If you are searching for how to translate BEC codes, how to translate Broad Economic Categories, how to translate BEC Revision 5 descriptions, or how to preserve end-use and global value-chain meaning in multilingual trade statistics, the crucial starting point is that BEC is a statistical classification, not a loose vocabulary list. Translate the human-readable category description; preserve the code, revision, hierarchy and dimension logic that make the classification analytically useful.
This matters in international trade statistics, national accounts, global value-chain research, import and export analysis, government dashboards, economic reports, product classification, services statistics and multilingual data systems. A translation can read beautifully and still be economically wrong if intermediate consumption becomes final consumption, if “generic” and “specified” inputs collapse into one term, if goods and services are mixed, if primary and processed products are confused, or if BEC Rev. 4 language is silently applied to BEC Rev. 5 data.
This guide explains how to translate BEC Revision 5 safely. It shows how the UN classification separates broad economic category from end-use, adds services, distinguishes processing, specification and durability, and supports analysis of contemporary production networks. It also explains how to keep BEC distinct from HS, SITC, CPC and COICOP. Conceptually, it connects to eduKateSG’s vocabulary and How English Works ecosystems because the challenge is fundamentally semantic: the translator must preserve how technical words carve an economic system into meaningful contrasts.
1. BEC means Classification by Broad Economic Categories
BEC is maintained within the United Nations statistical classification system. Its purpose is to provide broad product groupings that support analysis of trade and economic use. The classification does not merely rename individual products. It organizes them so economists can ask questions about consumption, investment, intermediate inputs and wider production structures. That analytical purpose must guide translation. When a target term sounds natural but changes the economic role of the category, it is not an acceptable equivalent.
2. Revision 5 is a major structural redesign
BEC Revision 5 was adopted in 2016 and published by the United Nations in 2018. It differs significantly from Revision 4. The UN explicitly notes that Rev. 5 fully separates broad economic categories from end-use categories, adds services to the product dimension and introduces a specification dimension. Translation therefore requires version awareness. Reusing older BEC terminology without checking the new structure can turn a correct Rev. 4 phrase into a misleading Rev. 5 label.
3. BEC is not SITC
SITC organizes merchandise trade for analytical purposes, while BEC provides broad economic categories and dimensions designed to support other kinds of economic interpretation. BEC historically developed in relation to SITC, but the classifications are not interchangeable. A SITC code should not be translated into a BEC code by linguistic resemblance. If a project needs correspondence between them, use official statistical mappings. Language translation and classification mapping are separate operations.
4. BEC is not the Harmonized System
The Harmonized System is a detailed customs-oriented commodity classification used widely for merchandise-trade collection. BEC Rev. 5 can be defined in relation to HS categories, but it serves a different analytical purpose. The same traded good can therefore have an HS identity and a BEC role. Translators should preserve both where they coexist. Replacing a BEC descriptor with the nearest HS wording can erase the economic dimension that the BEC code was chosen to express.
5. BEC is not the Central Product Classification
CPC is a broad product classification covering goods and services. BEC Rev. 5 uses product-classification relationships, including CPC for services, to support its own structure. That does not make the two classifications equivalent. CPC asks what product is being classified; BEC adds analytical perspectives such as broad economic category and end use. When translating data that contains both, label the fields explicitly so users can tell product identity from economic-use grouping.
6. BEC is not COICOP
COICOP classifies individual consumption according to purpose. BEC includes final consumption as an end-use dimension but is not a household-consumption classification. Confusing the two can happen because both use words such as consumption, goods and services. The translator needs to read the methodological role, not just the vocabulary. A familiar word can carry a narrower statistical meaning depending on the classification system around it.
7. The first major idea is broad economic category
BEC Rev. 5 organizes products into eight broad economic categories. These cover agriculture and food; mining, fuels, chemicals and related products; construction and household-related products; textiles and apparel; transport and travel-related products and services; ICT, business, finance and media; health, culture, sport and education; and government and other activities. The wording should be translated as an analytical umbrella, not as a claim that every product inside a category is produced by one industry.
