CHAPTER 5. International Integration
- Christian Beddies, Enrique Gelbard, James McHugh, Laure Redifer, and Garbis Iradian
- Published Date:
- November 2005
Armenia’s high growth rates over the past three years have been fueled by fast-growing exports, donor inflows, remittances, and FDI. Strong export-led growth is rather surprising in a country that lacks natural resources, has been subject to a trade blockade from two important neighbors, and has poor transportation routes. This chapter analyzes changes in Armenia’s trade patterns in recent years, the role of government policies, the success of the diamond industry, and the costs and consequences of the trade blockade.
The main ingredients of Armenia’s trade success are a liberal trade and investment regime, a well-educated but low-cost labor force, a committed diaspora that promotes investment, and enabling but not overly interventionist government policies. Given high transport costs to major markets, Armenian exports are concentrated on lightweight, high-value products that can be flown out of the country. If the political issues that have led to the trade blockade with Turkey and Azerbaijan could be resolved, the positive effect on Armenian trade and income levels are expected to be significant.
A. Trade and External Debt
As noted in Chapter 1, the conflict between Armenia and Azerbaijan led to the closure by Turkey and Azerbaijan of their borders with Armenia.62 Under pressure from the European Union, Russia, and the United States, there have been numerous attempts to work out a peace settlement, but no agreement has been reached to date. If an agreement were to be reached, some infrastructure to facilitate trade exists, such as unused road networks with Turkey on the western Armenian border.
|(Annual percent change)|
|Merchandise exports, f.o.b.||4.9||9.9||24.8|
|Merchandise imports, c.i.f.||34.6||–0.5||11.6|
|Exports of goods and services||14.3||10.8||20.6|
|Imports of goods and services||30.6||0.7||11.6|
|(In percent of GDP)|
|Foreign direct investment1||2.1||5.5||4.6|
|Real GDP growth (annual percent change)||5.3||5.5||11.7|
Armenia’s north and south borders, while open, are not hospitable. The main road from the Islamic Republic of Iran is treacherous, traversing a mountain range that is frequently impassable in winter. The single main transport road into Armenia from Georgia is passable year-round. The main rail line into Armenia runs through Georgia from Russia, but unrest in northern Georgia has often rendered large swaths of it unusable. Imports from Russia, notably oil, must therefore skirt northern Georgia by ship on the Black Sea and reconnect with the rail line further south. As a result, a high share of trade is conducted by air.
In contrast to existing political and physical barriers to trade, Armenia has one of the most liberal trade regimes in the world. Most imports are duty-free, and consumer goods have a flat tariff of 10 percent. The trade-weighted average tariff rate is 3.0 percent. There are no trade processing fees or nontariff barriers. One complicating factor is the prevalence of discretionary practices by customs on the valuation of imports. Such practices raise import costs. Armenia joined the World Trade Organization (WTO) in February 2003 and belongs to the Black Sea Economic Cooperation (BSEC).
Armenia’s current account deficit narrowed to an estimated 5.8 percent of GDP in 2004, from 21.3 percent of GDP in 1998, as a result of booming exports and strong import substitution activity. Over the same period, the trade deficit narrowed from more than 30.5 percent of GDP to 15.5 percent of GDP (Figure 5.1). The still relatively large trade deficit is (partially) offset by private sector remittances within the current account.63 The level of remittances has been particularly high since 2003 because of high income growth in Russia.
Figure 5.1.Current Account, 1994–2004
Source: Armenian authorities.
During the past four years, Armenia managed to bring its external debt position to sustainable levels by lowering its net present value of debt-to-exports ratio below 100 percent. Looking ahead, stress tests reveal that such a level is resilient to a variety of shocks through 2015.64 As discussed in Chapter 3, a key challenge after 2015 will be the cost of decommissioning the nuclear power plant. Booming exports and debt swap agreements with Russia and Turkmenistan in 2003 were instrumental in cutting debt ratios by almost one-half in just a few years (Table 5.2).
|Debt service to exports|
|Debt service to central government revenue|
|Public External Debt to GDP|
|Net Present Value of Public External Debt to Exports|
B. Changes in Trade Patterns
The dismantling of the Soviet Union and the concomitant change in relative prices and demand caused a drastic reorientation of Armenia’s trade in the past decade. The share of trade with the European Union, the United States, and Israel increased rapidly, while trade with the CIS countries and the Islamic Republic of Iran decreased substantially (Figure 5.2).
