or region. Our exercise illustrates how
an isolated measure affecting two specific chapters in the HS classification can generate negative effects across
regions and sectors in a country. Interesting examples are found in Di Comite and Potters (2014) and Zhang
(2016). In this line of research, our analysis could be extended in order to better embed our interregional framework
in the WIOD, as well as to develop a multi‐scale CGE to capture the full system‐wide effects. The recent contribution
of Chen et al., 2018 highlighted the asymmetric impacts of Brexit on the regional economies of Europe and revealed
only modest impact on Spain. However, the adverse economic consequences of a US‐China and or US‐EU trade war
could be much more significant than those of Brexit and would require a more detailed analysis.

ACKNOWLEDGEMEN TS
This paper was developed in the context of two research projects: The C‐intereg Project (www.c‐intereg.es) and
ECO2016‐79650‐P from the Spanish Ministry of Economics and Innovation.

RE FE R ENC E S
Arto, I., Rueda‐Cantuche, J. M., Cazcarro, I., Amores, A. F., Dietzenbacher, E., & Kutlina‐Dimitrova, Z. (2018). EU exports to the
world: Effects on employment. Luxembourg: Publications Office of the European Union.
Bown, C. & Keynes S. (2018). Trade talks, 24, podcast of the Peterson Institute for International Economics. https://piie.com/
experts/peterson‐perspectives/trade‐talks‐episode‐24‐trump‐administration‐views‐trade‐national
Chen, W., Los, B., McCann, P., Ortega‐Argilés, R., Thissen, M., & van Oort, F. (2018). The continental divide? Economic expo-
sure to Brexit in regions and countries on both sides of The Channel. Papers in Regional Science, 97, 25–54. https://doi.
org/10.1111/pirs.12334
Crescenzi, R., Luca, D., & Milio, S. (2016). The geography of the economic crisis in Europe: national macroeconomic condi-
tions, regional structural factors and short‐term economic performance. Cambridge Journal of Regions, Economy and
Society, 9(1), 13–32. https://doi.org/10.1093/cjres/rsv031
Cuadrado‐Roura, J. R., & Maroto, A. (2016). Unbalanced regional resilience to the economic crisis in Spain: a tale of special-
isation and productivity. Cambridge Journal of Regions, Economy and Society, 9(1), 153–178. https://doi.org/10.1093/
cjres/rsv034
Cuadrado‐Roura, J. R., Martin, R., & Rodríguez‐Pose, A. (2016). The economic crisis in Europe: urban and regional conse-
quences. Cambridge Journal of Regions, Economy and Society, 9(1), 3–11. https://doi.org/10.1093/cjres/rsv036
Dadush, U. (2018). U.S. steel and aluminum tariffs: How should the EU respond? Bruegel blog, 9 March. URL: http://bruegel.
org/2018/03/u‐s‐steel‐and‐aluminum‐tariffs‐how‐should‐the‐eu‐respond/
 17577802, 2019, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/rsp3.12198 by Benemerita Universidad Autonoma De Puebla, Wiley Online Library on [12/02/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
542 LLANO ET AL.

