Economic effects of isolating Russia from international trade due to its ‘special military operation’ in Ukraine

ABSTRACT

The international community has reacted with surprising speed and unity to Russia’s ‘special military operation’ on Ukrainian territory through commercial and financial sanctions to achieve its economic isolation. This military action will change the relations between Russia and most world countries in ways that cannot yet be foreseen. This study analyzes the short-term effects of international trade interruptions on the economy, considering different isolation scenarios. The hypothetical extraction method and a multi-regional input-output model are used to simulate the economic effects on the production of 189 countries. The results show that the most affected country is Russia, with a drop in production of 10.1% in the scenario with sanctions from the European Union and 14.8% when the sanctions are also applied by Australia, Canada, Japan, United States, and the United Kingdom. The European countries with the greatest geographical proximity and strong trade flow with Russia suffer a significant drop in their production, including Lithuania, Latvia, Estonia, Finland, Hungary, and Poland. In Russia, the most affected economic sectors are Re-export & Re-import and Mining & Quarrying. Finally, the estimated impacts are a lower bound since the effects associated with financial sanctions, exchange rates, commodity prices, among others, are not considered.

KEYWORDS:

• MRIO model
• input-output
• war economics
• international trade

1. Introduction

The ‘special military operation’ on Ukraine¹ triggered a set of immediate economic impacts on the price of the main products exported by Russia, such as oil, gas, coal, wheat, aluminum, among others. At the same time, there was a drop in world stock markets, a depreciation of the ruble, and a rise in the price of gold as it is a haven asset for investors in times of high uncertainty. The economic and financial sanctions of European and other Western countries in retaliation for this military action were rapid, coordinated, progressive, and stronger than had been anticipated. The economic sanctions focused on specific sectors of the Russian economy, but Germany still does not restrict the purchase of gas due to its heavy energy dependence. The financial sanctions aimed to freeze Russian assets and prevent Russian banks from accessing the SWIFT system. These sanctions have not stopped the conflict but seek to make it difficult to finance this military operation. Despite this, Russia has claimed that it has sufficient financial resources to ensure financial stability in the face of sanctions and external threats. The duration of this warlike conflict is unpredictable at the moment. However, the different adverse effects will last for years, and they will have repercussions both in Russia and in the countries that impose them. The International Monetary Fund (IMF) has stated that this conflict will seriously impact the world economy due to increased energy and raw materials prices, inflationary pressures, and interruptions in the supply chain.² In addition, the IMF states that the countries close to the countries in conflict will face severe problems due to the interruption of trade flows and the growing influx of refugees.

Table 1. Percentage variation (%) in total production by country.

Source: Own elaboration.

To date, there is no economic quantification of the effects of the sanctions, which makes political decision-making in this conflict difficult, both for the countries that adopt them and for the sanctioned country. Perhaps if this information had been available earlier, those who decided on this ‘special military operation’ would have refrained from carrying it out. Therefore, this study aims to approximate the economic impact of isolating Russia from international trade as a sanction for its ‘special military operation,’ considering two sanction scenarios. The first scenario simulates the elimination of international trade flows between Russia and countries of the European Union. The second scenario simulates the elimination of international trade flows between Russia and countries of the European Union, Canada, the United States, Japan, Australia, and the United Kingdom. Specifically, the effects on the total and sectoral production of the sanctioned country, the sanctioning countries, and the rest of the world are quantified. A Multi-Regional Input-Output (MRIO) model and the hypothetical extraction method are used for the above. To the best of our knowledge, the novelty of this research consists of being the first empirical study that tries to measure -in the short term- the potential economic effects derived from commercial sanctions -as a means of deterring military operations- using the hypothetical extraction method in the context of the multi-regional input-output analysis.

The question raised in this research could be answered with other more sophisticated methods, such as New Quantitative Trade Models (NQTM) (Eppinger et al. 2021; Caliendo and Parro 2015) or computable general equilibrium (CGE) models specialized in global trade (Cheong and Turakulov 2021; Gopalakrishnan et al. 2021). However, the models mentioned have high computational complexity, high data requirements for their calibration, difficulty in estimating key parameters, and problems to be replicated. In contrast, the method proposed in this study is simple to implement since the required data are easy to obtain, and the model is a linear approximation of the economy that can be solved with matrix algebra. The above facilitates the rapid estimation of the economic costs of trade sanctions, as evidenced by the realization of this study just a couple of weeks after the war in Ukraine began.

The remainder of the paper is organized as follows. Section 2 provides a brief literature review of the MRIO model and the hypothetical extraction method. Section 3 introduces the utilized methods and the source of data, followed by a description of the two scenarios to be analyzed. Section 4 presents and critically evaluates the derived results. Section 5 summarizes and makes some concluding observations.