Figure 5.2.Direction of Trade, 1996–2004
Source: IMF International Financial Statistics.
The composition of exports changed radically, and export growth has been unprecedented. Exports of industrial goods such as machinery, rubber, chemicals, and electronics virtually disappeared, and a new set of activities gradually emerged. At the same time, exports became more concentrated. Since 1999, a radical transformation of the diamond industry has led to a boom in diamond processing, with diamond exports increasing by 450 percent. Other dynamic sectors were agricultural products, foodstuffs, textiles, minerals, base metals, and instruments (Figure 5.3). While export growth began to pick up in the late 1990s, most of the growth took place between 2000 and 2004, when merchandise exports grew by 130 percent.
Figure 5.3.Composition of Exports, 1994–2004
Source: Armenian authorities.
Exports of services also rose in recent years, but their value remains well below that of merchandise exports. There has been an increase in exports of information technology products (mainly software) which benefited from Armenia’s human capital base. Tourism receipts have improved somewhat, but they remain hampered by an inadequate transportation infrastructure.
The composition of imports also changed because of the shocks in the early 1990s and the subsequent surge of the diamond industry. Imports grew by 256 percent during 1994–2004, although the growth rate slowed after 2000 owing to import substitution. Import growth was particularly strong in raw diamonds (to service the diamond processing industry), machinery, equipment, paper, and chemical products. Imports of foodstuffs and minerals decreased as a share of total imports, reflecting import substitution by emerging domestic activities (Figure 5.4).
Figure 5.4.Composition of Imports, 1994–2004
Source: Armenian authorities.
The observed increase in imports of services reflects transport costs associated with higher imports. Telecommunications services did not expand because of perverse monopoly rights granted to a single foreign investor (communication costs remain an impediment to business in Armenia and a constraint on trade). Some of these monopoly rights (for mobile services) have been recently rescinded, although a deeper restructuring of the sector is needed to bring costs towards competitive levels.
C. Investment, Exports, and the Role of the Government
Armenia has experienced relatively high levels of domestic and foreign investment in recent years. FDI averaged 5 percent of GDP during 1999–04, compared to an average of 4 percent for other CIS countries. These levels of investment are relatively high given the location of the country, its small domestic market, and the trade blockade. Several factors seem to have contributed to this outcome, namely the improvement in the business environment in response to prudent economic policies and deregulation, the role of the diaspora in intermediating investments and exports, and Armenia’s low level of wages and skilled labor force.
Foreign investment in export sectors was channeled toward human capital-intensive activities with high value-added, low volume and weight, and minimal use of raw materials. Besides diamond processing, the most recent investments have been in energy, software, jewelry making, watches, processed food, wine and liqueurs, and textiles. In addition, a large project related to airport improvement is under way, and the state copper company will be privatized in late 2005 with mandatory investment clauses for the next few years.65
In light of the success of Armenia’s exports so far, we explore below the extent to which dynamic (compared to static) factors such as increasing returns to scale and the domestic market played a role in the process. A skilled labor force and a long-standing tradition of artisanship in diamond polishing are Armenia’s sources of comparative advantage. Modern trade theory emphasizes the importance of economies of scale and market size in shaping comparative advantage (see Feenstra, 2004). While most of Armenia’s emerging exports (diamonds, foodstuffs, textiles, minerals, base metals, and instruments) are subject to economies of scale at the firm level, the size of the exporting firms in these sectors has not changed much over the last few years and is lower than in developed countries. Furthermore, the size of the domestic market is still too small to make a difference on unit costs and affect comparative advantage. Therefore, Armenia’s export boom seems to be associated with its traditional sources of comparative advantage (a high-skilled, low-wage labor force) and with an appropriate incentive structure (i.e., enabling tax and trade regimes).
