We study the situation when a commodity-exporting economy is under sanctions and cannot use its accumulated fx-reserves or attract new fx-debt to smooth import restrictions amid slower decline of income flow from a commodity-export. Our study attempts to determine the adequate response of the economy, depending on assumptions about the possibilities of import substitution within a structural dynamic general equilibrium model calibrated on Russian data. We use a modified version of the Ramsey–Cass–Koopmans’ model to evaluate aggregate and structural changes in the economy in the shorter and longer­ run. We consider several scenarios with different assumptions about the efficiency of import substitution, which is defined along two dimensions (for consumers as well as for producers). The results show that trade restrictions make import substitution optimal, but only in those sectors where such substitution is relatively more effective. Limited labour resources in the economy are compensated with higher capital intensity of production in the optimistic and neutral scenarios. Reallocation of resources to build up the neces­sary capital intensity calls for temporarily higher saving­ rates. As a result, GDP may be higher, but consumption may be lower comparing to the baseline. The results mean that effectiveness of import substitution should be taken into account in decision making on industrial policy. If efficiency of import-substitution is asymmetric and biased to goods for final consumption relative to goods for investments, the structure of imported goods becomes biased to the latter. The results imply higher relative price of consumer goods.

sanctions, financial autarky, structural changes, import substitution, total factor productivity, neoclassical growth model.

In 2022, the Russian economy faced the so-called “new reality,” which was formed as a result of a series of trade and financial restrictions. First there came big financial shocks:

• • The West imposed financial sanctions that provoked capital controls measures by the Bank of Russia. As a result, the financial account got almost closed. Import financing became mostly possible directly from export earnings;
• • The inability to use the accumulated gold and foreign exchange reserves/net foreign assets (NFA) of the economy to finance consumption or investment. 1

The result of these shocks, amplified by growing uncertainty, was a significant weakening of the ruble in Q1 2022, which, however, quickly found a new equilibrium (Fig. 1). The Bank of Russia played an important role in the stabilization tightening of the monetary policy: the key rate increased from 8.5% to 20.0%.

Fig. 1.

Current account and exchange rate, Russia, QI 2017 – QI 2024.

Source: Bank of Russia.

In Q2 and Q3 2022, trade balance shocks came to the fore. On the import side these changes were: firstly, the curtailment of the activities of foreign companies represented in Russia and the associated reduction of their imports of materials for production and intermediate products; secondly, the logistics problems of domestic companies with the physical delivery of imports and with payments for imports. It took time for Russian business to set up the logistics. These forces led to a decrease in the value of imports.

On the export side, logistical challenges (physical delivery as well as payments) for the main Russian export goods also were raised, but for some goods they were compensated by rising prices on the world market.2 So, the total negative effect on exports turned out to be weaker than on imports. As a consequence, the current account grew strongly (see Fig. 1). The results of the shocks at this stage were the strengthening the ruble, which overshot the pre-shock level, and an active reshaping of the foreign trade logistics.3

In 2023, the situation reverted to the opposite scenario: the restoration of imports and introduction of new export-oriented sanctions led to a decrease in the current account. An additional factor of the imports increase in 2023 was the growth of domestic demand, financed by government spending and commercial credit. As a result, the economy is approaching the second half of 2024 with a higher level of demand than before the shocks, amid a tight monetary policy. The share of investment and government consumption in the structure of demand have increased.4

To understand further transformation of the new reality and assess future ­dynamics of the Russian economy, it is important to consider two types of heighten­ing risks: tightening restrictions on Russian imports and decrease in the value of exports.

The first risk is tightening restrictions on Russian imports associated with (secondary) sanctions pressure on Russia’s trading partner countries.5 From a medium-term perspective, the import recovery that has been observed since the second half of 2022 may turn out to be temporary. Thus the scenario of a permanent reduction in imports due to its quantitative restrictions becomes relevant. This scenario is considered by Turdyeva (2024) within the framework of a static CGE model.

For modeling purposes, we consider a scenario of physical restriction of all imports by 50% with the 2019 level taken as the base. The imports remaining after the restriction are represented by consumer and/or investment goods. The specific structure of imports in the new conditions is determined endogenously in the new equilibrium.

