Measuring Smartness: Understanding the
Economic Impact of Targeted Sanctions
Daniel P. Ahn
Deputy Chief Economist
U.S. Department of State
Rodney Ludema
Georgetown University

U.S. Department of State
Office of the Chief Economist
Working Paper 2017-01

The Office of the Chief Economist (OCE) Working Paper Series allows the OCE to
disseminate preliminary research findings in a format intended to generate
discussion and critical comments. Papers in the OCE Working Paper Series are
works in progress and subject to revision. View and opinions expressed do not
necessarily represent official positions or policy of the OCE or Department of State.
Comments and suggestions for improvement are welcome and should be directed to
the authors.

 About OCE:
The Office of the Chief Economist (OCE) was created in 2011 as an essential tool for
advancing economic statecraft as a U.S. foreign policy priority. Our work supports
the Department of State’s capacity and resources to engage effectively with our
partners in a world where economic factors are central to diplomacy. The Chief
Economist reports to the Secretary of State through the Under Secretary for
Economic Growth, Energy, and the Environment.
Our mission in OCE has three parts: strategic advice, capacity building and outreach.
Our primary role is to provide expert analysis and timely advice to the Secretary and
other senior leaders across the Department. OCE monitors and explains current
economic and financial trends in key countries and regions. We both analyze
contemporaneous, high-priority issues where economics and foreign policy
intersect and engage in deeper research into longer term global economic trends.
Contact OCE at: OCE@state.gov

 Measuring Smartness:
Understanding the Economic Impact of Targeted Sanctions
by
Daniel P. Ahn† and Rodney D. Ludema‡*
December 2016

Abstract
While broad economic sanctions and trade embargoes have long been used as instruments
of foreign policy, targeted sanctions focusing on specific individuals, entities, and
transactions are relatively new and less understood. Using detailed firm-level data, this
paper provides empirical estimates of the impact of targeted sanctions, focusing on the
case of sanctions deployed by the United States and the European Union against Russia
after its intervention in Ukraine in 2014 as a natural experiment. Our main finding is the
average sanctioned company or associated company loses about one-third of its operating
revenue, over one-half of its asset value, and about one-third of its employees relative to
their non-sanctioned peers. We also survey the literature which suggests that sanctions
have had a relatively smaller impact on Russia’s macroeconomy compared to oil prices.
Together, these results indicate that the sanctions are quite “smart,” in the sense of hitting
the intended targets with relatively minimal collateral damage.



U.S. Department of State and Johns Hopkins University, School of Advanced International Studies.
Email: pahn4@jhu.edu
‡Georgetown University, Department of Economics and School of Foreign Service and U.S.
Department of State. Email: ludemar@georgetown.edu
* The Office of the Chief Economist (OCE) Working Paper Series allows the OCE to disseminate
preliminary research findings in a format intended to generate discussion and critical comments.
Papers in the OCE Working Paper Series are works in progress and subject to revision. Views and
opinions expressed do not necessarily represent official positions or policy of the OCE or the U.S.
Department of State. Comments and suggestions for improvements are welcome and should be
directed to the authors.

 I. Introduction
While broad economic sanctions and trade embargoes have long been used
as instruments of foreign policy, targeted sanctions focusing on specific individuals,
entities, and transactions are relatively new. A prominent example of this new
approach to sanctions policy is the targeted sanctions program of the United States
and the European Union (EU), imposed in response to Russia’s annexation of Crimea
and its alleged use of force in eastern Ukraine. The U.S. and EU targeted a select list
of Russian individuals and companies starting in March 2014. Over the next two
years, the list of targets grew in scope and was met by Russian counter-sanctions on
agricultural imports.
Understanding the impact of these targeted sanctions is essential to
assessing their efficacy. Clouding the picture, however, is the fact that the conflict in
2014 roughly coincided with a series of powerful macroeconomic shocks, especially
a dramatic decline in the price of oil (Russia’s main export), which jolted both the
Russian and world economies.
The difficulty inherent in attributing Russia’s poor economic performance
following sanctions to a single cause has allowed for a wide range of conflicting
claims regarding the economic costs of the sanctions to Russia and to neighboring
economies (principally members of the EU). Opponents of sanctions, in particular,
claim that sanctions have caused little pain to the specific targets, while inflicting
untold economic damage on the Russian people and on neighboring countries. Such
assertions pose a challenge for the sanctions policy, which was intended to be

1

 “smart,” in the sense of hitting those responsible for the offending policy with
minimal collateral damage.
The purpose of this paper is to fill this information gap with empirical
evidence. While several credible institutions provided model-based estimates of the
potential impact of sanctions in the months after they went into effect, and others
have chronicled the collapse of the Russian economy and its trade with the rest of
the world, this paper uses pre- and post-sanctions data to measure the actual effect
of sanctions and counter-sanctions from the surrounding macroeconomic shocks.
To measure the “smartness” of the sanctions’ impact, we proceed in two
steps. First, we examine whether the sanctions hit in the intended targets. We do
this using data on individual firms from Bureau Van Dijk’s ORBIS database. To our
knowledge, this paper represents the only paper that uses detailed firm-level data to
understand the economic impact on the targets themselves. Our main finding is that
sanctioned companies or those associated with sanctioned individuals are indeed
harmed by sanctions relative to non-sanctioned peer companies. On average, a
sanctioned or associated company loses an estimated one-third of its operating
revenue, over one-half of its asset value and about one-third of its employees after
being targeted compared to non-sanctioned companies or those not associated with
sanctioned companies.

