Biopiracy

While the term biopiracy was only coined in 1993, the act is as old as civilisation. For example, in an early recorded case from 1482 BC, Ancient Egypt’s Queen Hatshepsut seized thirty-one frankincense-yielding Boswellia trees during an East African raid. The practice accelerated during the Age of Exploration and the subsequent colonial period from the 15th to the 18th century, as European powers extensively collected and geographically diffused a vast array of animals and plants, including spices and tea. These actions laid the groundwork for what we now recognise as biopiracy (Robinson, Reference Robinson2010).

Hevea brasiliensis

Over 200 plant species worldwide contain the hydrocarbons needed for rubber production. Starting in 1860, research into the merits of different latex-producing plants concluded that Hevea brasiliensis is the one whose latex leads to the best quality rubber (Royal Botanic Gardens, Kew, 1898; Santos, Reference Santos1980: 229), rendering it the only member of the genus Hevea of commercial importance (Musgrave and Musgrave, Reference Musgrave and Musgrave2012). Even with the development of synthetic rubber, natural rubber from Hevea brasiliensis still remains the preferred material for the manufacture of pneumatic tyres, which consume about 75% of global production of natural rubber (Schultes, Reference Schultes1993: 480).

Hevea’s natural habitat is along the Amazon River’s right bank in South America, within the countries of Bolivia, Brazil, Colombia, Ecuador, French Guiana, Guyana, Peru, Suriname and Venezuela. However, over three-fifths of the habitat is in Brazil (in the states of Amazonas and Pará), with the remaining fraction shared among the other eight countries. (See Figure A4.) The conditions for Hevea’s growth are constant heat, 1,800 millimetres of rain annually, and a maximum altitude of 800 m above sea level (Dean, Reference Dean and Brandão1989: 33).

The rubber industry before vulcanisation: Pre-1839

As documented by French travellers (Musgrave and Musgrave, Reference Musgrave and Musgrave2012), the use of rubber in the Amazon basin before the 19^th century was confined mainly to the indigenous production of waterproof bags, coats, and shoes, and toys, particularly toy balls (Figure A5). During this period, Hevea trees, like their surrounding rainforest, were fundamentally conceived of as common-pool resources (Melby, Reference Melby1942).

The low price of rubber in the 1830s facilitated industry entry. However, the instability of rubber products (e.g., raincoats, shoes) caused firms to exit quickly. The products in fact remained sensitive to temperature changes, which would alter their consistency: brittle in cold winter, sticky and molten in the heat of summer (Schultes, Reference Schultes1993: 480–481). For instance, in 1835, the Roxbury India Rubber Company collapsed within six months after customers returned US$20,000 worth of goods that melted in the summer heat (Guise-Richardson, Reference Guise-Richardson2010: 361).

Brazil’s rubber industry boom after vulcanisation: 1850–1910

Rubber production post-vulcanisation started picking up from around 1850, beginning in Brazil. The first production increase came from a steamboat charter firm granted monopoly rights to navigate the Amazon River and its tributaries: the Companhia de Navegação do Amazonas, also known as Companhia do Amazonas. This charter facilitated the transport of rubber to both Manaus and Belém – see Figure A4. A further step came in December 1866, when the Brazilian government issued a decree opening the river and certain tributaries to international merchant ships, once the rubber industry had begun to show potential (Melby, Reference Melby1942: 453, 456). While initially subsidised by the Brazilian government, the Companhia soon turned profitable.

The institutional structure of production in the wild

Although parts of the rainforest were genuinely owned by the barons, the Brazilian institutional structure of production did not, in general, benefit from standardised techniques, a capital-intensive apparatus, and well-defined rights over Amazonian natural resources. It simply entailed the harvesting of wild rubber, which depended on the geographically uneven and widely distributed Hevea trees in the rainforest as well as on rudimentary extraction methods that were both damaging to the trees and very labour-intensive. Basically, production and transaction cost minimisation did not properly enter the logic of its economic organisation (Barham and Coomes, Reference Barham and Coomes1994).

Barons ran the industry during this period of Brazilian dominance.Footnote ¹ As their individual businesses grew, so did their demand for labour. Initially, labour was mostly drawn from the Indigenous Peoples whose lands were seized and whose villages were destroyed for rubber extraction (C. Ross, Reference Ross2017: 102; Schultes, Reference Schultes1993: 482). However, this labour pool was insufficient. Attempts to import labour from Europe, Japan, and other parts of the Americas proved unsuccessful. The labour shortage was naturally offset in 1879 following the drought of 1877–1878 (Grande Seca). The drought resulted in tens of thousands of displaced individuals, mostly from the state of Ceará in northeast Brazil (Figure A4), that the barons quickly recruited for labour (Furtado, Reference Furtado, de Aguiar and Drysdale1963: 141–148; Melby, Reference Melby1942: 453).

