88992 JUNE 2014 • Number 151 Information for Export Survival: An Analysis of Georgian Export Performance and Survival in International Markets Jose-Daniel Reyes, Gonzalo Varela, and Miles McKenna Successful entry into export markets and the subsequent survival of export flows are crucial if a country is to grow and diversify its export base. The accumulated experience of firms that export a particular product or serve a specific destina- tion can provide valuable information. When more of this export-relevant information is available to exporters, export flows have better chances of survival in new markets. However, many developing countries face two problems when it comes to acquiring this information. First, and most important, is that firms do not have incentives to share information. Information constitutes part of their competitive advantage, and sharing it may weaken their competitiveness. The second major problem is that this type of information is often absent, lacking, or underutilized. There is no clear mapping of available resources or activities that are actually profitable. Discovery is costly, and chances of export survival are lower. This note focuses on the role of information in the survival of Georgia’s exports, unveiling a robust association between product-specific and destination-specific information and export survival. These results suggest that there are gains to be made through fostering greater interfirm dialogue—especially dialogue that generates market- and product- specific information—to increase the changes of exports’ survival in Georgia. How Does Information Help Export Survival? the path to development. For individual firms, discovering this is often costly. They need to learn the characteristics of Entering and surviving in a new market is tricky business. As foreign demand (tastes, willingness to pay, volume, and so the literature points out, exporting can be an extremely risky activity, full of potential pitfalls—particularly for exporters in forth), the intricacies of the exporting activity (dealing with lower-income countries (Besedes and Prusa 2004; Brenton, customs regulations, freight forwarders, insurance compa- Saborowski, and Uexkull 2010). Determining the key factors nies), and calculate actual production costs (technologies, for export survival is therefore tremendously valuable for minimum efficient scale, sources of inputs). Without this in- transitioning economies. From a policy perspective, creating a formation, firms must learn through trial and error, which more enabled export sector requires a better understanding often results in inefficiencies, including potentially enormous of these factors and their relationships, including the role of sunk discovery costs. When discovery costs are high and the export-related information in improving the chances for suc- ability to profit from the new discovery is low, those exporters cessful entry and survival. that do become successful have little incentive to share infor- Informational failures are major obstacles to strong ex- mation on their exporting experiences. This results in the un- port performance and, more generally, to economic develop- dersupply of export information. ment. As argued by Hausmann and Rodrik (2003), learning For new firms, the accumulated experience of incum- what one is good at producing is an important milestone on bent competitors exporting the same products, or to the same 1 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise country, is a source of valuable information. Broader dissemi- ketplace. Table 1 shows yearly firm survival rates in exporting nation of this information reduces discovery costs, and can activities, by cohort, from 2003 to 2012. The decline in the help firms survive longer in export markets. This can then re- first-year survival probabilities is clear, with 2006 being par- sult in increased export earnings, job creation, and therefore, ticularly tough on exporters. In 2003, around 54 percent of more generally, overall economic growth. When information new exporters were surviving after one year of operation in is proven to increase an export’s survival chances, it is also international markets. This survival rate dropped to 37 per- confirming the presence of a spillover effect. For transitioning cent in 2011. The Russian trade embargo in 2006, a preamble economies like Georgia, enabling positive spillover effects to the 2008 war, clearly impacts the survival rates in 2006 and should be a priority. 