2. Literature review

The input-output analysis was developed by Leontief (1936) to analyze the interdependence of the productive sectors in an economy. Isard (1951) expanded the Leontief model to represent interregional linkages in the case of more than one region. Isard’s interregional input-output model (IRIO) makes it possible to measure the effect of an increase in final demand in one region on the levels of sectoral production in another and how the latter affects sectoral production in the first region. However, the IRIO model has detailed data requirements for interregional transactions and assumes that interregional trade relations remain constant, so it has had few real-world applications (Miller and Blair 2021). The previous limitations inspired simplifications to obtain a consistent estimate of intra- and interregional transactions through the so-called MRIO model developed by Chenery (1953) and Moses (1955). The MRIO model was extended by Polenske (1970) and subsequently implemented widely scale in the United States by Polenske (1980). Tukker and Dietzenbacher (2013) point out a clear distinction between interregional and multiregional models. In the IRIO model, it is required to know the production flow from sector i in region m to sector j in region r. In contrast, in the MRIO model, it is only necessary to know the production flow from sector i to sector j in region r, regardless of the region where that production comes. The lower information requirement in the MRIO model is achieved through the inclusion of a parameter that indicates the proportion of the production flow that region r receives from sector i in region m, and that is applied uniformly for each sector j of destination. Therefore, an MRIO model should be considered a special case of an IRIO model (Tukker and Dietzenbacher 2013).

A MRIO model has become a standard tool for analyzing international trade. For example, Carpa and Martínez-Zarzoso (2022) examine the impact of the global value chain on income distribution. Guei (2021) discusses the effects of the global value chain and technological innovation on exports. Ayadi et al. (2021) examine how the COVID-19 shock affects different European countries, considering their regional integration and exposure to the global value chain. Croft, West, and Green (2018) analyze the heterogeneity of production at the sub-national level in global trade flows. This type of model is also used to analyze the relationship between trade flows and the environment, including the determinants of emissions, carbon footprint, energy footprint, land use, virtual water flows, among others (Lin, Zhou, and Chen 2022; Wang and Liu 2021; Liu, Liu, and Liang 2021; Usubiaga-Liaño, Arto, and Acosta-Fernández 2021; Zhong et al. 2021; Yang, Qu, and Cai 2020; Qasemipour et al. 2020; Lenzen, Sun, and Faturay 2018; Kander, Jiborn, and Moran 2015). Another novel application is developing a collaborative platform based on an MRIO model that allows countries to monitor progress towards the Sustainable Development Goals (Lenzen, Geschke, and West 2022).

On the other hand, the hypothetical extraction method allows evaluating the importance of the economic sectors in the country’s economy within an input-output approach. This method has also been used to simulate shocks in international trade patterns between multiple countries or regions. For example, Giammetti et al. (2022) study the effect of three deglobalization scenarios on EU regional economies, considering different interruption levels of imports and exports in intermediate inputs. Tsekeris (2021) hypothetically extracts the regions of the United Kingdom within the economic structure of the European Union to assess the effects of Brexit, for which he assumes that the sale and purchase of products or inputs among these regions ceases. Giammetti, Russo, and Gallegati (2020a) apply the hypothetical extraction method to identify those sectors most affected by Brexit. Bonet-Morón et al. (2020) partially extract the economic flows from the intermediate consumption and final demand matrices to study the regional and sectoral economic impact of lockdown measures in Colombia. Giammetti et al. (2020b) analyze the role of the domestic value chain in the transmission of the economic effect of Covid-19 in Italy. Haddad et al. (2020) use the partial hypothetical extraction method to evaluate the regional and sectoral economic costs of different strategies of lockdown measures in Brazil. Sanguinet et al. (2021) analyze partial lockdown measures in Brazil to determine the regional impacts on integration in supply chains. Finally, Boundi-Chraki (2017) analyzes the effects of Mexico’s economic integration and dependence in the context of the North American Free Trade Agreement.

3. Material and methods

3.1. Multi-regional input-output (MRIO) model

As mentioned earlier, an MRIO model expands the Leontief model to analyze the productive interdependencies between M regions (countries) and N economic sectors (Miller and Blair 2021). This model has certain limitations since it assumes constant technical coefficients and trade coefficients, which prevents a substitution between inputs in the face of changes in relative prices. Despite the above, it is a handy tool for analyzing international trade flows caused by a particular economic shock in the short term.

