Global Policy Forum

5. The expert group considered basic approaches to developing a possible methodology for assessing the consequences actually incurred by third States as a result of preventive or enforcement measures, taking into account the particular problems and needs of developing countries confronted with the special economic problems arising from carrying out such measures. In this context, general conceptual frameworks for the impact analysis, typology of effects and sources of hardship and applicable methods and procedures were discussed at the meeting.

General Conceptual Framework

6. The expert group took note of the conceptual framework for identifying and assessing the effects of UN-imposed preventive or enforcement measures on third States, as outlined in the previous reports of the Secretary-General. It was also appraised of the comments provided by other international organizations, in particular the international financial and trade institutions, on such assessments. In this light, the group felt that any useful methodology for impact assessment should be related to the political purpose of imposing sanctions on the target State, the economic realities and particular situations of third countries and the need for international cooperation and mutual assistance in bearing the costs of implementing sanctions, in order to ensure their effectiveness and minimize collateral damage, particularly adverse effects on third States.

7. It was recognized that the actual impact of sanctions on individual third States and their policy options may differ substantially from case to case, depending on the specific nature of the sanctions regime and the particular situation of the country in question, including the structure and intensity of its suspended links with the target State. Therefore, developing a common methodology for impact assessment would require establishing general guidelines, criteria or procedures for the identification and proper categorization of various effects, the selection of applicable methods for estimating the incurred losses and costs and the designing feasible and practical measures of relief and international assistance. Once these essential and interrelated elements are clearly defined, a common methodology or modalities for action may be formulated and, if accepted, may be applied on a case-by-case basis.

Sources of Hardship

8. Based on recent experiences of the practical application of Article 50 of the Charter of the United Nations, the expert group considered the main sources of hardship and the variety of adverse effects actually incurred by third States, in particular developing countries, as a result of imposing preventive or enforcement measures. In principle, a proper identification and categorization of effects is essential for any methodology of impact assessment. However, these effects could be grouped in different ways, as direct and indirect, trade and financial, economic and social, primary and secondary, temporary and long-lasting. An illustration of various categories of effects follows.

9. In general economic terms, various types of losses and costs confronting third States may be classified into two broad categories: "direct" and "indirect" effects of sanctions. Direct effects are evaluated as the income foregone and losses incurred stemming directly from the cancellation of contracts and/or severance of economic relations with the target country. Examples include suspended sales or outstanding orders for contracted deliveries; interrupted shipments, payments or other transactions; and disrupted production of jointly operated facilities. In most cases, these adverse effects are related to the balance of payments on current account, particularly exports and imports in goods and services, but they may also refer to capital account activities such as investment projects. On the other hand, indirect effects, which are largely the induced effects of the former, represent mainly the negative impact on domestic variables, I.e. output, investment, employment and the budget. They may include disrupted production due to the absence or hi her cost of sanctioned supplies, suspended financial inputs and services, foregone profit tax or tariff revenues, lost jobs or income and the ensuing increase in social expenditures. Obviously, the latter category of damages is more difficult for a quantitative assessment.

10. Depending on the nature of sanctions imposed on a target country, specific costs for third States associated with the balance of payments may be further grouped into external-sector categories: (1) those related to trade with the target country; (ii) those related to financial links with the target country; and (iii) other costs associated with special relationship with the target country (e.g. joint ventures). Sectoral categorization is made more difficult because of the enormous expansion of trade in services in recent years and a vast variety in the structure and intensity of links between senders and targets of sanctions. Nevertheless, identification of main sources of hardship in the trade and financial sectors may be particularly useful for the purpose of impact assessment.

11. Typically, trade embargoes entail forfeiture of foreign exchange earnings which may be badly needed to finance essential imports. On the export side, the incurred consequences may include: undelivered regular merchandise exports with no alternative markets readily available (exports foregone); outstanding orders for contracted future delivery for which production is already in progress; and suspended sales of services to the target State (e.g. engineering or construction projects, activities in the areas of transportation, communication, maintenance, packaging and others). Similarly, losses on the import side may include: undelivered regular merchandise imports, particularly if no alternative suppliers are available (imports foregone); lost or suspended imports on concessionary terms, especially at subsidized prices; outstanding orders for future imports for which payment has already been made; and terminated or suspended import of services.