2007. But how do these survival rates in Georgia compare A recent implementation of the World Bank’s Trade and with similar countries? Competitiveness Diagnostic (TCD) Toolkit looked at how Georgian products faced lower survival rates in interna- Georgian exporting firms fared in international markets, pay- tional markets than comparable countries. Figure 1 shows ing particular attention to their survival patterns (Reyes and the Kaplan-Meier survival function for Georgia, Armenia, Varela 2013). Lithuania, the Former Yugoslav Republic of Macedonia, Slo- vakia, and the Czech Republic. For 1999–2011, the proba- Understanding Export Survival in Georgia bility of a Georgian export flow (a particular product being Survival of firms in export markets is a challenge in Georgia, exported to a particular destination) surviving one year is particularly after the Russian Federation trade embargo of close to 32 percent, almost 2 percentage points above that of 2006. However, not all firms necessarily should survive. A an Armenian export flow, but 21 percentage points below thriving export sector requires strong dynamism stemming that of the Czech Republic’s flows. An alternative way of from exporters testing new markets or new products (or looking at export survival is to see the mean length of time both), as well as new firms trying to become exporters. While over which an export flow is active. This analysis indicates some of these firms go on to become profitable, others fail. that Georgian flows, on average, are active for slightly more Failure, or exit, is important in the process of churning— than one year, while those of Lithuania average 2.84 years, where resources are reallocated from less efficient to more ef- Slovakia 3.35 years, and those of the Czech Republic 3.5 ficient uses, a constant process Schumpeter called “creative years. Figure 1 also reveals that the longer the partnership destruction.” Firm churning can provide a measure of a coun- with the buyer (and hence the longer the export flow), the try’s competitiveness in international markets. The policy ob- more likely the export flow is to survive an extra year and the jective, therefore, should not be to just ensure high levels of less likely it is to exit. This is likely related to the fact that in- export survival, but rather to generate the conditions where formation costs decline as exporting experience accumu- firm churning helps to drive competitiveness, export values, lates (Araujo and Ornelas 2007). and job creation. According to Georgian producers, the amount of in- In Georgia, survival rates are actually declining, which is formation available regarding business opportunities in- preventing firms from greater integration into the global mar- country is scarce and constrains their growth potential. In agriculture, for example, more than half of Table 1. Survival Rates in Exporting Activity by Cohort (percentage of surviving firms) farmers operate on less than one hectare. With the land highly divided, there is little Year 2003 2004 2005 2006 2007 2008 2009 2010 2011 knowledge of production potentials or of 2003 100 opportunities for marketing specific prod- 2004 54 100 ucts, and even less about which goods could 2005 38 40 100 be candidates for exporting. To what extent 2006 28 25 52 100 do exporters like these benefit from infor- 2007 21 15 6 17 100 mational spillovers from other exporting firms? To shed light on this issue, the deter- 2008 17 12 5 12 34 100 minants of export survival were examined 2009 15 10 3 8 23 33 100 using a detailed data set of Georgian export 2010 13 9 2 6 17 23 38 100 transactions by country of destination and 2011 13 6 2 5 13 18 24 40 100 product (at the HS6 level) for about 1,900 2012 11 5 2 4 9 12 15 25 37 exporting firms per year, over the period Source: Authors’ calculations based on data from Geostat. 2003–12.1 The analysis asks the following Note: This table shows the share of firms that survive in international markets after the year of entry. questions: 2 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise Figure 1. Survival Rates of Export Relationships at the Sector turally dissimilar and geographically distant—have the low- Level, 1999–2011 est survival probabilities. 1.0 Export flows to China seem to be relatively resilient; GEO ARM these are only 3.6 percentage points less likely to survive than survival probability 0.8 CZE LTU MKD SVK flows to ECA countries, with one-year survival rates of just 0.6 below 50 percent. However, digging deeper, many of Geor- 0.4 gia’s exports to East Asia and Pacific (EAP) countries tend to 0.2 have low survival rates. This is likely due to the nature of the products being traded, with many being “hazardous” from 0 0 2 4 6 8 10 12 the point of view of export survival—meaning typically more analysis time differentiated, non-resource-based products.3 These types of Source: Authors’ calculations products are prone to short lifecycles, making it more likely that it is the specific product traded to EAP countries that ex- • Is the overall volume of exports to a particular destination, plains the low survival rates to the region. and hence, the amount of information available in Geor- Combining information on survival rates by firm desti- gia about exporting to that destination, associated with nation and type of product with the export structure by des- the likelihood of survival of a firm’s export flow to that tination is revealing. The export basket to EAP contains more same destination (involving any product)? of these hazardous