Di Comite, F., & Potters, L. (2014). Modelling knowledge creation, investment decisions and economic growth in a spatial
CGE setting. Investigaciones Regionales, 29, 47–75.
Dietzenbacher, E., Los, B., Stehrer, R., Timmer, M. P., & de Vries, G. J. (2013). The construction of World Input‐Output Tables
in the WIOD Project. Economic Systems Research, 25, 71–98. https://doi.org/10.1080/09535314.2012.761180
Feenstra (2015). Advanced international trade: Theory and evidence (2nd ed.). Long Island City, NY: Princeton University Press.
Francois, J., & Baughman, L. M. (2001). Estimated economic effects of proposed import relief remedies for steel. Washington,
DC: CITAC Foundation.
Francois, J., & Baughman, L. M. (2003). The unintended consequences of US steel import tariffs: A quantification of the impact
during 2002. Washington DC: CITAC Foundation.
Fratesi, U., & Rodríguez‐Pose, A. (2016). The crisis and regional employment in Europe: What role for sheltered economies?
Cambridge Journal of Regions, Economy and Society, 9(1), 33–57. https://doi.org/10.1093/cjres/rsv032
Gaalya Micah, S. (2015). Uganda's trade and revenue effects with the EAC countries, DRC and Sudan. Modern Economy, 6,
338–357.
García, G., & Pérez, J. (2005). Ports economic impact. Literature review and alternative proposal. Journal of Maritime Research,
I(3), 85–104.
Gómez‐Abella, D., Pereira‐Villa, C. M., & Gaitán‐Guerrero, L. A. (2013). Impacto de un TLC en el comercio entre Colombia y
Turquía. Cuadernos de Administración, 26(4), 159–199.
Isard, W. (1951). Interregional and regional input‐output analysis: A model of space economy. Review of Economics and Sta-
tistics, 33, 318–328. https://doi.org/10.2307/1926459
Jammes, O., & Olarreaga, M. (2005). Explaining SMART and GSIM. Washington D.C.: The World Bank.
Kabir, M., & Salim, R. (2011). Analysing potential effects of preferential liberalisation in some Asian emerging economies.
International Economic Journal, 25(2), 191–213. https://doi.org/10.1080/10168737.2011.586805
Khorana, S., Kimbugwe, K., & Perdikis, N. (2009). Assessing the welfare effects of the East African Community Customs
Union's transition arrangements on Uganda. Journal of Economic Integration, 24(4), 685–708. https://doi.org/
10.11130/jei.2009.24.4.685
Kindleberger, C. (1973). The world in depression, 1929–1939. Berkeley: University of California Press.
Laird, S. (1997). Quantifying commercial policies. In J. Francois, & K. Reinert (Eds.), Applied methods for trade policy analysis: A
handbook. Cambridge, UK: Cambridge University Press. https://doi.org/10.1017/CBO9781139174824.004
Laird S. & Yeats, A. (1986). The UNCTAD trade policy simulation model: A note on the methodology, data and uses.
UNCTAD Discussion Paper No. 19, Geneva.
Llano, C. (2004a). Economía espacial y sectorial: El comercio interregional en el contexto de un modelo Multirregional para la
economía española. Instituto de Estudios Fiscales. Ministerio de Economía y Hacienda. Colección Investigaciones:
N°1/04.
Llano, C. (2004b). The interregional trade in the context of a multiregional input‐output model for Spain. Estudios de
Economía Aplicada, 22(3), 539–577.
Llano, C. (2009). Efectos de desbordamiento interregional en España: Una estimación a través del modelo input‐output inter-
regional. Investigaciones Regionales, 16, 181–188.
Llano, C., De la Mata, T., Diaz‐Lanchas, J., & Gallego, N. (2017). Transport‐mode competition in intra‐national trade: An
empirical investigation for the Spanish case. Transportation Research Part a, 95, 334–355.
Makochekanwa, A. (2014). Welfare implications of COMESA‐EAC‐SADC tripartite free trade area. African Development
Review, 26(1), 186–202. https://doi.org/10.1111/1467‐8268.12074
Miller, R. E., & Blair, P. (2009). Input‐output analysis: Foundation and extensions (2nd ed.). Cambridge, UK: Cambridge Univer-
sity Press. https://doi.org/10.1017/CBO9780511626982
Munemo, J. (2013). Trade between China and South Africa: Prospects of a successful SACU‐China free trade agreement.
African Development Review, 25(3), 303–329. https://doi.org/10.1111/j.1467‐8268.2013.12031.x
Pereira Villa, C., Gómez Abella, D., & Omar Herrera, L. (2012). The Colombia‐Canada Free Trade Area: A partial equilibrium
simulation. Semestre Económico, 15(3), 15–42. https://doi.org/10.22395/seec.v15n31a1
Pérez, J., Dones, M., & Llano, C. (2009). An interregional impact analysis of the EU Structural Funds in Spain (1995–1999).
Papers in Regional Science, 88(39), 509–529. https://doi.org/10.1111/j.1435‐5957.2008.00212.x
Pérez, J., Llano, C., & García, G. (2009). Valoración de las tablas Input‐output interregionales de la economía española. Revista
del ICE, 848, 37–65.
Piermartini, R., & Teh, R. (2005). Demystifying modelling methods for trade policy. Working Paper. WTO Discussion Paper,
No. 10. URL: http://hdl.handle.net/10419/107045
 17577802, 2019, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/rsp3.12198 by Benemerita Universidad Autonoma De Puebla, Wiley Online Library on [12/02/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
LLANO ET AL. 543