Barons advanced expenses for tapping equipment and rainforest transport. Tappers (seringueiros) then autonomously sourced rubber in the rainforest to repay these debts. However, repayment was typically impossible because barons controlled the sale of both the rubber and essential tapper supplies (e.g., food, tobacco). Barons ensured that the arrangement was always in their favour by underpaying the tappers for rubber while simultaneously overcharging for the necessities. Tappers remained poor while barons grew wealthier (Bradford Burns, Reference Bradford Burns1965: 412; Schultes, Reference Schultes1993: 482).

Latex could only be tapped between August and January because the rainy season lasted for the other six months. Tappers lived in shacks by the river and would venture into the forest to collect the latex. After collection, tappers smoked the liquid latex by their shacks, transforming it into balls or sheets of rough rubber ready for shipping (Figure A1, left column). The balls and sheets were sent by boat to Manaus or to Belém, both now hubs, especially Manaus for its pivotal location, for the export of rubber to European and US markets (Frank and Musacchio, Reference Frank, Musacchio, Topik, Marichal and Frank2006: 277–278).

Given the production structure in the wild, the tappers had little time to hunt and fish and would often suffer from malnutrition (Schultes, Reference Schultes1993: 482). Moreover, deep in the rainforest, they were also vulnerable to diseases (yellow fever, malaria), and medical care was only located in distant urban areas (Frank and Musacchio, Reference Frank, Musacchio, Topik, Marichal and Frank2006: 280). Their mortality rate was very high because of these challenges and the harsh working conditions. At the same time, Ceará’s persisting drought assured a continual labour supply for the replacement of the ill and deceased (Melby, Reference Melby1942: 453).

Notwithstanding assured labour turnover, production in the wild could not guarantee a stable rubber supply. The latex flow would vary with rainfall level, trees’ forest distribution, and how much tapping each tree had already endured. Hevea is moreover vulnerable to endemic diseases, particularly the South American leaf blight, caused by the fungus Microcyclus ulei, which – even today – can severely damage or kill a tree or even an entire forest. Consequently, predicting rubber supply was impossible (Dean, Reference Dean and Brandão1989; Melby, Reference Melby1942: 467).

The blight’s endemic distribution in South America meant that plantation cultivation in the Amazon was biologically precluded. Henry Ford’s well-funded Fordlândia plantation in Pará during the 1920s–1930s later confirmed this: Fordlândia failed because of the blight despite significant capital investment and modern agronomic techniques (C. Ross, Reference Ross2017: 104; Tucker, Reference Tucker, McNeill and Pomeranz2015: 425). Plantation cultivation was thus not a viable Amazonian alternative to wild harvesting.

Consequences of the boom

Figure 1 shows Brazil’s share in world rubber production during 1850–1910. Brazil clearly dominated world rubber production (63%), and particularly superior quality rubber production (at least 80% from Hevea).

Figure 1.

Brazil’s dominance of the natural rubber market, 1850–1910. Source: Data from Jünger (Reference Jünger and Aguti1942: 273), authors’ calculations.

Figure 1. Long description

The bar graph compares the percentage of Hevea brasiliensis and all rubber types from 1850 to 1910. The x-axis represents the years 1850, 1880, 1895, and 1910. The y-axis represents the percentage, ranging from 0% to 100%. The graph features two data series: one in blue for Hevea brasiliensis and one in orange for all rubber types. In 1850, both data series reach 100%. By 1880, Hevea brasiliensis is at approximately 85%, while all rubber types are at around 75%. In 1895, Hevea brasiliensis is at about 80%, and all rubber types are at approximately 70%. By 1910, Hevea brasiliensis is at around 85%, and all rubber types drop to about 60%. All values are approximated.

One consequence of this rubber boom was that, between 1870 and 1910, Manaus emerged as the rubber capital of the world, where, in 1910, a Rubber Exchange was established at the headquarters of the Commercial Association of Amazonas (Associação Comercial do Amazonas, ACA). Significantly, the Exchange had a direct telegraph line with London (Bradford Burns, Reference Bradford Burns1965).

A second consequence of the boom is that the tappers continually increased their tapping area, slowly claiming more uninhabited territory for Brazil through territorial sovereignty from mere occupation of land (uti possidetis, ‘as you possess’). One notable occupation was in Acre, where the tapper settlement led to Brazil taking over an area of over 187,981 Km² previously regarded as part of Bolivia (Bradford Burns, Reference Bradford Burns1965: 412; Melby, Reference Melby1942: 461).

