1.the atlas ofM a p p i n g P a t h sECONOMIC COMPLEXITYT o P r o s p e r i t yHausmann, Hidalgo et al.

2. the atlas ofM a p p i n g P a t h sECONOMIC COMPLEXITYT o P r o s p e r i t yHausmann, Hidalgo et al. 3. THE ATLAS OF ECONOMIC COMPLEXIT YMAPPING PATHS TO PROSPERITYAUTHORS:Ricardo Hausmann | Csar A. Hidalgo | Sebastin Bustos | Michele CosciaSarah Chung | Juan Jimenez | Alexander Simoes | Muhammed A. YldrmACKNOWLEDGMENTSThe research on which this Atlas is based began around 2006 with the idea of theproduct space. In the original paper published in Science in 2007, we collaboratedwith Albert-Laszlo Barabasi and Bailey Klinger. The view of economic development ofcountries as a process of discovering which products a country can master, a processwe called self-discovery, came from joint work with Dani Rodrik and later also withJason Hwang. We explored different implications of the basic approach in paperswith Dany Bahar, Bailey Klinger, Robert Lawrence, Francisco Rodriguez, Dani Rodrik,Charles Sabel, Rodrigo Wagner and Andrs Zahler. Throughout, we received significantfeedback and advice from Lant Pritchett, Andrs Velasco and Adrian Wood.We want to thank the dedicated team that runs Harvards Center for InternationalDevelopment (CID) for helping bring the Atlas to life: Marcela Escobari, Jennifer Gala,Irene Gandara Jones, Aimee Fox, Adriana Hoyos, Andrea Carranza, Anne Morriss andCatalina Prieto. We are also indebted to the Necsys team at the MIT Media Lab and toSandy Sener. We thank the leadership at Harvard Kennedy School and the MIT MediaLab who were early enthusiasts of our work.The editorial design of this book was produced by DRAFT. We would like to especiallyacknowledge the contributions of Francisca Barros and Beltrn Garca.ISBN-10: 0615546625ISBN-13: 9780615546629 4. the atlas ofM a p p i n g P a t h s ECONOMIC COMPLEXITY T o P r o s p e r i t y| Ricardo Hausmann | Csar A. Hidalgo | Sebastin Bustos | Michele Coscia || Sarah Chung | Juan Jimenez | Alexander Simoes | Muhammed A. Yldrm | 5. PRE F A C EOver the past two centuries, mankind hasaccomplished what used to be unthink-able. When we look back at our long list ofachievements, it is easy to focus on the mostaudacious of them, such as our conquest ofthe skies and the moon. Our lives, however,have been made easier and more prosper-ous by a large number of more modest, yetcrucially important feats. Think of electricbulbs, telephones, cars, personal computers, antibiotics, TVs,refrigerators, watches and water heaters. Think of the manyinnovations that benefit us despite our minimal awarenessof them, such as advances in port management, electricpower distribution, agrochemicals and water purification.This progress was possible because we got smarter. Duringthe past two centuries, the amount of productive knowledgewe hold expanded dramatically. This was not, however, anindividual phenomenon. It was a collective phenomenon. Asindividuals we are not much more capable than our ances-tors, but as societies we have developed the ability to makeall that we have mentioned and much, much more. Modern societies can amass large amounts of produc-tive knowledge because they distribute bits and pieces of itamong its many members. But to make use of it, this knowl-edge has to be put back together through organizations andmarkets. Thus, individual specialization begets diversity atthe national and global level. Our most prosperous modernsocieties are wiser, not because their citizens are individu-ally brilliant, but because these societies hold a diversity ofknowhow and because they are able to recombine it to createa larger variety of smarter and better products. 6. The social accumulation of productive knowledge has notmost part, it is not available in books or on the Internet.been a universal phenomenon. It has taken place in some It is embedded in brains and human networks. It is tacitparts of the world, but not in others. Where it has hap-and hard to transmit and acquire. It comes from years ofpened, it has underpinned an incredible increase in livingexperience more than from years of schooling. Productivestandards. Where it has not, living standards resemble thoseknowledge, therefore, cannot be learned easily like a songof centuries past. The enormous income gaps between richor a poem. It requires structural changes. Just like learningand poor nations are an expression of the vast differences in a language requires changes in the structure of the brain,productive knowledge amassed by different nations. Thesedeveloping a new industry requires changes in the patternsdifferences are expressed in the diversity and sophistication of interaction inside an organization or society.of the things that each of them makes, which we explore in Expanding the amount of productive knowledge availabledetail in this Atlas. in a country involves enlarging the set of activities that the Just as nations differ in the amount of productive knowl-country is able to do. This process, however, is tricky. Indus-edge they hold, so do products. The amount of knowledge tries cannot exist if the requisite productive knowledge isthat is required to make a product can vary enormouslyabsent, yet accumulating bits of productive knowledge willfrom one good to the next. Most modern products require make little sense in places where the industries that requiremore knowledge than what a single person can hold. No-it are not present. This chicken and egg problem slowsbody in this world, not even the saviest geek nor the mostdown the accumulation of productive knowledge. It alsoknowledgeable entrepreneur knows how to make a com- creates important path dependencies. It is easier for coun-puter. He has to rely on others who know about batterytries to move into industries that mostly reuse what theytechnology, liquid crystals, microprocessor design, softwarealready know, since these industries require adding modestdevelopment, metallurgy, milling, lean manufacturing andamounts of productive knowledge. By gradually adding newhuman resource management, among many other skills. knowledge to what they already know, countries economizeThat is why the average worker in a rich country works in on the chicken and egg problem. That is why we find em-a firm that is much larger and more connected than firmspirically that countries move from the products that theyin poor countries. For a society to operate at a high level already create to others that are close by in terms of theof total productive knowledge, individuals must know dif- productive knowledge that they require.ferent things. Diversity of productive knowledge, however, isThe Atlas of Economic Complexity attempts to measure thenot enough. In order to put knowledge into productive use,amount of productive knowledge that each country holds. Oursocieties need to reassemble these distributed bits through measure of productive knowledge can account for the enor-teams, organizations and markets. mous income differences between the nations of the world Accumulating productive knowledge is difficult. For theand has the capacity to predict the rate at which countries 7. will grow. In fact, it is much more predictive than other well-of increasing complexity. This process involves trial and er-known development indicators, such as those that attempt toror. It is a risky journey in search of the possible. Entrepre-measure competitiveness, governance and education. neurs, investors and policymakers play a fundamental roleA central contribution of this Atlas is the creation of ain this economic exploration.map that captures the similarity of products in terms ofBy providing rankings, we wish to clarify the scope of thetheir knowledge requirements. This map provides pathsachievable, as revealed by the experience of others. By track-through which productive knowledge is more easily accu-ing progress, we offer feedback regarding current trends. Bymulated. We call this map, or network, the product space,providing maps, we do not pretend to tell potential explor-and use it to locate each country, illustrating their currenters where to go, but to pinpoint what is out there and whatproductive capabilities and the products that lie nearby.routes may be shorter or more secure. We hope this will em-Ultimately, this Atlas views economic development as a power these explorers with valuable information that willsocial learning process, but one that is rife with pitfalls andencourage them to take on the challenge and thus speed updangers. Countries accumulate productive knowledge bythe process of economic development.developing the capacity to make a larger variety of productsRicardo Hausmann Csar A. HidalgoDirector, Center for International Development at Harvard University,ABC Career Development Professor, MIT Media Lab,Professor of the Practice of Economic Development, Harvard Kennedy School, Massachusetts Institute of Technology (MIT),George Cowan Professor, Santa Fe Institute.Faculty Associate, Center for International Development at Harvard University. 8. We thank the many individuals who, early on, understood the potential impact of research oneconomic growth, and shared our teams vision. The generosity of these supporters made this work feasibleand now makes it available to individuals, organizations and governments throughout the world. T h e a u th o r s w a n t t o a c k n o w l e d g e th e g e n e r o u s s u p p o rt o f :| Alejandro Santo Domingo | MPower Foundation | Standard Bank | Anonymous Donor | 9. c o n t e n t sPA R T 1WHAT, WHY AND HOW? 12SECTION 1 What Do We Mean by Economic Complexity? 17SECTION 2 How Do We Measure Economic Complexity?24 SECTION 3 Why Is Economic Complexity Important?30 SECTION 4 How Is Complexity Different from Other Approaches? 41SECTION 5 How Does Economic Complexity Evolve?53 SECTION 6 How Can This Atlas Be Used?56 SECTION 7 Which Countries Are Included in This Atlas? 10. PA R T 2 COMPLEXITY RANKINGS60 RANKING 1Economic Complexity Index66 RANKING 2Expected Growth in Per Capita GDP to 202072 RANKING 3 Expected GDP Growth to 202078 RANKING 4Change in Economic Complexity (1964-2008)84 RANKING 5 Expected Contribution to World GDP Growth to 2020PA R T 3 COUNTRY PAGES 92 How to Read the Country Pages 96 Albania 350Zimbabwe 11. PA R T 1 WHAT, WHY AND HOW? 12. SEC T I O N 1What Do We Mean by Economic Complexity? 