8. Category 1 combines agriculture, forestry, fishing, food, beverages and tobacco
This label illustrates why BEC names are broad. It collects related product families into an economic category used for analysis. Translators should preserve each major component and avoid shortening the label to simply “food” if the official structure is being presented. A shortened dashboard label can be useful, but it should remain a display alias linked to the full canonical description rather than replacing the underlying statistical label.
9. Category 2 covers mining, quarrying, refinery, fuels, chemicals, electricity, water and waste treatment
This is another composite category. Translation must keep distinctions such as mining versus quarrying, fuels versus electricity, and water versus waste treatment visible where the full label is required. The category demonstrates that BEC is not a simple product dictionary: it creates broad analytical spaces. A translator should avoid reinterpreting the category according to one country’s ministry structure or industrial terminology.
10. Category 3 connects construction, housing, appliances and furniture
The combination makes sense analytically even though the items span different product forms and uses. This is exactly why the code matters. A natural-language reader may wonder why furniture appears beside construction. The answer lies in the classification’s economic design. Translation should preserve that design rather than reorganizing the label into a more culturally familiar grouping.
11. Category 4 covers textile, apparel, footwear and jewellery
Textile vocabulary is especially prone to confusion between material, intermediate input and finished consumer product. Translators should keep “textile” broad enough to include relevant product classes while distinguishing apparel and footwear as downstream forms. If the target language uses different words for fabric, textile material and textile industry, choose the term that fits the classification scope rather than the most common retail word.
12. Category 5 brings together transport equipment and services, travel and postal services
This category shows why BEC Revision 5’s inclusion of services matters. The analytical field extends beyond physical vehicles to transport-related services and travel. A translator who assumes BEC is only a goods classification can misread the structure. Keep “transport equipment” distinct from “transport services,” and preserve “travel” as a statistical services concept rather than translating it only as tourism marketing language.
13. Category 6 includes ICT, business, finance and media
These knowledge-intensive areas contain terminology that changes quickly. Acronyms such as ICT may have official target-language equivalents or may remain internationally recognizable. Check institutional practice before inventing an acronym. The translation should also distinguish financial services, business services and media-related products rather than flattening them into a generic “services” label. Broad does not mean semantically careless.
14. Category 7 includes health, sport, cultural activities and education
This grouping combines human-capability and social-service domains. When translating “cultural activities,” avoid narrowing the term to arts alone if the statistical category is broader. Likewise, “education” here functions as a product or service grouping within BEC rather than a school-level classification such as ISCED. Context prevents overlap with other statistical families.
15. Category 8 covers government and other
Residual and government-related categories require particular care because “other” is always relative to the classification structure. It does not mean “unimportant” or “uncertain.” Translate residual wording consistently and preserve the parent context. If a user interface shows only “Other,” provide enough metadata or navigation so readers know which universe it is residual to.
16. The second major idea is the product dimension
BEC Revision 5 distinguishes goods and services at the product dimension. This is a major change from earlier structures because services are explicitly incorporated. Translation should use a stable pair of terms for goods and services across the entire dataset. Switching between “goods,” “products,” “merchandise” and “articles” can blur statistical distinctions, especially when “product” is being used as an umbrella term that includes services.
17. “Goods” is narrower than “products” in statistical language
Everyday English often uses “product” to mean a physical good. Statistical frameworks can use product more broadly to include goods and services. A translator must preserve that distinction. If the target language has one common word for both, add a clarifying phrase where necessary. This is a classic vocabulary problem: the dictionary gives several possible equivalents, but the classification defines the intended conceptual scope.
18. “Services” in BEC follow statistical definitions
Services are not simply “everything intangible.” The System of National Accounts and related classifications provide technical definitions and distinctions, including knowledge-capturing products and different forms of service production. Translators should rely on the statistical context rather than everyday intuition. A label such as “computer services” may need to remain distinct from software products, licensing or telecommunications depending on the underlying mapping.