An MRIO model with M regions (countries) and N economic sectors in each region can be represented by the following equation: (1) $\mathit{x}=\mathit{A}x+\mathit{f}$(1) Where $\mathit{x}$ represents a column vector of production with N×M elements, $\mathit{f}$ is a column vector of final demand with N×M elements. The production vector can be partitioned into the production vectors of each region $\mathit{x}=({\mathit{x}}^{1\mathbf{\prime }},\dots ,{\mathit{x}}^{\mathit{m}\mathbf{\prime }},\dots ,{\mathit{x}}^{\mathit{M}\mathbf{\prime }})$. Likewise, the final demand vector can be partitioned as $\mathit{f}=\left({\mathit{f}}^{1\mathbf{\prime }},\dots ,{\mathit{f}}^{\mathit{m}\mathbf{\prime }},\dots ,{\mathit{f}}^{\mathit{M}\mathbf{\prime }}\right)$. On the other hand, $\mathit{A}$ is a matrix with dimension NM×NM that includes the matrix of technical coefficients for each region (${\mathit{A}}^{\mathit{m}m}$) and the matrix of trade coefficients between region m and r (${\mathit{A}}^{\mathit{m}r}$). In addition, $a_{ij}^{mm}$ is the technical coefficient that denotes the flow of inputs from sector i to sector j per unit of output from sector j in region m. In contrast, $a_{ij}^{mr}$ is the trade coefficient denoting the flow of inputs from sector i in region r to sector j per unit of output from sector j in region m. (2) $\mathit{A}=\left[\begin{array}{cc}\begin{array}{cc}\begin{array}{cc}{\mathit{A}}^{11}& \dots \end{array}& {\mathit{A}}^{1\mathit{m}}\\ \begin{array}{cc}⋮& \ddots \end{array}& \end{array}& \begin{array}{cc}\dots & {\mathit{A}}^{1\mathit{M}}\\ & ⋮\end{array}\\ \begin{array}{cc}\begin{array}{cc}{\mathit{A}}^{\mathit{r}1}& \end{array}& {\mathit{A}}^{\mathit{r}m}\\ \begin{array}{cc}\begin{array}{c}⋮\\ {\mathit{A}}^{\mathit{M}1}\end{array}& \begin{array}{c}\\ \dots \end{array}\end{array}& \begin{array}{c}\\ {\mathit{A}}^{\mathit{M}m}\end{array}\end{array}& \begin{array}{cc}& {\mathit{A}}^{\mathit{r}M}\\ \begin{array}{c}\ddots \\ \dots \end{array}& \begin{array}{c}⋮\\ {\mathit{A}}^{\mathit{M}M}\end{array}\end{array}\end{array}\right]$(2) (3) ${\mathit{A}}^{\mathit{m}m}=\left[\begin{array}{cc}\begin{array}{cc}\begin{array}{cc}{a}_{11}^{mm}& \dots \end{array}& a_{1j}^{mm}\\ \begin{array}{cc}⋮& \ddots \end{array}& \end{array}& \begin{array}{cc}\dots & a_{1n}^{mm}\\ & ⋮\end{array}\\ \begin{array}{cc}\begin{array}{cc}{a}_{i1}^{mm}& \end{array}& a_{ij}^{mm}\\ \begin{array}{cc}\begin{array}{c}⋮\\ a_{n1}^{mm}\end{array}& \begin{array}{c}\\ \dots \end{array}\end{array}& \begin{array}{c}\\ a_{nj}^{mm}\end{array}\end{array}& \begin{array}{cc}& a_{in}^{mm}\\ \begin{array}{c}\ddots \\ \dots \end{array}& \begin{array}{c}⋮\\ a_{nn}^{mm}\end{array}\end{array}\end{array}\right]$(3) (4) ${\mathit{A}}^{\mathit{m}r}=\left[\begin{array}{cc}\begin{array}{cc}\begin{array}{cc}{a}_{11}^{mr}& \dots \end{array}& a_{1j}^{mr}\\ \begin{array}{cc}⋮& \ddots \end{array}& \end{array}& \begin{array}{cc}\dots & a_{1n}^{mr}\\ & ⋮\end{array}\\ \begin{array}{cc}\begin{array}{cc}{a}_{i1}^{mr}& \end{array}& a_{ij}^{mr}\\ \begin{array}{cc}\begin{array}{c}⋮\\ a_{n1}^{mr}\end{array}& \begin{array}{c}\\ \dots \end{array}\end{array}& \begin{array}{c}\\ a_{nj}^{mr}\end{array}\end{array}& \begin{array}{cc}& a_{in}^{mr}\\ \begin{array}{c}\ddots \\ \dots \end{array}& \begin{array}{c}⋮\\ a_{nn}^{mr}\end{array}\end{array}\end{array}\right]$(4)

The MRIO model represents a system of NM equations (N sectors and M regions) that can be written in matrix form as follows: (5) $\mathit{x}=(\mathit{I}-\mathit{A}{)}^{-1}\mathit{f}$(5) Equation (5) allows expressing the production value of each sector and region (country) as the inverse Leontief matrix ($(\mathit{I}-\mathit{A}{)}^{-1}$) multiplied by the final demand vector ($\mathit{f}$).

3.2. Hypothetical extraction method

In the input-output model, an increase in the production of one sector causes an increase in the demand for inputs from other sectors. Still, the initial increase in production also generates a greater availability of goods used as inputs by other sectors. These two effects are known as backward linkages and forward linkages, respectively. The first methods to obtain productive linkages and determine the key sectors were developed by Rasmussen (1956), Hirschman (1958), and Chenery and Watanabe (1958). Later, Paelinck, de Caevel, and Degueldre (1965), Miller (1966), Strassert (1968), and Schultz (1977) proposed the hypothetical extraction method (HEM) as an alternative to Rasmussen, Chenery-Watanabe, and Hirschman methods. The HEM evaluates the importance of a sector in a country’s economy, considering its linkages with the rest of the sectors (Cella, 1984). For the above, the original economic structure is compared with the counterfactual economic structure if that sector is removed from the technical coefficients matrix and final demand vector (Tokito, Kagawa, and Hanaka 2020).