12. Financial difficulties experienced by third States as a result of sanctions comprise those related to. suspended capital flows, interrupted current transfers and overdue debt servicing. Examples include: lost profits on investments or other income, such as remittances from migrant workers; confiscation, seizure or conversion of savings and assets (as possible countermeasures by the target); loss of concessionary loans, credits or grants; investment foregone; and tourist revenues foregone. Although lower interest receipts reduce foreign exchange earnings in the current period, unlike lost export receipts, these overdue payments can be capitalized and added to the affected country's foreign claims on the target State. However, unreceived payments on account of debt servicing by the target country may take the form of discontinued or suspended delivery of critical commodities (e.g. Iraq's outstanding foreign debt undertaken against future oil deliveries).

13. With regard to estimating the economic effects of sanctions on third States, it should be noted that there may be offsetting factors in both trade and financial sectors. For instance, part of the exports intended for the target State may be absorbed by the domestic economy or redirected to other markets and there may also be savings on interest, amortization and other payments owed to the target country when financial measures preclude their transfer. Therefore, export losses should be estimated net of the value of redirected exports and net of the cost of imports that the affected country would have used to produce the corresponding exports. Similarly, only incremental costs of obtaining replacement imports from other sources (given by the price or interest payments differential) should be included in the overall estimates. However, a proper assessment of the incremental impact would require a reliable baseline estimate of what would have happened in the absence of sanctions.

14. Apart from the economic effects of sanctions on third States, there may also be social costs resulting, for instance, from the return of large numbers of migrant workers; the influx of refugees; and the loss of employment or income, all of which would require additional public expenditures (e.g. to provide housing, health, education and other social services to the affected population groups) and may strain the social fabric, particularly in developing countries. In,a longer term, diverting already limited and stretched resources from development objectives to cover sanctions-related costs in the affected developing countries would put an inordinate burden on their vulnerable economies, thus perpetuating poverty, unemployment and social inequity.

15. In addition, a miscellany of secondary effects may stem from particular features of the sanctions regime, geographical factors and monitoring and enforcement requirements, thus affecting traditional economic interaction in the region. For example, the imposition of restrictions on transshipment through a target State which plays an important role in transport and communication links in the region would seriously impede the neighbouring countries' external economic relate' ns not directly involving the target State. Thus, a specific feature of the Yugoslav sanctions regime was that the bulk of damages reported by the neighbouring countries were those related to transportation disruptions, including losses in exports (e.g. perishable agricultural products), additional expenses for or lower receipts from transport-related services and higher payments for imports owing to widespread rerouting and extensive delays at border-crossing points, as well as monitoring and enforcement costs. However , damages of this type are most difficult for an independent assessment.

Methods of Impact Assessment 16. Based on relevant special studies, the expert group reviewed a number of specific methods that may be applicable for assessing the incidental impact of multilateral economic sanctions on non-target States. These include: (1) time series analysis of balance of payments changes; (2) stratified sample survey of firms or other affected entities; (3) gravity model of bilateral trade flows; (4) regression analysis of income shocks-, and (5) analytic hierarchy process for perception surveys. Main features of these methods are summarized and analyzed below. 17. Time Series Analysis of Balance of Payments Changes. In this method, expert observers examine time series data for traded goods and services and relevant financial flows before and after the imposition of sanctions, with a view to assessing the direct effects of sanctions in reducing merchandise exports and imports as well as financial transfers of the concerned third State. The result is an estimate of the impact of sanctions on the balance of payments, either in absolute or relative terms. In principle, the scope of the impact analysis can be expanded to estimate the indirect, second-round effects of sanctions on domestic variables, such as output, investment, employment, and the budget. These estimates may be also subdivided by economic sector or branch of industry, agriculture and services.