The second risk is a decrease in the value of exports. In addition to the tighten­ing Western sanctions, in the medium term, the global energy transition may become an important negative risk factor for Russian exports. A key element of this will be the introduction of cross-border carbon emissions regulation in the EU, and then likely in trading partner countries in Asia. Such regulation would mean lower prices for Russian exports, much of which are energy-intensive. In the model scenario, the export restrictions are weakly expressed (minus 10% with the 2019 level taken as the base). In three years, restrictions reach 50% of exports. The resulting positive trade balance is not spent on consumption and/or investment, which is partly due to the difficulties to increase imports.6

The goal of this paper is to assess optimal aggregate and structural changes in the economy due to the effect of these risks. In the paper we model the optimal adjustment of a small open commodity-exporting economy to a new equilibrium in the context of the emergence of a new reality: restrictions on imports and a reduction in exports.⁷ We use a modified Ramsey model with two types of goods — a consumer good and an investment good.⁸ Each type of good is produced domestically by “packaging” two products — domestic and imported. These products could be called “intermediate.” Imported intermediate goods can be produced within the country as well. Such domestic products are labelled as “import substituting goods.”

In the pre-sanction model world, economic agents make decisions when the economy is embedded in global production and trade chains. The fact that a part of investment and consumer goods is imported in such conditions reflects the economic inefficiency of their domestic production (the inefficiency of import substitution). Economic rationale for this may be economy of scale, for example. The presence of global brands in the domestic market, capable of providing low prices due to their large scale of production, serves as a natural economic barrier for import substitution of such goods. The exception is goods, in which the country has some comparative advantage. In other words, the “old reality” in the model is defined in such a way that the need for domestic production of import-substituting goods does not arise: resources from commodity exports are used to finance the import of consumer and investment goods, which are not economically efficient to produce within the country.

We address the following issues in the paper:

• • What adjustment of key macroeconomic parameters is needed to maximize the level of household consumption in a scenario of imports and exports restrictions? How do the structural parameters of the model economy change in this scenario? — These are the issues of optimal allocation of resources.
• • How do assumptions about the effectiveness of import substitution affect optimal adjustment? The effectiveness of import substitution has two dimensions in the paper — effectiveness in consumption and in production. Effectiveness in consumption is the ability of one piece of an import-substituting product to replace one piece of the original import in consumption. For example, the ability of a light bulb or car to work for one year without breaking down, if the original import has such a characteristic. Effectiveness in production is productivity, measured in terms of total factor productivity (TFP), of using import-substituting goods in production. For example, with the same operational reliability, the ability of import-substituting chips to produce the same number of operations per second as that of the imported original. By assumption, the effectiveness of import substitution is fixed (there is no learning by doing or economy of scale).

The simulation results show that trade restrictions (through an increase in the price of imports due to additional logistical costs or a decrease in export earnings­ to finance imports) makes import substitution optimal, but only in those sectors (goods) where such substitution is relatively more effective. These are goods in which prior to the new reality the economy was closer to the efficient production frontier, represented by the global/foreign producers. In the calibrated economy, this is the case for the production of consumer goods. Regarding capital­ goods (machinery, equipment), our results show that the optimal strategy is to look for channels of preserving their imports as long as possible.

Import substitution requires a reallocation of limited labor resources to the production of goods that were previously imported. If domestic analogues are comparable in their characteristics to original goods, then the optimal strategy­ involves the accumulation of capital in sectors, producing these goods, and an increase of capital-to-labor ratio to preserve active import substitution. But this optimal strategy is associated with a permanently lower level of consumption. Import substitution requires a buildup of capital as well as its permanent reproduction, and this limits the medium-term possibilities of creating final goods, including import-substituting goods for consumption in an environment when a country cannot increase export of such goods. As a result, GDP may even grow, but consumption will be below the base level (in the absence of shocks), that is, the share of consumption in GDP will decrease.

Asymmetry in the efficiency of import substitution between sectors leads to distortions in the structure of imported and produced goods. The low effectiveness of import substitution of capital goods requires a reorientation of imports towards investment goods. In this case, active import substitution in the consumer goods sector turns out to be optimal, but at the expense of resources reallocation within the sector from the production of original domestic goods.9 If import substitution is very ineffective in both sectors, then the economy “eats up” capital and degrades.

The results mean that effectiveness of import substitution should play an important role in decision making on industrial policy in Russia in order for the economy to optimally adjust to the evolving new reality. Widescale import substitution is generally not optimal.