Regarding the extent of government intervention as a factor behind the recent export boom, models of trade under imperfect competition point to a possible beneficial effect of industrial policy via subsidies or trade protection, although these results are often contingent on model specification and the empirical evidence on the matter remains ambiguous (see Lancaster, 1996; and Perdikis and Kerr, 1998). These models suggest that, under certain market imperfections or externalities, government intervention can serve as a catalyst to jump-start private sector activities that investors may otherwise avoid. In the case of Armenia, and except for a rather minor degree of government support to the diamond sector (see below), the overall environment has not been characterized by policy interventions (e.g., subsidies, import protection, monopoly rights). Foreign investors enjoy temporary exemptions from profit taxes (until 2007), but such incentives are similar to those present in most developing countries. Beyond profit tax exemptions, Armenia’s tax rates are relatively low and uniform, and there are no subsidies or other advantages granted to exporting firms.
D. Diamond Processing
Lack of market incentives during the Soviet era prevented diamond processing from flourishing in Armenia. In the late 1960s, a large diamond processing factory was established in Armenia to exploit domestic know-how and service the Soviet market.66 However, the factory was a victim of inefficient state management and its costs were relatively high by international standards. By the late 1990s, the state-owned factory was still operating, thanks to Armenia’s traditional expertise in diamond polishing and cutting, combined with labor costs that were six to eight times lower than in other countries.
Since the late 1990s, a favorable domestic and international environment were the key factors behind the successful transformation of Armenia’s diamond industry. The industry drew many investors, including firms from Belgium and Israel, with output increasing from US$100 million in 2000 to more than US$250 million in 2004. By 2004, approximately 60 companies engaged in diamond processing. The largest 15 companies account for 90 percent of total output, employ 4,200 people, and pay an average salary 1.5 times the average for the country. Nearly all output is sold abroad, and its value-added is about 10 percent.
The government did not actively subsidize the sector, although by ensuring a stable supply of raw materials it may have provided a positive signal to some investors. In 1998, the government signed an agreement with Russia that ensured a minimum volume of raw diamonds to be supplied to Armenia. The agreement sought to assure investors of a stable supply of raw diamonds, but it did not specify prices. Prices were supposed to be negotiated directly between the importer and the Russian suppliers. In the end, the contract was never fully used as many firms developed their own supply sources, although it may have played a useful signaling role that some investors needed to commit resources to a remote country like Armenia.
The only tax revenue the industry generates is income taxes from its employees. Imports are duty-free and there are no value-added taxes because almost all output is exported. Under the Kimberley certification process, raw diamond import values are assessed only on transaction values. In fact, profits in diamond processing are difficult to ascertain given that more than half of the incoming product is turned into waste in the polishing process. The international standard is a turnover tax, but this is also difficult to implement, and the authorities decided not to apply it in Armenia.
Looking ahead, the sector is unlikely to grow as quickly as in recent years. Export growth already fell in 2004 owing to higher prices for raw diamonds. The higher price of raw diamonds caused many of the smaller firms that rely mostly on manual labor to stop operations. The larger companies remain competitive as they enjoy better prices for their large purchases of inputs and use newer technologies.
E. Closed Borders and Transport Costs
Opening Armenia’s border with Turkey and Azerbaijan would increase the volume of trade, change its composition and orientation, and lower tradable goods’ prices. Expanded rail and road routes would lower transport costs and enable both trade of heavier items and tourism.67 These developments would be extremely positive for Armenia and eastern Turkey, and to a lesser extent for Azerbaijan. The proximity of the large Turkish market and Armenia’s lower wages would provide incentives for new investments and the development of new export activities. The impact on import prices would also be significant. For instance, energy prices in Armenia are higher than elsewhere in the region, and so opening up new opportunities in the sector (particularly for natural gas imports from Azerbaijan to replace some Russian imports) could lower import prices.