According to Dietzenbacher & Lahr (2013), the traditional HEM procedure extracts sector j from the economic system to measure the output decrease. The extraction of sector j implies that each row and column of the matrix $\mathit{A}$ will be equal to zero, obtaining a new matrix ${\overline{\mathit{A}}}_{\left(j\right)}$. Dietzenbacher & Lahr (2013) highlight that the deliveries made by the extracted j sector will be covered by imports. On the other hand, ${\overline{\mathit{f}}}_{\left(j\right)}$ is the reduced final demand vector when the sector j has been removed. Instead of deleting element j in vector f, it can be replaced by zero. Finally, output in the reduced economy is found as follows: (6) ${\overline{\mathit{x}}}_{\left(\mathit{j}\right)}=(\mathit{I}-{\overline{\mathit{A}}}_{\left(j\right)}{)}^{-1}{\overline{\mathit{f}}}_{\left(j\right)}$(6) So, $T_j=i^{\mathbf{\prime }x}-i^{\mathbf{\prime }}{\overline{\mathit{x}}}_{\left(\mathit{j}\right)}$ measures the economy’s loss (decreased output) if sector j disappears (Miller and Blair 2021).

Dietzenbacher, van der Linden, and Steenge (1993) extended this method to quantify inter-regional linkages in a multi-region input-output table when one region is removed. In the latter case, the technical coefficients and the final demand vector of the removed region are eliminated to compare the original and counterfactual economic structure. More recently, Dietzenbacher, van Burken, and Kondo (2019) proposed an adaptation to the HEM and pointed out that this method has certain limitations when used from a global perspective since it should model how imports from other countries replace imports from the suppressed country. In a situation of international sanctions, it may be challenging to carry out this substitution in the short term. In addition, there is not enough information to determine from and toward what countries and what proportion this substitution will be made. Therefore, the hypothetical extraction method serves as a short-term approximation to determine the most affected countries and sectors in the context analyzed in this study.

3.3. Available data

In this study, an MRIO model is calibrated with the Eora database consisting of a multi-region input-output table for 2015 (see Casella et al. 2019; Lenzen et al. 2013). Eora includes 26 productive sectors³ and 189 countries (see Table A1 in the Appendix), valued at basic prices in dollars.

3.4. Scenarios

In this study, two scenarios of economic sanctions against Russia are proposed. In the first scenario, the countries of the European Union suppress their international trade flows entirely with Russia. In the second scenario, international trade flows between Russia and a block of countries including the European Union, Australia, Canada, the United States, Japan, and the United Kingdom are suppressed. Once the initial shock in production caused by eliminating imports and exports of intermediate goods in each country is obtained, the final demand vector is modified to get the final shock in all the countries. The latter implies that exports of final goods to Russia from countries applying sanctions and exports of final goods to countries applying sanctions from Russia are eliminated from the final demand vector. It is necessary to clarify that these two scenarios do not include financial shocks and effects on exchange rates and commodity prices since they go beyond the type of analysis that can be modelled with an MRIO model and a hypothetical extraction method.

4. Results

The proposed methodology allows estimating the effects on production caused by removing Russia from international trade controlling the impact of other shocks or financial sanctions. Scenario 1 considers sanctions only from the countries of the European Union, and scenario 2 assumes joint sanctions of the European Union, Australia, Canada, Japan, the United States, and the United Kingdom. In both scenarios, the initial shock is presented due to the elimination of international trade between the countries involved, and the final shock also includes the fall in final demand.

Table 1 shows the percentage variation in the total production of each country included in the Eora database. The results show that the most affected country is Russia (RUS), with an initial drop in production of 8.2% in Scenario 1 and 12.0% in Scenario 2 and a final drop in production of 10.1% in Scenario 1 and 14.8% in Scenario 2. The non-sanctioned countries in which production falls sharply are those members of the European Union that have greater geographical proximity and trade flows with Russia, such as Lithuania (LTU), Latvia (LVA), Estonia (EST), Finland (FIN), Hungary (HUN), Poland (POL), Bulgaria (BGR), Slovenia (SVN), Slovakia (SVK), and Sweden (SWE). Also, countries of the European Union somewhat further away from Russia are significantly affected, including Luxembourg (LUX), Austria (AUT), the Netherlands (NLD), Belgium (BEL), Italy (ITA), Ireland (IRL), among others. Despite Germany’s commented dependence on Russian gas, the economic effects are less than in other European countries, with variations in the production of 0.7% or 1.2% depending on whether the initial or final shock is evaluated. The vast majority of countries are only marginally affected in their production.