A third consequence was tree damage. The standard method of gouging trunks to collect latex in bowls damaged and sometimes killed trees, lowering yields (C. Ross, Reference Ross2017: 102). That the barons did not widely adopt less damaging techniques reflects the extractive logic of the institutional structure of production: with debt-peonage securing labour turnover and the Amazon’s vast tree population providing a buffer, there was no conservation incentive.

Finally, the rubber boom boosted Brazilian public revenues. The government collected an ad valorem export tax of 20% for each kilogram of rubber. For example, during 1909–1911, rubber provided, on average, a substantial 80% of all public revenue received in the state of Amazonas (Bradford Burns, Reference Bradford Burns1965: 416; Melby, Reference Melby1942: 462). Rather than strengthening formal institutional capacity, the rubber boom reinforced patterns of institutional fragility typical of resource windfalls (M. Ross, Reference Ross2001).

The advent of the automobile industry in the 1890s surged world rubber demand. Brazil’s rubber exports more than doubled from 1890 to 1910 (Figure A6). However, Brazil’s market dominance faced challenges as its institutional structure of production in the wild eventually proved inefficient (Dean, Reference Dean and Brandão1989).

Appropriation of the seeds

British naval officer Clement Markham was one of the first to recommend appropriating rubber seeds. During two years of travel in Peru, he made two key observations. First, he noticed native cinchona trees – yielding quinine – and resolved to transplant them to treat tropical diseases affecting British soldiers and their families. Second, he witnessed rubber trees being severely over-tapped to the point of dying. Years later, in about 1865, while transplanting the cinchona trees from Peru to India and Ceylon (Sri Lanka), he found that the rubber variety native to India, Ficus elastica, was also being tapped until the tree died. He then realised that wild rubber production would not be able to meet industrial demand. Markham then wrote to the India Office with his recommendation (Loadman, Reference Loadman2005: 82–83).

With growing global demand for rubber and constrained supply, the idea of domesticating rubber by establishing tree plantations – cultivating rubber commercially – began to take hold in other parts of the world. One of the first to propose rubber cultivation within Brazil was Brazilian João Martins da Silva Coutinho, who suggested it first in 1861, and again in 1867 during the Universal Exposition in Paris (Dean, Reference Dean and Brandão1989: 34; Loadman, Reference Loadman2005: 92).

Given the environmental conditions needed, the British looked to their tropical colonies like India, Ceylon, and British Malaya (Malaysia). In 1876, Joseph Hooker, the director of London’s Royal Botanic Gardens, Kew, advocated for the transplantation of Hevea brasiliensis to the Asian colonies (Dean, Reference Dean and Brandão1989: 36–65).

Cultivating rubber posed unique challenges, however. The first was obtaining Hevea seeds. Charles Farris, a Brazilian contact of James Collins, the curator of the museum of the British Pharmaceutical Society, managed to smuggle 2,000 seeds from Brazil to Britain in 1873. This brought on the second challenge: guaranteeing a successful transplant, as only 12 of the 2,000 seeds germinated. Half of these died, and only six seedlings were sent to the Calcutta Botanic Garden. But none survived once they were there. This first transplant attempt through biopiracy failed (Musgrave and Musgrave, Reference Musgrave and Musgrave2012).

Yet the British Empire was undeterred. Still judging there to be rent opportunities, its efforts to transplant Hevea in its Eastern colonies persisted. This is when Henry Alexander Wickham, an adventurous British subject whose most previous entrepreneurial ideas had failed, entered the scene.

Wickham, at the time, lived in Santarém, near the Tapajós River (Figure A4) in the state of Pará, and had significant knowledge of botany and of the local environment (Jackson, Reference Jackson2008). He received instructions and the promise of payment from Hooker to transport an unlimited quantity of Hevea seeds to Britain. In 1876, he successfully smuggled more than 70,000 seeds, with approximately 2,700 surviving the journey, for which he was paid £740 (Musgrave and Musgrave, Reference Musgrave and Musgrave2012). This event marked the initial stages of Hevea as a commercial crop beginning in Ceylon (with 1,919 seedlings) and Singapore (with 100 seedlings), before spreading to other British colonies in Asia (Dean, Reference Dean and Brandão1989: 51–65).

The exact coordinates of the appropriation remain unknown. However, sources agree that the appropriation occurred southwest of Santarém, near the confluence of the Tapajós and Amazon Rivers (Figure A4). Under prevailing legal interpretations of the period, this forest territory, as most others, was formally considered terra nullius (belonging to no one), albeit under Brazilian jurisdiction – the fact that the territory was the longtime home to Indigenous Peoples under their own informal ownership systems without written land titles was not factored in. Further, there were no physical restrictions to access the forest (a capacity challenge for any country over extremely large frontier territories), and no formal legal prohibitions and export controls on the seeds (and other natural resources). Crucially, at the time of Wickham’s appropriation in 1876, no Brazilian law prohibited the export of rubber seeds or seedlings. The appropriation therefore occurred with the seeds, neither protected property nor contraband under the legislation of the period. The forest’s ‘contents’ were in a situation of symmetry with its territory: by implication, they were considered res nullius – things belonging to no one.