These estimates, which are large and appear highly

statistically robust, suggest targeted sanctions do have a powerful impact on the
targets themselves.
Second, we examine the collateral damage. In particular, we consider impact
the sanctions on Russian GDP and on its imports from the EU. In contrast to the

2

 firm-level approach, which fully controls for macroeconomic shocks with time
varying fixed effects, it is not possible to control for all conceivable confounders in
the aggregate analysis. Instead, we dial back our ambition and consider only how
much of the post-sanction performance of the Russian economy can be explained
either directly or indirectly by falling oil prices, with the residual capturing the
combined effect of sanctions and any other exogenous factors. We find that oil price
volatility explains the vast majority of the decline in Russia’s GDP and import
demand, with very little left to be explained by sanctions or other factors. Thus
either sanctions had only a small negative effect on these variables or other positive
factors largely cancelled out the effect of sanctions. The most plausible candidate
factor would be the Russian policy response, which is probably not exogenous to
sanctions.
Finally, we find that sanctions and countersanctions have had a small effect
on the economies of most EU countries. Adding together the impacts of sanctions
and countersanctions on exports – the main vulnerability – gives a median impact
across EU countries of just -0.13 percent of GDP. The reasons for this are: (1)
Russia generally accounts for a small share of total EU countries’ exports; and (2)
most of the decline in Russian imports is explained by lower oil prices and trend
factors.
Economic sanctions are meant to signal international disapproval, deter
further aggression, and create leverage in negotiations with the targeted country
aimed at reversing the offending policies. Whether sanctions will ultimately

3

 accomplish these goals is a key question but is beyond the scope of this paper.1 If
success depends on the sanctions delivering focused impact with minimal collateral
damage, our results suggest the targeted sanctions are acting as advertised.
The remainder of the paper is as follows. Section II reviews the literature on
the economic impact of sanctions against Russia. Section III provides a description
of the targeted sanctions program. Section IV presents the empirical methodology
and analysis and results of the firm-level approach, along with various robustness
checks. Section VI considers the collateral damage to the broader economic
performance of the Russian and EU economies. Section V concludes.

See Harrell (2015) for further discussion on the policy implications drawn from the Russian
sanctions experience.
1

4

 II. Literature Review
The literature, both theoretical and empirical, on economic sanctions is vast.
However, most empirical studies cover the period when policymakers invoked
comprehensive sanctions involving broad country-wide trade embargoes rather
than selective sanctions against specific targets.
The academic literature on these comprehensive sanctions, such as those
against Iraq after the 1990-91 Gulf War, has tended to give only mixed support for
the effectiveness of sanctions. Many studies have focused on the secondary effects of
sanctions on corruption and humanitarian consequences without matching political
dividends. Recent surveys are provided in Hufbauer et al. (2007) and Drezner
(2011).
Empirical studies of targeted sanctions are much fewer, partially due to the
relatively short history of targeted sanctions programs and the related paucity of
examples involving purely targeted sanctions programs instead of becoming part of
a broader comprehensive sanctions program. Indeed, U.S.-EU targeted sanctions
against Russia in the wake the Ukraine crisis represents a fairly rare example of
purely targeted sanctions against a middle-income economy fairly well integrated
into the global economic and financial system.
Economists attempting to empirically estimate the impact of sanctions on
Russia face the challenge of disentangling the impact of sanctions from the
confounding effects of the broader political uncertainty stemming from the Ukraine
crisis and the dramatic drop in oil prices discussed above. However, most studies

5

 conclude that oil prices were far more important in explaining Russia’s post-2014
macroeconomic weakness, with a relatively small effect ascribed to sanctions.
An IMF (2015) report, using a generic macroeconomic model, forecasted that
sanctions could reduce Russia’s real output by about 1 to 1.5 percent of GDP, via
weaker investment and consumption. A World Bank (2015) study similarly argues
that sanctions against Russia and counter-sanctions may have affected investment
and consumption but does not provide any specific numbers. Neither of these
studies attempts to directly measure the economic impact of sanctions.
Dreger et al. (2015) used a VAR model featuring oil prices, the ruble
exchange rate, and a sanctions news index to argue that the oil price drop was the
primary driver of the ruble depreciation but that sanctions news surprises may have
had some impact on the ruble’s conditional volatility. The study does not consider
the effect on GDP growth or imports.
Tuzova and Qayub (2016) presents another reduced-form VAR model
featuring a variety of Russian macroeconomic variables including GDP, the real
exchange rate, inflation, fiscal and consumption expenditures, and external trade to
argue that oil prices were the main cause of Russia’s poor macroeconomic outlook.
Moret et al. (2016) also looked at country-level trade data and compared
trading volumes between Russia and the EU pre- and post-crisis to conclude that the
Baltic states suffered the greatest relative losses, similar to what we find Section V of
this paper.2
In another chapter of this volume, Crozet and Hinz (2016) use detailed monthly
trade data at both the country level and firm level for France to estimate the impact
of sanctions on trading activity. They find large impacts on trade (as much as $60
2