products than the basket to other coun- • Is the overall volume of exports of a particular product, tries in ECA, to developed partners, to the Middle East and and hence, the amount of information available in the North Africa (MENA) region, or to South Asia (figure 3). economy about exporting that product (controlling for Building on these results, figure 4 plots survival rates for com- the country’s comparative advantages) associated with modities and for differentiated goods.4 It shows that the dif- the likelihood of survival of a firm’s export flow of that ference in survival is substantial, with commodities being product (to any destination)? close to 10 percentage points more likely to survive the first • Are flows that start small less likely to survive past the first year in export markets? Figure 2. Georgian Export Survival Rates by Destination, 2003–12 • Are flows from firms with a broader product or destina- export survival probability .5 China EAP tion scope more likely to survive past the first year? ECA EU27 .4 NEC Russian Fed. The Value of Destination-Specific .3 South Asia United States Information .2 Results suggest that a 100 percent increase in exports to a spe- .1 cific destination is associated with an increase in the probabil- 0 ity of survival of a firm-product-destination export flow to 0 2 4 6 8 10 that specific destination by slightly more than one-third of a years active Source: Authors’ calculations. percentage point. While these results are not economically large, they are statistically well defined. Cultural and geographic ties between trading partners Figure 3. Export Bundles by Region and Type of Product (averages for 1999–2011) are important to the survival of new export flows. Trading with neighbors, both in the cultural and geographic sense, is 90 resource-based less costly when looking at implied transaction costs.2 The 80 non-resource-based probability of a Georgian product’s export survival is high- 70 est when being traded to other countries in the Europe and 60 percent Central Asia (ECA) region. Figure 2 shows export survival 50 probabilities by region of destination. Georgian firms with 40 flows to other countries in ECA have an almost 50 percent 30 20 chance of lasting one year, and a more than 20 percent 10 chance of lasting two years. On the other hand, flows to the 0 EU-27 region have a lower chance of surviving: less than a 40 d P A C A ia A pe EA EC LA EN SS As percent probability of surviving one year, and less than 20 lo M h ve ut percent of surviving a second year. Flows to Latin America, de So the Caribbean, and sub-Saharan Africa—relatively more cul- Source: Authors’ calculations. 3 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise Figure 4. Georgian Export Survival Rates by Type of Product: the initial export flow and reveals a positive association be- Commodity/Differentiated, 2003–12 tween the two variables. The average initial value of a Geor- .5 gian firm’s export flow is roughly US$77,000 dollars. If a firm commodity export survival probability differentiated starts exporting with a flow double that average, its export is .4 about a quarter of a percentage point more likely to survive .3 than firms starting with an order of average size. One interpretation for this phenomenon is that when .2 importers are uncertain about the capacity of the exporter to .1 meet an order, or when exporters face cost uncertainty, both exporters and importers may start with small orders. Both 0 sides seek to update their information about each other 0 2 4 6 8 10 through trial and error. The case of a Georgian firm exporting years active Source: Authors’ calculations. licorice roots to Japan helps illustrate this point. The Geor- gian firm exported 20 tons to Japan in 2012, but the flow was year. This result is in line with previous findings in the litera- disrupted because the firm could not meet the importer’s ture on export survival.5 steadily increasing demand for roots. The exporter was uncer- tain of just how many of these roots were available in the The Value of Product-Specific Information country, which prevented it from committing to delivering The amount of information available in the market about ex- larger orders.8 This does not mean that exporters should start porting a particular product is systematically and positively with large orders. Instead, it suggests that the size of the initial associated with the likelihood of surviving past the first year flow may be associated with the uncertainty of the capacity of in export markets. the exporter to meet the needs of the importer. Because Georgian firms tend to export more of the prod- Another interpretation for this result is that larger firms ucts for which the country has a comparative advantage, it is tend to receive larger orders. This could happen for many rea- natural to find that export flows of these products tend to sur- sons—reputation and production capacity being two major vive longer. At the same time, these products are exported in