19. The third major idea is end use
BEC Rev. 5 separates broad economic category from SNA end-use dimension. This makes it possible to ask whether products within an economic category are used for intermediate consumption, gross fixed capital formation or final consumption. That separation is one of the most important improvements of Revision 5. The translation must therefore keep these three end-use concepts distinct and stable across all categories.
20. Intermediate consumption is not ordinary “consumption”
In national accounts, intermediate consumption refers to goods and services used up or transformed in production. Translating it with an everyday word meaning household consumption would be wrong. Use the accepted national-accounts term in the target language where one exists. If a general audience needs help, add a plain-language explanation such as “inputs used in producing other goods and services,” but keep the formal statistical term intact.
21. Gross fixed capital formation is an investment concept
Gross fixed capital formation refers to acquisition and creation of fixed assets in the national-accounts framework. It is not simply “capital goods” in casual business language. A translator should not shorten the phrase unless the target statistical community has an established abbreviation or equivalent. The word “gross” and the concept of fixed capital are both meaningful. Losing them can turn a precise SNA concept into a vague idea of investment.
22. Final consumption is distinct from intermediate use
Final consumption concerns goods and services used to satisfy individual or collective needs rather than as inputs into further production. Translators should preserve that contrast. A product can be physically identical yet have different end-use classification depending on how it is used. This is why BEC mapping sometimes requires empirical judgement: language alone may not reveal the actual economic use.
23. End use can be ambiguous
Some products have multiple plausible uses. A computer can be purchased by a household for final consumption, by a business as fixed capital or as part of another product. BEC methodology recognizes such challenges. Translation should not pretend that the label alone resolves them. If a dataset contains an allocated end-use code, preserve that code; if classification is still being assigned, use the relevant statistical rules and evidence rather than a translator’s intuition.
24. The fourth major idea is processing
BEC Rev. 5 distinguishes primary and processed forms for relevant goods. This dimension helps analysts understand where a product sits in a value-added chain. Translating “primary” as merely “important” would be disastrous. Here it concerns processing state. The target term should clearly express unprocessed or primary-form status according to the classification’s definitions.
25. “Processed” does not always mean “finished”
A processed good can still be an intermediate input. The adjective describes transformation, not necessarily final consumer readiness. Translators should therefore avoid target words that imply “complete,” “retail-ready” or “final” unless the classification says so. The distinction between processed intermediate goods and final goods is central to value-chain analysis.
26. The fifth major idea is specification
Revision 5 introduces a specification dimension that distinguishes generic and specified products in relevant contexts. This helps analysts study global value chains because highly specific intermediate inputs can reflect coordinated production relationships. Translation must preserve the technical contrast. “Specified” does not merely mean “described in detail”; it refers to the degree to which a product is made or used for particular production relationships or end products.
27. Generic products are not low-quality products
“Generic” can sound evaluative in everyday language, especially in pharmaceuticals or branding. In BEC it functions analytically. A generic intermediate may be standardized, widely traded or suitable across many uses. The translation should choose a neutral statistical term and avoid connotations of inferior quality. This is a good example of why vocabulary knowledge must be domain-specific.
28. Specified products can indicate tighter value-chain relationships
A specified processed intermediate may be designed for a particular industry, model or production relationship. This makes the category useful in studying global value chains. Translators should preserve the sense of specificity without turning it into “customized” if the official definition is broader. Review nearby examples and explanatory notes before settling on terminology.
29. The sixth major idea is durability
For relevant final-consumption goods, BEC distinguishes durable and non-durable categories. Durability here is an economic classification concept, not a marketing claim about quality. A “durable” good is not necessarily better made; it is expected to be used repeatedly or over a longer period according to statistical conventions. Translate the category using established economic terminology.
30. BEC Rev. 5 changes durability treatment from earlier revisions
Earlier BEC structures included semi-durable distinctions in ways that do not simply carry over into Revision 5. This reinforces the need to keep revision metadata visible. A legacy bilingual glossary may contain correct older terms that no longer map directly onto the new structure. Do not migrate them automatically. Review the current hierarchy and dimensions.