6

 Lacking from all of this literature is the use of firm-level data to estimate
empirically the impact on the economic performance of the targets themselves. The
closest our paper is Stone (2016), which uses an event study methodology to study
the impact of sanctions news events on the asset prices of sanctioned sectors and
the 11 largest energy firms and banks. The paper finds a negative impact on the
asset prices of targeted sectors compared to non-targeted sectors but no significant
difference between targeted and non-targeted firms within a sector.
Compared to the literature, our paper is unique in several respects. First,
ours is the first to use detailed firm-level data to estimate empirically the impact on
the real performance, such as operating revenue and employment, of targeted firms
themselves. Second, we consider a comprehensive sample of firms, including all of
targeted private firms (not only publically traded financial and energy firms) as well
as firms associated with sanctioned individuals. Third, we find strong evidence that
sanctioned firms are indeed harmed by sanctions relative to their non-sanctioned
peers.

billion collectively) between Russia and many EU economies. The authors recognize
that most of this effect is due to oil prices, Russian economic underperformance, and
political uncertainty. The effect from targeted sanctions appears to be at most $3
billion.

7

 III. Description of Targeted Sanctions
III.a. Overview of U.S. and EU Targeted Sanctions Policy against Russia
On March 6, 2014, President Obama declared a national emergency and
issued the first of four Executive Orders to deal with the threat posed by the
situation in Ukraine, including the actions of the government of the Russian
Federation. Issued in March and December of 2014, these Executive Orders
provided the authority for various agencies of the U.S. Government, including
Treasury, State, Commerce, and others, to impose targeted sanctions on primarily
Russian entities.3 The U.S. Department of Treasury’s Office of Foreign Asset Control
is the primary entity responsible for implementing targeted sanctions. The targets
are determined by the Secretary of the Treasury, in consultation with the Secretary
of State, after a careful investigation and vetting process.
U.S. targeted sanctions with respect to the Ukraine/Russia crisis broadly fall
into two categories:


SDN Sanctions: Blocking sanctions against individuals and entities on the
List of Specially Designated Nationals and Blocked Persons (SDN) List



SSI Sanctions: Sectoral sanctions against entities operating in the financial,
energy, and defense sectors of the Russian economy listed on the Sectoral
Sanctions Identification (SSI) List

Designated SDN entities and individuals face asset freezes and travel bans in the
United States. And unless otherwise authorized or exempt, all trade and financial

These are E.O. 13660 (March 6, 2014), E.O. 13661 (March 16, 2014), E.O. 13662 (March 20, 2014),
and E.O. 13685 (December 19, 2014).
3

8

 transactions and other activities by U.S. persons (individuals or entities) with these
designated SDN individuals and entities are prohibited.4
Meanwhile, the SSI entities represent those entities for which U.S. persons
are prohibited from engaging in certain transactions. Notably, U.S. persons cannot
transact in or issue debt of longer than 30 days maturity or acquire new equity with
targeted companies in the Russian financial sector or the Russian defense sector.
Similar restrictions also apply to debt of longer than 90 days maturity with targeted
companies in the Russian energy sector. Furthermore, U.S. persons are prohibited
from transacting in certain technology and services related to deep-water, Arctic
offshore, or shale oil activity with the Russian energy sector.
The EU also developed a targeted sanctions policy in the form of EU Council
Regulations starting in March 2014.5 From these EU Council Regulations, the
European Union maintains a categorization of targeted sanctions similar to that of
the United States in response to the events in Ukraine/Russia in 2014:


Restricted Measures List: Asset freezes and visa bans apply to entities and
individuals designated on the EU Restricted Measures List.



Sectoral Sanctions List: The EU also prohibits EU nationals and companies
from transacting in equity or debt instruments with a maturity exceeding 30
days with entities on the EU Sectoral Sanctions List.
Similar to that of the U.S. sectoral sanctions, the EU imposed an embargo on

trade in arms and dual-use goods/technology with Russia, covering all items on the
50% Rule: Transactions with an entity that is 50 percent or more owned, whether individually or in
the aggregate, directly or indirectly, by an SDN designated entity or individual, is also blocked,
regardless of whether the entity itself is listed.
5 EU Council Regulations No. 269/2014, 284/2014, 433, 2014, 833/2014, and No 960/2014.
4

