Skip to content

Brazil and Portugal · Printed matter · 2019–2025

Portugal in Brazil’s printed-matter trade: two shares, one widening bilateral gap

In 2025 Portugal received 4.146% of Brazil’s HS4901 exports and originated 0.606% of its imports. Brazil’s bilateral export/import ratio rose from 1.213 in 2019 to 2.516 in 2025, while both partner shares fell. This broad customs basket is not a measure of fiction, language or reading demand.

Published
Share
Share this page
EmailCite & reuse this study
Key findings as text
Reusable imagesPortugal in Brazil’s printed-matter trade: two shares, one widening bilateral gap — 4.146%: Portugal share of Brazil’s 2025 HS4901 exports. Destination basis; denominator $44,584,821, all reported partners.Download sharing image (JPG) ↓

Key findings

4.146%

Portugal share of Brazil’s 2025 HS4901 exports

Destination basis; denominator $44,584,821, all reported partners.

0.606%

Portugal-origin share of Brazil’s 2025 HS4901 imports

Origin basis; denominator $121,227,811, all reported partners.

2.516

2025 bilateral FOB export/import ratio

$1,848,406 divided by $734,682; not either country’s dependence.

7 of 7

Years with positive Brazilian bilateral FOB balance

Complete 2019–2025, broad HS4901 merchandise basket.

Two denominators, one broad commodity basket

Portugal accounted for 4.146% of Brazil’s exports of printed books, brochures and related printed matter in 2025, but only 0.606% of Brazil’s imports of that basket. The first percentage divides Portugal-bound exports by exports to all reported destinations. The second divides Portugal-origin imports by imports from all reported origins. Neither percentage describes Portugal’s own dependence on Brazil, and the two denominators must not be combined into a single market share.

The basket is Harmonized System heading 4901, not a measure of literary exchange in its entirety. It includes books alongside brochures, leaflets and similar printed matter, and its boundaries exclude several products that a general reader might regard as books. This analysis uses the broad customs category consistently rather than assigning a more attractive but unsupported label such as novels, Portuguese-language publishing or cultural consumption. Every monetary comparison uses Brazil-reported nominal US dollars on a free-on-board basis.

The central finding is a contrast between bilateral asymmetry and national partner shares. Brazil’s recorded exports to Portugal exceeded its Portugal-origin imports in every year from 2019 through 2025. The export/import ratio increased between the endpoints even though Portugal’s share of each corresponding Brazilian trade flow decreased. Those statements are compatible because a bilateral ratio and a national partner share answer different questions. Understanding the denominators is more informative than treating either result as evidence of a change in reading preferences.

What the customs category includes

The extraction retains all three observed Brazilian NCM codes within heading 4901. Code 49011000 covers books, brochures and similar printed matter in single sheets, whether or not folded. Code 49019100 covers dictionaries and encyclopaedias, including serial instalments. Code 49019900 covers other books, brochures and similar printed matter. All three appear in both trade directions in each of the seven annual observations. No narrower subcode is silently substituted for the complete heading.

Several exclusions are important. Newspapers and periodicals belong to heading 4902, while children’s picture, drawing and colouring books belong to 4903. Printed music and cartographic products have their own headings. Predominantly advertising publications fall outside 4901 under the chapter notes. This does not mean that every illustrated book or every book intended for children is excluded: the specific customs definition, rather than the age of a presumed reader, determines inclusion. Electronic books, subscriptions delivered as services and copyright licensing are not measured by this merchandise basket.

These boundaries limit interpretation but do not make the category unusable. They provide a consistent, publicly observable grouping for comparing declared merchandise flows. The defensible question is how Portugal fits within Brazil’s trade in that grouping. It is not how much Brazilian fiction Portugal bought, how many Portuguese titles Brazilian readers consumed, or which country’s publishers benefited. Those questions require information about titles, language, transactions and rights that the selected public aggregates do not contain.