18. In the process, several methodological and practical difficulties may arise, thus imparting a potential bias to the estimates. Typically, in assessing the overall welfare implications of sanctions, the main challenges to be addressed are: (1) how to separate the effects of sanctions from other factors causing economic hardship; (ii) how to avoid double-counting of export, output and revenue losses; (iii) how to net out redirected exports and the import component of lost production; and (iv) how to estimate the incremental costs of acquiring the imports from other sources. In many cases, therefore, overcoming or minimizing these methodological limitations would require an assessment to focus primarily on the balance of payments impact of sanctions and associated external financing needs of an affected country. In addition, reliable and up-to-date macroeconomic data is essential for a credible effort to measure the effects of sanctions on third countries.

19. Stratified Sample Survey of Firms. This method may only be feasible if the country in question possesses data on a recent census or comprehensive survey of firms (or other affected entities, e.g. households) conducted prior to the imposition of sanctions, so that it can identify a stratified random sample of firms (entities) to query regarding the impact of sanctions and establish the benchmark experience of firms (entities) prior to the sanctions episode. This is essential to avoid a potential bias from focusing on self identified or officially-noticed " wounded" firms (entities). Once the required data are obtained, the next step is to perform a regression analysis of post-sanction vs. pre-sanction firm (entity) performance expressed as a function of firm (entity) characteristics. 20. In the case of affected firms, the dependent variable is the change in firm performance represented by decline in either firm sales or employment since the imposition of sanctions. Firm characteristics suggested as explanatory variables may be classified into two groups: (1) those that reflect the degree of the firm's reliance on(or intensity of its interaction with) the target country, a notable example of which is the percentage of trade with the target country in the firm's total sales and purchases prior to the sanctions; and (ii) those that reflect particular features of the third country's firm itself, e.g. firm size measured by its sales or number of employees. Large (low) values of the coefficients for firm characteristics in group (1) above (which are usually valued on a scale from 0 to 1) indicate a high (low) dependency of the firm on the target country, resulting in a corresponding decline in the firm's total sales or employment. On this basis, national losses or costs can be estimated by applying the coefficients to mean values of firm reliance on purchased inputs from and export markets in the target country.

21. While this method provides for impact analysis at the macroeconomic level of individual firms or other affected entities in the third State, it would not identify offsetting sales growth (or employment growth) enjoyed by the firms (entities) that did not rely heavily on the target country for inputs or markets. Nor would it capture the decline in confidence which may grip the entire economy of a neighbouring country as an indirect consequence of sanctions imposed on the target State. Moreover, this method does not distinguish the effect of sanctions from other adversities, unless supplemented by a separate estimate intended to disentangle the effects.

22. Gravity Model of Bilateral Trade Flows. This is a model specifically designed to measure the dynamics of economic interaction, primarily bilateral trade flows, between any two countries, e.g. the sender (third) State and the target State. Applying a common statistical technique, known as "ordinary least squares" regression analysis, to the "gravity model" allows the researcher to isolate the direct and indirect effects of sanctions on bilateral trade flows while holding other factors constant. In this method, large data sets are used to estimate parameters that describe the basic forces which determine the magnitude of bilateral merchandise trade (imports plus exports). Additional independent variables are then added to reflect the existence, duration and strength of economic sanctions. Thus, the obtained parameters for both sets of independent variables in the regression can be used to calculate the predicted percentage reduction in bilateral merchandise trade resulting from sanctions. Although the focus is on trade in goods, this method would also capture effects of financial sanctions to the extent that they reduce trade by denying investment, foreign exchange or credit to the target country and by possibly raising the cost of credit to the affected third State.

23. However, in practical application of this method, some important caveats should be duly noted. In the first instance, the use of predicted percentage coefficients to assess a specific episode requires prior classification of the severity of sanctions in question. Thus, taking an average of parameter estimates for the years 1985, 1990 and 1995, a recent study of the impact of sanctions on the United States' exports found that "limited" sanctions reduced bilateral trade by 27 percent from the level that would otherwise have been observed; "moderate" sanctions resulted in the decline of bilateral trade by 36 percent; and "extensive" sanctions accounted for a 91 percent drop in exports. Moreover, the model does not permit calculation of offsetting trade flows that may have been induced by bilateral trade losses with the target country; nor does it purport to calculate losses in domestic output cascading from lost trade. Furthermore, as with all statistically estimated parameters, the predicted trade losses have an associated error term of some ten percentage points to the mean prediction.