13. MAPPING PATHS TO PROSPERITY | 15What are things made out of? One way atomic cocktail, and who together with their colleagues atof describing the economic worldthe toothpaste factory, can deposit it into a product that weis to say that things are made with can use.machines, raw materials and labor. We owe to Adam Smith the idea that the division of laborAnother way is to emphasize thatis the secret of the wealth of nations. In a modern reinter-products are made with knowledge. pretation of this idea, the division of labor is what allowsConsider toothpaste. Is toothpasteus to access a quantity of knowledge that none of us wouldjust some paste in a tube? Or dobe able to hold individually. We rely on dentists, plumbers,the paste and the tube allow us tolawyers, meteorologists and car mechanics to sustain ouraccess knowledge about the properties of sodium fluoridestandard of living, because few of us know how to fill cavi-on teeth and about how to achieve its synthesis? The true ties, repair leaks, write contracts, predict the weather or fixvalue of a tube of toothpaste, in other words, is that it mani- our cars. Many of us, however, can get our cavities filled, ourfests knowledge about the chemicals that facilitate brush-cars repaired and our weather predicted. Markets and orga-ing, and that kill the germs that cause bad breath, cavitiesnizations allow the knowledge that is held by few to reachand gum disease.many. In other words, they make us collectively wiser. When we think of products in these terms, markets takeThe amount of knowledge embedded in a society, how-on a different meaning. Markets allow us to access the vast ever, does not depend mainly on how much knowledge eachamounts of knowledge that are scattered among the peopleindividual holds. It depends, instead, on the diversity ofof the world. Toothpaste embeds our knowledge about the knowledge across individuals and on their ability to com-chemicals that prevent tooth decay, just like cars embody bine this knowledge, and make use of it, through complexour knowledge of mechanical engineering, metallurgy, elec-webs of interaction. A hunter-gatherer in the Arctic musttronics and design. Computers package knowledge about in- know a lot of things to survive. Without the knowledge em-formation theory, electronics, plastics and graphics, whereas bedded in an Inuit, most of us would die in the Arctic, as hasapples embody thousands of years of plant domestication been demonstrated by the number of Westerners who haveas well as knowledge about logistics, refrigeration, pest con-tried and failed. Yet, the total amount of knowledge embed-trol, food safety and the preservation of fresh produce.ded in a hunter-gatherer society is not very different from Products are vehicles for knowledge, but embedding that which is embedded in each one of its members. The se-knowledge in products requires people who possess a work- cret of modern societies is not that each person holds muching understanding of that knowledge. Most of us can be ig-more productive knowledge than those in a more traditionalnorant about how to synthesize sodium fluoride becausesociety. The secret to modernity is that we collectively usewe can rely on the few people who know how to create this large volumes of knowledge, while each one of us holds only 14. a few bits of it. Society functions because its members formbaseball, you must know how to catch, field and bat, but youwebs that allow them to specialize and share their knowl- do not need to know how to give financial advice or fix cavi-edge with others. ties. On the other hand, to perform the function of baseball We can distinguish between two kinds of knowledge: ex- player, knowing how to catch a ball is not enough (you mustplicit and tacit. Explicit knowledge can be transferred easilyalso be able to field and bat). In other words, in allocatingby reading a text or listening to a conversation. Yesterdays productive knowledge to individuals, it is important thatsports results, tomorrows weather forecast or the size of thethe chunks each person gets be internally coherent so thatmoon can all be learned quickly by looking them up in a he or she can perform a certain function. We refer to thesenewspaper or on the web. And yet, if all knowledge had this modularized chunks of embedded knowledge as capabili-characteristic, the world would be very different. Countriesties. Some of these capabilities have been modularized atwould catch up very quickly to frontier technologies, and the the level of individuals, while others have been grouped intoincome differences across the world would be much smaller organizations and even into networks of organizations.than what we see today. The problem is that crucial parts of For example, consider what has happened with under-knowledge are tacit and therefore hard to embed in people.graduate degrees, which in the US take four years of study.Learning how to fix dental problems, speak a foreign lan- This norm has remained constant for the last four centuries.guage, or run a farm requires a costly and time-consuming During the same period, however, knowledge has expandedeffort. As a consequence, it does not make sense for all of usenormously. The university system did not respond to theto spend our lives learning how to do everything. Because itincrease in knowledge by lengthening the time it takes tois hard to transfer, tacit knowledge is what constrains the get a college degree. Instead, it increased the diversity ofprocess of growth and development. Ultimately, differencesdegrees. What used to be a degree in philosophy, split intoin prosperity are related to the amount of tacit knowledgeseveral branches, one being natural philosophy, which laterthat societies hold.split into physics, chemistry and biology and later into other Because embedding tacit knowledge is a long and costly disciplines such as ecology, earth sciences and psychology.process, we specialize. This is why people are trained forThe Bureau of Labor Statistics Standard Occupation Clas-specific occupations and why organizations become good at sification for 2010 lists 840 different occupations, includingspecific functions. To fix cavities you must be able to identi- 78 in healthcare, 16 in engineering, 35 kinds of scientists fy them, remove the decayed material and replace it. To playin coarse categories such as economists, physicists and 15. chemists five types of artists, and eight kinds of design- cure the fabric, cut it, sew it, pack it, brand it, market it anders. We can all imagine a much more nuanced classification distribute it. In a firm that manufactures shirts, expertisein our respective fields. For instance, we could distinguish in each of these knowledge chunks will be held by differ-between economists that specialize in labor, trade, finance, ent people. And shirts require all of them. Moreover, youdevelopment, industrial organization, macro and econo- need to finance the operation, hire the relevant people, co-metrics, among others. If we did this further disaggrega-ordinate all the activities and negotiate everybodys buy-in,tion for all occupations, we would easily go into the tens which in itself require different kinds of knowhow. We canof thousands. The only way that society can hold all of thesay that putting together this operation requires know-whoknowledge we have is by distributing coherent pieces of it and know-where. Know-who can be thought of as knowl-among individuals. It is the way the world adapts to expand- edge of who has the requisite chunks of knowledge, anding knowledge. know-where as knowledge of where the people and orga- Most products, however, require more knowledge than nizations that have this knowledge are located. To makecan be mastered by any individual. Hence, those products shirts, you can import the fabric and access the knowledgerequire that individuals with different capabilities interact. about looms and threading that is embedded in a piece ofAssume that a person has the capacity to hold an amount of cloth. Yet some of the knowledge required cannot be acce-tacit knowledge equal to one personbyte. How can you makessed through shipped inputs. The people with the relevanta product that requires 100 different personbytes? Obvious-knowledge must be near the place where shirts are made.ly, it cannot be made by a micro-entrepreneur working onIn fact, just as knowhow is modularized in people in theher own. It has to be made either by an organization with at form of individual capabilities, larger amounts of knowhowleast 100 individuals (with a different personbyte each), or are modularized in organizations, and networks of organi-by a network of organizations that can aggregate these 100 zations, as organizational or collective capabilities. For ex-personbytes of knowledge. How can a society hold a kilo-,ample, to operate a garment plant you need power and wa-mega- or giga-personbyte? Only through a deep division ofter. You need to be able to move raw materials in and shiplabor, in which individuals become experts in small pieces the final product out. Workers need access to urban trans-of the available knowledge and then aggregate their person-portation, day care centers and health facilities. To be ablebytes into peoplebytes through organizations and markets.to operate, the plant manager needs all of these services to For example, to make a shirt you need to design it, pro-be locally available. This implies that others must be aggre- 16. 18 | THE ATLAS OF ECONOMIC COMPLEXITYgating the personbytes required to generate power, provide that have it retire or die.clean water, and run a transportation system. The relevantSaid differently, countries do not simply make the prod-capabilities to perform all of these functions reside in orga- ucts and services they need. They make the ones they can.nizations that are able to package the relevant knowledgeTo do so, they need people and organizations that possessinto transferable bundles. These are bundles of knowhowrelevant knowledge. Some goods, like medical imaging de-that are more efficiently organized separately and trans-vices or jet engines, embed large amounts of knowledgeferred as intermediate inputs. We can think of these bun-and are the results of very large networks of people and or-dles as organizational capabilities the manufacturer needs.ganizations. By contrast, wood logs or coffee, embed much Ultimately, the complexity of an economy is related toless knowledge, and the networks required to support thesethe multiplicity of useful knowledge embedded in it. For operations do not need to be as large. Complex economiesa complex society to exist, and to sustain itself, people whoare those that can weave vast quantities of relevant knowl-know about design, marketing, finance, technology, human edge together, across large networks of people, to generateresource management, operations and trade law must bea diverse mix of knowledge-intensive products. Simplerable to interact and combine their knowledge to make prod- economies, in contrast, have a narrow base of productiveucts. These same products cannot be made in societies that knowledge and produce fewer and simpler products, whichare missing parts of this capability set. Economic complex-require smaller webs of interaction. Because individualsity, therefore, is expressed in the composition of a coun- are limited in what they know, the only way societies cantrys productive output and reflects the structures that expand their knowledge base is by facilitating the interac-emerge to hold and combine knowledge.tion of individuals in increasingly complex webs of orga- Knowledge can only be accumulated, transferred andnizations and markets. Increased economic complexity ispreserved if it is embedded in networks of individuals and necessary for a society to be able to hold and use a largerorganizations that put this knowledge into productive use. amount of productive knowledge, and we can measure itKnowledge that is not used, however, is also not transferred,from the mix of products that countries are able to make.and will disappear once the individuals and organization 17. SECTION 2How Do We Measure Economic Complexity? 