Portugal as an export destination

Brazil’s HS4901 exports to Portugal were $1,195,595 in 2019 and $1,848,406 in 2025. Over the same endpoints, Brazil’s exports of the basket to all reported destinations increased from $22,116,040 to $44,584,821. Portugal-bound value therefore increased, while Portugal’s share decreased from 5.406% to 4.146%. The national denominator grew faster than the bilateral numerator; a declining share is not equivalent to a declining amount.

The intervening observations matter. Portugal’s export share was 3.213% in 2020, 4.612% in 2021, 2.674% in 2022, 3.682% in 2023 and 3.523% in 2024. The series is not a steady decline. It contains a fall, a rebound, another fall and later increases. Reporting only the endpoints would conceal that uneven path, while reporting only the latest increase would conceal the lower share relative to 2019. The figure and annual table retain every year.

Portugal ranked ninth among the reported export partner categories in 2019 and eighth in 2025. Across the whole period its rank ranged from seventh to tenth. Rank here means one plus the number of partners with a strictly larger annual FOB value. It is a descriptive ordering within this commodity basket, not an assessment of diplomatic importance or commercial opportunity. A rank can improve while share falls because the values of other partners also change.

Portugal as an origin of imports

Brazil recorded $985,554 in HS4901 imports originating in Portugal in 2019 and $734,682 in 2025. The corresponding all-origin Brazilian import totals were $93,759,541 and $121,227,811. Portugal’s share consequently declined from 1.051% to 0.606%. Unlike the export comparison, the endpoint numerator fell while the denominator rose. This describes the declared origin of merchandise, not the location of the seller, rights owner or author.

The import shares between those endpoints were 0.535% in 2020, 1.027% in 2021, 0.701% in 2022, 0.774% in 2023 and 0.665% in 2024. Again, the sequence is uneven rather than monotonic. The 2021 share approached the 2019 level before declining in later observations. Portugal ranked fourteenth among import origins in 2019 and eighteenth in 2025, with annual ranks ranging from fourteenth to nineteenth across the study period.

An origin-based series should not be presented as a complete account of goods purchased from businesses in Portugal. A transaction involving a Portuguese intermediary can involve merchandise originating elsewhere. Conversely, a Portuguese-origin product can reach Brazil through another commercial route. The aggregate classification does not identify those arrangements. Retaining the phrase Portugal-origin imports prevents the geographical label from becoming an unsupported statement about contractual relationships or the nationality of firms.

A bilateral surplus throughout the period

Subtracting Portugal-origin imports from Portugal-bound exports produces a positive Brazilian FOB balance in every observed year. The balance was $210,041 in 2019, $264,935 in 2020, $288,871 in 2021, $479,707 in 2022, $502,435 in 2023, $797,242 in 2024 and $1,113,724 in 2025. These are balances for one merchandise heading and one partner relationship, not Brazil’s overall trade balance with Portugal.

The ratio of exports to imports supplies a complementary description. It was approximately 1.213 in 2019 and 2.516 in 2025: the latter means that Brazil reported about $2.52 of Portugal-bound exports for each dollar of Portugal-origin imports within this basket. The ratio was not steadily increasing. It rose to about 1.683 in 2020, fell to 1.414 in 2021, rose to 1.806 in 2022, fell to 1.621 in 2023 and rose to 2.237 in 2024.

The balance and ratio express different aspects of asymmetry. A dollar balance measures the absolute difference, whereas a ratio scales exports by imports. Neither measures profitability, welfare or the fairness of exchange. Both are sensitive to the selected product boundary, and both would change if additional merchandise headings or publishing services were added. Keeping the boundary fixed permits comparison without suggesting that this balance captures the broader cultural relationship.