factors. Therefore, these large firms may also have better sur- larger volumes, regardless of any informational spillover ef- vival prospects in export markets than those of smaller firms. fect. Controlling for the role of comparative advantage (using However, results continue to hold when controlling for scope revealed comparative advantage indicators) isolates the infor- of the firm in terms of destination and products, which are mational spillover effect on export survival. As expected, the likely correlated with the size of the firm.9 estimated effect of the product-specific informational vari- Survival and Diversification Interact in able on the likelihood of survival more than halves after con- Complex Ways trolling for revealed comparative advantage and firm- and flow-specific characteristics, but it remains statistically signifi- The scope of products exported by a firm and the number of cant at 1 percent. This suggests that firms benefit in the form destinations it exports to (that is, how diversified firms are of a greater probability of survival from the accumulated ex- along the product and destination dimensions) both matter porting experience of a given product. From an economic for survival. Flows from firms that reach more destinations view, and similar to the case of destination-specific informa- have a better chance of survival, as they likely have better in- tion, the effects are relatively small.6 Taking the most conser- formation on these markets. However, the converse holds for vative estimate (column 6 in Table 2), the results suggest that flows from firms that have a relatively wider product scope. If if the value of all firms’ exports of product p were to double, product specialization is associated with higher productivity, the probability of export survival for firms selling that same then firms that concentrate on fewer products may be rela- product would increase by about a quarter of a percentage tively more likely to survive in export markets. point, a proportional increase of 0.63 percent.7 Policy Implications Trial and Error as a Road to Discovery: The The main findings from this analysis have important policy Role of Smaller flows implications. As mentioned earlier, incumbent exporters do Flows that start small tend to have shorter life spans in export not have incentives to share valuable information regarding markets. Analysis shows that the initial value of the export destination markets or products with entrants. This informa- flow is positively associated with its survival outlook. Figure 5 tion was probably costly to obtain in the first place, and it plots the probabilities of surviving one year against the size of could erode their competitive edge, and ultimately their prof- 4 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise Table 2. Determinants of Export Survival Past the First Year, Georgia Dependent variable: Survivor = 1 if f, d, p lasts more than 1 year. (1) (2) (3) (4) (5) (6) Value of all exports of product ‘p’ 1.98e-06*** 1.61e-06*** 1.61e-06*** 1.55e-06*** 1.09e-06*** 7.63e-07*** (1.36e-07) (1.44e-07) (1.44e-07) (1.44e-07) (1.34e-07) (1.47e-07) Value of all exports to destination ‘d’ 7.83e-08* 8.83e-08** 1.07e-07** 9.29e-08** (4.33e-08) (4.27e-08) (4.20e-08) (4.21e-08) Destination scope 0.00466*** 0.00794*** 0.00735*** (0.000391) (0.000427) (0.000436) Product scope -0.00083*** -0.00079*** (4.52e-05) (4.52e-05) Initial export value 3.39e-05*** (8.16e-06) Revealed comparative advantage 0.000154*** 0.000155*** 0.000149*** 0.000105*** 9.32e-05*** (2.12e-05) (2.12e-05) (2.11e-05) (2.05e-05) (2.10e-05) Dummy East Asia Pacific -0.0408** -0.0372* -0.0368* -0.0452** -0.0352* -0.0320 (0.0194) (0.0199) (0.0199) (0.0184) (0.0194) (0.0199) Dummy Europe and Central Asia (non-EU-27) 0.0183 0.0219 0.0181 0.0321* 0.0362** 0.0358** (0.0170) (0.0170) (0.0172) (0.0170) (0.0167) (0.0167) Dummy EU-27 -0.0268* -0.0266* -0.0266* -0.0251 -0.0112 -0.0118 (0.0159) (0.0160) (0.0160) (0.0159) (0.0164) (0.0163) Dummy Latin America and Caribbean -0.0649*** -0.0629*** -0.0628*** -0.0734*** -0.0497*** -0.0548*** (0.0164) (0.0168) (0.0168) (0.0144) (0.0187) (0.0180) Dummy Middle East and North Africa -0.0290* -0.0295* -0.0293* -0.0241 -0.00694 -0.00728 (0.0163) (0.0162) (0.0162) (0.0167) (0.0182) (0.0181) Dummy other developed -0.0438*** -0.0413*** -0.0411*** -0.0446*** -0.0305** -0.0301** (0.0143) (0.0146) (0.0146) (0.0141) (0.0152) (0.0152) Dummy Russian Federation -0.00736 -0.00586 -0.0102 -0.00190 0.0131 0.0111 (0.0180) (0.0182) (0.0179) (0.0186) (0.0198) (0.0196) Dummy South Asia -0.00346 0.00397 0.00449 0.0135 0.0129 0.0162 (0.0224) (0.0233) (0.0234) (0.0243) (0.0238) (0.0241) Dummy sub-Saharan Africa -0.0767*** -0.0743*** -0.0740*** -0.0765*** -0.0455** -0.0469*** (0.0126) (0.0130) (0.0130) (0.0124) (0.0177) (0.0174) Dummy United States -0.0371** -0.0338** -0.0365** -0.0330** -0.0216 -0.0276   (0.0160) (0.0164) (0.0161) (0.0164) (0.0174) (0.0170) Observations 33,500 33,500 33,500 33,500 33,500 33,500 Region of destination dummies Yes Yes