18. 20 | THE ATLAS OF ECONOMIC COMPLEXITY H ow do we go from what a country makes to words. So we can expect the diversity of words (products) what a country knows? If making a productthat a player (country) can make to be strongly related to the requires a particular type and mix of knowl- number of letters (capabilities) that he (it) has. Long words edge, then the countries that make the will tend to be rare, since they can only be put together by product reveal having the requisite knowl- players with many letters. Hence, the number of players edge (see Technical Box 2.1). From this simple that can make a word tells us something about the variety observation, it is possible to extract a few of letters each word requires: longer words tend to be less implications that can be used to construct ubiquitous, while shorter words tend to be more common. a measure of economic complexity. First, Similarly, ubiquitous products are more likely to require fewcountries whose residents and organizations possess morecapabilities, and less ubiquitous products are more likely toknowledge have what it takes to produce a more diverserequire a large variety of capabilities.set of products. In other words, the amount of embeddedDiversity and ubiquity are, respectively, crude approxi-knowledge that a country has is expressed in its productive mations of the variety of capabilities available in a countrydiversity, or the number of distinct products that it makes.or required by a product. Both of these mappings are af-Second, products that demand large volumes of knowledge fected by the existence of rare letters, such as Q and X. Forare feasible only in the few places where all the requisite instance, players holding rare letters will be able to put to-knowledge is available. We define ubiquity as the number of gether words that few other players can make, not becausecountries that make a product (Figure 2.1). Using this termi- they have many letters, but because the letters that theynology, we can observe that complex products those thathave are rare. This is just like rare natural resources, suchcontain many personbytes of knowledgeare less ubiqui-as uranium or diamonds. Yet, we can see whether low ubiq-tous. The ubiquity of a product, therefore, reveals informa-uity originates in scarcity or complexity by looking at thetion about the volume of knowledge that is required for its number of other words that the makers of rare words areproduction. Hence, the amount of knowledge that a country able to form. If these players can only make a few otherhas is expressed in the diversity and ubiquity of the prod- words, then it is likely that rarity explains the low ubiquity.ucts that it makes. However, if the players that can make these rare words are, A game of scrabble is a useful analogy. In scrabble, play- in general, able to put together many other words, then iters use tiles containing single letters to make words. For in-is likely that the low ubiquity of the word reflects the factstance, a player can use the tiles R, A and C to construct thethat it requires a large number of letters and not just a fewword CAR or ARC. In this analogy, each product is repre-rare ones.sented by a word, and each capability, or module of embed- Diversity can therefore be used to correct the informa-ded knowledge, is represented by a letter. We assume that tion carried by ubiquity, and ubiquity can be used to cor-each player has plenty of copies of the letters they have. Ourrect the information carried by diversity. We can take thismeasure of economic complexity corresponds to estimatingprocess a step further by correcting diversity using a mea-what fraction of the alphabet a player possesses, knowing sure of ubiquity that has already been corrected by diversityonly how many words he or she can make, and how manyand vice versa. In fact, we can do this an infinite numberother players can also make those same words. of times using mathematics. This process converges after a Players who have more letters will be able to make morefew iterations and represents our quantitative measures of 19. MAPPING PATHS TO PROSPERITY | 21 F igure 2 . 1 :Graphical explanation of diversity and ubiquity. Diversity (kc,0): :D i v ersity Diversity is related to the number of X-RAY MACHINESDiversity is related a country is connectedproducts thatto the number of NETHERLANDS (NLD)products that ato. This is equal to the number of country is connected to.This is equal tolinks that this country has thatthe number of links in thenetwork. In this example, using a this country hasof the 2009 data, the In thissubset in the network. diversityexample, using a subset ofisthe 2009 data,of Netherlands 5, that of Argentinais 3, and that of Gana is 1. the diversity of Netherlands is 5, that of Argentina is 3, and that of Gana is 1. MEDICAMENTS ARGENTINA (ARG) CREAMS AND POLISHESCHEESE Ubiquity (kp,0):GHANA (GHA)Ubiquity is is :related to the number of U b i q uity Ubiquity is is related to the number of FROZEN FISH countries that a product is connected to. countries that a product is connected to. Thisisis equalthe number of linksof links that This equal to to the number that thisproduct hashas innetwork. In this In this this product in the the network. example, using a subset of the 2009 data, example, using a subset of the 2009 data, the ubiquity of Cheese is 2, that of Fish is the ubiquity of Cheese is 2, that of Fish is 3 3 and that of Medicaments is 1. and that of Medicaments is 1. 20. 22 | THE ATLAS OF ECONOMIC COMPLEXITYF igure 2 . 2 : Map of the World colored according to ECI Ranking. Rank 1128complexity. For countries, we refer to this as the Econom- 28 other countries (placing Pakistan in the 60th percentileic Complexity Index (ECI). The corresponding measure for of countries in terms of the average ubiquity of their prod-products gives us the Product Complexity Index. Technicalucts), while Singapore exports products that are exportedBox 2.2 presents the mathematical definition of these twoon average by 17 other countries (1st percentile). Moreover,quantities and Ranking 1 lists countries sorted by their ECI.the products that Singapore exports are exported by highlyFigure 2.2 shows a map of the world colored according to a diversified countries, while those that Pakistan exports arecountrys ECI ranking. exported by poorly diversified countries. Our mathematical Consider the case of Singapore and Pakistan. The popula-approach exploits these second, third and higher order dif-tion of Pakistan is 34 times larger than that of Singapore. At ferences to create measures that approximate the amountmarket prices their GDPs are similar since Singapore is 38 of productive knowledge held in each of these countries. Ul-times richer than Pakistan in per capita terms. Under thetimately, what countries make reveals what they know (seeclassification we use in this Atlas, they both export a simi-Information Box 2.1).lar number of different products, about 133. How can prod-Take medical imaging devices. These machines are made inucts tell us about the conspicuous differences in the levelfew places, but the countries that are able to make them, suchof development that exist between these two countries? as the United States or Germany, also export a large numberPakistan exports products that are on average exported byof other products. We can infer that medical imaging devices 21. MAPPING PATHS TO PROSPERITY | 23I n f o r m a t i o n B o x 2 . 1 : A n a lt e r n a t i v eway t o u n d e r s ta n d o u r m e a s u r e sof economic complexityare complex because few countries make them, and thosethat do tend to be diverse. By contrast, wood logs are exportedUnderstanding the measures of economic complexity described in this At-by most countries, indicating that many countries have thelas can be sometimes challenging. Analogies, however, can help get our mindsknowledge required to export them. Now consider the casearound what the economic complexity index is able to capture.of raw diamonds. These products are extracted in very few Think of a particular country and consider a random product. Now, askplaces, making their ubiquity quite low. But is this a reflection yourself the following question: If this country cannot make this product, inof the high knowledge-intensity of raw diamonds? Of coursehow many other countries can this product be made? If the answer is manycountries, then this country probably does not have a complex economy. Onnot. If raw diamonds were complex, the countries that wouldthe other hand, if few other countries are able to make a product that thisextract diamonds should also be able to make many other country cannot make, this would suggest that this is a complex economy.things. Since Sierra Leone and Botswana are not very diversi- Let us illustrate this with a few examples. According to our measures, Ja-fied, this indicates that something other than large volumespan and Germany are the two countries with the highest levels of economicof knowledge is what makes diamonds rare (see Information complexity. Ask yourself the question: If a good cannot be produced in Japanor Germany, where else can it be made? That list of countries is likely to beBox 2.2 on Product Complexity).a very short one, indicating that Japan and Germany are complex economies. This Atlas relies on international trade data. We made Now take an opposite example: if a product cannot be made in Mauritania orthis choice because it is the only dataset available that has Sudan, where else can it be made? For most products this is likely to be a longa rich detailed cross-country information linking countries list of countries, indicating that Sudan and Mauritania are among the worldsleast complex economies.to the products that they produce in a standardized clas-This analogy is useful to understand the difference between economic com-sification. As such, it offers great advantages, but it doesplexity and the level of income per capita of a country. Two countries that havehave limitations. First, it includes data on exports, not pro-high levels of economic complexity, but still low levels of per capita income areduction. Countries may be able to make things that they doChina and Thailand. Ask yourself the question, if you cannot produce it in Chinanot export. The fact that they do not export them, however, or Thailand, where else can you produce it? That list of countries will tend to berelatively short. The comparison becomes starker if we restrict it to countriessuggests that they may not be very good at them. Coun-with a similar level of per capita income, like Iran, Peru and Venezuela, countriestries may also export things they do not make. To circum- that do not make things that many