It is important to contextualize the previous results based on international trade flows. In 2019, the top ten destination countries for Russian exports were China (13.4%), the Netherlands (10.5%), Germany (6.6%), Belarus (5.1%), Turkey (5.0%), Korea (3.8%), Italy (3.4%), Kazakhstan (3.3%), United Kingdom (3.1%), and United States (3.1%).⁴ In contrast, the ten countries that sent the largest proportion of their exports to Russia were Belarus (41.3%), Armenia (27.5%), Lithuania (14.0%), Uzbekistan (13.6%), Georgia (13.1%), Kazakhstan (9.7%), Latvia (9.2%), Moldova (9.0%), Estonia (8.4%), and Finland (5.5%). However, the direct trade flows between Russia and the other countries do not fully explain the results obtained in Table 1. It is also essential to highlight the indirect trade links with other affected countries. For example, Poland sends a relatively small fraction of its exports to Russia but has close trade links with the Baltic countries (Lithuania, Latvia, and Estonia), heavily affected by sanctions on Russia.

Table 2 shows the variations in the production of the most affected economic sectors by country. The most affected sector is Re-export & Re-import in Russia (RUS), Spain (ESP), Slovakia (SVK), Croatia (HVR), Bulgaria (BGR), Latvia (LVA), Luxembourg (LUX), and Lithuania (LTU). Other severely affected sectors in Russia are Mining and Quarrying, Hotels and Restaurants, Public Administration, Metal Products, Petroleum, and Chemical and Non-Metallic Mineral Products. In countries with geographic and commercial proximity to Russia, affected sectors include the following: Electrical and Machinery, Metal Products, and Food & Beverages in Latvia (LVA); Recycling, Metal Products, Electrical and Machinery, and Food & Beverages in Lithuania (LTU); and Fishing and Textiles & Wearing Apparel in Finland (FIN). The ranking of the most affected sectors depends on the proposed scenario and whether the initial or final shock is considered, but the most affected sectors and countries do not significantly change.

Table 2. Percentage variation (%) in the production of the most affected economic sectors and countries.