9

 EU common military and dual-use lists, as well as on the export of certain energyrelated equipment and technology for offshore deep-water, Arctic, or shale oil
exploration and production.
III.b. Identifying U.S. and EU Sanction Targets
This section discusses the number of targets selected by the U.S. and the EU
for targeted sanctions. Figure 1 shows as of July 1, 2016 the overview of entities or
individuals explicitly sanctioned by the U.S. and EU authorities, beginning on March
17, 2014.
Altogether, the U.S. has designated 111 individuals and 82 entities on its SDN
List as related to its Russia/Ukraine-related sanctions program. Also, the U.S. has
explicitly designated 136 entities on its SSI List facing sectoral sanctions. Meanwhile,
as of July 1, 2016, the EU had placed 149 individuals and 37 entities on its EU
Restricted Measures List and 20 entities onto its EU Sectoral Sanctions List. Figure 2
shows the timeline of these sanctions.
We have classified sanctioned individuals into two categories: political and
business figures. A political figure is an individual whose primary occupation
appears to be political rather than commercial in nature, such as a legislator, a
government official, or a militia commander. However, a named individual would be
classified as a business figure (even if he/she had worked in public service) if the
individual could be identified as being “associated” with a company according to the
Bureau van Dijk standardized positions database. These associations include being
part of corporate management, on the board of directors, or a major shareholder.
According to this classification, of the 111 individuals on the U.S. SDN List, 87

10

 appear to be purely political figures, while 24 appear to have business associations.
Hence, of the U.S. SDN individuals, about one-fourth are business figures.
Meanwhile, the EU sanctions against individuals are skewed more heavily toward
political figures. Of the 149 individuals on the EU Restricted List, only six have
business associations.
We subsequently identified 269 companies as being “associated”, either in
the past or present, with a U.S. or EU sanctioned individual. Figures 3 and 4
presents a Venn diagram of the space of U.S. and EU sanctioned entities by category,
including associations with sanctioned individuals.
As discussed above, this tabulation captures those entities or individuals
explicitly listed by the U.S. and EU governments as facing sectoral restrictions.
However, both the U.S. and EU follow a 50% ownership rule whereby those
subsidiaries 50% or more owned, directly or indirectly, by an explicitly sanctioned
entity, also face the same sanctions. The United States appears to be more forwardleaning in explicitly identifying subsidiaries that should also face sanctions
according to this rule.6
Having described the U.S. and EU targeted sanctions policy against primarily
Russian targets in response to the crisis, we turn next to estimating the economic
impact of these sanctions on the targets themselves.

In ongoing research, the authors are investigating how much overlap there would be if we could
include those companies implicitly sanctioned via this ownership rule, and whether those companies
that are implicitly sanctioned also face the same economic effects as those explicitly sanctioned.
6

11

 IV. Firm-Level Economic Impact of Targeted Sanctions
In this section, we provide our methodological specification and the results of
our empirical analysis of the effects of targeted sanctions.
IV.a. Data
Our firm-level approach requires a careful examination of the sanctions lists
of the U.S. and the EU along with firm-level data covering sanctioned and nonsanctioned companies. Our firm-level data come from the Bureau Van Dijk (BvD)
Orbis database,7 and we track a universe of 78,381 companies.
These include 433 specific companies identified as being sanctioned in
Section III that are also present in the BvD database. The remainder is a control
group of peer companies, constructed by collecting all companies that share the
same home country and sector of business operation as the sanctioned companies in
the global BvD database.8
For each company, we track its home country location, sector of business
operation (according to the 4-digit NAICs code specification), and its total Operating
Revenue, Total Assets, and Number of Employees for the years 2013, 2014, and
2015. We also track the status of the firm, whether it remains active or whether it
has become bankrupt, liquidated or dissolved, or changed to some other non-active
status.
IV.b. Empirical Methodology

More information about the database can be found on Bureau van Dijk’s homepage at
www.bvdinfo.com.
8 Bureau van Dijk assigns a unique identification number that tracks a company through name and
ownership changes.
7

12

 Our econometric specification is a standard difference-in-differences
approach as follows:
ln 𝑦𝑖𝑠𝑐𝑡 = 𝛼𝑖 + 𝜆𝑠𝑡 + 𝜃𝑐𝑡 + 𝛽𝑑𝑖𝑡 + 𝜖𝑖𝑠𝑐𝑡
where the subscript i denotes company identification, s denotes sector, c denotes
country, and t denotes time period.
The left-hand side dependent variable 𝑦𝑖𝑠𝑐𝑡 tracks the particular firm’s
financial metrics of total Operating Revenue, Total Asset value, and Number of
Employees. Also, we construct a dummy variable which equals 1 if the firm is active
in that year, versus 0 if the firm loses its active status due to a bankruptcy,
liquidation, etc. The variables 𝛼𝑖 capture company fixed-effects, λst capture sectortime fixed effects, θct capture country-time fixed effects, and dit are the sanction
treatment dummies. Our sanctions dummies dit capture those periods when firm i
faces any of our three categories of targeted sanctions by either U.S. or EU
authorities.
Table 1: Cross-Sectional Regression Results