Why the two national shares do not measure reciprocity

Portugal’s export share is consistently higher than its import-origin share in these Brazilian records. That observation does not mean that Portugal is more dependent on Brazilian books than Brazil is on Portuguese books. To calculate Portugal’s dependence, one would need Portugal’s own corresponding total imports and exports under compatible definitions. This study deliberately does not introduce partner-reported mirror statistics, which would require a separate reconciliation of reporting, origin, destination and valuation conventions.

The distinction is visible in the size of the Brazilian denominators. In 2025, the all-origin import total of $121,227,811 was much larger than the all-destination export total of $44,584,821. A given bilateral amount would therefore occupy a different percentage of each flow even if the bilateral amounts happened to be equal. A percentage-point gap between the two shares is not a bilateral balance, and multiplying either share by the wrong total would produce a meaningless value.

For the same reason, no average of export and import shares is used as a headline index of connection. Such an average would mix denominators without adding explanatory content. The analysis keeps the four underlying amounts visible: bilateral exports, bilateral imports, national exports and national imports. Readers can then distinguish changes in Portugal’s recorded values from changes in the size and composition of Brazil’s wider trade basket.

Concentration beyond the bilateral relationship

Portugal is only one part of a wider partner distribution. To describe that distribution, the analysis calculates the combined share of the five largest partner categories and a Herfindahl–Hirschman index for each direction and year. The index sums squared partner shares and multiplies the result by 10,000. Larger values indicate a more concentrated distribution of recorded FOB value; they are not measures of publishing-market competition among firms.

In 2019, the five largest export destinations accounted for 48.155% of the basket, compared with 75.016% for the five largest import origins. In 2025 the corresponding shares were 52.070% and 70.991%. The import distribution remained more concentrated than the export distribution by this measure in every observed year. The HHI also placed imports above exports throughout the period, rising between the endpoints from approximately 1,362 to 1,815 for imports and from 699 to 870 for exports.

The two concentration measures need not move together. The import top-five share fell between the endpoints while its HHI rose, because HHI is sensitive to how value is distributed within the leading group and across the remaining partners. A top-five share alone cannot reveal that redistribution. These results concern the complete national basket; they do not show that Portugal caused concentration to rise or that any individual supplier gained market power.

FOB dollars are not revenue earned by publishers

Both trade directions are measured using the same FOB value field. This is important because the Comex Stat API also provides import freight, insurance and CIF measures. Comparing export FOB with import CIF would introduce an additional valuation difference. Using FOB on both sides keeps the recorded merchandise values on the selected common basis, excluding freight and insurance from the bilateral comparison.

The dollars are nominal amounts reported in the underlying trade records. They are not converted to constant purchasing power or adjusted to a reference year’s price level. A change in declared value can reflect changes in quantities, product composition, declared prices and other features of transactions. The aggregate series does not identify the contribution of each component. It therefore cannot establish that real consumption or the economic value of literature changed by the same percentage.

Nor is customs value the same thing as a publisher’s net revenue or profit. The selected data do not identify publishers, production costs, retail margins, authors' payments, royalties or the contractual terms of rights sales. The study makes no claim about which firms or creators received the recorded amounts. Its use of the word surplus refers only to the arithmetic difference between these two merchandise flows, not a surplus accruing to a named industry participant.

Net kilograms cannot be read as book copies

The extraction also retains net mass so that the monetary aggregations can be checked against a second source field. Brazil’s Portugal-bound exports weighed 134,434 net kilograms in 2019 and 140,974 in 2025. Portugal-origin imports weighed 46,975 kilograms and 151,207 kilograms in those years. These are measurements of mass, not the number of books, editions, titles or readers represented by the transactions.

The distinction matters because physical products within the heading can differ substantially in format and weight. A shipment of dictionaries, a shipment of thin booklets and a shipment of folded sheets cannot be converted into comparable copy counts using their combined kilograms. Dividing value by mass would produce a unit value for a changing product mixture, not a retail book price. No such price or copy-count estimate is used in the headline findings.