Wickham’s export to Britain ultimately worked thanks to the support of the British Consul and the pretext that the seeds were ‘exceedingly delicate botanical specimens specially designated for delivery to Her Britannic Majesty’s own Royal Gardens of Kew’ (Wickham, Reference Wickham1908: 54). Nevertheless, the success of the entire appropriation remained speculative. For at the time, little was known about the possibility of cultivating rubber as a crop (Resor, Reference Resor1977: 348). As Wickham (Reference Wickham1908: 45) himself recalled: ‘[d]ead weight of inertia, not to say opposition, prevailed. The idea of cultivating a ‘jungle forest tree’ was looked upon as not less than visionary’.

The Brazilian response to Wickham’s actions was initially one of scorn. There was a popular belief that Hevea seeds would not grow in Asia. Then, when the trees in Ceylon thrived, the Brazilians claimed that they would not yield rubber; and when the trees started to yield, the Brazilians claimed that the yield would be inferior. The Brazilians only realised the serious consequences for their own industry once the Asian rubber proved to be of the best quality and much cheaper to produce than in the wild (Resor, Reference Resor1977: 344). Thereafter, Wickham was looked upon with contempt and considered a thief in Brazil while, in Britain, he earned the title of ‘Sir’ (Supplement to the London Gazette , 5 June, 1920: 6315).Footnote ²

After Wickham’s appropriation, it took two decades for commercial cultivation to materialise as initially the plants in Asia were interspersed among other crops to supplement income (C. Ross, Reference Ross2017: 106). Hevea plantations became more viable once the rise of the automobile industry sharply increased rubber demand for tyres, pushing prices significantly upward. The first commercial harvest from British Asia appeared in 1900. It constituted merely 3 out of a total of 53,933 tons or 0.006% of market share. Brazil instead produced 26,750 tons of wild Hevea, accounting for about 50% of market share (Weinstein, Reference Weinstein1983: 218).

By 1913, the tables turned: Britain had produced 47,618 tons of rubber, surpassing Brazil’s 39,560. As Figure 2 illustrates, Britain overtook Brazil’s market share in less than four years (1910–1914). By 1919, Britain’s dominance was unequivocal, claiming over 90% of the market with a production of 381,860 tons out of a total 423,495, while Brazil’s share shrunk to less than 9%, amounting to 34,285 tons (Weinstein, Reference Weinstein1983: 218).Footnote ³ The remarkable overtaking meant that as British controlled Asian plantations gained dominance, Brazil’s loss of market power approached its maximum, while contemporaneously Britain’s post-vulcanisation return increased.

Figure 2.

Brazil’s fall and Britain’s rise in rubber production %, 1900–1922. Sources: Data from Jünger (Reference Jünger and Aguti1942: 273) and Santos (Reference Santos1980: 236), authors’ calculations.

Figure 2. Long description

A line graph showing the percentage of rubber production by Brazil and Britain from 1900 to 1922. The x-axis represents the years from 1900 to 1922, and the y-axis represents the percentage of rubber production. The blue line represents Brazil, and the orange line represents Britain. Key points include 1900, where British Asia produces just 0.006 percent, and Brazil feels untouchable. In 1913, British production surpasses Brazil, driven by auto industry demand. By 1919, the British Empire controls over 90 percent of the market, and Brazil falls below 9 percent. The timeline at the bottom highlights significant events: seed appropriation in 1876, the state of Amazonas imposing export duties in 1884, the Federal Rubber Defence Law in 1912, and a total ban on seed exports in 1918. All values are approximated.

The British Empire ended soon after World War II. Indonesia and India became independent in the 1940s, while Malaysia followed in 1957. By this point, Hevea plantations were firmly established in these and other Asian countries. Figure A7 illustrates the major producers 1960-1980, highlighting Indonesia, Malaysia, and Thailand along with China, India, and Sri Lanka. Another notable observation is that no single country consistently contributes more than 45% to world rubber production, indicating greater competition and a secure supply of rubber.

Brazilian protectionism after the biopiracy

Brazil protected the rubber industry only years after Wickham’s appropriation. In 1884, the Amazonas state placed high export duties on seeds and seedlings. From May 1913, the Asian markets began to take over the rubber market, and the price of rubber declined, leading to economic stagnation in the Amazon. The barons accused the federal government of focusing on coffee production and neglecting the larger economic picture (Bradford Burns, Reference Bradford Burns1965: 419). In fact, even during the rubber boom of 1890–1920, coffee remained Brazil’s largest export crop by a significant share (de Abreu and Bevilaqua, Reference de Abreu, Bevilaqua, Cárdenas, Ocampo and Thorp2000: 35).