other can.vent this issue we require that countries export a fair share At the opposite end of this comparison, there are countries with high levelsof the products we connect them to (see Technical Box 2.1). of per capita income but relatively low levels of economic complexity. ExamplesSecond, because the data is collected by customs offices, itof this are Qatar, Kuwait, Oman, Venezuela, Libya and Chile. These countriesare not rich because of the productive knowledge they hold but because of theirincludes only goods and not services. This is an importantgeological luck, given the large volumes of natural resources based wealth.drawback, as services are becoming a rising share of inter- Ask yourself the question; if you cannot build it in Chile or Venezuela, wherenational trade. Unfortunately, the statistical efforts of mostelse can you build it? The fact that there are many countries where it would becountries of the world have not kept up with this reality.possible to produce many things that are not being made in Chile or Venezuela,Finally, the data does not include information on non-trad- including countries with a similar level of income such as Hungary or the CzechRepublic, indicates that the level of economic complexity of these countries isable activities. These are an important part of the economiclow, despite their fairly high level of income.eco-system that allows products and services to be made.In fact, as we show in this Atlas, the gap between a countrys complexityOur current research is focused on finding implementableand its level of per capita income is an important determinant of future growth:solutions to these limitations, and we hope we will be able countries tend to converge to the level of income that can be supported by theto present them in future versions of this Atlas. knowhow that is embedded in their economy. 22. 24 | THE ATLAS OF ECONOMIC COMPLEXITYt e c h n i c a l b ox 2 . 1 : M e a s u r i n g e c o n o m i c c o m p l e x i t y: If we define, as a matrix that is 1 if country produces product , and and rewrite this as :otherwise, we can measure diversity and ubiquity simply by summing over therows or columns of that matrix. Formally, we define: whereTo generate a more accurate measure of the number of capabilities available We note (7) is satisfied when .This is the eigenvector ofin a country, or required by a product, we need to correct the information that which is associated with the largest eigenvalue. Since this eigenvector isdiversity and ubiquity carry by using each one to correct the other. For coun- a vector of ones, it is not informative. We look, instead, for the eigenvector asso-tries, this requires us to calculate the average ubiquity of the products that itciated with the second largest eigenvalue. This is the eigenvector that capturesexports, the average diversity of the countries that make those products and the largest amount of variance in the system and is our measure of economicso forth. For products, this requires us to calculate the average diversity of the complexity. Hence, we define the Economic Complexity Index (ECI) as:countries that make them and the average ubiquity of the other products thatthese countries make. This can be expressed by the recursion: where < > represents an average, stdev stands for the standard deviation and We then insert (4) into (3) to obtainAnalogously, we define a Product Complexity Index (PCI). Because of the symmetry of the problem, this can be done simply by exchanging the index of countries (c) with that for products (p) in the definitions above. Hence, we de- fine PCI as: where 23. MAPPING PATHS TO PROSPERITY | 25I n f o r m at i o n B o x 2 . 2 : T h e w o r l d s m o s t a n d l e a s t c o m p l e x p r o d u c t s Table 2.2.1 and Table 2.2.2 show respectively the products that rank highest products, on the other hand, are raw minerals or simple agricultural products.and lowest in the complexity scale. The difference between the worlds mostThe economic complexity of a country is connected intimately to the com-and less complex products is stark. The most complex products are sophistica- plexity of the products that it exports. Ultimately, countries can only increaseted chemicals and machinery that tend to emerge from organizations where atheir score in the Economic Complexity Index by becoming competitive in anlarge number of high skilled individuals participate. The worlds least complex increasing number of complex industries. TA B L E 2 . 2 . 1 : T O P 5 P R O D U C T S B Y C O M P L E X I T Y Product Code (SITC4)Product Name Product CommunityProduct Complexity Index7284Machines & appliances for specialized particular industriesMachinery 2.278744Instrument & appliances for physical or chemical analysisChemicals & Health2.217742Appliances based on the use of X-rays or radiation Chemicals & Health2.163345Lubricating petrol oils & other heavy petrol oilsChemicals & Health2.107367Other machine tools for working metal or metal carbide Machinery 2.05 T able 2 . 2 . 2 : B ottom 5 products by complexity Product Code (SITC4) Product NameProduct CommunityProduct Complexity Index3330Crude oilOil -3.002876Tin ores & concentratesMining-2.632631Cotton, not carded or combed Cotton, Rice, Soy & Others-2.633345Cocoa beansTropical Agriculture-2.617367Sesame seeds Cotton, Rice, Soy & Others-2.58t e c h n i c a l b o x 2 . 2 : W h o m a k e s w h at ?When associating countries to products it is important to take into accountthe size of the export volume of countries and that of the world trade of prod-ucts. This is because, even for the same product, we expect the volume of ex-ports of a large country like China, to be larger than the volume of exports of asmall country like Uruguay. By the same token, we expect the export volume ofWe use this measure to construct a matrix that connects each country toproducts that represent a large fraction of world trade, such as cars or footwear,the products that it makes. The entries in the matrix are 1 if country exportsto represent a larger share of a countrys exports than products that account for product with Revealed Comparative Advantage larger than 1, and o otherwise.a small fraction of world trade, like cotton seed oil or potato flour.Formally we define this as thematrix, whereTo make countries and products comparable we use Balassas definition ofRevealed Comparative Advantage or RCA. Balassas definition says that a coun-try has Revealed Comparative Advantage in a product if it exports more than itsfair share, that is, a share that is equal to the share of total world trade thatthe product represents. For example, in 2008, with exports of $42 billion, soy-beans represented 0.35% of world trade. Of this total, Brazil exported nearly $11billion, and since Brazils total exports for that year were $140 billion, soybeansis the matrix summarizing which country makes what, and is used toaccounted for 7.8% of Brazils exports. This represents around 21 times Brazilsconstruct the product space and our measures of economic complexity forfair share of soybean exports (7.8% divided by 0.35%), so we can say that countries and products. In our research we have played around with cutoffBrazil has revealed comparative advantage in soybeans.values other than 1 to construct thematrix and found that our results areFormally, if represents the exports of country in product , we canrobust to these changes.express the Revealed Comparative Advantage that country has in product as: Going forward, we smooth changes in export volumes induced by the pricefluctuation of commodities by using a modified definition of RCA in which thedenominator is averaged over the previous three years. 24. SEC T I O N 3Why Is Economic Complexity Important? 25. MAPPING PATHS TO PROSPERITY | 27as we have argued, economic complexity re-economic complexity is greater than what we wouldflects the amount of knowledge that is em- expect, given their level of income, tend to grow fasterbedded in the productive structure of an than those that are too rich for their current level ofeconomy. Seen this way, it is no coincidence economic complexity. In this sense, economic complex-that there is a strong correlation between ity is not just a symptom or an expression of prosperity:our measures of economic complexity andit is a driver.the income per capita that countries areTechnical Box 3.1 presents the regression that we use to re-able to generate.late economic complexity to subsequent economic growth. Figure 3.1 illustrates the relationship be- The equation is simple. We regress the growth in per capitatween the Economic Complexity Index (ECI) and Income per income over 10-year periods on economic complexity, whilecapita for the 128 countries studied in this Atlas. Here, we controlling for initial income and for the increase in realseparate countries according to their intensity in natural re- natural resource income experienced during that period. Wesource exports. We color in red those countries for whichalso include an interaction term between initial income pernatural resources, such as minerals, gas and oil, represent at capita and the ECI. The increase in the explanatory power ofleast 10% of GDP. For the 75 countries with a limited relative the growth equation that can be attributed to the Economicpresence of natural-resource exports (in blue), economic Complexity Index is at least 15 percentage points, or morecomplexity accounts for 75 percent of the variance in in-than a third of the variance explained by the whole equa-come per capita. But as the Figure 3.1 illustrates, countriestion. Moreover, the size of the estimated effect is large: anwith a large presence of natural resources can be relatively increase of one standard deviation in complexity, which isrich without being complex. If we control for the income something that Thailand achieved between 1970 and 1985,that is generated from extractive activities, which has more is associated with a subsequent acceleration of a countrysto do with geology than knowhow, economic complexity can long-term growth rate of 1.6 percent per year. This is overexplain about 73 percent of the variation in income across and above the growth that would have been expected fromall 128 countries. Figure 3.2 shows the tight relationship mineral wealth and global trends.between economic complexity and income per capita thatThe ability of the ECI to predict future economic growthemerges after we take into account a countrys natural re- suggests that countries tend to move towards an incomesource income. level that is compatible with their overall level of embedded Economic complexity, therefore, is related to a countrys knowhow. On average, their income tends to reflect theirlevel of prosperity. As such, it is just a correlation of things embedded knowledge. But when it does not, it gets correctedwe care about. The relationship between income and com-through accelerated or diminished growth. The gap betweenplexity, however, goes deeper than this. Countries whose a countrys level of income and complexity is the key vari- 26. 