Views on the constraint caused by the closed borders are mixed, but on balance, the effects are expected to be significant, especially considering the political uncertainty regarding the unresolved territorial dispute with Azerbaijan. Discussions with government officials suggest a profound impact, with wide-reaching effects that would effectively reinvent the structure of the economy. Two empirical models can be estimated to gauge the extent to which trade is constrained: the openness model and the gravity model. Based on the model estimated by Freinkman, Polyakov, and Revenco (2004), Table 5.3 displays the simulation results for 2004 from two typical openness models that attempt to explain trade levels as a function of GDP per capita, population, and regional dummy variables. The two models differ with respect to their explanatory variables: one uses the trade-to-GDP ratio, the other uses the exports-to-GDP ratio.68
|Actual trade-to-GDP ratio||54.2|
|Predicted trade-to-GDP ratio||115.2|
|Actual exports-to-GDP ratio||20.6|
|Predicted exports-to-GDP ratio||52.3|
The gravity model has also been used extensively in the trade literature, primarily with a view to analyzing the extent and direction of trade among countries.69 The model has performed relatively well empirically, and despite some initial controversy regarding its theoretical underpinnings, its basic specification has been shown to be consistent with modern trade theory (Helpman and Krugman, 1985; Deardoff, 1998). Recent versions of the model have been expanded to estimate the impact of institutions, common currencies, and the trade potential of being a member of key international organizations (de Groot and others, 2004; Frankel and Rose, 2002). We used Kukharchuk and Maurel’s (2004) model to estimate Armenia’s trade potential (assuming open borders and appropriate trade routes) and the impact of institutional constraints on trade. Two important features of this model are the inclusion of CIS countries in the sample and the introduction of measures of institutional quality in the trade equation. The basic specification of the model is:
where i and j denote country pairs, and T and Y are the natural logarithm of bilateral trade flows and national income, respectively. D is the distance between i and j, VOL measures exchange rate volatility, and Dummy represents binary variables for different trading blocks, namely the European Union (EU), the Commonwealth of Independent States (CIS), and the Central and Eastern European Countries (CEEC).70INS represents the institutional environment and is captured by five variables: trade restrictions, foreign investment restrictions, restrictions on financial and banking transactions, property rights, and shadow market activities.71 The data panel used in the estimation has 13,712 observations from 42 countries over an eight-year period since 1994. The simulation results for Armenia, displayed in Table 5.4, suggest that its “potential” level of trade could be 39 percent higher than the actual trade.72 The increase is somewhat larger than the one suggested in a previous study by Polyakov (2001).73
|Actual level of trade (2004)||1,945|
|Predicted level of trade||2,707|
|Difference between predicted and actual trade (in percent)||39.2|
|Increase in total trade in response to improved institutional environment||34.8|
|Selected scores for institutional environment1||Armenia||CIS||EU|
Scores range from 1 to 5, with 5 being the worst environment.
Scores range from 1 to 5, with 5 being the worst environment.
While both the openness and the gravity models suggest that opening Armenia’s borders and improving trade routes would lead to a much higher level of international trade, the role of the institutional environment to facilitate trade and investment should not be overlooked. The gravity model used above also allows us to gauge the effect on trade volumes of an improvement in the institutional environment.74 We simulated in Table 5.4 the effect on Armenia’s trade of an improvement in the institutional environment to the level of EU countries. The results show that the increase in trade is almost as high as the difference between actual and “potential” levels.
F. Concluding Comments
Armenia’s tradable sector has thrived in recent years as a result of macroeconomic stability, improvements in the business environment, help from the Armenian diaspora, and a low-wage, highly skilled labor force. A successful process of import substitution and export-led growth has been taking place in an environment of low and rather uniform tax rates and minimal government intervention. The government did not grant subsidies or other advantages to firms (e.g., import protection, monopoly rights, or export subsidies), although it is possible that the diamond sector benefited initially through a government-engineered supply contract that ensured a stable supply of raw materials from Russia.