The barons lobbied the government to push for the valorisation of rubber. As an emergency measure, Amazonas’ state government passed a law for intervention in the rubber industry. Under further political pressure, the Brazilian federal government passed the Rubber Defence Law of 1912, which ‘encouraged the creation of plantations, the improvement of transportation, and the attraction of immigrants, and ordered a fifty per cent reduction of export taxes’ (Bradford Burns, Reference Bradford Burns1965: 420–421). In 1918, a complete export ban on seeds and seedlings was introduced (Loadman, Reference Loadman2005: 92). The Bank of Brazil bought rubber, hoping to restrict supply and thereby raise prices. However, the valorisation schemes were unsuccessful, the legislation was ineffective, and the Bank of Brazil neared insolvency (Bradford Burns, Reference Bradford Burns1965: 420–421).

Technological shock and institutional change

We can now trace the transition from a common-pool resource production structure to a commercial one. By grafting some elementary formalism onto the factual and historical narrative, we can see how the transition was moulded by both technological and institutional interactions. We use game theory to organise the changing incentives of Brazilian and British actors as a heuristic guide to the narrative.

Imagine a simple game that reflects the sequential strategic choices made by Brazil (whose actors are first movers) and Britain (whose actors are second movers) concerning access to Hevea seeds. The periods of interest are pre-and post-vulcanisation. For ease of exposition, let us split the game periods into two figures: the game starts during pre-vulcanisation with Figure 3a, when the industrial uses of rubber had yet to manifest, and continues with post-vulcanisation in Figure 3b, when the industrial uses of rubber are common knowledge.

Figure 3.

(a). Individual strategic choices and societal outcomes pre-vulcanisation. (b). Individual strategic choices and societal outcomes post-vulcanisation.

Figure 3. Long description

The flowchart depicts the strategic choices and societal outcomes related to rubber production before and after the technological innovation of vulcanization. Before vulcanization, rubber is a low-value common-pool resource. Brazil observes its low commercial value and decides to either block or allow access. If Brazil blocks access, Britain’s strategy involves either not appropriating or appropriating the resource. If Brazil allows access, Britain’s strategy also involves either not appropriating or appropriating the resource. Post-vulcanization, rubber gains high commercial value. Brazil updates its strategy, deciding to either block or allow access. Britain’s strategy again involves either not appropriating or appropriating the resource. The flowchart highlights the transition from common-pool extraction to organized commercial production, with Brazil introducing new formal rules.

Brazil’s strategies are either to deny access – Block access – or to permit access – Allow access – to the seeds. Britain’s strategies are either to engage in the appropriation of the seeds – Appropriate – or to refrain from it – Do not appropriate. Payoffs for both countries are in brackets, with Brazil’s payoff on the left and Britain’s payoff on the right.

Following events, we begin by considering why Brazil originally overlooked appropriation (Figure 3a). Initially, the repertoire of rubber applications was modest. We saw that the Indigenous Peoples in the Amazon, Brazil included, had mostly limited uses for rubber. The implication is that there was no significant demand or commercialisation of rubber during pre-vulcanisation. Brazil in practice tacitly managed the resource through informal rules in use rather than formal rules in form (well-defined property rights, legislation). Recall the common-pool nature of Hevea.

The payoffs reflect the limited uses of rubber before vulcanisation. Brazil’s α (α ≥ 0) reflects the value of market share when rubber had a relatively low value, little or no economic rents to seize, and Britain did not appropriate the seeds; but α − ϵ (ϵ ≥ 0) is the loss from British seed appropriation. To Block access incurs a cost, c (c ≥ 0), reflecting public administrative and logistical burdens of managing a frontier resource – already high in a vast region characterised by sparse state presence. The payoff for Brazil to Block access is thus α − c (when Britain does not appropriate) and α − c − ϵ (when Britain appropriates notwithstanding the restriction).

β is the payoff for Britain without appropriation, regardless of whether Brazil allows or blocks access. When Britain appropriates while Brazil allows access, its payoff is β − ω; ω ≥ 0 reflects the cost of appropriation – the seeds might not survive the journey or grow well in Asia. The cost to Britain of appropriating when Brazil does not allow access is given by p, which may be a threat of a penalty, such as jail time or a pecuniary fine, where p ≥ 0, resulting in a payoff of β − p − ω for Britain.