28 | THE ATLAS OF ECONOMIC COMPLEXITYFIGURE 3.1: Shows the relationship between income per capita and the Economic Complexity Index (ECI) for countries where natural resource exports are larger than 10% of GDP (red) and for those where natural resource exports are lower than 10% of GDP (blue). For the latter group of countries, the Economic Complexity Index accounts for 75% of the variance. Countries in which the levels of natural resource exports is relatively high tend to be significantly richer than what would be expected given the complexity of their economies, yet the ECI still correlates strongly with income for that group.NORQAT 105 DNKKWT ARE NLD CHEIRL USAFRA FINSWEAUSCANBELGBR AUT DEUESPITASGPGRCHKG JPNNZLISR SVN OMNTTOSAUPRT SVK KOR CZE LBYLVA EST HRVRUS LTUHUNVENPOLGAB CHLURY 104 KAZBRATUR ROU MEXGDP per capita in USD [2009] BWAARG LBNMYSMUS BGR ZAF PAN AZE DZA CUB CRI SRB BLRAGO PER JAM MKDCOL IRN DOM BIHECUNAMALB TUN UKR THA TKM SLV JORCOG MAR GEO CHNPRYGTMMNG SYR IDN BOL LKA EGYHND MDA SDNNGA PHL MRTPNG YEM ZMBGHA SENNIC VNM 103CMR TJKCIVUZB PAK KGZ INDLAOKEN MLI KHM BGD TZAGIN MDGUGA MOZ ETH ZWER2 = 0.75LBRMWI-3-2-101 2 Economic Complexity Index [2008]FIGURE 3.2: Shows the relationship between economic complexity and income per capita obtained after controlling for each countrys natural resource exports. After including this control, through the inclusion of the log of natural resource exports per capita, economic complexity and natural resources explain 73% of the variance in per capita income across countries. 1.5 SVNIncome per capita controlling for initial income and proportion 1 FRA CHEof natural resource exports per capita in logs [2008]DOM NLDITAAZE GRCQAT ESP IRLDEU USA AUT JPNKWTLBN GBRCOG MWILBY BEL NOR FIN SAU ARECRI DNKSWE .5PRTAGO SGPKOR BGD OMN JORHRV HUNNZL KHMTURHKGSVKCZEAUS CAN SRBMUS EST ROUGAB URYLVA LTU POL ISR PRY TTO LKA KAZ MDAPAN THADZA SLVBLR 0 VEN ARGBRA CHLRUSCHNCUBTKMBWA ALBZAF MEXSENCOL BGRMYSIRN MAR MKD PHLBIHPER ECUJAMPAK NAM UKRO GTMKEN PNG HND GEO MNG TUN NGA BOL -.5 CIV MRT SYREGY IDN IND NCC YEM MDG SDNETH ZMBUGAGIN CMRGHA LAO UZB VNMMLIKGZTJKTZA R2 = 0.73 -1MOZLBR ZWE -2 -101 2Economic Complexity Index controlling for initial income and proportionof natural resource exports per capita in logs [2008] 27. MAPPING PATHS TO PROSPERITY | 29FIGURE 3.3: Shows the relationship between the annualized GDP per capita growth for the period between 1998 and 2008 and the Economic Complexity Index for 1998, after taking into account the initial level of income and the increase in natural resource exports during that period (in constant dollars as a share of initial GDP). CHN TKM .05LVABLR Growth in per capita GDP, controlling for initial income and growth KHMESTLTURUS GEO UKRCUB BGRKAZ in natural resource exports [1998-2008] ALB ROU TTO TJKKORSVKSGP SVNBIHHKG VNM INDPOLMDA IRLGRCPANHRV ETH HUN CZEDOMLKAJOR AGOTUNTHAMUSUGA BGD MNG LAOUZB FIN SWECRI PERMYSIDNMOZ 0IRN MAR TZA EGY MKDGBRAUT KGZAUSNZL LBN CAN TUR NLD ESPPHLZAF BEL SLV OMN NIC GHA USA DEUCHL ARGISR BRACHEAREFRAECU DNK MEX JPNHND NGA NORPRT URYPAK COL AZE MLIZMBITAJAM GTMSEN VENDZALBRCMRSAUPRYSYR KENMRTYEMMDG BOLGIN COGGABPNG -.05 CIV-10 1 Economic Complexity Index controlling for initial income and growth in natural resource export [1998]able that we use here to estimate the growth potential ofa measure of export diversification. Controlling for standardcountries (Figure 3.3).measures of export concentration, such as the Herfindahl- It is important to note what the Economic ComplexityHirschman Index, does not affect our results. In fact, neitherIndex is not about: it is not about export-oriented growth,openness nor export concentration are statistically signifi-openness, export diversification or country size. Although cant determinants of growth after controlling for the ECIwe calculate the ECI using export data, the channel through(see Technical Box 3.2).which it contributes to future growth is not limited to its Finally, the ECI is not about a countrys size. The ability ofimpact on the growth of exports. Clearly, countries whosethe ECI to predict growth is unaffected when we take intoexports grow faster, all other things being equal, will neces- account a countrys size, as measured by its population,sarily experience higher GDP growth. This is simply becausewhile the population itself is not statistically significant (seeexports are a component of GDP. However, as Technical BoxTechnical Box 3.2).3.2 shows, the contribution of the ECI to future economic In short, economic complexity matters because it helpsgrowth remains strong after accounting for the growth in explain differences in the level of income of countries,real exports.and more important, because it predicts future economic The ECI is also not about openness to trade: the impact ofgrowth. Economic complexity might not be simple to ac-the ECI on growth is essentially unaffected if we account forcomplish, but the countries that do achieve it, tend to reapdifferences in the ratio of exports to GDP. And the ECI is not important rewards. 28. 30 | THE ATLAS OF ECONOMIC COMPLEXITYTechnical Box 3.1: The Growth regressionTo analyze the impact of the Economic Complexity Index (ECI) on future eco-an increase in the ECI of one standard deviation would accelerate growth by 2.3nomic growth we estimate two regressions where the dependent variable is the percent per year in a country at the 10th percentile of income, by 1.6 percent inannualized growth rate of GDP per capita for the periods 1978-1988, 1988-1998a country at the median income, and by 0.7 percent for countries in the 90thand 1998-2008. In the first of these equations we do not include ECI and use percentile. The variables contained in Column 2 jointly account for 43.4 percentonly two control variables: the logarithm of the initial level of GDP per capita inof the variance in growth rates. The difference between these two regressionseach period and the increase in natural resource exports in constant dollars as aindicates that the ECI increases the regressions R2 in 15 percentage points. Thisshare of initial GDP. The first variable captures the idea that, other things equal, represents over a third of the explained fraction of the 43.4 percent of the vari-poorer countries should grow faster than rich countries and catch up. This isance that the equation explains as a whole.known in the economic literature as convergence. The second control variable The estimates of the second column of Table 3.1.1 are used to forecast thecaptures the effect on growth of increases in income that come from naturalgrowth in GDP per capita and rank countries according to their growth poten-resource wealth, which complexity does not explain. In addition, we include atial (See Table 3.1.1). To predict average annualized growth between 2008 anddummy variable for each decade, capturing any common factor affecting all2020 we make two assumptions. First, we assume a worldwide common growthcountries during that decade, such as a global boom or a widespread financialterm for the decade, which we take to be the same as that observed in the 1998-crisis. Taken together, these variables account for 28.5 percent of the variance 2008 period. Changing this assumption would affect the growth rate of allin countries growth rates. This is shown in the first column of Table 3.1.1.countries by a similar amount but would not change the rankings. Second, weIn addition to the above mentioned variables, the second regression includes assume that there will be no change in the real value of natural resource exportsthe effect of economic complexity on growth. We do this by adding two addition-as a share of initial GDP. This implies that the we assume that natural resourceal terms: the ECI at the beginning of the decade and an interaction term between exports in real terms in the next decade will remain at the record-high levelsthe ECI and the initial level of GDP per capita. The interaction attempts to cap-achieved in 2008. This assumption may underestimate the effect on countriesture the idea that the contribution of economic complexity to future economicwhose volumes of natural resource extraction will increase significantly andgrowth depends on the level of per capita income. The second column of Table over-estimate the growth in countries that will see their natural-resource export3.1.1 shows that economic complexity is strongly associated with future eco- volumes declines. A higher (lower) constant dollar price of natural resource ex-nomic growth. The negative coefficient on the interaction term indicates thatports would improve (reduce) the projected growth performance of countries bythe impact of complexity on growth declines with a countrys level of income.an amount proportional to their natural resource intensity.For example, according to the estimation in Column 2, and using data for 1998,T able 3 . 1 . 1 Annualized growth in GDP pc (by decade)(1978-1988, 1988-1998, 1998-2008)VARIABLES(1) (2)-0.00017 -0.00638***Initial Income per capita, log(0.001)(0.001)Increase in natural resource exports0.03960*** 0.03682***- in constant dollars (as a share of initial GDP) (0.008)(0.010)Initial Economic Complexity Index (ECI)0.04430*** (0.009)[ ECI] X [Income per capita, log]-0.00371*** (0.001)Constant0.03036*** 0.08251***(0.008)(0.011)Observations 291 291R2 0.2850.434Year FEYes Yes 29. MAPPING PATHS TO PROSPERITY | 31T e c h n i c a l B ox 3 . 2 : E c o n o m i c c o m p l e x i t y:t h e v o l u m e a n d c o n c e n t r at i o n o f e x p o r t s a n d c o u n t r y s i z eThis box explores the robustness of the impact of the Economic Complex- ECI on growth remains strong and significant. Column 3 introduces export asity Index on growth. While the ECI is constructed using export data, its rela-a share of GDP. We use this as a measure of openness. Column 4 includes thetionship with future growth is not driven by export volumes or concentration. Herfindahl-Hirschman index as a measure of export concentration. Column 5To show this, we start with our basic growth equation (Table 3.2.1, column 1).includes the log of initial population as a measure of size. This is equivalent toColumn 2 adds to this equation the increase in the real value of the exportsintroducing total GDP, given that we are already controlling for GDP per capita.of goods and services in the decade in question as a fraction of initial GDP. The contribution to growth of the variables introduced in columns 3, 4 and 5 areExports are a component of GDP, and therefore, we expect them to contribute estimated to be very close to zero, are not statistically significant and do notto growth. Nevertheless, after including the increase in exports, the effect of affect the ability of the ECI to predict future economic growth.TA B L E 3 . 2 . 