31. BEC Revision 5 includes services because the economy changed
Modern trade is not only trade in physical merchandise. Business, finance, ICT, transport, travel and other services are central to international production and consumption. BEC Revision 5 reflects that reality. Translation teams working from old commodity-focused habits must expand their terminology resources accordingly. A single glossary built for manufactured goods will not be enough for services, intellectual work and digital activities.
32. Global value chains are a key analytical use
The UN revision work explicitly highlights the growing importance of global value chains. Distinguishing intermediate from final products, primary from processed forms and generic from specified inputs helps analysts examine how production is distributed across countries. Translators should preserve this logic because a small lexical collapse can erase the very distinction the classification was revised to measure.
33. “Global value chain” should be translated consistently
The phrase has established equivalents in many languages. Use the institutional or academic term that readers expect, not a literal phrase that sounds like a physical chain. If the target audience includes non-specialists, explain that the concept concerns stages of production and value creation spread across firms and countries. Keep the formal term available for search and comparability.
34. BEC is multidimensional even when the code looks compact
A BEC Rev. 5 code can reflect several analytical dimensions at once. Translators should resist reading it as a simple numbered list. The hierarchy combines category, product type and use-related dimensions. If the system exposes decomposed fields, preserve them. If only a composite code is available, maintain documentation that shows how its positions map to concepts.
35. Never translate digits in the canonical code
Localized interfaces can display local numeral glyphs for ordinary quantities, but classification identifiers should remain canonical machine-readable values unless an official encoding explicitly says otherwise. A code is not a sentence. Keep it stable in databases, APIs and exports. If the interface visually localizes numerals, retain the original code in accessible metadata or a copyable canonical field.
36. Do not infer a BEC code from the translated label alone
A label can be ambiguous without context, and multiple categories can contain similar nouns. Classification requires the underlying product mapping and dimensional rules. Translators should not “repair” a missing code by choosing the label that sounds closest. Preserve the missing value or send it to a statistical classifier with the necessary source information.
37. Correspondence tables are data tools
BEC is designed to relate to classifications such as HS and CPC. Correspondence tables are therefore essential. They should be versioned and treated as structured data, not copied manually into a bilingual document. A correspondence row may express a relationship that cannot be summarized by one translated phrase. Keep mapping logic separate from language resources.
38. Mapping HS to BEC is not linguistic translation
A user searching “translate HS code to BEC” is usually asking for a statistical crosswalk. The correct workflow identifies the HS edition, finds the official correspondence to BEC Rev. 5 and applies any required end-use assumptions. Only after the BEC record is established should its label be translated. This distinction answers search intent more accurately and avoids unsafe code guessing.
39. Mapping CPC to BEC services needs version awareness
BEC Rev. 5 was defined with reference to CPC 2.1 for services. Newer CPC revisions can require updated correspondence work. A translation should never assume that a service label from CPC Version 3 automatically maps to the same BEC code without checking current tables or methodology. Classification versions evolve independently.
40. Search intent: “what is my BEC code?”
This question is often really a classification request. The user needs product information, possibly HS or CPC source codes, and an understanding of end use. A translation article can explain the process but should not imply that one English product name uniquely determines a BEC code. Ask the data, not the wording alone. For production systems, use official correspondence resources and documented statistical rules.
41. Search intent: “translate BEC code description”
Here the correct task is language translation. Keep the code and revision fixed, consult the full hierarchy, identify every active dimension, and render the label in the target language. If an official UN translation exists for that language, prefer or reconcile with it. Add plain-language help separately rather than replacing the formal descriptor.
42. Search intent: “BEC intermediate vs final consumption”
This is a conceptual distinction that translation must make visible. Intermediate consumption feeds production; final consumption satisfies final needs. The same everyday word for “use” may be insufficient. A strong target glossary records the formal statistical terms and explanatory notes. Search-oriented content should define the contrast directly because users often encounter these labels in databases without the full manual beside them.