Brazil’s dominant strategy is to Allow access to avoid the costs incurred when blocking access. Britain’s dominant strategy is Do not appropriate, to avoid the costs involved in taking seeds whose commercial value is low. Thus, the Nash Equilibrium is [α, β], where Brazil will Allow access and Britain Does not appropriate. The interaction between the two players repeats under low incentives to deviate, represented by the loop (see the ‘*’ sign) at the ‘Allow access → Do not appropriate’ branch of the game tree in Figure 3a. This repeated-play equilibrium embodies a stable common-pool resource regime where an informal structure of production persists (i.e., there is no institutional change). This corresponds to rules in use without rules in form, reflecting a sort of residual incentive alignment between Brazil and Britain from the absence of rivalry over low-value common-pool resources.

This equilibrium dovetails with the known record preceding vulcanisation, in which Brazil made no attempt to prevent appropriation, and Britain no attempt to appropriate. However, the equilibrium changed dramatically after vulcanisation increased the commercial viability of rubber – the shock from vulcanisation increased the nominal value of rubber to both Brazil and Britain. From this point onward in the game, Brazil and Britain faced altered payoffs and new incentives, as shown by Figure 3b, which continues the game.

For Brazil, the payoff of the rubber industry increased to A > α, reflecting the greater value from promising new industrial trajectories. The cost to Brazil to Block access remains c. For Britain, there is now a large expected return to Appropriate the seeds, which exceeds the return of appropriating the seeds before vulcanisation, B ^e > β, even after accounting for the cost to appropriate the seeds when Brazil allows access.

Brazil’s strategy to Allow access and Britain’s strategy to Appropriate yields the payoff A − δ for Brazil, where δ ≥ 0 is the value of Brazil’s loss of market power to Britain (as plantation output grew, Brazil’s loss of market power approached near-total levels), and B ^e is the payoff for Britain. Brazil’s strategy to Allow access and Britain’s strategy of Do not appropriate yields a payoff of [A, b]. Here, Brazil reaps the payoff of the increased value of the rubber industry without any costs, A; Britain’s payoff is instead b, where B ^e > b > β, reflecting the seeds’ increased value, possible rents, and limits of their returns to Britain should they remain under Brazilian control.

Brazil’s strategy to Block access and Britain’s strategy to Appropriate yields a payoff for Brazil of A − c − δ, which is the payoff of the rubber industry for Brazil, A, less the cost of preventing access, c,Footnote ⁴ and less the loss of market power to Britain, δ. For Britain, B ^e − p is the payoff from gaining control of the seeds less the costs of appropriating the seeds, p, when Brazil blocks access. Brazil’s strategy to Block access and Britain’s strategy of Do not appropriate yields a payoff for Brazil of A − c. The latter means retaining the rubber industry, A, less the cost of preventing access, c. Britain’s payoff is b, which reflects the increased value of the rubber seeds with limits to their returns to Britain if the seeds remain under Brazil’s control.

Given the cost to Block access, the dominant strategy for Brazil is still to Allow access. Within the immediate logic of the formal game, this result is driven purely by Brazil’s high enforcement costs, c, and limited administrative capacity to monitor access across its vast Amazon region. However, the broader historical record explains why there was little political urgency to overcome these capacity limits. At the time, Brazil downplayed both the possibility that the seeds could grow outside the Amazon and the quality of any resulting yield. Both these beliefs reflect informational limits more than irrationality – a situation that, given previous repeated interactions with Britain, can be likened to a ‘self-sustaining system of shared beliefs’ (Aoki, Reference Aoki2001: 10). Because Brazil did not envision that the loss of market power to Britain, δ, would be very large, the institutional vacuum remained unchallenged.

Whether or not Britain appropriates, the payoffs for Brazil are lower when attempting to Block access. The rational behaviour of Brazil is to always Allow access. For Britain, provided that the expected value of appropriating the seeds less the cost of appropriating the seeds is greater than the value of the seeds under Brazil’s control (B ^e − p > b), the dominant strategy is to Appropriate. This result is historically accurate given our understanding of Wickham’s actions and the steep increase in the value of rubber from vulcanisation. Together, this means that the Nash Equilibrium is [A − δ, B ^e ], in which Brazil chooses, albeit passively, to Allow access and Britain chooses to Appropriate the seeds. This equilibrium also aligns with history.

Once the appropriation proved successful, Brazil’s institutional response, as we learned, was to introduce rules in form (such as tightening export controls) (Loadman, Reference Loadman2005: 92). Figure 3b represents these macro-level responses with the loop (the ‘**’ sign) at the ‘Appropriate’ outcome. Crucially, this loop does not imply a literal repetition of the initial seed-access game, as the successful appropriation was largely a one-off shock. Instead, it represents endogenous institutional feedback: the act of appropriation triggered the emergence of the ‘Commercial plantation structure’ (a shift in the institutional structure of production), prompting ‘Brazil introduces new formal rules’ (rules in form), which in turn redefined the broader strategic environment for subsequent resource governance.