Posen, A. (2018). The post American world economy. Foreign Affairs, 97(2), 28–39.
Presidential Proclamation on Adjusting Imports of Steel into the United States. March 8, 2018. URL: https://www.
whitehouse.gov/presidential‐actions/presidential‐proclamationadjusting‐imports‐steel‐united‐states/?utm_source=link
Read, R. (2005). The political economy of trade protection: The determinants and welfare impact of the 2002 US emergency
steel safeguard measures. The World Economy, 28, 1119–1137. https://doi.org/10.1111/j.1467‐9701.2005.00722.x
Sadni Jallab, M., Lahsen Abdelmalki, L., & Sandretto, R. (2007). The free trade agreement between the United States and
Morocco: The importance of a gradual and assymetric agreement. Journal of Economic Integration, 22(4), 852–887.
https://doi.org/10.11130/jei.2007.22.4.852
Tian, W., & Yu, M. (2014). Measuring the impact of trade protection on industrial production size. Review of Development
Economics, 18(2), 231–253. https://doi.org/10.1111/rode.12081
Timmer, M. P., Dietzenbacher, E., Los, B., Stehrer, R., & de Vries, G. J. (2015). An illustrated user guide to the World Input–
Output Database: The case of global automotive production. Review of International Economics, 23, 575–605. https://doi.
org/10.1111/roie.12178
Timmer, M. P., Los, B., Stehrer, R. & de Vries, G. J. (2016). An anatomy of the global trade slowdown based on the WIOD
2016 release. GGDC research memorandum 162, University of Groningen.
US International Trade Commission (2001). Steel: Investigation no. TA‐201‐73, volumes I to III, publication 3479. Washington
DC: USITC.
US International Trade Commission (2003a). Steel: Monitoring developments in the domestic industry (TA‐204‐9), volumes I to II,
publication 3632. Washington DC: USITC.
US International Trade Commission (2003b). Steel‐consuming industries: Competitive conditions with respect to steel safeguard
measures, investigation no. 332–452. Washington DC: USITC.
USTR (2017). 2017 trade policy agenda. URL: https://ustr.gov/sites/default/files/files/reports/2017/AnnualReport/
AnnualReport2017.pdf
WITS (2019). World integrated trade solutions. URL: https://wits.worldbank.org/WITS/WITS/Restricted/Login.aspx
WTO‐UNCTAD, (2012). A practical guide to trade policy analysis. URL: https://www.wto.org/english/res_e/publications_e/
practical_guide12_e.htm
Zhang, W. (2016). Economic globalization and interregional agglomeration in a multi‐country and multi‐regional neoclassical
growth model. Investigaciones Regionales, 34, 95–121.

How to cite this article: Llano C, Pérez J, Steinberg F, Hewings GJD. Global and regional effects of the US
tariffs on iron, steel and aluminium: A SMART combination of models with a focus on Spain. Reg Sci Policy
Pract. 2019;11:525–547. https://doi.org/10.1111/rsp3.12198
 17577802, 2019, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/rsp3.12198 by Benemerita Universidad Autonoma De Puebla, Wiley Online Library on [12/02/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
544 LLANO ET AL.

APPENDIX

TABLE A1 WIOD structure

WIOD initial branches detail WIOD aggregated branches

A01 Crop and animal production, hunting and related r1 Crop and animal production, hunting and related
service activities service activities+ Forestry and logging+ Fishing
and aquaculture
A02 Forestry and logging
A03 Fishing and aquaculture
B Mining and quarrying r2 Mining and quarrying
C10‐C12 Manufacture of food products, beverages and r3 Manufacture of food products, beverages and
tobacco products tobacco products
C13‐C15 Manufacture of textiles, wearing apparel and r4 Manufacture of textiles, wearing apparel and
leather products leather products
C16 Manufacture of wood and of products of wood r5 Manufacture of wood and of products of wood
and cork, except furniture; manufacture of and cork, except furniture; manufacture of
articles of straw and plaiting materials articles of straw and plaiting materials
C17 Manufacture of paper and paper products r6 Manufacture of paper and paper products+
Printing and reproduction of recorded media
C18 Printing and reproduction of recorded media
C19 Manufacture of coke and refined petroleum r7 Manufacture of coke and refined petroleum
products products
C20 Manufacture of chemicals and chemical products r8 Manufacture of chemicals and chemical products
+Manufacture of basic pharmaceutical products
C21 Manufacture of basic pharmaceutical products and
and pharmaceutical preparations
pharmaceutical preparations
C22 Manufacture of rubber and plastic products r9 Manufacture of rubber and plastic products
C23 Manufacture of other non‐metallic mineral r10 Manufacture of other non‐metallic mineral
products products
C24 Manufacture of basic metals r11 Manufacture of basic metals
C25 Manufacture of fabricated metal products, except r12 Manufacture of fabricated metal products, except
machinery and equipment machinery and equipment+ Manufacture of
machinery and equipment n.e.c.
C26 Manufacture of computer, electronic and optical
products
C27 Manufacture of electrical equipment r13 Manufacture of computer, electronic and optical
products+ Manufacture of electrical equipment
C28 Manufacture of machinery and equipment n.e.c.
C29 Manufacture of motor vehicles, trailers and semi‐ r14 Manufacture of motor vehicles, trailers and semi‐
trailers trailers+ Manufacture of other transport
equipment
C30 Manufacture of other transport equipment
C31_C32 Manufacture of furniture; other manufacturing r15 Manufacture of furniture; other manufacturing+
Repair and installation of machinery and
C33 Repair and installation of machinery and
equipment
equipment
D35 Electricity, gas, steam and air conditioning supply r16 Electricity, gas, steam and air conditioning supply+
Water collection, treatment and supply+
E36 Water collection, treatment and supply
Sewerage; waste collection, treatment and
E37‐E39 Sewerage; waste collection, treatment and disposal activities; materials recovery;
disposal activities; materials recovery; remediation activities and other waste
remediation activities and other waste management services
management services