43. Search intent: “BEC generic vs specified”
Users may mistake this for a language-style distinction. It is an economic classification dimension connected to product specificity and value-chain analysis. Translate both terms as a controlled pair and explain them together. Pairs matter: the meaning of one is partly defined by its contrast with the other.
44. Search intent: “BEC primary vs processed”
This pair describes processing stage. Translators should check whether the target economic-statistics community uses a term closer to “primary,” “unprocessed” or “raw” for the formal category. Avoid words that imply agricultural primary production unless the BEC definition is that narrow. Again, official terminology and hierarchy should outrank dictionary habit.
45. Spreadsheet workflows can destroy dimensional relationships
A BEC dataset may have separate columns for code, description, broad category, product dimension and end-use dimension. Sorting or filtering only one translated column can detach those attributes. Use immutable row identifiers and perform translations in joined tables or controlled localization systems. After import, verify entire records, not isolated codes.
46. Duplicate translations can hide distinct categories
If two different BEC records receive identical target descriptions, that may be legitimate in rare cases but should trigger review. The source might distinguish goods from services, intermediate from final use or generic from specified products with a small modifier the target lost. Automated duplicate-label checks are therefore useful semantic QA.
47. Translation memories need dimensional context
A phrase such as “final consumption” may recur hundreds of times. Translation memory can improve consistency, but a label containing several dimensions may differ by one adjective. Store context fields such as parent category and code. Reviewers should see the full BEC path before accepting high matches. A classification termbase is more valuable than raw sentence memory alone.
48. Machine translation should receive controlled terminology
Machine translation can handle many repeated descriptors efficiently when the glossary protects terms such as intermediate consumption, gross fixed capital formation, final consumption, primary, processed, generic, specified, durable and non-durable. Without these controls, an engine may produce stylistic variants that sound natural but destroy category consistency. The goal is not variety; it is stable meaning.
49. Generative AI should not assign end use from a product name alone
An AI model may infer that a machine is capital equipment or that food is final consumption, but real products can have multiple uses. Production classification needs source evidence and statistical rules. AI can flag likely cases for review, explain dimensions or help draft glossary notes. It should not silently overwrite an official BEC assignment.
50. Human reviewers should compare sibling codes
The most reliable linguistic review often happens horizontally. Compare all labels under one broad category and ask whether each distinction remains visible. Then review vertically from broad category to detailed code and ask whether the target narrows logically. This catches errors that sentence-by-sentence proofreading misses.
51. Worked example: an ICT input
Consider a specialized component used in a larger technology system. Its BEC treatment can involve the ICT broad category, goods product dimension, intermediate consumption, processed status and possibly a specified rather than generic role. A translation that says only “ICT product” loses most of the analytical meaning. Preserve the dimensional labels or keep the composite code linked to explanatory metadata.
52. Worked example: household furniture
A furniture item purchased by a household may fall within a broad construction, housing, appliances and furniture category and be treated as final consumption with durability characteristics. The translator should not infer its classification from the English word “furniture” alone, but once the code is known, the target description should make the economic role understandable without altering the code.
53. Worked example: transport service
A freight service and a piece of transport equipment can both relate to category 5 but differ in the product dimension and use. Translating both simply as “transport” would erase a critical goods/services distinction. Structured interfaces should show enough of the hierarchy to prevent this collapse, especially in multilingual dashboards with shortened labels.
54. Worked example: a dual-use product
A product used both by households and businesses may not have a universally obvious end-use allocation. The statistical compiler may rely on national accounts or proportionality assumptions. The translator should preserve the assigned category and translate its label; it should not change the end-use classification because one target-language word feels more consumer-oriented.
55. Worked example: a generic intermediate
A widely traded standardized input may be classified as generic rather than specified. In some languages, the ordinary translation of “generic” suggests a low-cost copy. That connotation would be misleading. The reviewer chooses the established statistical term and, if necessary, adds a note explaining that generic means broadly applicable rather than inferior.