Over the medium term, more will need to be done for the country to further develop its export sectors and remain competitive. The opening up of Armenia’s borders could have a significant role in fostering trade and regional integration. For the time being, however, the current situation should be viewed as an opportunity to improve the existing trade and investment environment. The latter can be achieved by strengthening the quality of Armenia’s institutions, especially by making tax and customs agencies more efficient, improving the financial sector, ensuring an independent judiciary, and adopting effective anticorruption practices at all levels of government.
CamposNauro and YukoKinoshita2003 “Why Does FDI Go Where it Goes? New Evidence from the Transition Economies” IMF Working Paper 03/228 (Washington: International Monetary Fund).
DeardoffAllan1998 “Determinants of Bilateral Trade: Does Gravity Work in a Neoclassical World” inThe Regionalization of the World Economyed.J.Frankel (Chicago: University of Chicago Press).
De GrootHenriGert-JanLindersPietRietveld and UmaSubramanian2004 “The Institutional Determinants of Trade Patterns” Kyklos Vol. 57 Fasc. 1 pp. 103–24.
FeenstraRobert2004Advanced International Trade: Theory and Evidence (Princeton, New Jersey: Princeton University Press).
FrankelJeffrey and AndrewRose2002 “An Estimate of the Effect of Common Currencies on Trade and Income” Quarterly Journal of EconomicsMay pp. 437–66.
FreinkmanLevEvgenyPolyakov and CarolinaRevenco2004 “Armenia’s Trade Performance and the Effect of Closed Borders: A Cross-Country Perspective” Armenian Journal of Public Policy Vol. 1 No. 2 pp. 185–212.
GiragosianRichard2003 “Geopolitics and the Formation of Foreign Policy in the South Caucasus: An Examination of Armenia, Azerbaijan, and Georgia” Working Paper 03/19 (Washington: Armenian International Policy Research Group).
GaribaldiPietroNadaMoraRatnaSahay and JerominZettelmeyer2002 “What Moves Capital to Transition Economies?” IMF Working Paper 02/64 (Washington: International Monetary Fund).
HausmanDavidA.Kraay and PZoido-Lobaton2002 “Governance Matters II: Updated Indicators for 2001–02” World Bank Policy Research Working Paper No. 2772 (Washington: World Bank).
HelpmanEllhanan and PaulKrugman1985Market Structure and Foreign Trade (Cambridge, Massachusetts: MIT Press).
KukharchukOxana and MathildeMaurel2004 “Russia’s Accession to the WTO: The Potential for Trade Increase” Journal of Comparative Economics Vol. 32 pp. 680–699.
LancasterKevin1996Trade Markets and Welfare (Brookfield, Vermont: Elgar, distributed by Ashgate).
LimEwe-Ghee2001 “Determinants of, and the Relation Between, Foreign Direct Investment and Growth: A Summary of the Recent Literature” IMF Working Paper 01/175 (Washington: International Monetary Fund).
MathieuPaul2004 “The Sustainability of Public External Debt in the Five Highly-Indebted CIS Countries: A Background Paper for the CIS-7 Initiative” EBS/04/56 4/29/04 (Washington: International Monetary Fund).
PerdikisNicholas and WilliamKerr1998Trade Theories and Empirical Evidence (Manchester, England: Manchester University Press).
PolyakovEvgeny2001Changing Trade Patterns after Conflict Resolution in South Caucasus (Washington: World Bank).
Republic of Armenia—Fifth Review Under the Poverty Reduction and Growth Facility and Request for Extension of the ArrangementIMF Country Report No. 04/136 (Washington: International Monetary Fund).
RossAndrew2005 “Which International Institutions Promote International Trade” Review of International Economics (forthcoming).
SamuelianThomas2003 “Implementing Country Competitiveness as a Standard for Law and Policy Making” Group Working Paper 03/16 (Washington: Armenian International Policy Research Group).
ShiellsClinton2003 “FDI and the Investment Climate in the CIS Countries” PDP/03/5 (Washington: International Monetary Fund).
While export data is relatively accurate, data on imports and remittances are not reliable. Unofficial estimates suggest they may be two or three times higher than the official statistics. This would imply, all else equal, that Armenia has a current account surplus. Most of the unrecorded inflows are thought to be in the form of “pocket money” from Armenians working in Russia.
Debt calculations until 2015 are presented in Mathieu (2004) and in IMF (2004).
Recent literature on FDI highlights two main modalities of participation: “horizontal” investments geared to the size of the potential market or “vertical” intra-industry integration that parcels out the chain of production among different countries (Garibaldi, 2002; Lim, 2001; and Campos and Kinoshita and others, 2003). The latter has been the predominant form of activity in the last two decades in most countries, and while data limitations prevent a disaggregated analysis for Armenia, most emerging activities seem to have this feature (e.g., diamonds, textiles).
Armenia has a long history of artisanship in diamonds and jewelry.
Transport costs can be estimated by the ratio of freight costs to the value of exports, which in the case of Armenia is 9.3 percent. The corresponding ratio for the European Union is 1.5 percent (Polyakov, 2001). However, the Armenian figure is pushed downward by a trade structure that favors low-weight, high-value items.
The models were estimated with data for 149 countries from the World Development Indicators database covering the period 1994–2001. The two-stage least squares procedure was used to deal with possible endogeneity between GDP and the openness measures. The first model is specified as Trade/GDP = 119.8 + 12.7*ln (GDP per capita) – 14.5*ln (population) + 23.7 (CIS dummy). All of the above are statistically significant at the 99 percent confidence level. The second model is specified as Exports/GDP = 32.7 + 8.3*ln (GDP per capita) + –6.2*ln (population) + 11.6 (CIS dummy). In the latter model, all variables except the dummy and the intercept are statistically significant at the 99 percent confidence level. The dummy and the intercept are significant at the 95 percent confidence level.
The gravity model specifies bilateral trade as a function of the countries’ relative sizes (measured by GDP or population) and their distance from each other. The name is derived from Newton’s analysis of gravity in physics, where the gravity force between two objects is proportional to their masses and inversely related to their distance.
Trade, income, and distance data are from the Direction of Trade Statistics, the Fund’s World Economic Outlook database, and the Chelem-CEPII dataset. VOL is calculated as the standard deviation of the ratio between the monthly nominal exchange rate and its annual average (International Financial Statistics).
The values for these variables are taken from the Heritage Foundation’s Indices of Economic Freedom (www.heritage.org/research/features/index) and are measured on a 1–5 scale, with a value of 1 representing the optimal environment. The variables are scaled as deviations from their mean.
The estimated equation (using generalized least squares) is Ln Tij = 0.87Yi + 0.93Yj – 1.02Dij – 0.13VOLij + 2.49 Dummy (CIS) – 0.04 Tradej – 0.08 Tradei – 0.01Investmenti – 0.03Investmentj – 0.07 Bankingj – 0.03 Bankingi – 0.12 Property rightsj – 0.07 Property rightsi – 0.11 Shadow Economij – 0.09 Shadow Economyi + 2.49 CIS+ 0.44 EU + 0.28 CEEC + έij. All but two variables are statistically significant at the 99 percent confidence level. The remaining two variables (foreign investment and CEEC dummy) are statistically significant at the 90 percent level.
Polyakov (2001) also used a gravity model to estimate the costs of the blockade, and the results suggested that opening the borders and improving transportation networks could double Armenian exports and increase GDP by about 30 percent. The model contains a traditional gravity equation that includes income, distance, and population as explanatory variables.
As noted in previous chapters, much remains to be done to improve governance and the business environment in Armenia. A recent World Bank survey of investment climate indicates that the tax system, political and policy uncertainty, corruption, and the functioning of the judiciary remain key impediments to doing business in Armenia (World Bank, 2004).