Country	Sector	Scenario 1		Country	Sector	Scenario 2
		Initial shock	Final shock			Initial shock	Final shock
ESP	Re-export & Re-import	−95.8	−184.8	RUS	Re-export & Re-import	−300.9	−288.0
SVK	Re-export & Re-import	−117.3	−147.0	ESP	Re-export & Re-import	−95.8	−185.2
RUS	Re-export & Re-import	−127.8	−113.4	SVK	Re-export & Re-import	−117.3	−147.2
RUS	Hotels and Restraurants	−77.5	−60.7	RUS	Mining and Quarrying	−78.4	−75.4
RUS	Mining and Quarrying	−59.1	−56.8	RUS	Hotels and Restraurants	−91.6	−71.9
RUS	Public Administration	−56.0	−48.5	RUS	Public Administration	−64.2	−54.5
HRV	Re-export & Re-import	−23.7	−31.5	HRV	Re-export & Re-import	−23.7	−31.2
BGR	Re-export & Re-import	−31.5	−28.6	RUS	Metal Products	−30.0	−30.4
LVA	Re-export & Re-import	−12.4	−21.1	BGR	Re-export & Re-import	−31.5	−28.3
LVA	Electrical and Machinery	−14.1	−20.6	LVA	Re-export & Re-import	−12.4	−21.1
LUX	Re-export & Re-import	−19.6	−20.0	LVA	Electrical and Machinery	−14.1	−20.6
LTU	Recycling	−16.2	−19.4	RUS	Petroleum, Chemical and Non-Metallic Mineral Products	−18.4	−20.2
LTU	Re-export & Re-import	−12.3	−16.7	LUX	Re-export & Re-import	−19.6	−19.7
LTU	Metal Products	−16.3	−16.1	LTU	Recycling	−16.2	−19.4
RUS	Metal Products	−15.2	−15.7	RUS	Electricity, Gas and Water	−16.6	−18.5
FIN	Fishing	−9.3	−15.4	RUS	Wood and Paper	−14.0	−17.7
LTU	Electrical and Machinery	−9.9	−15.0	LTU	Re-export & Re-import	−12.3	−16.7
RUS	Petroleum, Chemical and Non-Metallic Mineral Products	−12.6	−13.9	RUS	Transport	−11.2	−16.4
LVA	Petroleum, Chemical and Non-Metallic Mineral Products	−10.8	−12.9	LTU	Metal Products	−16.3	−16.1
RUS	Electricity, Gas and Water	−11.4	−12.7	FIN	Fishing	−9.3	−15.4
RUS	Transport	−8.2	−11.9	LTU	Electrical and Machinery	−9.9	−15.0
RUS	Wood and Paper	−9.3	−11.9	LVA	Petroleum, Chemical and Non-Metallic Mineral Products	−10.8	−12.9
LVA	Metal Products	−9.3	−11.2	RUS	Electrical and Machinery	−7.3	−12.2
FIN	Textiles and Wearing Apparel	−7.0	−11.0	LVA	Metal Products	−9.3	−11.2
SVN	Re-export & Re-import	−7.2	−10.3	FIN	Textiles and Wearing Apparel	−7.0	−11.0
LTU	Food & Beverages	−5.2	−9.8	RUS	Other Manufacturing	−6.4	−10.3
LTU	Petroleum, Chemical and Non-Metallic Mineral Products	−6.4	−9.7	RUS	Wholesale Trade	−5.8	−10.3
EST	Metal Products	−10.7	−9.2	RUS	Maintenance and Repair	−5.8	−10.3
LVA	Food & Beverages	−5.9	−9.1	RUS	Retail Trade	−5.8	−10.3
LTU	Agriculture	−4.6	−8.7	SVN	Re-export & Re-import	−7.2	−10.3
RUS	Electrical and Machinery	−5.2	−8.6	LTU	Food & Beverages	−5.2	−9.8
HUN	Recycling	−6.8	−8.6	LTU	Petroleum, Chemical and Non-Metallic Mineral Products	−6.4	−9.7
EST	Re-export & Re-import	−6.4	−8.6	RUS	Finacial Intermediation and Business Activities	−5.0	−9.3
EST	Petroleum, Chemical and Non-Metallic Mineral Products	−7.4	−8.3	EST	Metal Products	−10.7	−9.2
POL	Re-export & Re-import	−5.4	−8.1	LVA	Food & Beverages	−5.9	−9.1
LVA	Fishing	−5.4	−7.9	RUS	Textiles and Wearing Apparel	−5.0	−8.7
LTU	Wholesale Trade	−4.7	−7.5	LTU	Agriculture	−4.6	−8.7
LTU	Other Manufacturing	−4.5	−7.3	HUN	Recycling	−6.8	−8.6
RUS	Wholesale Trade	−4.0	−7.1	EST	Re-export & Re-import	−6.4	−8.6
RUS	Maintenance and Repair	−4.0	−7.1	RUS	Recycling	−4.7	−8.3
RUS	Retail Trade	−4.0	−7.1	EST	Petroleum, Chemical and Non-Metallic Mineral Products	−7.4	−8.3
EST	Food & Beverages	−3.9	−7.0	POL	Re-export & Re-import	−5.4	−8.1
SWE	Retail Trade	−4.0	−6.8	LVA	Fishing	−5.4	−7.9
RUS	Finacial Intermediation and Business Activities	−3.6	−6.8	LTU	Wholesale Trade	−4.7	−7.5
RUS	Other Manufacturing	−4.2	−6.8	LTU	Other Manufacturing	−4.5	−7.3

Source: Own elaboration.

Again, it is helpful to put the previous results in context. In 2021, Russia was the fifth largest trading partner of the European Union, while the European Union was Russia’s most important trading partner. That same year, exports of goods from Russia to the European Union totalled €158.5 billion, concentrated in minerals and fuels (62%), iron and steel (4.7%), wood (2%), and fertilizers (1.1%). In contrast, exports of goods from the European Union to Russia totalled €99.0 billion, which were led by machinery and equipment (19.7%), motor vehicles (9.0%), pharmaceuticals (8.1%), electrical equipment and machinery (7.6%), and plastics (4.3%). In 2020, imports of services from Russia to the European Union and exports of services from the European Union to Russia represented €8.9 and €20.5 billion, respectively.⁵ Consequently, Russia is a supplier of raw materials that imports industrial goods from Europe. Despite the above, the simulations carried out in this study reflect more diverse sectoral effects.

Although most of the findings in this study are quite reasonable, there is a counterintuitive result. Germany is one of the European countries most reluctant to suppress the purchase of Russian gas. Still, the estimated negative impacts for this country are less than those estimated for other Eastern European countries. There are several possible explanations to try to understand this situation. First, the hypothetical extraction method in the context of an MRIO assumes that the countries imposing the sanctions could import gas from any other country without any problem or restriction in the short term. This replacement of the supplier country is not easy in practice since Russian gas arrives through gas pipelines to Germany while importing gas by ship from other countries requires building regasification plants that could only operate in the long-term. Second, the input-output models assume that the production functions of each sector remain constant and that there are no substitution options among different domestic inputs. The preceding is not necessarily true since a country like Germany could replace thermoelectric generation with nuclear energy or renewable energies, transforming its production function in the energy sector. Third, increases in final demand from other gas, oil, or coal exporting countries are not explicitly considered since it would be necessary to know what percentage of the Russian fuels will be replaced by exports from these countries. Finally, it can be argued that wars condition countries’ strategic and economic decisions to modify their productive structure, which are not being considered in this study. In the case of Germany, a significant increase in the military budget was announced, generating a stimulus in the arms production sector with its consequent productive linkages, while public resources destined for social purposes will probably decrease, negatively affecting the linkages of the service sectors. All the above arguments conclude that the technical coefficients and elements of the final demand in an MRIO do not remain completely stable in the scenarios analyzed in this study. Despite this limitation, the proposed approach provides a valuable overview of the global effects of trade sanctions, identifies the most affected countries, and approximates the magnitude of each country’s impacts.