1Annualized growth in GDP pc (by decade) (1978-1988, 1988-1998, 1998-2008)VARIABLES(1) (2)(3)(4) (5)Initial Economic Complexity Index (ECI)0.04430***0.03005***0.04240***0.04143***0.04389***(0.009) (0.007) (0.008)(0.010)(0.009)[ECI] X [Income per capita, log] -0.00371*** -0.00244***-0.00345***-0.00354***-0.00381*** (0.001) (0.001)(0.001)(0.001)(0.001)Increase exports (goods and services)0.04549*** - in constant dollars (as a share of initial GDP)(0.007)Exports to GDP0.00009(0.000)Export Concentration-0.00890(0.008)Population, log 0.00168(0.001)Initial Income per capita, log -0.00638*** -0.00562***-0.00729***-0.00611***-0.00558*** (0.001) (0.001)(0.001)(0.001)(0.001)Increase in natural resource exports 0.03682*** 0.001690.03441***0.03699***0.03758*** - in constant dollars (as a share of initial GDP) (0.010)(0.005) (0.008)(0.010)(0.010)Constant 0.08251***0.06741***0.08616***0.08145***0.04878** (0.011) (0.011) (0.011) (0.011)(0.022)Observations 291260 284291 291R20.4340.584 0.449 0.4360.440Year FEYes YesYesYes YesRobust standard errors in parentheses *** p1Productnot exportedwith RCA>1Node size is proportionalto world trade2008 Export Opportunity Spectrum 33 Fraction of products with RCA > 1 22 Average Complexity of Missing ProductsAverage Complexity of Missing Products 22% of WT 11 6% of WT 00 2% of WT 0.8% of WT-1 -1 0.06% of WT-2 -2 0.987 0.988 0.9890.99 0.991 0.992 0.993 0.994-0.0200.020.040.06 DistanceOpportunity Gain2008 Evolution of export composition2.3 k1.8 k1.3 k 920 460 019681973 1978198319881993 199820032008 98. MAPPING PATHS TO PROSPERITY | 101 POPULATION 34 M / (33/3) GDP USD 171 B / (46/5)EXPORTS PER CAPITA USD 2,310 / (55/8) * Data are from 2008. Numbers indicate:Value (World Ranking / Regional Ranking). TOTAL EXPORTS USD 79 B / (40/5)GDPpc USD 4,974 / (67/9)EXPORTS AS SHARE OF GDP 46 % (36/8) Middle East and North Africa.ElectronicsMachineryAircraft BoilersShipsMetal productsConstr. matl. & eqpt.Home & officePulp & paper Chemicals & healthAgrochemicalsOther ChemicalsInor. salts & acidsPetrochemicalsLeatherMilk & cheeseAnimal FibersMeat & eggsFish & SeafoodTropical agric.Cereals & veg. oilsCotton/rice/soy & others Tobacco FruitMisc. AgricultureNot classified Textile & FabricsGarments Food Processing Beer/Spirits & cigs.Precious Stones CoalOil Mining1968 EXPORT TREEMAP Total Exports: 934.18 M / 228.33 B (0.41%) Total Exports: 934.18 M / 228.33 B (0.41%)1988 EXPORT TREEMAP Total Exports: 8.99 b / 2.79 T (0.32%) Total Exports: 8.99 B / 2.79 T (0.32%)3330 (84%) 1121 (8.5%)3330 (57%) 3413 (36%) 6727674967127283 0579 05652713 (1.3%)3345 05710812 (2.1%) 3352 0541 28207752 211665922008 EXPORT TREEMAP Total Exports: 79.36 B / 15.56 T (0.51%)Total Exports: 79.36 B / 15.56 T (0.51%) 3330 (58%) Crude petroleum 3414 (20%) petroleum gases3413 (20%) liquified hydrocarbons* Numbers indicate SITC-4 rev 2 codes. Parenthesis indicate percentage of total exports. Treemap Headers show: Total Trade/Total World Trade (share of world trade represented by the country). 99. 102 | THE ATLAS OF ECONOMIC COMPLEXITYangolaECONOMIC COMPLEXITY INDEX [2008] -1.793 / (127 / 25) EXPECTED GDPpc GROWTH * 0.54% / (127/25) *Expected annual average for the 2009-2020 period. 2008 Product Space Product exported with RCA>1 Product not exported with RCA>1Node size is proportionalto world trade2008Export OpportunitySpectrum 33 Fraction of products with RCA > 1 22 Average Complexity of Missing ProductsAverage Complexity of Missing Products 22% of WT 11 6% of WT 00 2% of WT 0.8% of WT-1 -1 0.06% of WT-2 -2 0.9945 0.995 0.9955 0.996 0.9965 0.997 0.9975-0.0200.02 0.040.06 DistanceOpportunity Gain2008Evolution ofexport composition3.6 k2.8 k 2.1 k 1.4 k7200 1968 19731978 1983 1988 1993 199820032008 100. MAPPING PATHS TO PROSPERITY | 103 POPULATION 18 M / (52/13)GDP USD 84 B / (61/3)EXPORTS PER CAPITA USD 3,601 / (45/2) * Data are from 2008. Numbers indicate: Value (World Ranking / Regional Ranking). TOTAL EXPORTS USD 65 B / (47/3)GDPpc USD 4,671 / (70/5) EXPORTS AS SHARE OF GDP 77 % (9/3) Sub-Saharan Africa.ElectronicsMachineryAircraft BoilersShipsMetal productsConstr. matl. & eqpt.Home & officePulp & paper Chemicals & healthAgrochemicalsOther ChemicalsInor. salts & acidsPetrochemicals Leather Milk & cheese Animal Fibers Meat & eggs Fish & Seafood Tropical agric. Cereals & veg. oils Cotton/rice/soy & othersTobaccoFruit Misc. Agriculture Not classifiedTextile & FabricsGarments Food Processing Beer/Spirits & cigs.Precious Stones CoalOilMining1968 EXPORT TREEMAP Total Exports: 289.36 M / 228.33 B (0.13%) Total Exports: 289.36 M / 228.33 B (0.13%) 1988 EXPORTTREEMAP Total Exports: 2.07 b / 2.79 T (0.1%)Total Exports: 2.70 B / 2.79 T (0.1%)0711 (51%) 2654 (3.6%)6673 (22%) 3330 (89%)6672 (7.5%) 2631 (3.2%) 0440 (4.3%) 2232 0573 2877 (0.98%) (0.84%) 2472 (2.5%) 0814 4242(1.6%)0548 0542 2924 072165752008 EXPORT TREEMAP Total Exports: 64.94 B / 15.56 T (0.42%) Total Exports: 64.94 B / 15.56 T (0.42%) 3330 (99%) Crude petroleum* Numbers indicate SITC-4 rev 2 codes. Parenthesis indicate percentage of total exports. Treemap Headers show: Total Trade/Total World Trade (share of world trade represented by the country). 101. 104 | THE ATLAS OF ECONOMIC COMPLEXITYargentinaECONOMIC COMPLEXITY INDEX [2008] 0.106 / (57 / 7) EXPECTED GDPpc GROWTH * 2.61% / (54/9)*Expected annual average for the 2009-2020 period. 2008 Product SpaceProductexportedwith RCA>1Productnot exportedwith RCA>1Node size is proportionalto world trade2008 Export Opportunity Spectrum Fraction of2.52.5 products with RCA > 1Average Complexity of Missing Products Average Complexity of Missing Products22 22% of WT1.51.511 6% of WT 2% of WT0.50.5 0.8% of WT00 0.06% of WT-0.5 -0.5 -1 -10.750.8 0.85 0.020.025 0.030.035 0.04Distance Opportunity Gain2008 Evolution of export composition1.8 k1.4 k1k 720360 0196819731978 19831988 1993199820032008 102. MAPPING PATHS TO PROSPERITY | 105 POPULATION 40 M / (30/4) GDP USD 327 B / (29/3) EXPORTS PER CAPITA USD 1,755 / (65/7)* Data are from 2008. Numbers indicate:Value (World Ranking / Regional Ranking). TOTAL EXPORTS USD 70 B / (44/4)GDPpc USD 8,189 / (53/7) EXPORTS AS SHARE OF GDP 21 % (93/12) Latin America and the Caribbean.ElectronicsMachineryAircraft Boilers Ships Metal products Constr. matl. & eqpt. Home & officePulp & paperChemicals & health Agrochemicals Other Chemicals Inor. salts & acids Petrochemicals Leather Milk & cheeseAnimal FibersMeat & eggs Fish & Seafood Tropical agric. Cereals & veg. oils Cotton/rice/soy & othersTobaccoFruitMisc. AgricultureNot classified Textile & Fabrics GarmentsFood ProcessingBeer/Spirits & cigs. Precious StonesCoal Oil Mining1968 EXPORT TREEMAP Total Exports: 1.54 b / 228.33 B (0.68%)Total Exports: 1.54 B / 228.33 B (0.68%)1988EXPORTTREEMAP Total Exports: 10.57 b / 2.79 T (0.38%)Total Exports: 10.57 B / 2.79 T (0.38%)0111 (12%) 0011 (2.4%) 0440 (13%)0813 (5.9%)0813 (16%)0440 (4.7%)0111 (3.9%)0149 (2.5%) 0344 0360(1.3%) (1.2%) 0115 (2.1%)0115034203434236 (4.2%) (0.9%) 0112 0141 4241 (1.8%) 4234 06162682 2681 (1.3%) (1.3%) (1.2%)6114 (5.2%) 2919 (1.5%) (1.5%)0149 (9.5%) 0412 (4%) 4232 (3.8%) 4236 (1.6%) 0430 0116 65122687 (1.3%) 0812 (3.3%) 0574 (4.4%) 0459 (2.9%) 41136783 6727 67820579 05852222 (6.7%) (0.9%) (0.87%) (1.8%) (1.4%)2681 (5.5%)2682 2111 (3.9%)5922 6745 6725 (3.2%)0452 0612 074142414249 6731 6746 6749 6841 (2.2%) (1.5%) 7525 212004592631 2221 0574 (1.5%) 0542121261142116 5112 5322 5831 0814 5322 8921 05794235 6782 6732(1.9%)(1.9%) (1.2%)68428481 6512 (1.6%) (0.93%)0546 7512 4233 0571 0572 08122919 29110545 28760611 5113247261232008 EXPORT TREEMAPTotal Exports: 70.04 B / 15.56 T (0.45%)Total Exports: 70.04 B / 15.56 T (0.45%) 0813 (13%) Oilcake4232 (8.8%) Soya bean oil7810 (5.1%) Cars7821 (3.4%) Trucks0111 (2.5%) Bovine0114& vansmeat061601160360 (1.1%) 0224 (0.89%)7849 (2.3%) Other7239 (1.9%) 6782vehicles parts Bulldozers, (0.74%)0344 03420240 angledozers & levellers parts 7831 2222 (8.3%) Soya beans 0440 (6.4%) Unmilled 0412 (4.6%) Other wheat & meslin, N.E.S.maizeunmilled6251 6114 (1.6%) Bovine &equine leather79243413 (2%) liquified 5989 (1.9%) 58391121 (1.2%)0579 (0.92%)hydrocarbonsChemicalWineproducts 4236 (2.7%) Sunflower 04300585 seed oil 2871 (2%) Copper 4311 05895913 5530(0.79%)(0.64%)09149710 (1.3%) 6841 (1.2%)Gold,Unwrought 0819 0460048804820572 04592221 1212 non-monetary aluminium & (0.8%)aluminium5831 0812 0422 alloys (0.58%)* Numbers indicate SITC-4 rev 2 codes. Parenthesis indicate percentage of total exports. Treemap Headers show: Total Trade/Total World Trade (share of world trade represented by the country). 103. 106 | THE ATLAS OF ECONOMIC COMPLEXITYaustraliaECONOMIC COMPLEXITY INDEX [2008] -0.321 / (79 / 12) EXPECTED GDPpc GROWTH * 1.23% / (114/15) *Expected annual average for the 2009-2020 period. 2008 Product Space Product exported with RCA>1 Product not exported with RCA>1Node size is proportionalto world trade2008Export OpportunitySpectrum33 Fraction of2.52.5 products with RCA > 122 Average Complexity of Missing ProductsAverage Complexity of Missing Products1.51.5 22% of WT11 6% of WT 0.50.500 2% of WT 0.8% of WT -0.5 -0.5 0.06% of WT -1 -1-1.5 -1.5 -2 -20.840.860.880.90.010.020.030.04 0.05 0.06 Distance Opportunity Gain2008Evolution ofexport composition8.8 k 7k5.2 k3.5 k 1.7 k0 1968 1973 1978198319881993 1998 20032008 104. MAPPING PATHS TO PROSPERITY | 107 POPULATION 21 M / (46/9) GDP USD 1 T / (14/3) EXPORTS PER CAPITA USD 8,819 / (28/3)* Data are from 2008. Numbers indicate:Value (World Ranking / Regional Ranking). TOTAL EXPORTS USD 189 B / (22/7) GDPpc USD 48,499 / (12/1)EXPORTS AS SHARE OF GDP 18 % (104/15) East Asia and Pacific.ElectronicsMachineryAircraft BoilersShipsMetal productsConstr. matl. & eqpt.Home & officePulp & paper Chemicals & healthAgrochemicalsOther ChemicalsInor. salts & acidsPetrochemicals Leather Milk & cheese Animal Fibers Meat & eggs Fish & Seafood Tropical agric. Cereals & veg. oils Cotton/rice/soy & others Tobacco FruitMisc. Agriculture Not classifiedTextile & Fabrics Garments Food ProcessingBeer/Spirits & cigs. Precious StonesCoal OilMining1968 EXPORT TREEMAP Total Exports: 3.73 b / 228.33 B (1.63%) Total Exports: 3.73 B / 228.33 B (1.63%) 1988 EXPORTTREEMAP Total Exports: 35.26 b / 2.79 T (1.26%)Total Exports: 35.26 B / 2.79 T (1.26%)2681 (32%) 2682 6851 (3.2%)2875 2879 3222 (16%) 2815 (6.9%) 2681 (13%) 0111 (6%)0012 (2.5%) (2.3%)(0.93%) 2874 0112 2876 (0.91%)5225 (2.3%)4113 6811 6861 2873 (1.5%) 2873 (4.7%)6861 6512 28776673 2872 2111 (1%) 0360 0224 (0.92%) 0611 (4.5%) 04600612 0574(1.6%) (1.5%)2877 2682 (2.9%)6512 2116 0240 2686 2816 03720111 (9.9%)0112 (2.6%) 02300224 02400488 0452 6841 (6.7%) 685128796821 0412 (4.1%) 0611 (2.4%) 88223345 2460 (2.3%) (0.94%) 0482 (1.6%) 5249 (1.4%)5922 6749 6822 0589 (2.4%) 9610 2871 28756842 (1.4%)2872 (1.4%) 5225 2631 0116 0149 2116 (2.5%) 2111 0360 6727 67126749 6725 1121 (1.6%)28202783 (1.1%) 2919 0589 4113057928746831 5241 3413 7132673104300579 05752008 EXPORT TREEMAP Total Exports: 188.89 B / 15.56 T (1.21%)Total Exports: 188.89 B / 15.56 T (1.21%) 3222 (28%) Other coal 9710 (8.5%) Gold, non-monetary 0111 (3%) Bovine meat0112 (0.82%) 01160224 (0.82%) 2681 (1.2%) Sheep or lambs greasy wool 3413 (6.2%) liquified hydrocarbons0240 03600412 (2.2%) Other wheat& meslin, unmilled 2815 (18%) Not agglomerated iron ore 2873 (3.9%) Aluminium 6841 (3.1%) Unwrought 2875 6851ore aluminium & aluminium (0.76%) alloys 0430 (0.75%)8720 89966832 2871 (2.4%) 6821 (1.6%) 6831 (1.4%) CopperUnwrought Unwrought nickel2877 6861copper && nickel alloys 2460 (0.68%)(1.2%) (0.62%) copper alloys1121 (1.5%) Wine 28602872 (0.95%)2820* Numbers indicate SITC-4 rev 2 codes. Parenthesis indicate percentage of total exports. Treemap Headers show: Total Trade/Total World Trade (share of world trade represented by the country). 