Our work contributes to the existing economic research studying the effect of sanctions on the dynamics of Russian economy, including its structural shifts. Earlier estimates have been obtained in the literature using semi-structural and structural models. Pestova and Mamonov (2019) employ a BVAR model to estimate the economic effects of financial sanctions imposed on Russia in 2014. It has been demonstrated that the sanctions led to a depression of GDP growth rate in Russia by 1.2 pps. In Gurvich and Prilepskiy (2015) the impact of financial sanctions on the Russian economy is evaluated by modelling capital flow components. The study shows a significant impact of sanctions on GDP. In a more recent study, Pestova et al. (2022) have evaluated the economic effects of sanctions on Russia using a SVAR model. Findings suggest that industrial production, consumption, and investments should decline short-term; Russian GDP was predicted to decrease by 12.5–16.5% in 2022. Using computable general equilibrium modelling, Mahlstein et al. (2022) explore the effects of trade embargo on Russia. According to the estimates, real GDP of Russia contracts by over 14%. Kumagai et al. (2022) utilize a computable geographic general equilibrium simulation model (IDE‑GSM) to evaluate the impact of sanctions on Russia and the global economy. Monastyrenko and Picard (2023) use a static Ricardian trade model to study the implications of trade embargoes which are (or may be) imposed against Russian fossil fuels (coal, gas, crude oil, and refined products of coal and oil) by the G7 countries, the EU countries, the OECD countries, and the union of three coalitions for the welfare and economic activity in Russia. The researchers find that each embargo results in Russia experiencing a trade export loss equivalent to a few percent of its GDP (with cumulative loss amounting to 10%). The demand from sanctioning countries is only partially substituted by the demand from China and India. As for the impacts on welfare, a total ban on all Russian fossil fuels by the coalition of OECD countries diminishes the Russian welfare (total utility) by more than 16%. The sanction sender countries also experience decline in their welfare; however, this adverse effect is lower on average (3.25% for EU-only coalition and less than 1% for other coalitions) but heterogenous across nations.

Imbs and Pauwels (2023) and Hosoe (2023) also concentrate on trade sanctions and reach qualitatively similar conclusions by employing various computable general equilibrium models of world trade.

Notably, all the aforementioned papers use static modelling framework whereas in this paper we resort to a dynamic (albeit highly stylized) model. In particular, it enables us to account for the long-run consequences of disrupted trade such as declining total factor productivity.

Existing estimates for Russia are consistent with the research studies­ conducted for other countries. Gutmann et al. (2021) have analyzed the impact of sanctions on economic growth in 162 countries from 1960 to 2016. The study identifies significant negative effects of international sanctions on GDP growth and its components (consumption, investment, and government spending), as well as on trade and foreign direct investment. In Ianchovichina et al. (2016) a global general equilibrium simulation model is employed to quantify the effects of lifting economic sanctions on Iran with and without strategic responses. It has been demonstrated that if economic sanctions on Iran are lifted, its average per capita welfare would grow by 3.7%, primarily due to the lifting of the oil embargo imposed by the EU. In Gharehgozli (2017) the economic costs of sanctions on Iran are evaluated using the synthetic control method: over three years, sanctions reduced GDP by 17.3%. A multi-regional computable general equilibrium model is used in Gharibnavaz and Waschik (2018) to evaluate the effects of sanctions on Iran. It has been demonstrated that sanctions led to an uneven welfare decline among households in the Iranian economy by 14–15%.

Comparing to existing papers, which focus on positive issues, we also address­ normative issues — try to find the optimal reallocation of resources in the sanctions-hit economy. We do this by considering a centralized equ­ilibrium set by a planner acting to maximize social welfare (flow of consumption). Our study attempts to determine the adequate response of the economy under the new reality­ of import and export restrictions, depending on assumptions about the possibilities of import substitution within a structural dynamic general equilibrium model calibrated on Russian data.

This paper seeks to explore the equilibrium and equilibrium dynamics in a modified Ramsey model with two types of goods — consumer and investment.¹⁰

The problem is formulated as follows. An agent acting in the public interest aims to maximise the discounted sum of the utility flow from consumption:

∑_t β^t ln C_t. (1)

Fixed volume of labour L, together with capital K_t, is used to create a domestic intermediate product Y_t with the Cobb–Douglas production function:

Y_t = A_t (L)^α (K_t)^{1– α}. (2)

The dynamics of capital stock is determined by the standard relationship:

K_t = (1 – δ)K_{t –1} + I_t, (3)

where I_t is gross investments.