(Continues)
 17577802, 2019, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/rsp3.12198 by Benemerita Universidad Autonoma De Puebla, Wiley Online Library on [12/02/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
LLANO ET AL. 545

TABLE A1 (Continued)

WIOD initial branches detail WIOD aggregated branches

F Construction r17 Construction
G45 Wholesale and retail trade and repair of motor r18 Wholesale and retail trade and repair of motor
vehicles and motorcycles vehicles and motorcycles
G46 Wholesale trade, except of motor vehicles and r19 Wholesale trade, except of motor vehicles and
motorcycles motorcycles
G47 Retail trade, except of motor vehicles and r20 Retail trade, except of motor vehicles and
motorcycles motorcycles

TABLE A2 WIOD 2014: Sectors

WIOD initial branches detail WIOD aggregated branches

H49 Land transport and transport via pipelines r21 Land transport and transport via pipelines
H50 Water transport r22 Water transport
H51 Air transport r23 Air transport
H52 Warehousing and support activities for r24 Warehousing and support activities for
transportation transportation
H53 Postal and courier activities r25 Postal and courier activities+
Telecommunications
I Accommodation and food service activities
J58 Publishing activities r26 Accommodation and food service activities
J59_J60 Motion picture, video and television programme r27 Publishing activities+ Motion picture, video and
production, sound recording and music television programme production, sound
publishing activities; programming and recording and music publishing activities;
broadcasting activities programming and broadcasting activities+
Computer programming, consultancy and
J61 Telecommunications
related activities; information service activities
J62_J63 Computer programming, consultancy and related + Legal and accounting activities; activities of
activities; information service activities head offices; management consultancy
K64 Financial service activities, except insurance and activities+ Architectural and engineering
pension funding activities; technical testing and analysis+
Scientific research and development+
K65 Insurance, reinsurance and pension funding,
Advertising and market research+ Other
except compulsory social security
professional, scientific and technical activities;
K66 Activities auxiliary to financial services and veterinary activities+ Administrative and
insurance activities support service activities
L68 Real estate activities
M69_M70 Legal and accounting activities; activities of head
offices; management consultancy activities
M71 Architectural and engineering activities; technical
testing and analysis
M72 Scientific research and development r28 Financial service activities, except insurance and
pension funding+ Insurance, reinsurance and
M73 Advertising and market research
pension funding, except compulsory social
M74_M75 Other professional, scientific and technical security+ Activities auxiliary to financial
activities; veterinary activities services and insurance activities

(Continues)
 17577802, 2019, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/rsp3.12198 by Benemerita Universidad Autonoma De Puebla, Wiley Online Library on [12/02/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
546 LLANO ET AL.

TABLE A2 (Continued)

WIOD initial branches detail WIOD aggregated branches

N Administrative and support service activities r29 Real estate activities
O84 Public administration and defence; compulsory r30 Public administration and defence; compulsory
social security social security
P85 Education r31 Education
Q Human health and social work activities r32 Human health and social work activities
R_S Other service activities r33 Other service activities
T Activities of households as employers; r34 Activities of households as employers;
undifferentiated goods‐ and services‐producing undifferentiated goods‐ and services‐
activities of households for own use producing activities of households for own use
U Activities of extraterritorial organizations and r35 Activities of extraterritorial organizations and
bodies bodies