The official NCM table labels the statistical unit for all three codes as net kilograms, but the API’s statistical-quantity field is not numerically identical to its net-mass field in every observed aggregate. The analysis preserves that distinction rather than silently substituting one for the other. Reported mass uses metricKG only. The presence of another quantity field does not provide a basis for describing quantities as individual books.

Complete years and explicit validation

The observation window comprises seven complete calendar years, from January 2019 through December 2025. The official availability response at extraction indicated data through September 2026, updated on October 6, 2026. Those partial-year 2026 observations are excluded. Mixing nine months of one year with twelve months of others would introduce a coverage difference unrelated to the bilateral question. Retrieval on October 30 is a provenance date, not an observation date.

The primary responses contain 3,450 export country-month records and 4,031 import country-month records. Every year contains all twelve months nationally, and no duplicate year-month-partner keys were found. For each direction, annual country-month sums match separate annual-total queries and separate NCM-year queries exactly for both FOB value and net mass. The three expected NCM codes are present in every annual product decomposition. These checks test consistency within the official source, not independence from that source’s underlying administrative records.

Explicit zero values are retained as reported. Missing partner-month rows are not converted into invented reported observations, and failed requests are never treated as zero trade. Annual partner totals sum the records returned for that year. Full national month coverage establishes calendar completeness of the extraction, but it does not promise that every customs declaration was error-free or that the official statistics can never be revised.

Continuity, provenance and the limits of the evidence

The historical classification check combines the Brazilian NCM2017 table reproduced in a ministry-hosted TIPI2018 document, the official MDIC 2017-to-2022 correlation table, current official definitions and the observed code membership in both flows. The three codes agree across those checks. The correlation table records changes to heading 4905 rather than 4901. Absence from that change table is not used alone as proof of continuity; the dated and current definitions provide the necessary accompanying evidence.

There are narrower provenance limits. The historical Brazilian PDF was readable through indexed official-source text, but a direct download returned a not-found response, so an original-byte fingerprint is not claimed for it. Successful API response bodies were retained, but the initial retrieval wrapper did not retain their successful HTTP headers. These gaps should remain visible instead of being replaced by invented archival precision. Original numerical responses and the selected reproducible inputs nevertheless support the calculations reported here.

The statistical reuse basis is also described precisely. MDIC’s own open-data plan identifies the Brazilian trade database, including export/import records and code tables, as an open-data product. The package uses attributed published aggregates and original calculations, without reproducing private records or adapting an entire official publication. A website content footer is not presented as a separately verified dataset licence, and no claim of official endorsement accompanies Alisson Bento’s analysis.

What this comparison establishes

Taken together, the series establishes a specific pattern: Portugal occupied a larger share of Brazil’s exports than of its imports within HS4901, Brazil recorded a positive bilateral FOB balance throughout the period, and the bilateral export/import ratio was higher in 2025 than in 2019. At the same time, Portugal’s share of each national flow was lower at the later endpoint. Keeping both facts avoids mistaking bilateral asymmetry for increasing national concentration on Portugal.

The surrounding partner distributions add context rather than a causal explanation. Brazil’s import origins were more concentrated than its export destinations under both reported measures, while Portugal remained outside the five largest partners in either direction. That combination makes it inappropriate to describe this bilateral relationship as dominating the national basket. It is a measurable part of a much broader geography of merchandise exchange.

The strongest conclusion is therefore deliberately bounded. These public customs aggregates describe the direction, relative weight and composition of a broad printed-matter trade relationship. They do not identify what was read, why it was purchased, the language of its text or the nationality of its creative ownership. A study of literary circulation could use these findings as context, but would need additional, differently defined evidence before making those further claims.

Portugal’s shares of Brazil’s HS4901 trade

Separate all-partner denominators; printed books, brochures and related matter; nominal USD FOB.