The country imports foreign intermediate goods Imp_t in exchange for export (endowment), Ex_t.11

The produced intermediate product Y_t is used in several ways:

• • as an intermediate product (input) C _t ^Y in the production of final consumer goods C _t ;
• • as an intermediate product (input) I _t ^Y for the production of capital goods I _t ;
• • as a substitute C _t ^S for imported intermediate consumer goods C _t ^imp used in the production of final consumer goods C _t ;
• • as a substitute I _t ^S for imported investment intermediate goods I _t ^imp used in the production of final investment goods I _t :

Y_t = C_t^Y + I_t^Y + C_t^S + I_t^S. (4)

Therefore, using labour and capital, the economy has the potential to produce import-substituting intermediate consumer and capital goods, but will not necessarily do so. It depends on the solution of the optimization problem, whether it is profitable to allocate resources this way. We assume that before the introduction of restrictions, the producers do not spend resources on import substitution due to its inefficiency. To put it simply, we calibrate the efficiency indicators of import substitution as follows: even if we did impose a restriction “the output of import-substituting products is equal to zero” in times of a full economic openness, the restriction would not impact the economy’s equilibrium parameters (the restriction is “non-binding”).

An imported product Imp_t is an intermediate good, a portion of which C_t^imp is used in the production of final consumer goods, while the remainder I_t^imp is used in the production of investment goods (physical capital):

Imp_t = C_t^imp + I_t^imp. (5)

The final consumer good is derived by “packaging” of two types of products­ without any labour or capital costs: intermediate consumer products and imported consumer goods. The creation of a final consumer good can be described as follows: consumption necessitates the merging of an imported good with a domestic one. For instance, this could involve establishing a network of car dealerships (domestic service) that sell imported cars (imported goods), factories that assemble cars from imported parts, or a facility for packaging imported medicines. Prior to “packaging,” imported consumer goods can be substituted by domestic equivalents. The production function, or “packaging,” of the final consumer goods is a constant elasticity of substitution (CES) function:

$C_t=A_c{\left[{\gamma }_C^{\frac{1}{{\eta }_C}}{\left(C_t^Y\right)}^{\frac{{\eta }_c-1}{{\eta }_C}}+{\left(1-{\gamma }_C\right)}^{\frac{1}{{\eta }_C}}{\left(C_t^{imp}+k_C{C}_t^S\right)}^{\frac{{\eta }_C-1}{{\eta }_C}}\right]}^{\frac{{\eta }_C}{{\eta }_C-1}}$.(6)

According to the model calibration, a domestic intermediate product C_t^Y and an imported intermediate product C_t^imp are imperfect substitutes. Imported consumer goods and import-substituting goods are perfect substitutes but adjusted for the quality of the domestic import-substituting product.

The parameter κ_C will play a significant role in the subsequent evaluation of the aggregate and structural effects of the economic adjustment to the new reality­. This parameter determines the comparability of imports and their domestic equivalents. Let κ_C be equal to 0.1. Therefore, to replace one unit of imported goods, we need to use ten units of its equivalent. In economic terms, this can be explained as follows: imported goods last longer than domestic ones. Or, from a consumer’s perspective, imported goods are of higher quality (more delicious, etc.).¹²

A capital good is obtained by “packaging” two products without labour or capital costs: an intermediate capital product, which requires labour and capital, and an imported capital good. Before packaging, imported capital goods can be substituted by domestic equivalents:

$I_t=A_I{\left[{\gamma }_I^{\frac{1}{{\eta }_I}}{\left(I_t^Y\right)}^{\frac{{\eta }_I-1}{{\eta }_I}}+{\left(1-{\gamma }_I\right)}^{\frac{1}{\eta I}}{\left(I_t^{imp}+k_I{I}_t^S\right)}^{\frac{{\eta }_I-1}{{\eta }_I}}\right]}^{\frac{{\eta }_I}{{\eta }_I-1}}$. (7)

In the absence of specific restrictions, an imported product Imp_t is exchanged on the foreign market for an export product that is constantly available and requires no production costs Ex (endowment).¹³ The condition for exchange with the external world is defined as:

P_t^ex Ex_t + B_t – (r + 1)B_{t –1} = P_t^im Imp_t, (8)

where: P_t^ex, P_t^im are the externally set real export and import prices; r — is the constant interest rate on foreign assets/liabilities.14

A key assumption concerns net foreign assets (NFA, or net external debt, if B_t > 0):