5. Conclusions

This study uses the hypothetical extraction method and an MRIO model to simulate the economic effects of isolating Russia from international trade. The possibility of having this type of information might have prevented Russia’s ‘special military operation,’ considering the high economic costs that this country faces and will face. Now it is no longer possible to avoid this conflict, but at least this study could be an input for politicians to improve their decisions and quickly solve this situation. Alternatively, the methodology proposed in this study can become a standard tool to predict the effect of international trade sanctions and estimate the costs that a possible country aggressor would face, which could help discourage future armed conflicts.

The results show that Russia would have a drop in production of 10.1% if international trade sanctions were applied only by the countries of the European Union and 14.8% if countries such as Australia, Canada, Japan, United States, and the United Kingdom join to the sanctions. These sanctions also generate relevant impacts in European countries that are geographically close and have strong trade flows with Russia, such as Lithuania, Latvia, Estonia, Finland, Hungary, Poland, Bulgaria, Slovenia, Slovakia, Sweden, among others. However, the economic effects on Germany are relatively low despite its heavy reliance on Russian gas. At the sectoral level, the most relevant declines in the Russian economy are seen in Re-export & Re-import, Mining and Quarrying, Hotels and Restaurants, Public Administration, Metal Products, Petroleum, and Chemical and Non-Metallic Mineral Products. The Re-export & Re-import, Electrical and Machinery, Metal Products, and Food & Beverages sectors are also strongly affected in other European countries.

The proposed methodology is simple, transparent, and easy to replicate. In addition, it is useful for simulating other scenarios of international trade sanctions. This application is novel since, as was mentioned before, commercial sanctions have not been simulated in the literature on countries involved in military conflicts as a deterrent mechanism. However, this methodology only allows the identification of short term impacts associated with the suspension of international trade flows among the countries involved. The economic effects of the displaced population, financial sanctions, duration of the conflict and sanctions, destruction of productive capacity, change in international trade patterns, new strategic and commercial alliances, among many other elements, are not being simulated. Therefore, the results obtained in this study can only be interpreted as a lower and short term bound for all the economic impacts that this military conflict will trigger.

The future implications of these trade sanctions are impossible to identify with the MRIO model and the HEM method since they are associated with the strategic autonomy of the European Union, which will change the global supply chain. The diversification of energy providers, choosing alternative energy sources (nuclear energy, coal, among others), and promoting renewable energies are key to this autonomy. Russian gas’s widespread availability and low cost led to excessive energy dependence on the European Union. Still, it involves various security risks that many countries are no longer willing to take. For example, the European Commission has published a plan to reduce the European Union’s dependence on Russian gas by two-thirds this year and end its reliance on Russian fuels well before 2030.⁶ At the present, Russia remains dependent on European energy markets. It has accepted increasing trade sanctions without cutting off gas supplies since most of the pipelines from its territory flow west (nine to Europe and only one to China).⁷ In the future, it is logical that Russia sees China as an eventual destination to divert the gas exports it currently supplies to Europe. Still, for that to be possible, it first requires the construction of gas pipelines, such as the Power of Siberia project 2.⁸

Acknowledgements

The author is grateful to the ANID (Regular FONDECYT project no. 1220010) for the financing it provided for this research. Two anonymous reviewers are also acknowledged for their recommendations.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by ANID / FONDECYT Regular: [Grant Number 1220010].

Notes

1 Or ‘Putin's war’.

2 https://www.imf.org/en/News/Articles/2022/03/05/pr2261-imf-staff-statement-on-the-economic-impact-of-war-in-ukraine

3 Agriculture; Fishing; Mining and Quarrying; Food & Beverages; Textiles and Wearing Apparel; Wood and Paper; Petroleum, Chemical and Non-Metallic Mineral Products; Metal Products; Electrical and Machinery; Transport Equipment; Other Manufacturing; Recycling; Energy; Construction; Maintenance and Repair; Wholesale Trade; Retail Trade; Hotels and Restaurants; Transport; Post and Telecommunications; Financial Intermediation and Business Activities; Public Administration; Education, Health and Other Services; Private Households; Others; Re-export & Re-import.

4 https://wits.worldbank.org/countrysnapshot/en/RUSSIA

5 https://ec.europa.eu/trade/policy/countries-and-regions/countries/russia/#:∼:text = In%202021%2C%20Russia%20was%20the,in%20goods%20with%20the%20world.