This process exemplifies what North (Reference North1990) describes as endogenous institutional change, where shifting relative payoffs and beliefs lead actors to modify the rules that structure their interactions. Through repeated play, these adaptive revisions create a new equilibrium configuration of incentives and behaviours, consistent with the conception of institutions as equilibria sustained by shared expectations and feedback (Greif and Laitin, Reference Greif and Laitin2004). The resulting institutional change – emerging from a previously informal and weak governance setting – together with the diffusion and successful cultivation of Hevea abroad, produced a broader ‘Transition from common-pool extraction to organised commercial production’. Over time, these adjustments reconfigured incentives, monitoring, and enforcement, transforming rules in use into rules in form.

The equilibrium shift across the two figures, which, recall, together represent the same game, maps directly onto induced institutional innovation. The technological shock of vulcanisation altered relative factor valuations, raising the value placed on securing access to (the shadow price of) reliable rubber supply. In turn, this restructured the payoff environment, inducing Britain to establish plantations under new institutional structures of production and Brazil to introduce export controls, albeit belatedly. Critically, however, the inducement crossed a geopolitical boundary. Britain’s institutional capacity – coordinated scientific establishments, colonial administrative networks, and willingness to bear appropriation risk – allowed it to seize rents from a reorganisation of production that Brazil’s formal-rule vacuum had inadvertently created.

Unveiling positive societal consequences

The Brazilian experience, therefore, underscores how a technological shock can expose the limits of informal institutions and trigger formal institutional adaptation – albeit lagging – in the context of a common-pool. Let us now turn from individual payoffs, as depicted by the entries in the brackets of the game in Figure 3a and b, to societal payoffs, represented by summing the payoffs contained in each pair of brackets.

Before vulcanisation, the overall payoff for society was α + β. After the technological shock and the subsequent institutional shift to plantations, the realised payoff for society in [A−δ, B ^e ] is (A−δ) + B ^e . While comparing these two moments of play confirms a broad historical improvement to social welfare, isolating the specific effect of Britain’s appropriation requires a narrower baseline: the post-vulcanisation world without appropriation, where societal payoffs would be A + b. The shift to plantations was Kaldor-Hicks positive-sum because the gains from British appropriation exceeded both Britain’s foregone baseline and the loss in Brazil’s market power: B ^e − b > δ. The remainder of this subsection defends this claim empirically by documenting Brazil’s actual losses and comparing the observed trajectory to counterfactual ones.

Accounting for gross effects, the localised loss is equivalent to the collapse of the Amazonian rubber market from Asian competition, which led to a deep and lasting Brazilian decline (Weinstein, Reference Weinstein1983: 213–261). The gains to GDP per capita between 1840 and 1910 – from US$ 49 to US$ 329 – dissipated. Fiscal revenues, once buoyed by a 10–20% export tax on rubber, dried up (Coomes and Barham, Reference Coomes and Barham1994: 243). The decline coincided with growing destitution among rubber tappers (Weinstein, Reference Weinstein1983: 239–243), many of whom had endured debt peonage throughout the boom years (Coomes and Barham, Reference Coomes and Barham1994: 237).

To better focus the effect of the biopiracy on welfare, let us consider four counterfactuals that test Brazil’s dominance. The first counterfactual (CF1) posits that Hevea remains Amazon-bound indefinitely. The previously identified supply constraints (six-month tapping season, high labour mortality, dispersed tree distribution, and rudimentary extraction methods) suggest that Brazilian wild production could not have accommodated the demand surge driven by the automobile industry. We would have witnessed a sharp increase in prices, with detrimental effects on downstream industrial development (construction, medical, tyre, etc.), at least until synthetic rubber production matured in the 1930s–1940s. Brazil maintains its rents, but global welfare is lower than in the actual trajectory.

CF2 posits that Wickham’s mission fails, but a later transplantation succeeds (through, e.g., another British attempt, Dutch botanical networks in Indonesia or French Indochina or a private initiative induced by high rubber prices). CF2 is the most realistic alternative because multiple parties sought supply security, and botanical knowledge was diffusing globally. We thus have convergence to the actual trajectory, albeit with a lag: Brazil keeps its rents longer; the rest of the world endures higher prices longer.

CF3 posits that Brazil domesticates Hevea, as some had indeed proposed. However, as established earlier, Amazon plantation cultivation was biologically precluded by the endemic leaf blight. Brazil could not have domestically replicated the successful outcome of the British commercial plantations in the Southeast Asian colonies.

CF4 is more speculative: it posits Coasean bargaining. Brazil, aware that domestic cultivation was biologically precluded (CF3), establishes and enforces formal property rights over the seeds soon after vulcanisation. It then bargains for transferring access rights to the resource until mutually beneficial terms are reached. Rather than suffering uncompensated appropriation, Brazil formally licenses the non-rivalrous genetic resource for cultivation in blight-free Asia in exchange for royalties. Global output still expands. However, Brazil captures a permanent share of the plantation rents, demonstrating that its localised loss was a formal institutional failure rather than an inevitable biological destiny.