B_t ≤ B_lim. (9)

We sequentially perform several types of calculations. As a starting point for model calculations, we use data on the Russian economy in 2019. We chose this year instead of 2022 because the volume and structure of the economy were not distorted by the pandemic effects. To determine the basis for comparison — the dynamics of the economy in the absence of restrictions, we make a simplifying assumption. We assume that in a completely open economy (open financial account), the amount of NFA in the model economy in 2019 is at a level corresponding to the long-run equilibrium. From 2019 onwards, the model economy stopped accumulating or spending NFA. Since the volume of NFA was positive in 2019 ($490 billion), the economy could use NFA revenues to finance the foreign trade deficit, i.e. increase imports.15 This implies a change in the consumption of imported goods, as well as a corresponding restructuring of production and fixed capital stocks.16 For other scenarios, we assume this equation (9) is met, but for a different reason: according to our description of the new reality, the model economy remains with NFA equal to the external public debt, which continues to be served, but positive NFA do not accumulate and the external debt does not increase.¹⁷

Ultimately, the planner’s problem is to maximise (1) under constraints (2) – (9) setting variables C_t, Y_t, I_t, Imp_t, K_t, C_t^Y, I_t^Y, C_t^imp, B_t. In the model, for some scenarios­, we assume that the production function parameter responsible for TFP could decrease with an increased share of import substitution in GDP:¹⁸

A_t = ω × e^–θ(C_t^{S +I}_t^S)/GDP. (10)

The following simple observation serves as empirical evidence of the following premise: in the absence of import restrictions, firms import a large number of intermediate­ goods. This implies that for firms it is more advantageous (from a cost or quality­ perspective) than acquiring domestic equivalents (import substitution).¹⁹

We examine two mechanisms of import substitution efficiency:

• • Quality comparability with imported goods from a consumption/investment perspective. Lower comparability implies increased consumption/investment of domestic equivalents;
• • Cost escalation (reduction in total factor productivity) due to use of domestic import substitutes in production. The endogeneity of TFP implies that the more an economy produces import-substituting goods, which are subsequently packaged for final consumption or investment, the farther it is from the global frontier of maximum possible productivity. This premise is backed by simple logic: when a firm has the option to either import or substitute imports, and it chooses to import, it indicates that this option is more beneficial for it (as it allows for cost reduction). Consequently, when there are no alternatives and only a less favoured option remains, choosing it ends up with cost escalation (and productivity decline).

GDP measured by the output method is composed of the value added at all stages of production, including intermediate and final goods. Since no added value is created in the production of final goods (there are no labour and capital costs), the GDP in such an economy will be made up of the value added in the production of intermediate goods. None of these goods are exported. Exports should be added as another component of GDP — this represents the flow of commodity income created without labour with specific capital, which remains constant in further calculations:²⁰

GDP_t = Y_t + Ex_t. (11)

For the optimization problem defined by the objective function (1), constraints (2) – (9), and non-negativity constraints on the control variables, we construct a Lagrangian and compute its first derivatives with respect to control variables and dual variables. For simplicity, we omit the full list of first-order conditions (which also includes “complementary slackness” conditions for removing inequality-type constraints), considering only a few of them. Under the conditions of the Baseline Scenario (see below) for all variables, except B_t, C_t^S, I_t^S, the solution will be internal­, meaning the first-order conditions are the corresponding Lagrangian derivatives equalling zero. From the system of equations and “complementary slackness” conditions obtained in this way, it easily follows that if 1/β > 1 + r, then B_t = B_lim.²¹

Furthermore, it follows from the first-order conditions that for κ_C ≠ κ_I both variables C_t^S, I_t^S cannot be positive simultaneously.²² Numerical modelling can be used to confirm that:

$\frac{\partial L}{\partial I_t^S}=0, \frac{\partial L}{\partial C_t^S}=0,$ (13)