105. 108 | THE ATLAS OF ECONOMIC COMPLEXITYaustriaECONOMIC COMPLEXITY INDEX [2008] 1.807 / (5/4) EXPECTED GDPpc GROWTH * 2.10% / (81/5)*Expected annual average for the 2009-2020 period. 2008 Product SpaceProductexportedwith RCA>1Productnot exportedwith RCA>1Node size is proportionalto world trade2008 Export Opportunity Spectrum2.5 2.5 Fraction of products with RCA > 12 2 Average Complexity of Missing Products Average Complexity of Missing Products 22% of WT1.5 1.5 6% of WT 2% of WT 1 1 0.8% of WT 0.06% of WT0.5 0.50 00.54 0.560.58 0.6 0.62 0.64 -0.07 -0.065 -0.06-0.055-0.05 -0.045 -0.04DistanceOpportunity Gain2008 Evolution of export composition21 k16.8 k12.6 k 8.4 k 4.2 k0 19681973 19781983198819931998 20032008 106. MAPPING PATHS TO PROSPERITY | 109 POPULATION 8.3 M / (77/11) GDP USD 415 B / (25/11) EXPORTS PER CAPITA USD 20,631 / (12/7) * Data are from 2008. Numbers indicate: Value (World Ranking / Regional Ranking). TOTAL EXPORTS USD 172 B / (27/11)GDPpc USD 49,739 / (11/8) EXPORTS AS SHARE OF GDP 41 % (43/4) Western Europe.ElectronicsMachineryAircraftBoilers ShipsMetal productsConstr. matl. & eqpt. Home & office Pulp & paperChemicals & health AgrochemicalsOther ChemicalsInor. salts & acids PetrochemicalsLeather Milk & cheese Animal FibersMeat & eggs Fish & Seafood Tropical agric. Cereals & veg. oilsCotton/rice/soy & others TobaccoFruit Misc. Agriculture Not classifiedTextile & Fabrics GarmentsFood ProcessingBeer/Spirits & cigs. Precious StonesCoalOilMining1968 EXPORT TREEMAPTotal Exports: 1.94 b / 228.33 B (0.85%) Total Exports: 1.94 B / 228.33 B (0.85%)1988 EXPORT TREEMAP Total Exports: 31.40 b / 2.79 T (1.12%)Total Exports: 31.40 B / 2.79 T (1.12%)7499 7411 7281 7491 6746 (3.9%)6623 (2.6%) 8472 (2.6%)6560 (2.2%) 65227132 (4.3%)6418 7492 7281 62512482 (2.7%)6732 6343 6412 (2.9%)(1.8%)(1.3%)7452 73617731 8219 7493 7272 78516770 8731 8510 (2.4%) 641564136727 (2.8%) 84716517 74497436 7491 6911 69978812 6631 72127284 (2.5%) 6842 (1.9%)(1.7%)(0.8%) 7259 87436428 64116577 8947 7788 6353 8842665267826991 6992 6716 6822 6783 (1.4%) (1.2%) 64248921 2671 (1.6%) 562167827721 7234 7499 7162 3510 (1.4%)66336760(1.2%) (1%)(0.9%)(1%)89607283(1.3%)79116733 6745812453356419 513754168939 5831 (1.2%) 6658 (0.95%)69917442 6954 6731 7822 (1.1%)2482 (6.4%) 2483 (3.5%) 6842 641264150011 (2.9%) 6512 2782 (1.4%) (1.2%)26716785 67605832 6210(1.5%) (3.3%)(1.4%) 5839 874276118947 7638 67466745 67835834 0111 6114 2518 02406841 6998 (1.8%)(1.5%)(1.7%)(0.86%)51485621 0011(1.3%)6822 67276633 6996 6341 6413 0118 27898510 8462 8431 65226517 6421 6416 66186658 8972 8124 7761 67496744 6716 (1.4%)6732 (3.5%) 7731 6953 7852 8998 8459 8465 (0.98%) 6924 (1.2%) 89286421 6623 6252 7915 6560 84636516 0585 6343(1.1%) 88428983 771269118993 2665 (1%) 0484 8960 6960 67702008 EXPORT TREEMAP Total Exports: 172.26 B / 15.56 T (1.11%) Total Exports: 172.26 B / 15.56 T (1.11%) 7810 (5%) Cars 7849 (3.5%) Other 7234 (1.4%) 6991 (1.3%) 71397224 79196842 (1.6%) 2482 (1.4%) 3510 67327731 (0.84%) 6415vehicles partsConstructionBase metal (0.86%) WorkedWorked wood& mininglocksmiths aluminium & of coniferousmachinery wares N.E.S. aluminium6997 (0.78%) 6412 74937492 6954 alloys 63436924663364137162 (1.1%) 7239 8219 (1.2%)6911AC electric (0.89%)Furniture parts 6973motors & 7491 7499 N.E.S.generators 8211 6353 (1.2%)7452 7281 6782 7272 7188 Builders` carpentry 7132 (3%) Internal7284 (2.5%) 6416 7832 7758 & joinery combustion engines forSpecialized industry motor vehiclesmachinery & parts7361 5417 (3.2%) Medicaments1110 (1.4%)8931 011372837436 7842 7413Non-alcoholic N.E.S6785 7259 6418beverages N.E.S.7416 8121 8939 (1.1%)09800223 7821 (1.8%) 7721 (1.8%)7442 (1.6%) (0.6%) Trucks & vans Switchboards,Lifting & 7783 6760Miscellaneous relays & fuses loadingarticles of plastic72126421machinery5821 (0.58%) 61147431 5416 (1.7%) Glycosides &5831 2820 8972 vaccines(0.84%)(0.5%) 7643 (1.1%) 7788 8124 6749 (1.7%) Other6746 (0.77%) 7711 (0.74%) worked iron/steel8928 sheets5839 77128942 6745 621069986821 (0.84%) (0.82%) 7763 7415 (0.69%) 674487425832 7915* Numbers indicate SITC-4 rev 2 codes. Parenthesis indicate percentage of total exports. Treemap Headers show: Total Trade/Total World Trade (share of world trade represented by the country). 107. 110 | THE ATLAS OF ECONOMIC COMPLEXITYazerbaijanECONOMIC COMPLEXITY INDEX [2008] -1.223 / (117/27)EXPECTED GDPpc GROWTH * 1.25% / (112/27)*Expected annual average for the 2009-2020 period.2008ProductSpace Product exported with RCA>1 Product not exported with RCA>1Node size is proportionalto world trade2008Export OpportunitySpectrum 33 Fraction of products with RCA > 1 22 Average Complexity of Missing ProductsAverage Complexity of Missing Products 22% of WT 11 6% of WT 00 2% of WT 0.8% of WT-1 -1 0.06% of WT-2 -20.955 0.96 0.965 0.97 0.975 0.98 0.985 0 0.020.040.06 DistanceOpportunity Gain2008Evolution ofexport composition5.5 k4.4 k3.3 k2.2 k 1.1 k0 1994 1996 1998 200020022004 2006 2008 108. MAPPING PATHS TO PROSPERITY | 111 POPULATION 8.7 M / (76/11) GDP USD 46 B / (74/16)EXPORTS PER CAPITA USD 5,507 / (37/7) * Data are from 2008. Numbers indicate:Value (World Ranking / Regional Ranking). TOTAL EXPORTS USD 48 B / (55/10) GDPpc USD 5,329 / (65/17) EXPORTS AS SHARE OF GDP 103 % (4/1) Eastern Europe and Central Asia.ElectronicsMachineryAircraft BoilersShipsMetal productsConstr. matl. & eqpt.Home & office Pulp & paperChemicals & health AgrochemicalsOther ChemicalsInor. salts & acidsPetrochemicalsLeather Milk & cheeseAnimal FibersMeat & eggsFish & SeafoodTropical agric.Cereals & veg. oilsCotton/rice/soy & others Tobacco FruitMisc. AgricultureNot classified Textile & Fabrics GarmentsFood ProcessingBeer/Spirits & cigs. Precious Stones Coal Oil Mining1994 EXPORT TREEMAP Total Exports: 53.45 M / 4.30 t (0.001%) Total Exports: 53.45 M / 4.30 T (0.001%)2001 EXPORT TREEMAP Total Exports: 2.34 b / 5.90 T (0.04%) Total Exports: 2.34 B / 5.90 T (0.04%)2631 (61%) 6592 (8.5%)6521 (5.3%) 3330 (93%) 2634 0577 12116513 (2.9%) 1222 6841 (3.8%) 56235113 (0.94%) 7239 2820 (2.1%) 2882 (3.1%) 7832 2116 7415(1.2%) 21125831 6821 (1.6%) 68618973 2681 89600571057223202008 EXPORT TREEMAP Total Exports: 47.78 B / 15.56 T (0.31%)Total Exports: 47.78 B / 15.56 T (0.31%) 3330 (98%) Crude petroleum* Numbers indicate SITC-4 rev 2 codes. Parenthesis indicate percentage of total exports. Treemap Headers show: Total Trade/Total World Trade (share of world trade represented by the country). 109. 112 | THE ATLAS OF ECONOMIC COMPLEXITYbangladeshECONOMIC COMPLEXITY INDEX [2008] -0.887 / (103/4) EXPECTED GDPpc GROWTH * 2.40% / (65/4)*Expected annual average for the 2009-2020 period.2008ProductSpaceProductexportedwith RCA>1Productnot exportedwith RCA>1Node size is proportionalto world trade2008 Export Opportunity Spectrum 3 3 Fraction of products with RCA > 1 2 2 Average Complexity of Missing Products Average Complexity of Missing Products 22% of WT 1 1 6% of WT 0 0 2% of WT 0.8% of WT 0.06% of WT-1-1-2-2 0.86 0.88 0.90.92 0.940.960.01 0.02 0.03 0.04 0.050.06 DistanceOpportunity Gain2008 Evolution of export composition100 80 60 40 200 1994 1996 19982000 2002 2004 20062008 110. MAPPING PATHS TO PROSPERITY | 113 POPULATION 160 M / (7/3)GDP USD 80 B / (62/3) EXPORTS PER CAPITA USD 104 / (123/4)* Data are from 2008. Numbers indicate: Value (World Ranking / Regional Ranking). TOTAL EXPORTS USD 17 B / (71/3) GDPpc USD 497 / (120/4) EXPORTS AS SHARE OF GDP 21 % (95/1) South Asia.ElectronicsMachineryAircraft BoilersShips Metal products Constr. matl. & eqpt. Home & office Pulp & paperChemicals & healthAgrochemicals Other Chemicals Inor. salts & acidsPetrochemicals Leather Milk & cheese Animal Fibers Meat & eggs Fish & Seafood Tropical agric. Cereals & veg. oilsCotton/rice/soy & others TobaccoFruit Misc. AgricultureNot classifiedTextile & FabricsGarments Food ProcessingBeer/Spirits & cigs.Precious Stones Coal OilMining1968 EXPORT TREEMAP Total Exports: 731.88 M / 228.33 B (0.32%) Total Exports: 731.88 M / 228.33 B (0.32%) 1988 EXPORTTREEMAPTotal Exports: 1.25 b / 2.79 T (0.04%)Total Exports: 1.25 B / 2.79 T (0.04%)2640 (26%) 6513 (8.4%) 6521 (4.6%) 8441 (15%)8439 (8.1%) 0360 (14%)6114 (5.3%) 6522 (4%)6592 6116 6113 (3.1%)(1.1%) (3.3%)65492631 (15%)6545 (9.9%)6581 (8.4%) 85106581 (4.1%) 8423 (3%)8463 84312111(2.3%) 034203500118 2633 8435 (3.9%) 6545 (8.5%) 2640 (4.6%) 5621 84598429 8434 8421(2.5%) 6118 (3.7%) 03600350 8947 (1%)(0.99%)0422 (3.9%)2634(1.3%) (0.76%) 29118462 (3.1%) 8451 8443 8452 (1.3%)8433 2681 2682 (0.93%) 0741 (2.1%) 6113 6114 2919 651905452008 EXPORT TREEMAPTotal Exports: 16.58 B / 15.56 T (0.11%)Total Exports: 16.58 B / 15.56 T (0.11%) 8462 (21%) Knitted undergarments of cotton 8423 (14%) Mens trousers 8439 (10%) Other women outerwear 0360 (2.8%) Fresh, chilled, frozen or salted crustaceans & molluscs 6114 (1.3%) Bovine & equine8459 (6.5%) Other knitted outerwear6584 (2.7%) Linens8429 (2.2%) 8435leather & furnishing textileOther men (1.7%) articlesouterwear Blouses 8451 (17%) Knitted jerseys, pullovers & cardigans 2640 (1%) Raw processed jute and other fibres 8510 (1.2%) 84638434 6581 84658441 (6.3%) Mens undershirt Footwear 6582 8431 8484 8452 5621 (0.76%)(0.72%)* Numbers indicate SITC-4 rev 2 codes. Parenthesis indicate percentage of total exports. Treemap Headers show: Total Trade/Total World Trade (share of world trade represented by the country). 111. 