56. Worked example: a specified intermediate
An auto component designed for a particular model can illustrate the specified dimension. The target term should communicate product specificity and production linkage without implying that every item is individually customized. A clear explanatory example helps readers understand why the distinction matters for global value-chain analysis.
57. Build a BEC translation termbase
The minimum termbase should include broad economic category, goods, services, intermediate consumption, gross fixed capital formation, final consumption, primary, processed, generic, specified, durable, non-durable, product, end use, value chain, correspondence table and residual-category qualifiers. Store definition, approved target term, source, revision and examples. This turns repeated translation decisions into governed terminology rather than personal memory.
58. Use official multilingual editions where available
The UN classification family provides BEC Revision 5 information in multiple official languages. Those versions should anchor terminology for those languages. If a local statistical office uses a different conventional term, document the choice and audience. The goal is not blind copying but interoperability: readers should be able to connect the translated label to the international category.
59. Record the source and revision
A robust translated dataset stores “BEC Revision 5,” the source publication or code list, retrieval date, translation version and reviewer. If official corrections or correspondence tables change later, analysts can reconstruct what was used. Classification translation is a maintained data product, so provenance matters as much as style.
60. Cite the authoritative UN source
Technical reviewers should consult the United Nations Statistics Division BEC page and the official Classification by Broad Economic Categories Revision 5 publication. These sources explain the redesign, dimensions, structure and relationships to HS and CPC.
61. Keep this owner distinct from CPC Version 3
eduKateSG’s CPC owner should continue to answer product-classification questions. This BEC page answers how broad economic categories and end-use dimensions travel across languages. A user asking “what service is this?” belongs primarily in CPC; a user asking whether a product is intermediate, final, capital-related, generic or specified belongs in BEC analysis. Clear ownership prevents cannibalisation.
62. Keep this owner distinct from COICOP
COICOP owns household and individual-consumption purpose. BEC final consumption is one end-use dimension within a trade-oriented economic classification. The pages can link because the vocabulary overlaps, but neither should absorb the other. Search users benefit when the site explains why similar words belong to different statistical systems.
63. Connection to vocabulary learning
BEC demonstrates how technical vocabulary works through contrasts. “Intermediate” means more clearly beside “final.” “Generic” means more clearly beside “specified.” “Primary” means more clearly beside “processed.” A strong vocabulary learner does not memorize isolated definitions; they learn the semantic field. Translators of classifications do the same thing at professional scale.
64. Connection to How English Works
BEC labels compress several grammatical relationships into noun phrases. “Gross fixed capital formation,” “final consumption,” “processed intermediate goods” and “specified intermediate services” are compact structures where adjective order and noun modification carry conceptual hierarchy. Translators should unpack the grammatical relationship before rebuilding it. This linguistic analysis supports technical accuracy without turning the page into a competing grammar hub.
65. Connection to the master translation architecture
This article belongs beneath eduKateSG’s Master Art of Translation — The Technical Translation System. The master owns broad process, terminology governance and QA. This page owns one specialist intent: BEC Revision 5 economic-category and end-use meaning.
66. Release checklist
Before release, confirm the dataset is really BEC Revision 5; codes are unchanged; broad economic category, product dimension and end-use dimension remain distinct; primary and processed terminology is consistent; generic and specified are not mistranslated as quality judgements; durable and non-durable retain statistical meaning; services are not dropped; HS and CPC mappings are treated as correspondences rather than translations; and every translated label remains attached to the correct code and hierarchy.
Then review grammar, readability, search language and explanatory notes. The target should be understandable to readers who search for “translate BEC code,” but precise enough for statisticians who need the same analytical category after translation.
67. Final rule: translate the economic label, preserve the analytical dimensions
BEC Revision 5 is powerful because it lets trade be viewed through several lenses at once: broad economic area, goods or services, end use, processing, specification and durability. Translation succeeds when those lenses survive intact. Change the language so people can understand the system; do not change the code or the economic relationships that make the system useful.