$I_t^S\times \frac{\partial L}{\partial I_t^S}=0, C_t^S\times \frac{\partial L}{\partial C_t^S}=0,$ (14)

where L represents the Lagrangian.²³

Hereinafter, when calibrating, we use the characteristics of the Russian economy in 2019: Russia’s exports, according to the data of Rosstat’s System of National Accounts (SNA), amounted to ₽32 trillion ($419 billion according to the Bank of Russia), the NFA level was $491 billion according to the Bank of Russia (₽29 trillion at the exchange rate of ₽60 per US dollar). We assume the real interest rate to be 2.5%, so the annual income in equilibrium will be 0.025 × ₽29 trillion. The capital stock was assumed to be ₽250 trillion, which is the average of two Rosstat estimates.24 The labour stock was assumed to be equal to the average annual size of the labour force in 2019, at 75.4 million people. Collectively, this represents the resources available to the economy when it is in equilibrium (prior to shocks). The economy had a surplus in foreign trade, continuing to accumulate NFA. The list of calibrated parameters, their values, and sources of calibration are provided in Appendix A (Tables A1–A2). The model operates on a quarterly frequency, but we present all results on an annual basis.

Firstly, we calculate the basic equilibrium and the path of convergence to this equilibrium from the starting point (2019). We call it the Baseline Scenario. In the basic equilibrium, the economy ceases to accumulate net foreign assets, maintaining them at the 2019 level. The problem is solved in all cases under the assumption of perfect foresight: at the moment of the shock, economic agents are aware of the entire future trajectory of exports, imports, and NFA. Since NFA in 2019 were positive, the economy annually received interest income from them, which could be directed towards consumption and investment. (For simplicity, we disregard a more complex problem as a basic option: the search for the best possible NFA size. This would require additional assumptions about future trajectories of income from commodity exports.) Our basic equilibrium is a hypothetical trajectory of macroeconomic indicators that corresponds to the maximum level of public welfare, under the condition of cessation of NFA accumulation from 2020. The main parameters of the basic equilibrium and other scenarios are given in Appendix A (Table A3). All subsequent scenarios will be presented in relation to the baseline equilibrium.

First, we calculate an alternative that involves only closure of the financial account, without imposing restrictions on the export-import flows — this is Scenario A or “reduction of NFA.” In comparison to the Baseline Scenario, the only change in such a long-run equilibrium is the NFA size. NFA shift from positive to negative, at a level equivalent to the size of the public external debt.25 We compute the equilibrium and the optimal trajectory towards it. This scenario allows us to examine the economic and structural implications solely associated with the inability to expend NFA and earn income from them.26

We then compute the Shock Scenario in three variations of the model para­meters, which are responsible for the effectiveness of import substitution. In this scenario, not only does the financial account “close” and NFA decrease, but export and import flows also decline. Furthermore, the decrease in imports is more abrupt and significant than that of exports.

Therefore, the shock that we examine for the three variations of the model parameters is, firstly, characterised by restrictions on capital flows and the freezing of NFA, and secondly, by limitations on imports and exports. As a result, the economy­ loses the ability to use both the income from previously accumulated NFA and the NFA reserve itself to finance imports. Hence, the results of Scenario A are taken into account in subsequent calculations as part of the Shock Scenario.

In each option of the scenario, we introduce an import volume restriction (without restricting the import structure) and an export value restriction. These restrictions ensure that the economy maintains a positive trade balance after they are imposed, but it does not accumulate NFA (as per the first part of the Shock Scenario). Economic agents would like to use the export resources, though reduced, for imports, but they are unable to do so due to sanctions on import. Therefore, the central planner is challenged to find a steady state under these restrictions, where the economy formally has a positive trade balance but does not use the NFA stock: does not spend it to repay external debt (debt principal) or does not accumulate assets, but only uses a portion of it to pay interest on the external public debt.²⁷ In such circumstances, we examine three alternatives for the efficiency of import substitution:

• • Scenario B (Optimistic) assumes the ease of creating full equivalents of imports and no TFP losses when using import-substituting goods in production.
• • Scenario C (Neutral) suggests incomplete import substitution in terms of quality and the presence of asymmetry in the ability to replace imports: it’s more difficult to replace investment goods than consumer goods; in addition, there are losses in total factor productivity due to the use of domestic import equivalents. 28
• • Scenario D (Pessimistic) suggests low import substitution in terms of quality with symmetry of import substitution in the investment and consumer sectors. ²⁹

Scenario conditions for each scenario are presented in Appendix A (Table A3).

The dynamics in the Baseline Scenario are presented in Appendix B. An economy that ceases to accumulate NFA when having positive trade balance has the potential to significantly increase consumption. It is optimal to direct additional resources primarily towards consumption and, whatever remains, towards investments. Once the accumulation of NFA from the foreign trade surplus ceases, a “permanent income” emerges, which can be properly used to enhance the standard of living over a long-term horizon. In the Baseline, net investments are positive and facilitate the accumulation of capital, an increase in capital intensity, leading to a higher level of production and income. In this case, due to the ability to import goods and investments, there is no need for import substitution. Moreover, in the Baseline Scenario, the proportion of imports in investments is increasing.