13

 Table 1 displays our headline results, which shows the coefficients from
regressing the log Operating Revenue, log Assets, and log Employee Count on our
sanctions dummy d_Sanc. Only for d_Active, our dummy indicating firm active status,
do we show the results of a probit regression. The number of observations can vary
across regression variables due missing or non-reported data for many companies.
We note that this result arises after controlling for both sector-time and
country-time fixed effects, which should eliminate oil factors and other factors that
may apply to companies in particular sectors or countries. The sectors are
determined by the 4-digit NAICS core code, providing a high degree of granularity.
The robustness of our results to these controls lends confidence in our results.
Table 2 converts the log coefficients in Table 1 into the estimated average
impact on the firm performance metrics. We find that targeted sanctions do have a
statistically significant negative impact on firm-level financial health relative to nonsanctioned peer companies. After facing targeted sanctions, a company or
associated entity on average faces a 2 percent increased likelihood of losing its
active status. Also, its operating revenue falls by about 30 percent, total assets by 55
percent, and employee count by about 30 percent compared to non-sanctioned or
non-associated companies. These results are highly statistically significant. The
significance is at the 1% confidence level for all of our financial health variables with
the exception of the status dummy variable, which is significant at the 5% level.
Table 2: Headline Sanctions Impact on Sanctioned and Associated Companies
Higher Probability of Bankruptcy

Estimated Average Impact on Indicators of Financial Health
Fall in Operating Revenue
Fall in Total Assets

Sanctions on Average
Target Company
2.3%**
28.5%***
55.3%***
* indicates significance at 10% confidence level, ** at 5% confidence level, *** at 1% confidence level

14

Fall in Number of Employees
30.1%***

 Although the magnitudes of the estimated losses are large, these results
should be interpreted with caution. In particular, one should not simply add the
firm-level losses of all targeted companies together to arrive at a macro-level impact.
The reasons are twofold. First, the effect does not necessarily apply uniformly to all
targeted companies, and may be smaller in proportional magnitude for larger-sized
targets. This is also in part because the largest companies tended to face only
sectoral sanctions which are deliberately designed not to have a large immediate
impact but affect their long-term health via their access to credit and technology.
This is one reason why the concentrated impact at the firm level does not
necessarily translate into a large macroeconomic impact, despite the target list
containing some of the largest Russian state-owned enterprises.
Second, the results capture the differential impact of sanctions on the
performance of targeted or associated companies compared to non-targeted
companies. They do not measure factors that might affect all firms equally in a
sector or a country. For example, if sanctions contributed to a depreciation in the
ruble, and this depreciation improved the performance of all firms, whether
specifically targeted or not, this impact of sanctions would not be reflected in the
performance differential exploited by the regression. For this reason, a separate
analysis needs to be done to capture macroeconomic impacts of the sanctions, as
discussed in the appendix.
IV.c. Robustness Checks
An important assumption underlying our difference-in-differences
methodology is the so-called “parallel trends” assumption, i.e., that targeted (or

15

 associated) firms would have experienced the same average change in performance
as their non-targeted peers (same sector and country) had they not been targeted.
This assumption could be violated if, for example, targeted companies tend to grow
more slowly than non-targeted peers for reasons other than sanctions.
It seems unlikely that there is an inherent bias for slow-growing firms to be
targeted. However, targeted firms were larger on average than their non-targeted
peers in 2013. If hypothetically larger firms tend to grow more slowly than smaller
firms, conditional on survival, then the parallel trends assumption may be violated,
and our estimates could be biased.
To address this concern, we use a matching estimator. For each targeted or
associated firm, a matching algorithm is used to find the non-sanctioned firm that is
most similar to the targeted firm in terms of industry, country, and size in 2013,
measured by assets.9 We then compare the average change in performance of all
such matched pairs between the pre- and post-sanction years. For robustness, we
consider two different matching estimators: a nearest-neighbor matching estimator,
which minimizes the Mahalanobis distance between matched pairs, and a
propensity score-matching estimator. The two estimators produce practically the
same results.
One complication in implementing this approach is that targeted or
associated firms are sanctioned at different times, some in 2014 and some in 2015.
Thus, we need to be careful that each matched pair is compared across the same
We use NAICs industries rather than 4-digit NAICs sectors to guarantee that the matching
estimation satisfies the common support assumption (i.e., that there are both sanctioned and nonsanctioned firms in each industry). For the same reason, we consider only two countries, Russia and
Ukraine, which comprise 97 percent of the companies in our sample.
9

16

 years, depending on when the targeted or associated firm was sanctioned. For firms
sanctioned in 2014 and their matched peers, we compare the difference in
performance between 2013 and the 2014-15 average. For firms sanctioned in 2015
and their matched peers, we compare the difference in performance between the
2013-14 average and 2015.
The results in Table 3, which show the average treatment effect (ATE), are in
line with our headline results reported in Table 1. In each case, the impact of
sanctions is to reduce the performance of targeted or associated firms relative to
their non-targeted peers. The results are statistically significant and the magnitudes
are consistently larger than in Table 1. This indicates that our difference-indifference estimates above are not only robust but likely conservative.

Table 3: Matching Estimation Results

Variables
ATE
Std.Err.
ATE
Std. Err.