114 | THE ATLAS OF ECONOMIC COMPLEXITYbelarusECONOMIC COMPLEXITY INDEX [2008] 1.116 / (21/5)EXPECTED GDPpc GROWTH * 3.99% / (4/1) *Expected annual average for the 2009-2020 period.2008ProductSpaceProductexportedwith RCA>1Productnot exportedwith RCA>1Node size is proportionalto world trade2008 Export Opportunity Spectrum Fraction of 3 products with RCA > 122.5Average Complexity of Missing Products Average Complexity of Missing Products 22% of WT 21.5 6% of WT1.51 2% of WT1 0.8% of WT 0.06% of WT0.50.5 00 -0.5 0.68 0.7 0.72 0.74 0.760.78 0.8-0.02-0.01 00.01 Distance Opportunity Gain2008 Evolution of export composition3.4 k2.7 k2k1.3 k680 01994199619982000 20022004 2006 2008 112. MAPPING PATHS TO PROSPERITY | 115 POPULATION 9.7 M / (74/10)GDP USD 61 B / (65/11)EXPORTS PER CAPITA USD 3,409 / (46/11)* Data are from 2008. Numbers indicate: Value (World Ranking / Regional Ranking). TOTAL EXPORTS USD 33 B / (58/11)GDPpc USD 6,277 / (61/16) EXPORTS AS SHARE OF GDP 54 % (24/6) Eastern Europe and Central Asia.ElectronicsMachinery AircraftBoilers Ships Metal products Constr. matl. & eqpt. Home & officePulp & paper Chemicals & healthAgrochemicals Other Chemicals Inor. salts & acids PetrochemicalsLeather Milk & cheeseAnimal FibersMeat & eggsFish & SeafoodTropical agric.Cereals & veg. oilsCotton/rice/soy & others Tobacco Fruit Misc. Agriculture Not classifiedTextile & Fabrics GarmentsFood ProcessingBeer/Spirits & cigs. Precious StonesCoalOil Mining1994 EXPORT TREEMAPTotal Exports: 942.42 M / 4.30 t (0.02%)Total Exports: 942.42 M / 4.30 T (0.02%)2001 EXPORT TREEMAP Total Exports: 6.25 b / 5.90 T (0.11%)Total Exports: 6.25 B / 5.90 T (0.11%)5623 (13%) 5156 (2%) 6841 (9.1%)6672 7224 (7.2%)7821 (7.4%)7224 (4%)6931 6259 7752 (3.1%) 6732 82198431 8510(2.8%)(1.5%)(2.6%) (2.3%)(1.2%) (1.1%) 5138 (1.5%) 84656252 6644 8459 8720 514868216770 7711 6624 2482 6353 6633(1.8%) 6725 (2.2%)6259 6770 7832 (2.1%) 7234 7491 6251 6572(1.3%)6514 (1.8%) 65317611 87106342 5112 7162 7758(1.1%)6931 (0.85%)5621 (5.8%)58312882 2890 52257831 6940 74226973 7861 2482 (2.7%) 8219 2665 (6.7%) 5922 61148124 (1%) (0.98%) 2666 65227783 7212 7132 7361 5629(1.4%) (1.5%) (1.1%) 67256544 785266526415 65178431 (2.5%) 84396514 (1.7%) 266602245623 (9%)5621 53346419 6546 0224 (1.3%) 0230 266589311124(1.9%)(1.3%) 63422471 (1.5%)(1.7%) 0011 0372 592206201123 0240 77526732(0.97%) (0.77%) 061226516544 0113 0251 6114 5823 6618 5823 (2.2%) 3224 05462631 5831 (1.7%) 5156 (1.7%)84298421(1.3%) 6416(1.2%)(0.95%) 6351 2820846565170579 33520111 6651 59816643 26722008EXPORTTREEMAP Total Exports: 32.96 B / 15.56 T (0.21%) Total Exports: 32.96 B / 15.56 T (0.21%) 7224 (6.1%) Wheeled tractors7832 (2.2%)7212 (1.3%) 6259 (2.1%) Other tires 69316732 (4.3%) Iron/steel rods 7731 (1.1%)66246973 Tractors for Harvesting &articles(1.2%)Electric wire semi-trailersthreshingmachines6633 (0.99%) 7234 (1%) 7861 Construction & 6770 (1.1%) 77116252 mining Not 6353 821163596416 7132machinery7442 7831 insulated 8219 (2%)7752 (1.9%)iron/steelFurniture partsRefrigerators & 24722471 7821 (5.8%) Trucks & vans (0.98%)6644 Internal wireN.E.S. freezers69966725 (2.9%) Iron/steel combustion 6911billets6343 engines for 7822 6940 motor vehicles 0240 (2.5%)0224 (2.3%)6514 6531 84657188Cheese & curdPreserved, (0.78%) 7431 concentrated or 5623 (20%) Potassic fertilizers 5989 (2.2%) Chemicalsweetened milk & 2666 productscream3413 (1.6%) liquified 2783hydrocarbons0230 (1.2%) Butter 5922 0223 5156 (0.86%)2665 (0.73%)06120111 (1.4%) 0113 0371 (0.96%) 6419 5831(0.97%) Refined sugar5823Polyethylene01498931 (1%) 6546 0142 6114* Numbers indicate SITC-4 rev 2 codes. Parenthesis indicate percentage of total exports. Treemap Headers show: Total Trade/Total World Trade (share of world trade represented by the country). 113. 116 | THE ATLAS OF ECONOMIC COMPLEXITYbelgiumECONOMIC COMPLEXITY INDEX [2008] 1.085 / (22/11) EXPECTED GDPpc GROWTH * 1.86% / (97/10) *Expected annual average for the 2009-2020 period.2008ProductSpace Product exported with RCA>1 Product not exported with RCA>1Node size is proportionalto world trade2008 Export Opportunity Spectrum Fraction of2.5 2.5 products with RCA > 1 Average Complexity of Missing Products Average Complexity of Missing Products2 2 22% of WT1.5 1.5 6% of WT1 1 2% of WT 0.8% of WT0.5 0.5 0.06% of WT0 0-0.5-0.50.59 0.60.61 0.62 0.630.64 0.65-0.05-0.04-0.03-0.02 -0.01 Distance Opportunity Gain2008 Evolution of export composition45 k36 k27 k18 k 9k019681973 1978198319881993199820032008 114. MAPPING PATHS TO PROSPERITY | 117 POPULATION 11 M / (66/8) GDP USD 505 B / (20/7) EXPORTS PER CAPITA USD 44,544 / (4/1)* Data are from 2008 . Numbers indicate:Value (World Ranking / Regional Ranking). TOTAL EXPORTS USD 477 B / (8/5)GDPpc USD 47,194 / (14/9)EXPORTS AS SHARE OF GDP 94 % (6/1) Western Europe.ElectronicsMachineryAircraft Boilers Ships Metal products Constr. matl. & eqpt.Home & officePulp & paper Chemicals & health AgrochemicalsOther Chemicals Inor. salts & acidsPetrochemicalsLeatherMilk & cheese Animal Fibers Meat & eggs Fish & Seafood Tropical agric. Cereals & veg. oilsCotton/rice/soy & others TobaccoFruitMisc. AgricultureNot classified Textile & FabricsGarments Food Processing Beer/Spirits & cigs.Precious Stones Coal Oil Mining1968 EXPORT TREEMAP Total Exports: 7.88 b / 228.33 B (3.45%) Total Exports: 7.88 B / 228.33 B (3.45%) 1988EXPORT TREEMAP Total Exports: 93.29 b / 2.79 T (3.34%)Total Exports: 93.29 B / 2.79 T (3.34%)6733 (4.9%)6746 (4.6%)7810 (9.7%)8822 (2.4%) 5629 6732 (4%)7810 (13%) 67276746 6749 6672 (8.4%) 68428219 (1.2%) (2.2%)(1.9%)(1.6%)(0.81%) 5232 6911 51553345 8211 3510 5622 2733 6842 66186733 (1.3%) 6744 6745 6343 6633 5137(1.3%) 6911(1%)692468226749 6725 6731 (1.4%) 8946 6419 5542 5621 664966336822 (1.2%) (0.85%) 6424 08196924 7821 7234 7783 78426725 6821 6861 5225 6649(2.4%)(1.2%)78316643 6412(1.5%) (1.3%) (0.9%) 6412 8960 8922 533581216770 693176116747 6783 72127449 0113 0224 0230 0730 6512 (2.4%) 01130114 05450730 77816727 (1.3%) 67456770 (1.6%)664464187431 74936644 61300111 0013 0565 0484 0484 8822 (1.4%) 23315417 8939 5831583201496821 (7.5%)6672 (6.3%)6998 0980(1.4%)(1.8%)(1%)2682 2686 02245113 0011 65146851 6935 2820 29265156 5148 58346595 (1.6%) 5823 65166861 6811 6522 6517 8472 04825824 5542 89216539(0.93%) (0.92%) 6519 5839 598956295621 65846651 6932 6573 53356594 659666122651 2665 52325913 5833 0819 84232008 EXPORT TREEMAPTotal Exports: 477.19 B / 15.56 T (3.07%) Total Exports: 477.19 B / 15.56 T (3.07%) 5156 (1.8%) 5416 (1.8%) 8720 (1.3%) Medical 8822 5145 7810 (9.1%) Cars 7821 (1.1%)7832 (0.8%)6727 (1.8%) 6749 (1.6%) HeterocyclicGlycosides &instruments N.E.S. Iron/steel coilsOther worked compound; nucleic vaccines iron/steel sheets acids 5111899674937234 (1.1%) 5137 5829 533151236746 (1.1%)6822 5148 (1.7%) Other 58397431 7784 nitrogen-function compounds 5146 516177836744 (0.98%) 6733 5989 (1.6%) Chemical58247842 7851 products 7781 23317436 69548124 5417 (12%) Medicaments 5831 (1.9%)Polyethylene 6672 (4.5%) Not mounted 3414 (2.6%) petroleum gases6842 (0.7%)64288510 (1.1%) diamondsFootwear6911 83105832 6812682184235833 659501135913 3352(0.52%) 0730 0546 893164215121 893955305419 56212820 0573 05850224 (0.59%) 5334 5629 0484 0980 1123 5542 8928 0579* Numbers indicate SITC-4 rev 2 codes. Parenthesis indicate percentage of total exports. Treemap Headers show: Total Trade/Total World Trade (share of world trade represented by the country). 115. 118 | THE ATLAS OF ECONOMIC COMPLEXITYboliviaECONOMIC COMPLEXITY INDEX [2008] -0.879 / (101/20) EXPECTED GDPpc GROWTH * 2.03% / (89/19)*Expected annual average for the 2009-2020 period.2008ProductSpace Product exported with RCA>1 Product not exported with RCA>1Node size is proportionalto world trade2008 Export Opportunity Spectrum 33 Fraction of products with RCA > 1 22Average Complexity of Missing Products Average Complexity of Missing Products 22% of WT 11 6% of WT 00 2% of WT 0.8% of WT 0.06% of WT-1 -1-2 -2 0.91 0.920.930.94 0.95 0 0.01 0.020.03 0.040.050.060.07 Distance Opportunity Gain2008 Evolution of export composition710568426284 142 01968197319781983 19881993 199820032008 116. MAPPING PATHS TO PROSPERITY | 119 POPULATION 9.7 M / (73/12)GDP USD 17 B / (99/18)EXPORTS PER CAPITA USD 714 / (87/17) * Data are from 2008. Numbers indicate:Value (World Ranking / Regional Ranking). TOTAL EXPORTS USD 6.9 B / (88/12) GDPpc USD 1,720 / (96/20) EXPORTS AS SHARE OF GDP 42 % (42/4) Latin America and the Caribbean.ElectronicsMachineryAircraft BoilersShipsMetal productsConstr. matl. & eqpt. Home & office Pulp & paperChemicals & health Agrochemicals Other Chemicals Inor. salts & acids PetrochemicalsLeatherMilk & cheese Animal FibersMeat & eggsFish & SeafoodTropical agric.Cereals & veg. oilsCotton/rice/soy & others Tobacco FruitMisc. AgricultureNot classified Textile & Fabrics GarmentsFood ProcessingBeer/Spirits & cigs. Precious StonesCoal Oil Mining1968 EXPORT TREEMAP Total Exports: 135.17 M / 228.33 B (0.06%) Total Exports: 135.17 M / 228.33 B (0.06%)1988 EXPORT TREEMAP Total Exports: 578.38 M / 2.79 T (0.02%) Total Exports: 578.38 M / 2.79 T (0.02%)2876 (50%) 2879 (14%) 07113414 (41%)2875 (10%)2876 (5.8%)0711 (3.7%) 0612 (1.7%) 0813 2222 0577 (1.3%) (1.3%) 23202874 (2.2%)6871 (7.3%) 2871 (4.1%)06129710 (9.1%) 2879 3330 (13%) (4.2%)0577 2483 (4%)6116611861142874 (7.4%) 2875 (2.2%) 687221202890 (2.6%) 211127412008 EXPORT TREEMAP Total Exports: 6.92 B / 15.56 T (0.04%) Total Exports: 6.92 B / 15.56 T (0.04%) 3414 (50%) petroleum gases2875 (12%) Zinc0813 (4.6%) Oilcake4232 (2.7%) Soya 4236 (1.8%)bean oil Sunflower seed oil 22222239 2874 (2.7%) Lead ore2876 (0.87%)2224 Tin0577 (1.4%)06120542Edible nuts 2890 (8%) Ores and precious metals 9710 (2.2%)Gold,0711non-monetary6871 (3.7%) Unwrought tin & alloys8973 (1.1%) 2483 (0.87%)65345121Preciousjewellery6513* Numbers indicate SITC-4 rev 2 codes. Parenthesis indicate percentage of total exports. Treemap Headers show: Total Trade/Total World Trade (share of world trade represented by the country). 117. 120 | THE ATLAS OF ECONOMIC COMPLEXITYbosnia and herzegovinaECONOMIC COMPLEXITY INDEX [2008] 0.597 / (38/12)EXPECTED GDPpc GROWTH * 3.59% / (8/4)*Expected annual average for the 2009-2020 period. 2008 Product SpaceProductexportedwith RCA>1Productnot exportedwith RCA>1Node size is proportionalto world trade2008 Export Opportunity Spectrum Fraction of2.52.5 products with RCA > 122 Average Complexity of Missing ProductsAverage Complexity of Missing Products 22% of WT1.51.5 6% of WT 11 2% of WT 0.50.5 0.8% of WT 0.06% of WT00-0.5 -0.5 -1 -1 0.72 0.74 0.760.78 0.8 0.82 0.005 0.010.0150.02 0.025DistanceOpportunity Gain2008 Evolution of export composition1.3 k1k780520260 01994199619982000 2002 20042006 2008 118. MAPPING PATHS TO PROSPERITY | 121 POPULATION 3.8 M / (107/20)GDP USD 19 B / (96/20)EXPORTS PER CAPITA USD 1,341 / (74/21)* Data are from 2008. Numbers indicate:Value (World Ranking / Regional Ranking). TOTAL EXPORTS USD 5.1 B / (95/21)GDPpc USD 4,906 / (68/18) EXPORTS AS SHARE OF GDP 27 % (73/18) Eastern Europe and Central Asia.ElectronicsMachineryAircraft BoilersShipsMetal productsConstr. matl. & eqpt.Home & office Pulp & paperChemicals & health AgrochemicalsOther ChemicalsInor. salts & acidsPetrochemicals Leather Milk & cheeseAnimal FibersMeat & eggsFish & SeafoodTropical agric.Cereals & veg. oilsCotton/rice/soy & othersTobacco FruitMisc. Agriculture Not classifiedTextile & Fabrics GarmentsFood ProcessingBeer/Spirits & cigs. Precious StonesCoal Oil Mining1994 EXPORT TREEMAPTotal Exports: 22.38 M / 4.30 B (0.001%) Total Exports: 22.38 M / 4.30 T (0.001%)2001 EXPORT TREEMAP Total Exports: 915.09 M / 5.90 T (0.02%) Total Exports: 915.09 M / 5.90 T (0.02%)8439 (12%)8510 (5.7%) 8946 (6.1%)7783 (3.9%) 6822 (2.6%)2483 (9%)8211 (4.8%)82196841 (19%) 2873(2.4%) 67166733 7493 7284 (2.1%)6732 (1.6%)6911 2471 63417436 (5.8%)(1.8%) (1.7%)2482 (5.3%)5225 3510 (1.5%)63536123 (3.9%) 8431 843584417643 (1.7%) 7428 7832 7416 2450(3.6%) (2.3%)(2.2%)24723223 8211 (2.3%)8219 8922 (5%) 1121 (3%) 6841 8510 (6.7%) 8424 8429 84312882 2783 8451 8424 8462 (1.2%) (1.2%) (1.5%) (1.4%) 69548946 7783 2111 (2.8%) (1%) 634127838462(1%)7499 6940 8452 8481 2483 (1.9%)74368434 (3.1%) 6911 6842 28828423 8421 8451 8422 8459 6123 (3%) 8439 (3%) (1.1%) 0484 52326589 2924 (2.4%) 011805616652 (3.1%) 84656716 (1.9%) 6783 67706516 (2.3%)6513 (1.7%)6521 65170819 (1.7%)8481 8459 6935 12126536(0.83%) 67332008 EXPORT TREEMAP Total Exports: 5.06 B / 15.56 T (0.03%) Total Exports: 5.06 B / 15.56 T (0.03%) 3510 (