In Scenario A, which involves restrictions only in a form of reduced NFA through asset freezing, the changes compared to the Baseline Scenario are not very significant (see Appendix C). The economy loses access to the available­ NFA but continues to service the public debt from NFA revenues (significantly less than the initial NFA amount). Meanwhile, an increase in external borrowing is impossible. Why do the results turn out to be close regardless of actual nullifying of the previously positive NFA? For a country that decides to stop accumulating NFA in the Baseline Scenario, the NFA amount no longer changes, and only interest income is spent on NFA. When a country loses access to such NFA, from the perspective of equilibrium dynamics, only the income from NFA changes — the country loses them (but payments for servicing the external public­ debt remain). In the Baseline Scenario, the economy did not plan to spend the foreign asset stock in the future, only the income from the stock. Therefore, when the economy loses access to the foreign asset stock, it only loses the income from these assets. Thus, the differences in the best possible dynamics in the asset freeze scenario compared to the basic equilibrium are minor.

We have modelled the effects of financial autarky (with net foreign assets­ frozen) and a significant reduction in imports, and later exports, for a commodity-exporting economy in general equilibrium using a modified centralised Ramsey model. We assume that imported goods are used for “packaging” along with domestic goods in the production of final consumer goods and investment goods. We highlight ­differences in the optimal dynamics that maximise a standard social welfare metric, which may arise due to variation in efficiency of import substitution. By this, we imply both the comparability of domestic substitutes with imports (the higher it is, the higher the efficiency is) for the consumers and the dependence of total factor productivity on the use of substituted intermediate goods.

Trade balance sanctions make import substitution preferable, but only in sectors where such substitution is relatively more efficient. In our study, this is the production of consumer goods. If domestic substitutes are comparable to the original products in their characteristics, the best strategy involves capital accumulation, increased capital–labour ratio, and intensive import substitution (in both sectors). This is associated with a decrease in the social welfare due to limited resources in the wake of decreasing exports. As a result, GDP may increase, but consumption will be below the baseline, i.e., the share of consumption in GDP will drop.

Asymmetry in the efficiency of import substitution between sectors leads to distortions in the structure of imports and production. Inefficient import substitution of capital goods necessitates a reorientation of imports towards investment goods. If import substitution is inefficient in both sectors, the economy “eats out” its capital stock and degrades.

Our model economy is centralized, but we can speculate on the inflation ­response in a decentralized formulation of it. Thus, in a market economy, a shift in import demand towards import of investments, as in the neutral Scenario C, needs a significant increase in the cost of production of domestic (import-substituting) investment goods relative to consumer goods. This ­contributes to the reorientation of the economy towards intensive import substitution of consumer goods, which should lead to an increase in wages and prices in the consumer sector. Therefore, the asymmetry in the efficiency of import substitution favouring the consumer sector is likely to generate unwelcome inflation effects.

There is room for further research in this area. Our findings may prove ­beneficial not only for a realistic comprehension of the potential paths of economic develop­ment in the new reality, but also when examining the green transition effects in economies heavily reliant on hydrocarbon exports.

Note: All values, unless otherwise stated, are in trillion rubles of 2019. The x-axis represents the year since “the new reality” beginning. GVA — Gross Value Added.

Sources: Rosstat; author’s calculations.

Note: All values, unless otherwise stated, are in percentage relative to the Baseline Scenario. The x-axis represents the year since “the new reality” beginning. GVA — Gross Value Added.

Sources: Rosstat; author’s calculations.

Note: All values, unless otherwise stated, are in percentage relative to the Baseline Scenario. The x-axis represents the year since “the new reality” beginning. GVA — Gross Value Added.

Sources: Rosstat; author’s calculations.

Note: All values, unless otherwise stated, are in percentage relative to the Baseline Scenario. The x-axis represents the year since “the new reality” beginning. GVA — Gross Value Added.

Sources: Rosstat; author’s calculations.

Note: All values, unless otherwise stated, are in percentage relative to the Baseline Scenario. The x-axis represents the year since “the new reality” beginning. GVA — Gross Value Added.

Sources: Rosstat; author’s calculations.