(1)
(2)
(3)
DlOpRev
DlAsset
DlEmp
Propensity Score Matching
-0.8816*
-1.7081*** -0.6123***
(0.501)
(0.084)
(0.053)
Nearest Neighbor Matching
-0.8503** -1.5226*** -0.4455**
(0.417)
(0.124)
(0.176)

Observations
2817
***p 0.01,**p<0.05*p<0.1

17

6950

6408

 V. Conclusion
In 2014, the United States and the European Union deployed targeted
sanctions against Russia in response to the crisis in Ukraine. However, at the same
time, macroeconomic shocks, notably a fall in the price of oil, obscured the economic
impact of sanctions. This allowed partisans to make conflicting claims on the
economic efficacy of sanctions, with critics claiming a limited impact on the targets
while inflicting large damage onto collateral bystanders and neighboring economies.
Using firm-level empirical data, we find strong evidence that sanctions
affected the financial health of the targeted firms or firms associated with
sanctioned individuals and entities. U.S. and EU targeted sanctions appear to be as
“smart,” in the sense of hitting the intended targets and those companies associated
with them with significant economic damage while causing minimal collateral
damage.

18

 References
Crozet, M. and J. Hinz, 2016, “Collateral Damage: The Impact of Russia Sanctions on
Sanctioning Countries’ Exports,” CEPII Working Paper, June 2016.
Dreger, C., J. Fidrmuc, K. Kholodilin, and D. Ulbricht, 2015, “The Ruble between the
Hammer and the Anvil: Oil Prices and Economic Sanctions”, DIW Berlin Discussion
Paper No. 1488.
Drezner, Daniel, 2011, “Sanctions Sometimes Smart: Targeted Sanctions in Theory
and Practice,” International Studies Review, Volume 13, pages 96-108.
Harrell, Peter, 2015, Lessons From Russia for the Future of Sanctions, Economic
Statecraft Series, Center for a New American Security.
Hufbauer, G., C. Jeffrey, J. Schott, K.A. Elliott, and B. Oegg, 2007. Economic Sanctions
Reconsidered, 3rd Edition. Washington, DC: Peterson Institute for International
Economics.
International Monetary Fund, Staff Report for the Article IV Consultation: Russian
Federation, July 2015.
Moret, E., T. Biersteker, F. Giumelli, C. Portela, M. Veber, D. Jarosz, C. Bobocea, 2016,
“The New Deterrent: International Sanctions against Russia over the Ukraine Crisis,”
Institute of International and Development Studies in Geneva.
Tuzova, Y. and Faryal Qayum, 2016, “Global oil glut and sanctions: The impact on
Putin’s Russia”, Energy Policy, Vol. 90, March 2016, pages 140–151.
Stone, M., 2016, “The Response of Russian Security Prices to Economic Sanctions:
Policy Effectiveness and Transmission,” Working paper.
World Bank, Russia Economic Report No. 33: The Dawn of a New Economic Era?
April 2015.

19

 VI. A Simple Model of the Impact of Sanctions against Russia’s Macro-economy
In this section, we consider impact the sanctions on Russian GDP and on its
imports from the EU, both of which declined substantially after the imposition of
sanctions. We are interested in the contribution of sanctions to these declines, as
opposed to other macroeconomic shocks, such as the major decline in oil prices.
The answer is critical to the question of collateral damage, which is the second of the
two criteria for “smartness.”
In principle, sanctions can directly impact a country’s aggregate supply, by
affecting firms’ access to credit or technology, or by creating uncertainty for
investors (both domestic and foreign), leading to capital flight and the
postponement of investments and durable purchases. Sanctions may also impact
aggregate demand, via increasing the cost of borrowing or increasing precautionary
savings. However, these effects may have been mitigated in the short-run by ruble
depreciation and other government policies –such as expansionary fiscal policy and
Russia’s own embargo on agricultural imports – which passed the burden through
to higher inflation and the budget deficit.
As discussed above, the main empirical problem is that, while all of these
macroeconomic consequences were present in the Russian economy in the
immediate aftermath of sanctions, all of them could be explained by other factors.
As seen in Figure 6, world oil price (Brent) fell from over $100 in the 2014-Q2 to
under $60 by the end of 2014, and declined further in the second half of 2015. A
common rule of thumb for oil exporters suggests a $40 drop in the world price of oil
should shrink energy-dependent Russia’s GDP by four to five percent.