Full chart and source
Export shares 2019–2025: 5.406, 3.213, 4.612, 2.674, 3.682, 3.523 and 4.146 percent. Import-origin shares: 1.051, 0.535, 1.027, 0.701, 0.774, 0.665 and 0.606 percent. Both use their own Brazilian all-partner FOB totals.

Scroll horizontally to inspect the chart. ↔

Chart and values

Scroll horizontally to see all columns.

Portugal shares and bilateral FOB asymmetry
YearExports to Portugal (USD)Imports originating in Portugal (USD)Export share (%)Import share (%)Balance (USD)X/M
20191,195,595985,5545.4061.051210,0411.213
2020653,043388,1083.2130.535264,9351.683
2021986,406697,5354.6121.027288,8711.414
20221,075,142595,4352.6740.701479,7071.806
20231,310,876808,4413.6820.774502,4351.621
20241,441,919644,6773.5230.665797,2422.237
20251,848,406734,6824.1460.6061,113,7242.516

Nominal USD FOB. Export share of Brazilian all-destination total; import share of Brazilian all-origin total. Ratio = X/M. All seven denominators positive; flags preserved in CSV.

Exports use destination; imports use country of origin. Complete 2019–2025, no 2026 partial data. No copy counts, literary-language inference or causal interpretation. MDIC/SECEX Comex Stat; original analysis Alisson Bento.

Download chart (SVG) ↓ · Download chart (PNG) ↓

Data & downloads

Scroll horizontally to see all columns.

Portugal shares and bilateral FOB asymmetry
YearExports to Portugal (USD)Imports originating in Portugal (USD)Export share (%)Import share (%)Balance (USD)X/M
20191,195,595985,5545.4061.051210,0411.213
2020653,043388,1083.2130.535264,9351.683
2021986,406697,5354.6121.027288,8711.414
20221,075,142595,4352.6740.701479,7071.806
20231,310,876808,4413.6820.774502,4351.621
20241,441,919644,6773.5230.665797,2422.237
20251,848,406734,6824.1460.6061,113,7242.516

Nominal USD FOB. Export share of Brazilian all-destination total; import share of Brazilian all-origin total. Ratio = X/M. All seven denominators positive; flags preserved in CSV.

Scroll horizontally to see all columns.

All-partner denominators and concentration
YearWorld exports (USD)World imports (USD)Export top-five (%)Import top-five (%)Export HHIImport HHI
201922,116,04093,759,54148.15575.016699.01,361.6
202020,324,92572,573,38456.69272.171814.51,441.7
202121,388,45467,901,20857.77274.070854.61,685.6
202240,213,25284,916,42757.62668.886844.11,365.1
202335,597,697104,487,16046.57473.557671.91,715.6
202440,927,60996,910,40654.71070.397837.01,752.4
202544,584,821121,227,81152.07070.991869.91,814.6

Same HS4901 scope. Top-five shares and HHI calculated separately by direction, retaining all partner categories. HHI scale 0–10,000; not firm-market concentration.

Evidence & method

How we calculated it

  1. Descriptive secondary analysis by Alisson Bento of MDIC/SECEX Comex Stat public aggregates, HS4901, 2019–2025; no mirror statistics.
  2. Portugal code 607 verified by official lookup. Exports by destination, imports by origin, nominal USD FOB in both directions; metricKG for net mass only.
  3. All country-month sums checked against separate annual-total and NCM-year queries for both FOB and kg. All 84 months and three NCM codes present per flow; unique keys and nonnegative integer cells.
  4. Shares use separate all-partner denominators; rank = 1 + number strictly greater; top-five share and HHI calculated across all partner categories; bilateral ratio undefined if M=0.
  5. Offline reproduction uses compact selected source JSON and Python 3 standard library. Every year and numerical flag retained; no 2026 partial observations.