TABLE A3 Countries explicitly included in the WIOD 2014: Countries

Code Country Code Country

AUS Australia JPN Japan

AUT Austria KOR Korea, Republic of
BEL Belgium LVA Latvia
BRA Brazil LTU Lithuania
BGR Bulgaria LUX Luxembourg
CAN Canada MLT Malta
CHE Switzerland MEX Mexico
CHN China NLD Netherlands
CYP Cyprus NOR Norway
CZE Czech Republic POL Poland
DNK Denmark PRT Portugal
EST Estonia ROU Romania
FIN Finland RUS Russia
FRA France SVK Slovakia
DEU Germany SVN Slovenia
GRC Greece ESP Spain
HRV Croatia SWE Sweden
HUN Hungary TWN Taiwan
IND India TUR Turkey
IDN Indonesia GBR United Kingdom
IRL Ireland USA United States
ITA Italy ROW Rest of World
 17577802, 2019, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/rsp3.12198 by Benemerita Universidad Autonoma De Puebla, Wiley Online Library on [12/02/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
LLANO ET AL. 547

TABLE A4 SIRIO structure: 2010

Code Region Coe Sector

AND Andalusia AA + BB Agriculture, animal production, fishing and aquaculture and

Forestry
ARA Aragon CA + CB + DF + EE Mining & quarrying, coke & petroleum; wlectricity and water
AST Asturias DA Manufacture of food products, beverages and tobacco products
BAL Balearic Islands DB + DC Manufacture of textiles, wearing apparel and leather products
CAN Canary Islands DD Manufacture of wood and of products of wood and cork
CTB Cantabria DE Manufacture of paper products and Printing
CYL Castile and León DG Manufacture of chemicals and pharmaceutical products
CLM Castilla‐La Mancha DH Manufacture of rubber and plastic products
CAT Catalonia DI Manufacture of other non‐metallic mineral products
VAL Valencian DJ Manufacture of basic metals
Community
EXT Extremadura DK Manufacture of metal products and machinery and equipment
GAL Galicia DL Manufacture of electrical equipment, computer, electronic and
optical
MAD Community of DM Manufacture of motor vehicles and other transport equipment
Madrid
MUR Region of Murcia DN Other manufacturing
NAV Navarre FF Construction
PVA Basque Country GG Wholesale and retail trade
RIO La Rioja HH Accommodation and food service activities
CYM Ceuta & Melilla II Transport and communication services
JJ Financial service activities
KK + OO + PP + QQ Other professional and personal services
LL + MM + NN Public administration, education and health
 17577802, 2019, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/rsp3.12198 by Benemerita Universidad Autonoma De Puebla, Wiley Online Library on [12/02/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
DOI: 10.1111/rsp3.12198

Resumen. El 8 de marzo, el Presidente de los Estados Unidos, Donald Trump, aprobó nuevos
aranceles a la importación de "hierro y acero" y de "aluminio" del 25% y el 10%,
respectivamente. El objetivo de este artículo es estimar el impacto que esta decisión podría
tener en las regiones españolas. La estrategia empírica de este estudio combina tres modelos
poderosos para abordar el efecto del comercio tanto en la economía global como, más
específicamente, en cada región española. En primer lugar, mediante el modelo de simulación
SMART, se estimó el efecto comercial para 131 países; a continuación, se calcularon los
efectos intersectoriales de dicha perturbación con la ayuda de la Base de Datos Input‐Output
Mundial (WIOD, por sus siglas en inglés); finalmente, se estimaron los efectos interregionales
e intersectoriales dentro de España, con la ayuda de una matriz interregional de input‐output.
Este último paso combina tanto el impacto inmediato del producto sobre el sector
metalúrgico en España, como el efecto global calculado mediante la WIOD. Los resultados
obtenidos muestran cómo los aranceles estadounidenses podrían provocar una pérdida total
de empleo de 185 000 personas en todo el mundo, mientras que en España podrían costar 3
500 puestos de trabajo.

抄録: 3月8日、ドナルド・トランプ米大統領は「鉄鋼」と「アルミニウム」の新関税率 (そ
れぞれ25%と10%)を承認した。本稿の目的は、この決定がスペインの地方に与える影響を推
計することである。今回の実証分析の方策として、グローバル経済と、特にスペインの各
地域全体の両方に対する貿易の影響に対応する3つの強力なモデルを組み合わせる。はじ
めに、SMARTシミュレーションモデルにより131カ国の貿易に対する影響を推計し、つぎに
世界産業連関表データベース (The World Input‐Output Database:WIOD)を用いて新関税率
によるショックの産業間の影響を計算する。最後に、地域間産業連関表を用いて、スペイ
ン国内の地域間および産業間の影響を推計する。この最後の段階で、スペインの鉄鋼業の
生産額に対する直接的なショックとWIODにより計算したグローバルな影響を統合する。結
果から、米国の関税率により全世界で合計185,000人の労働者(スペインでは3,500)が失業
する可能性が示される。

-------------------------------------------------------------------------------------------------------
© 2019 The Author(s). Regional Science Policy and Practice © 2019 RSAI