20

 But even before oil prices began falling, Russia’s growth had been on the
decline in recent years (Figure 7, black line) and growth projections were
downgraded beginning in late 2013 when the weak commitment of the Russian
regime to structural reform became clear. For example, the IMF nearly halved its
forecast for Russia’s 2014 growth rate between April 2013 and January 2014, well
before the Crimea aggression. Real GDP growth fell from 1.2 percent in 2013 to 0.7
percent in 2014, and -3.7 percent in 2015 before recovering slightly. From peak to
trough, real GDP fell about 5 percent. Import demand in U.S. dollars was about 37
percent lower.
VI.a. Strategy for Estimating the Macroeconomic Impact of Sanctions
Many factors can determine Russian macroeconomic variables such as GDP
growth and import demand. These include oil prices, sanctions and other factors,
such as consumer sentiment, investor expectations, exchange rates, and so on.
Suppose we can approximate the relationship with a linear function of the
form: 𝑔 = 𝛽0 + 𝛽1 𝑒 + 𝛽2 𝑠 + 𝛽3 𝑋, where g is the Russian macroeconomic variable of
interest, such as real GDP growth or import demand, e measures the change in oil
price (possibly including lags), s is a sanctions policy indictor, X is vector of other
factors, and {i} is a set of parameters.
Oil prices and sanctions are exogenous to the Russian economy, whereas X
contains many elements that are potentially influenced by oil prices and sanctions.
Similarly, we can represent X as 𝑋 = 𝛿0 + 𝛿1 𝑒 + 𝛿2 𝑠 + 𝛿3 𝑍, where 𝑍 a set of
exogenous factors uncorrelated with e or s.

21

 Putting these equations together gives a reduced-form equation of
𝑔 = (𝛽0 + 𝛽3 𝛿0 ) + (𝛽1 + 𝛽3 𝛿1 )𝑒 + (𝛽2 + 𝛽3 𝛿2 )𝑠 + 𝛽3 𝛿3 𝑍, which states GDP growth
is influenced both directly and indirectly by oil prices and sanctions and separately
by exogenous factors Z. Assuming the parameters of this model are stable, we can
estimate a regression, 𝑔𝑡 = 𝛾0 + 𝛾1 𝑒𝑡 + 𝜀𝑡 , over the decade prior to the imposition
of sanctions to obtain an unbiased estimate of the impact (direct and indirect) of oil
prices on growth, where 𝛾0 = (𝛽0 + 𝛽3 𝛿0 ) + 𝛽3 𝛿3 𝑍̅ , 𝛾1 = (𝛽1 + 𝛽3 𝛿1 ), and 𝜀𝑡 is a
mean-zero error.
Finally, letting T denote the sanctions period, 𝑔̂𝑇 = 𝛾̂0 + 𝛾̂1 𝑒𝑇 gives an
unbiased out-of-sample estimate of the oil component of growth in period T. It
+ 𝛽3 𝛿2 )𝑠𝑇 . In other words, the difference between
follows that 𝐸(𝑔𝑇 − 𝑔̂𝑇 ) = (𝛽2 ̂
actual growth during the sanctions period and our oil-based out-of-sample forecast
gives an unbiased estimate of the direct and indirect effects of sanctions on growth.
Of course, it is possible that regime shifts in the underlying coefficients may have
occurred, or there may have been deviations in other exogenous factors (i.e.,
deviations in Z) that have confounded the sanctions impact at exactly the same time.
While these are possibilities, they are unlikely.
VI.b. Impact on GDP growth
We begin with an extremely simple reduced-form model, which attempts to
explain quarterly variation in Russia’s real GDP growth between 2004 and 2014 as a
function of the log change in world oil prices (both current and one period lagged), a

22

 linear time trend (to capture Russia’s long term slowdown), a season dummy
(winter) and a global financial crisis dummy (2007-Q4 to 2009-Q2).10
We then use this model to conduct an out-of-sample prediction of Russia’s
quarterly GDP growth in 2014 and 2015, and ask how well this prediction matches
the actual growth, leaving the remainder as what is left to be explained by other
factors such as sanctions.
The regression (Table 4, Column 1) explains about three quarters of the
historical variation in Russia’s GDP, and all explanatory variables except the winter
season dummy are statistically significant.11 Figure 8 shows actual GDP growth
compared to that predicted by the model. Most of the deviation occurs in the first
half of the sample period. Indeed, if the model is estimated from 2007 onward,
instead of 2004, the overall model fit rises to over 85 percent. Of particular interest
is the out-of-sample prediction in 2014 and 2015. The model fits almost exactly in
2014 and actually under-predicts GDP growth in late 2015.
Hence, we find that oil prices manage to drive the majority of Russian
economic performance since 2004, with sanctions playing a secondary role
compared to oil prices at the macroeconomic level. From peak-to-trough, Russia’s
real GDP declined by about 5 percent, and therefore at most 20 percent of that, or 1
percent of GDP can be potentially explained by sanctions.12

The authors attempted multiple different specifications using the ruble-dollar exchange rate,
Russian consumer sentiment, and other variables and found that the results are almost identical with
no additional explanatory power, likely for the reasons discussed above.
11 The insignificance of the winter dummy is not surprising, given that quarterly real GDP is already
seasonally adjusted.
12 Interestingly, this 1 percent of GDP impact is similar to the IMF’s estimate of a 1-1.5 percent of GDP
impact from a completely different methodology.
10

23

 Table 4: Reduced-Form Macro Model

Of course, this is not definitive proof. It could be that, absent sanctions, the
model would have under-predicted post-sanction GDP growth even more than it
does. Yet given its historical accuracy, we find it implausible that the statistical
relationship between oil and Russian economic output would dramatically shift in
the out-of-sample forecast.
VI.c. Import Demand and EU Effects
The main economic vulnerability of EU countries to sanctions is their exports
to Russia, which have declined sharply since mid-2014. However, as with its GDP,
Russia’s imports from the EU could also be driven by oil prices and the deterioration
in Russia’s economic performance caused by domestic policies. One key

24

 determinant of a country’s imports is its national income, which we have already
seen is well explained by oil prices.
Another is the relative price of imports, which are affected by exchange rates
that are in turn affected by oil prices. Given that energy exports account for the vast
majority of Russia’s total export revenue, it is reasonable to assume that the
depreciation of the ruble in 2014 was largely driven by the oil price decline and
therefore indirectly captured in our regression.
Note that Russia’s recession and depreciation would likely affect its demand
for imports in general, not exclusively from the EU or other sanctioning countries.
The ruble declined strongly against all major currencies. The only exceptions were
CIS countries, which because of their dependency on exports to Russia, devalued
their own currencies to avoid losing ground. However, it is also possible that
sanctions, and certainly Russia’s countersanctions, had a discriminatory effect,
reducing Russia’s imports specifically from the EU (and other sanctioning countries).
In what follows, we repeat the exercise of the previous subsection for
Russia’s total imports to get an estimate of how much of the decline needs
explaining after oil prices are taken into account. From this we construct an
estimate of the non-discriminatory effect of sanctions. We then adjust our estimates
to account for counter-sanctioned agricultural products, which are clearly subject to
discrimination.13

It is also possible that non-counter-sanctioned products are affected by sanctions in a
discriminatory way beyond the scope of this study, and a more disaggregated approach, such as that
of Croznet and Hinz (2016) is required.
13

25

 The results are found in Table 4, Column 2, which is quite similar to the GDP
model. Figure 8 shows the out-of-sample prediction. Here we find that in the first
year of the sanctions period, the model predicts a smaller decline in Russia’s total
imports than actually occurred. Adding up the quarterly gaps between predicted
and actual, we can say that 80 percent of the difference in total imports between
2015 and 2013 are explained by oil. This amounts to about $26 billion in imports.
The rest is due to other factors, one of which could be sanctions.
For counter-sanctioned agricultural products, we know that Russia’s
embargo directly suppresses imports specifically from the EU (as well as the U.S.,
Norway and Australia). Nevertheless, our model is useful for determining how
much of the decline can be attributed to the embargo and how much would have
occurred anyway, as the result of lower oil prices. If we assume that agricultural
imports from the EU would have declined in proportion to all other imports absent
the counter-sanctions, and that 80 percent of that decline were driven by oil, then
the effect of counter-sanctions can be computed as a remainder.
As a final step we distribute the decline in imports to each EU country (plus
Norway, Iceland, and the United States) according its 2013 level of exports to Russia
and express the declines as a percentage of each EU country’s GDP. The results are
reported in Tables 9 and 10. The key result is that the impact of sanctions on the
exports of most EU countries is quite small.
We find that sanctions and countersanctions have had a small effect on the
economies of most EU countries. Adding together the impacts of sanctions and
countersanctions on exports – the main vulnerability – gives a median impact across

26

 EU countries of just -0.13 percent of GDP (Figure 9). The reasons for this are: (1)
Russia generally accounts for a small share of total EU countries’ exports; and (2)
most of the decline in Russian imports are explained by lower oil prices and trend
factors.
The countries that seem to be impacted the most by sanctions are the Baltics.
Lithuania has been hit the hardest, with a sanctioned-induced drop in exports of 2.7
percent of GDP. The exports of Estonia and Latvia fell by around one percent.
(Incidentally, these countries are also strong supporters of sanctions.) The hit on
exports for all other EU countries ranges from -0.01 percent to -0.3 percent of GDP.
To summarize, at the macroeconomic level, given that oil price fluctuations
alone can account for 80 percent or higher of the drop in Russian economic output
and import demand, sanctions appear to have had a second-order impact,
transmitting into relatively small spillover effects onto the economies of the
European Union.

27

 Figure 1: Overview of U.S. and EU Sanctions against Russia

Source: U.S. Treasury OFAC, Council of the European Union
Figure 2: Timeline of U.S. and EU Targeted Sanction Designations

Source: U.S. Treasury OFAC, Council of the European Union, BBC

28

 Figure 3: U.S. Sanctioned Entities by Type

Source: U.S. Treasury OFAC
Figure 4: EU Sanctioned Entities by Type

Source: Council of the European Union
29

 Figure 5: U.S. SDN and EU Restricted Individuals, by Type

Source: U.S. Treasury OFAC, Council of the European Union
Figure 6: World Oil Price

30

 Figure 7: Actual vs. Predicted Russian real GDP growth

Figure 8: Actual vs. Predicted Russian import demand growth

31

 Figure 9: Estimated Decline in EU Exports to Russia from Sanctions and
Counter-Sanctions, as a percent of GDP

32

 Figure 10: Estimated Decline in EU Exports to Russia from Counter-sanctions,
as a percent of GDP

33