What the data cannot tell us

  • HS4901 includes printed books, brochures and similar matter, excludes 4903 picture/drawing/colouring books and other headings; no pure-book/novel, language, reader, royalty or firm-nationality estimate.
  • Origin and destination differ conceptually; no Portuguese dependence or partner-mirror reconciliation. FOB nominal values are not real sales, profits or welfare; kg are not copies.
  • Administrative reporting, product composition and revisions may affect values; no causal pandemic, exchange-rate or policy attribution and no sampling confidence intervals.
  • Historical Brazilian NCM2017 PDF available as indexed official text but direct retrieval returned 404; no original-byte hash claimed. Successful API HTTP headers not retained; numerical bodies frozen.
  • Reuse basis is documented institutional open data and public statistical facts, not an invented dataset-specific CC licence; no official endorsement.

Original sources

  1. MDIC open data

    Source scope, file schema and latest period.

    Accessed October 30, 2026

  2. Comex Stat API documentation

    Documented read-only POST and site licence footer.

    Accessed October 30, 2026

  3. API specification

    Public query parameters and metric definitions; original bytes privately frozen.

    Accessed October 30, 2026

  4. MDIC FAQ

    FOB nominal valuation; country semantics; quantities; revisions.

    Accessed October 30, 2026

  5. Country codes

    Portugal 607 verified, not guessed.

    Accessed October 30, 2026

  6. NCM classification

    Search is substring: selected only headingCode 4901, excluding unrelated matches. Three NCM codes, net-kg statistical units.

    Accessed October 30, 2026

  7. Federal open data guidance

    General reuse guidance, not claimed dataset-specific licence.

    Accessed October 30, 2026

  8. NCM2017 official Mercosur member table

    Historical classification corroboration, not bilateral statistical input.

    Accessed October 30, 2026

  9. MDIC 2017–2018 Open Data Plan

    p. 8 explicitly lists raw Brazilian trade export/import, SH4 and code tables for opening; glossary specifies open reuse.

    Accessed October 30, 2026

  10. Brazilian NCM2017/TIPI2018

    p. 222 three 4901 codes and chapter exclusions; indexed official PDF text read via web, direct download returned 404; original bytes not frozen.

    Accessed October 30, 2026

  11. MDIC NCM2017–2022 correlation

    Frozen official change table: chapter 49 changes 4905, no 4901 entry; used alongside dated/current definitions, not alone.

    Accessed October 30, 2026

  12. Trade statistics decree

    Articles 9–10 revisions and free access; article 13 aggregate/privacy restrictions; not a named dataset licence.

    Accessed October 30, 2026

Cite & reuse this study

Alisson Bento. Portugal in Brazil’s printed-matter trade: two shares, one widening bilateral gap. Analysis of MDIC/SECEX Comex Stat, HS4901, 2019–2025. Sources accessed October 30, 2026. https://alissonbento.com/research/brazil-portugal-printed-matter-trade-2019-2025/

Statistical source: MDIC/SECEX Comex Stat. Original selection, calculations and explanation: Alisson Bento. No official endorsement. The MDIC 2017–2018 Open Data Plan explicitly identifies the Brazilian trade database, raw export/import, SH4 and code-correlation tables for opening (page 8) and describes open reuse in its glossary. The current data portal and API documentation link the same public statistics. Federal Governo Digital guidance describes the public-data free-use basis under Law 14.129/2021, subject to personal-data safeguards. This package contains only published product/country statistical aggregates, original calculations and code; no company-identifiable data or protected whole publications. No separately named current Comex Stat dataset-specific Creative Commons licence was located. The website content footer’s CC BY-ND 3.0 notice is not misrepresented as a distinct licence for these underlying statistical facts or for this article. Retain MDIC/SECEX attribution, source references and the distinction between official values and original analysis. This documented basis is not legal certification or permission to reuse logos, private records or third-party works.

Corrections or editorial questions

Pass it on.

Portugal in Brazil’s printed-matter trade: two shares, one widening bilateral gap | Alisson Bentoalissonbento.com

WhatsAppFacebookXTelegramRedditThreadsEmailSMS