These counterfactuals lead to a single conclusion: Brazil’s losses represent the end of a dominant market position that could not be permanently sustained. CF1 shows that wild extraction was too constrained to meet global demand. CF2 shows competitive diffusion would have ended Brazil’s dominance regardless. CF3 confirms Brazil could not pivot to domestic plantations given leaf blight. Finally, CF4 shows that even if Brazil had filled the institutional vacuum, the physical rubber industry would have still moved to Asia; yet Brazil could have captured a share of the plantation rents rather than losing them entirely. Evaluated jointly, these scenarios demonstrate that Brazil’s lost market power (δ) was an inherently temporary or purely institutional premium. Because this localised loss was finite, the vast expansion of global output and lower prices secured by the plantation shift indicates that the net welfare condition for a Kaldor-Hicks rather than Pareto improvement (B ^e − b > δ) is satisfied.

Comparing Brazilian and British institutional structures of production

The welfare verdict above in part rests on the differing productive capacities of the two institutional structures of production. Between 1870 and 1910, the elasticity of rubber supply from Brazil is estimated to be very inelastic: 0.22 (Fernandes, Reference Fernandes2009: 186). There were several reasons for this. Firstly, supply from the Amazon region was unpredictable because of the uneven spread of Hevea trees in the Amazon rainforest and inconsistent yields given Brazilian tapping methods. Secondly, the remoteness of rubber-producing regions in the rainforest created logistical difficulties for both transportation and communication. Thirdly, high mortality rates and a constant need to replace labour meant that the labour force was unable to consistently sustain rubber tapping. Fourthly, Hevea in the Amazon was susceptible to leaf blight, which could spread among the trees and further reduce yields (Dean, Reference Dean and Brandão1989; Frank and Musacchio, Reference Frank, Musacchio, Topik, Marichal and Frank2006). As we saw, the biological constraint of leaf blight made plantation cultivation in the Amazon unsustainable, compounding the institutional constraints of wild harvesting.

Growing plantations in Asia managed to increase the elasticity of supply and to address the challenges posed by Brazilian wild harvesting. First, concentrated planting and improved tapping methods increased both supply and its predictability. In 1888, Henry Nicholas Ridley, Director of the Singapore Botanic Gardens, developed a safer method of tapping the trees that would maintain a high yield for a longer time. Additionally, the tapping season in Asian plantations was much longer than the Amazonian one, which was limited by a six-month rainy season. Second, the logistical challenges were addressed: instead of finding and tapping wild trees spread out within the rainforest, the trees were cultivated densely on plantations. This reduced the cost of rubber production considerably. In 1910, the investment per worker for the British was £210, significantly less than Brazil’s £337 (a saving of about 38%). Third, concentrating production on plantations meant safer working conditions and access to medical care, which the rubber companies provided – attracting novel labour from India and China. Fourth, leaf blight is endemic to South America, but not Asia. The Asian plantations flourished, free of disease (Dean, Reference Dean and Brandão1989; Frank and Musacchio, Reference Frank, Musacchio, Topik, Marichal and Frank2006: 275–280).

The successful propagation of rubber trees outside the Amazonian native habitat demonstrates the potential for cultivation in various regions with suitable climates and soil conditions (Dean, Reference Dean and Brandão1989; Santos, Reference Santos1980). Notably, all ‘high-yielding cultivars of Hevea brasiliensis in the world originated from breeding programmes initially developed in Southeast Asia’ (Le Guen et al., Reference Le Guen, Doaré, Weber and Seguin2009: 673).

By establishing rubber plantations outside Brazil, Britain diversified the sources of, and globally decentralised, rubber production, reducing dependence on a single region. Following the fall of the British Empire and the loss of control over its Asian territories, competition in the rubber industry increased (Figure A7). The increased competition, as the Ostroms’ insights on polycentric governance suggest (Aligica, Reference Aligica2013), created a more resilient and adaptable institutional structure of production. The multiple centres of decision-making allowed for innovation and improved management, enhancing global resilience in the rubber supply chain. This diversification benefited all countries reliant on rubber, contributing to the stability and accessibility of rubber as a fundamental global resource.

Another positive consequence of the appropriation is environmental. After the Brazilian rubber market collapsed, the whole rubber network within the Amazon ended. The Amazon rainforest remained largely intact, avoiding overexploitation, even though the region had become more accessible from boats navigating the Amazon River and its tributaries during the rubber boom (Tucker, Reference Tucker, McNeill and Pomeranz2015: 425).Footnote ⁵