Yes Yes Yes Yes Initial year dummies Yes Yes Yes Yes Yes Yes Source: Authors’ calculations based on data from Geostat. Note: Robust standard errors in parentheses. its. Thus, entrants currently only benefit from the informa- materialized in Georgia. From a new exporter’s perspective, tion that ends up in the public domain through occasional understanding whether the main challenge to export survival leakages. The economically small effects identified in the is related to difficulties in a specific market, or with a specific above analysis are likely to reflect that only a fraction of the product, is essential to their business. With more detailed in- potential gains from export-related information sharing has formation, firms may not need to resort to trial and error as 5 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise Figure 5. Starting Export Value and Probability of Survival portization” of villages. Through the project, the association (1999–2011) has helped map the production capacity of more than 3,000 .8 export survival probability (1 year) .8 villages across the country.10 A separate public sector–led effort is using the Partnership Fund, with the help of the survival probability (1 year) .75 U.S. Agency for International Development (USAID), and .7 the Economic Prosperity Initiative to support farmers’ ex- fitted values ports of mandarins. The project has provided information .7 and financial resources to improve the packaging of fruits. It .6 has also introduced “consolidators” that buy from several .65 farmers and consolidate the produce to achieve a minimum efficiency of scale. These types of efforts are also being im- .5 .6 plemented in the production of bay leaves and hazelnuts. 0 1,000 2,000 3,000 4,000 The Italian firm Ferrero—with its lengthy experience ex- initial export value (in thousands) porting hazelnuts in Georgia—has helped prove that there is Source: Author’s calculations. sufficient international demand to make it profitable to ex- port these from Georgia. the dominant mechanism of cost discovery. Export “deaths” Evidence on the performance of publicly financed export and rapid exit from export markets would become less fre- promotion activities is, however, mixed (Lederman, Olarrea- quent. Therefore, from a policy maker’s perspective, support- ga, and Payton 2010). Inadequate funding, lack of strong lead- ing export survival is vital to promote broader economic ership, poor client orientation, and excess bureaucracy have growth and ensure greater export diversification. The greater been some of the problems associated with agencies engaged the export activity in a particular overseas market, or with a in export promotion in developing countries (Hogan, Keesi- particular type of product, the easier it should be to obtain ng, and Singer 1991). More recent evidence has shown that information. export promotion activities may be effective in terms of im- The evidence from Georgia supports the hypothesis that proving export performance. However, these agencies have increased information increases chances of exports’ survival. been more effective when helping circumvent trade barriers As the amount of export experience with a specific trading abroad or tackling asymmetric information problems (Leder- partner increases—as shown by the total exports to that trad- man, Olarreaga, and Payton 2010). For these reasons, and ing partner—the chance that the flow to that same market sur- given the scarcity of public funds in countries like Georgia, an vives the first year increases substantially. The same holds for evidence-based public-private dialogue is necessary to first the export experience with a particular product. As total ex- help fully understand the costs and benefits of export promo- ports of that product (to any destination) increase, the chance tion. Whether this comes in the form of information provi- of that product surviving one year also increases. sion or some other form, designing the most effective frame- The estimated spillover effects reported here are likely to work along with a monitoring and evaluation program to be only a fraction of the real value that firms attach to infor- reach the desired objectives of boosting trade competitiveness mation. The small size of the estimated spillovers in Georgia will be critical. should be taken as a lower bound of the benefits that firms About the Authors could realize by learning from the accumulated experience of other exporters serving similar markets or selling similar Jose-Daniel Reyes and Gonzalo Varela are Trade Economists, products. There is a sound, statistical relationship between Miles McKenna is a Research Analyst. All of the authors work information and export survival. It is one of many reasons— with the Trade & Competitiveness Global Practice of the World though not the silver bullet—why exports survive or die. Bank. Export promotion agencies? Notes In the light of these findings confirming the valuable role information can play in exports’ survival, it is