6 https://www.europarl.europa.eu/RegData/etudes/BRIE/2022/729300/EPRS_BRI(2022)729300_EN.pdf

7 https://www.energypolicy.columbia.edu/research/qa/qa-china-russia-energy-relations-will-new-oil-and-natural-gas-deals-help-russia-weather-economic

8 https://tass.com/economy/1375983

APPENDIX

Table A1. Acronyms of the countries.

Acronym	Country	Acronym	Country	Acronym	Country	Acronym	Country	Acronym	Country
AFG	Afghanistan	CHL	Chile	HUN	Hungary	MAR	Morocco	SVK	Slovakia
ALB	Albania	CHN	China	ISL	Iceland	MOZ	Mozambique	SVN	Slovenia
DZA	Algeria	COL	Colombia	IND	India	MMR	Myanmar	SOM	Somalia
AND	Andorra	COG	Congo	IDN	Indonesia	NAM	Namibia	ZAF	South Africa
AGO	Angola	CRI	Costa Rica	IRN	Iran	NPL	Nepal	SDS	South Sudan
ATG	Antigua	HRV	Croatia	IRQ	Iraq	NLD	Netherlands	ESP	Spain
ARG	Argentina	CUB	Cuba	IRL	Ireland	ANT	Netherlands Antilles	LKA	Sri Lanka
ARM	Armenia	CYP	Cyprus	ISR	Israel	NCL	New Caledonia	SUD	Sudan
ABW	Aruba	CZE	Czech Republic	ITA	Italy	NZL	New Zealand	SUR	Suriname
AUS	Australia	CIV	Cote dIvoire	JAM	Jamaica	NIC	Nicaragua	SWZ	Swaziland
AUT	Austria	PRK	North Korea	JPN	Japan	NER	Niger	SWE	Sweden
AZE	Azerbaijan	COD	DR Congo	JOR	Jordan	NGA	Nigeria	CHE	Switzerland
BHS	Bahamas	DNK	Denmark	KAZ	Kazakhstan	NOR	Norway	SYR	Syria
BHR	Bahrain	DJI	Djibouti	KEN	Kenya	PSE	Gaza Strip	TWN	Taiwan
BGD	Bangladesh	DOM	Dominican Republic	KWT	Kuwait	OMN	Oman	TJK	Tajikistan
BRB	Barbados	ECU	Ecuador	KGZ	Kyrgyzstan	PAK	Pakistan	THA	Thailand
BLR	Belarus	EGY	Egypt	LAO	Laos	PAN	Panama	MKD	TFYR Macedonia
BEL	Belgium	SLV	El Salvador	LVA	Latvia	PNG	Papua New Guinea	TGO	Togo
BLZ	Belize	ERI	Eritrea	LBN	Lebanon	PRY	Paraguay	TTO	Trinidad and Tobago
BEN	Benin	EST	Estonia	LSO	Lesotho	PER	Peru	TUN	Tunisia
BMU	Bermuda	ETH	Ethiopia	LBR	Liberia	PHL	Philippines	TUR	Turkey
BTN	Bhutan	FJI	Fiji	LBY	Libya	POL	Poland	TKM	Turkmenistan
BOL	Bolivia	FIN	Finland	LIE	Liechtenstein	PRT	Portugal	USR	Former USSR
BIH	Bosnia and Herzegovina	FRA	France	LTU	Lithuania	QAT	Qatar	UGA	Uganda
BWA	Botswana	PYF	French Polynesia	LUX	Luxembourg	KOR	South Korea	UKR	Ukraine
BRA	Brazil	GAB	Gabon	MAC	Macao SAR	MDA	Moldova	ARE	United Arab Emirates
VGB	British Virgin Islands	GMB	Gambia	MDG	Madagascar	ROU	Romania	GBR	UK
BRN	Brunei	GEO	Georgia	MWI	Malawi	RUS	Russia	TZA	Tanzania
BGR	Bulgaria	DEU	Germany	MYS	Malaysia	RWA	Rwanda	USA	USA
BFA	Burkina Faso	GHA	Ghana	MDV	Maldives	WSM	Samoa	URY	Uruguay
BDI	Burundi	GRC	Greece	MLI	Mali	SMR	San Marino	UZB	Uzbekistan
KHM	Cambodia	GRL	Greenland	MLT	Malta	STP	Sao Tome and Principe	VUT	Vanuatu
CMR	Cameroon	GTM	Guatemala	MRT	Mauritania	SAU	Saudi Arabia	VEN	Venezuela
CAN	Canada	GIN	Guinea	MUS	Mauritius	SEN	Senegal	VNM	Vietnam
CPV	Cape Verde	GUY	Guyana	MEX	Mexico	SRB	Serbia	YEM	Yemen
CYM	Cayman Islands	HTI	Haiti	MCO	Monaco	SYC	Seychelles	ZMB	Zambia
CAF	Central African Republic	HND	Honduras	MNG	Mongolia	SLE	Sierra Leone	ZWE	Zimbabwe
TCD	Chad	HKG	Hong Kong	MNE	Montenegro	SGP	Singapore

Source: The Eora Global Supply Chain Database.