24. Regression Analysis of Income Shocks. The logic behind this method is that the effect of a contraction in the target State's economy (i.e. a substantial income shock from discrete causes, such as sanctions) on the third State's income will vary according to the relative importance of the former in the total trade of the latter and will tend to become smaller over time. Thus, the first step is to collect data on a large number of adverse episodes and then devise a regression equation to estimate the induced change in the third country s gross domestic product (GDP). Accordingly, independent variables will take the form of interacted variables between (i) the share of bilateral trade with the target country in the third State's GDP prior to the shock and (ii) the percentage change (i.e. decline) in real GDP of the target country, taken at a certain time following the initiation of sanctions. It is expected that the larger the proportion of trade with the target country in the third State's GDP, the more important effect the drop in the GDP of the former will have on the output in the latter, especially in the initial period after the imposition of sanctions. However, as time goes on, the affected third state should find ways to adjust to its severed economic contacts with the target State. Therefore, the estimated parameters purport to indicate, in diminishing coefficient values, the current year and lagged year effects of income shocks.

25. This method seeks to provide a more comprehensive analysis, since the variables in the regression reflect both trade losses and domestic output losses, taking into account various external and internal offsets. However, the resulting calculations cannot be traced to particular events or affected entities in the episode in question, because the parameters are estimated as averages derived from a large number of similar but not identical episodes of income shocks.

26. Analytic Hierarchy Process for Perception Surveys. As a practical tool for conducting a perception survey of the impact of sanctions, this method seeks to: (1) capture intangible variables, such as political and social factors; (ii) measure the benefits and costs of several options, in order to avoid overstatement in judgments; iii) quantify inconsistent perceptions but still produce consistent ranking of priority; (iv) incorporate indirect and second-round effects; and (v) integrate the incurred consequences with policy responses or measures of assistance. While this method has solid theoretical and mathematical foundations, its practical application does not require long series data.

27. Basically, there are two major steps to be taken: the first step is to construct a hierarchy to synthesize the problems at hand; and the second step is to measure the priority ranking of each attribute in the hierarchy, by using a scaling range (e.g. I to 9). A typical hierarchy for impact analysis ("impact hierarchy") consists of several levels, including the one distinguishing the direct and indirect effects of sanctions. In designing the hierarchy for a particular case, there is ample flexibility for determining the number of attributes in each level as well as the number of levels in the hierarchy. However, the longer the hierarchy, the more time-consuming the calculation will be. Having identified the critical impact variables at the bottom level of the hierarchy, the next hierarchy ("response hierarchy") would have to fuse those variables with various factors considered crucial in scrutinizing the relevant measures of assistance to the affected third country.

28. Since human judgments are the main inputs in this method, there is a risk of overstatement (i.e. exaggerating the damaging effects of sanctions). To overcome such a predicament, a benefit/cost principle can be applied. To this end, two "response hierarchies" are to be constructed, one for the benefit and another for the cost, of taking alternative measures. These measures are catalogued at the bottom level of both hierarchies, in order to estimate both the benefits and the costs of each alternative measure of assistance. The benefit/cost ratio will then be used as the criterion to make the ultimate decision. The ratio of two ordinal scales is taken in this case due to the fact that the scales being used in measuring the priority ranking of each attribute in the hierarchy are ratio scales obtained through a pair-wise comparison. The accuracy of the final results is tested by measuring the degree of inconsistency calculated from the judgement inputs. The test of robustness may be conducted by running a sensitivity analysis, without requiring additional information.

29. While all of the above-mentioned methods may be useful in assessing the effects of sanctions on non-target States, the expert group felt that the choice of applicable method(s) will depend on particular circumstances of the affected State(s) in the context of specific sanctions regime(s). Ideally, as many of the methods as feasible should be attempted to satisfy as many analytical criteria as possible. However, the availability of data and the cost of analysis may impose serious limitations.