worth discuss- 1. The dataset was kindly provided by Geostat. For more de- ing the possibility of using public funds to increase the tails on the methodology, please contact the authors. amount of information regarding exporting specific prod- 2. This has been pointed out in the literature before (see Bren- ucts or serving specific markets. Some public and private ton, Saborowski, and Uexkull [2010]). initiatives that partially address this problem are already in 3. More recently, EAP’s imports of natural resource–based place. For example, the Georgian Farmers’ Association has products from Georgia increased substantially. Over 2006– implemented a new project based on what it calls the “pass- 11, EAP’s imports from Georgia represented 70 percent of its 6 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise total imports, compared to 54 percent over 1999–2005. This porters had expressed interest in investing to start processing is likely related both to a compositional shift in terms of quan- the licorice roots in Georgia, if local collectors could supply at tities, but also to an increase in the prices of natural resource– least 1,000 tons of the product. based products over the last five years. 9. Unfortunately, there is no information in the data set about 4. The classification follows Rauch (1999). the turnover or the number of employees of the firms. 5. For example, Besedes and Prusa (2004) find a 23 percent- 10. This information was provided by the Georgian Farmers’ age point difference in survival of commodities versus differ- Association during an interview held in Tbilisi in May 2013. entiated goods for U.S. export flows. References 6. Table 2 reports the results of estimating a probit model de- signed to test for informational spillovers on export survival Araujo, L., and E. Ornelas. 2007. “Trust-Based Trade.” Centre for on the export-transaction-based data set with data for 2003– Economic Performance Discussion Paper DP 0820, LSE. 12. This closely follows the approach of Cadot et al. (2013), Besedes, T., and T. J. Prusa. 2004. “Surviving the U.S. Import Mar- ket: The Role of Product Differentiation.” National Bureau of Brenton, Saborowski, and Uexkull (2010), and Rauch and Economic Research Working Paper 10319. Watson (2003). For further details of the methodology, please Brenton, P., C. Saborowski, and E. von Uexkull. 2010. “What contact the authors. Column 1 shows a basic specification, Explains the Low Survival Rate of Developing Country Export and columns 2–6 report results for models with added regres- Flows?” World Bank Economic Review 24 (3): 474–99. sors, which capture firm- and flow-specific characteristics. Cadot, O., L. Iacovone, D. Pierola, and F. Rauch. 2013. “Success and Failure of African Exporters.” Journal of Development Economics 7. The effect is calculated at the average value of exports of 101 (March): 284–96. product p. The reported coefficient is the marginal effect, giv- Hausmann, R., and D. Rodrik. 2003. “Economic Development As ing the change in the probability of survival, given an infini- Self-Discovery.” Journal of Development Economics 72: 603–33. tesimal change in the explanatory variable. To obtain the ef- Hogan, P., D. Keesing, and A. Singer. 1991. “The Role of Support fect of a 10 percent change in the value of all exports of p on Services in Expanding Manufactured Exports in Developing the probability of survival, the coefficient is multiplied by the Countries.” Economic Development Institute, World Bank, Washington, DC. size of the induced change in exports (US$362,000, in this Lederman, D., M. Olarreaga, and L. Payton. 2010. “Export Promo- case, which accounts for 10 percent of the average value of tion Agencies: Do They Work?” Journal of Development Econom- exports of a given product p). ics 91 (2): 257–65. 8. This information was provided by the managers of the ex- Rauch, J. E. 1999. “Networks Versus Markets in International porting firm. They claimed that a comprehensive herbal re- Trade.” Journal of International Economics 48 (1): 7–35. source assessment could help mitigate uncertainty with re- Rauch, J. E., and J. Watson. 2003. “Starting Small in an Unfamiliar Environment.” International Journal of Industrial Organization spect to the production potential of licorice roots, and wild 21: 1021–42. products in general, which would boost growth and employ- Reyes, J. D., and G. Varela. 2013. “Trade Competitiveness Diagnos- ment. In fact, the managers mentioned that the Japanese im- tic: The Case of Georgia.” Mimeo, World Bank, Washington, DC. The Economic Premise note series is intended to summarize good practices and key policy findings on topics related to economic policy. They are produced by the Poverty Reduction and Economic Management (PREM) Network Vice-Presidency of the World Bank. The views expressed here are those of the authors and do not necessarily reflect those of the World Bank. The notes are available at: www.worldbank.org/economicpremise. 7 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise