Domesticating the Machine

Every empire in this chronicle ran the same commercial architecture. The periphery supplies raw material; the core does the manufacturing; the periphery buys the finished goods back at the core's price. Rome ran it with grain and silver, the Atlantic system with sugar and bodies, the Congo Free State with rubber, and chapter 16 showed the postwar version running it with tariff schedules and structural adjustment instead of gunboats: cheaper, deniable, and effective. The one escape that has ever worked (the one every rich country used and most now discourage in others) is to domesticate the machine: to acquire, by whatever mix of protection, imitation, and patience it takes, the ability to make things yourself.¹

Under the rupture assumption of chapter 31, this stops being a development strategy and becomes a survival one. A world in crisis does not merely price the periphery out of finished goods; it stops shipping them altogether, as 2020 demonstrated in miniature when ninety-odd countries restricted exports of medical goods in a single year. The question this chapter answers is which manufacturing Rwanda must own outright by 2040, which it can share with the region, and what the record says about the resistance it will meet, because on that last point, Rwanda has already run the experiment.

The experiment already run: used clothes and AGOA

In March 2016 the East African Community's heads of state resolved to phase out imports of second-hand clothing and footwear by 2019: the cast-offs of the rich world, donated in Manchester or Houston, graded and baled by commercial exporters, and sold across the region by the kilogramme, a trade which had helped reduce East Africa's once-substantial textile industries to remnants. The logic was Hamilton's and List's, the same infant-industry logic every industrial country applied in its own ascent. Rwanda moved first and furthest, raising its duty on used garments from twenty cents to two and a half dollars a kilogramme, a rate designed not to tax the trade but to end it.

The response was immediate and instructive in its machinery. In March 2017 an American trade association, the Secondary Materials and Recycled Textiles Association, petitioned Washington on the ground that the phase-out threatened American jobs in the used-clothing export business. The Office of the United States Trade Representative opened an out-of-cycle review of the offending countries' eligibility under AGOA, the trade-preference programme advertised, since 2000, as helping Africa industrialise; and in March 2018 the White House gave sixty days' notice that the tariff must come down or the preferences would.²

Kenya, with actual apparel exports to America at stake (an industry worth some hundreds of millions of dollars a year under AGOA), folded. Tanzania and Uganda quietly slowed. Rwanda alone held the tariff, and in July 2018 the United States suspended Rwanda's AGOA benefits for apparel. The economic stakes were trivial: Rwandan apparel exports to America were then worth about a million and a half dollars a year, which is precisely why the episode is so legible. It was not about the money. It was a demonstration, by the core, that the periphery's role is not negotiable even at a scale too small to matter; and a demonstration, by Kigali, that it could absorb the penalty and proceed. The domestic garment sector has grown since. The chronicle files this beside chapter 11's Haiti: the price of refusing the assigned role, paid in full, and survivable, this time.

The episode also settles a definitional question this chapter needs settled. A preference that can be withdrawn the moment it is used for its advertised purpose is not a ladder; it is a leash, and the demonstration of 2018 is that the leash will be pulled over a rounding error. Any 2040 strategy built on the continuation of other people's goodwill has already failed the rupture test, which is why the rest of this chapter is built on capabilities instead.

The triage: survival manufacturing, not prestige manufacturing

Domestication under a deadline demands triage, and the test is the one chapter 33 used for the import ledger: what happens here if this stops arriving? Applying it sorts Rwandan industry into three tiers.

The record shows the first tier is not fantasy, because pieces of it exist, and the oldest proof is the dullest commodity in the table. Cement: CIMERWA's plant in Rusizi district, rebuilt in the mid-2010s to a capacity of roughly six hundred thousand tonnes a year and majority-owned since 2013 by South Africa's PPC, fires domestic limestone into the one industrial product Rwanda has never needed persuading to make.³ Cement is the classic first industry everywhere, and for a reason of physics rather than policy: it is heavy, cheap per tonne, and ruinous to haul far, so the weight-to-value arithmetic protects a domestic kiln more reliably than any tariff schedule could. Landlocked geography, a tax on almost everything else in this chronicle, is here a subsidy. The plant covers the bulk of national demand; construction booms still pull imports across the Ugandan and Tanzanian borders, which is what the regional tier is for.

Fertiliser is the half-built case: the OCP blending plant in Bugesera formulates for Rwandan soils, though (as chapter 33 noted) the feedstock dependency remains, and the ownership of that gap belongs to the energy programme, not the trade ministry.

Medicines are the sharpest case, because the continental dependency is close to absolute. Africa imports the large majority of the medicines it consumes and, before COVID, imported some ninety-nine per cent of the vaccines it administered; and the 2021 rupture demonstrated exactly what that means, when the core hoarded doses, outbid the poor on every order book, and blocked the TRIPS waiver at the WTO while the periphery waited.⁴ The BioNTech facility inaugurated in Kigali in December 2023, the first modular mRNA plant on the continent, having broken ground in mid-2022, is the emblematic answer, placed in Rwanda precisely because vaccine apartheid taught the same lesson this chronicle teaches.

The honest reading of that plant is a map of how long domestication actually takes. Shipping-container modules can express and formulate the vaccine, but a vaccine's value chain runs from plasmid DNA and lipids through fill-and-finish, cold chain, and, above all, regulatory approval, and much of that chain still lives elsewhere; a facility is geography, not yet sovereignty. Which is why the unglamorous half of the medicines file matters as much as the buildings: the Rwanda FDA's climb up the WHO's regulatory maturity ladder, the African Medicines Agency (treaty in force since 2021, with Rwanda selected to host it), and the African Union's target of sixty per cent of the vaccines used in Africa made in Africa by 2040. A medicine that cannot be licensed across borders cannot be made at continental scale, and the pharma park rising around the BioNTech site will stand or fall on paperwork as much as on chemistry.

Tools and repair complete the tier, and here the capacity already exists in embryo: the informal metalworking sector rebuilds engines and fabricates parts in every district town; the policy failure is that it does so uncredentialled and uncapitalised.

Assembling is not owning

Note what the triage leaves out of the first tier: the things development conferences applaud. A Mara phone was assembled in Kigali from October 2019 and marketed as the first smartphone made in Africa; Volkswagen has assembled cars in the city since 2018; Kigali hosts a growing fintech quarter. None of this is disparaged (exports earn the machines' price), but each case rewards the cold question: how much of the value, and how much of the knowledge, resides in the country once the ribbon is cut?

Run the question honestly. The Mara plant assembled imported boards, screens, and batteries around designs and chips owned elsewhere, and the venture has by most accounts struggled to sustain itself commercially since the launch coverage faded.⁵ Volkswagen's operation assembles semi-knocked-down kits at volumes counted in the low thousands, paired with a ride-hailing service; it earns its keep as a training floor, a standards regime, and a service network, which is to say as a toolkit entry point, provided nobody mistakes the entry point for the destination. The Kigali Special Economic Zone invites the same double reading: a real concentration of firms, jobs, and a rising share of manufactured output on one hand, and on the other the standing question every such zone must answer, which is whether knowledge leaks past the fence into the surrounding economy or stays gated inside it.

The distinction the chronicle wants on the record is between assembling and owning a supply chain. Assembly imports everything but the labour; ownership means the drawings, the jigs, the suppliers, and the failure analysis live in the country and survive the departure of any single foreign partner. Under the rupture test a phone assembly line ranks below a transformer workshop, and the chronicle's advice to planners is to fund accordingly and let prestige follow survival rather than precede it.

How domestication actually happens

The historical record on industrial catch-up is unusually clear, because every successful case used the same toolkit and every one is documented: protect the infant sector (as every now-rich country did, tariffs at levels that would today trigger arbitration); import the machines and the teachers, not the products; copy without apology until you can improve: the American nineteenth century was one long unlicensed reproduction of British technique; make the state the first customer; and export into a protected regional market before facing the world.⁶ Rwanda's existing kit maps onto this toolkit almost item for item: Made in Rwanda (the 2017 policy, with its procurement preferences, its import-substitution targets, and its branding arm) is the protection-and-first-customer instrument; the Kigali Special Economic Zone and its successors are the machine-importing arm; the polytechnic expansion, the vocational schools, and institutions like Carnegie Mellon Africa are the teacher-importing arm; and the AfCFTA (signed, fittingly, in Kigali in 2018) is the protected regional market, if it can be made real.

And because the first tier ends with the word repair, that word needs its concrete content, since it otherwise invites vagueness. It means a right-to-repair posture written into public procurement, so that no machine class enters the country without manuals, diagnostic access, training slots, and a committed spare-parts schedule as conditions of sale; standardisation of the national pump, mill, and motor fleet, so that one stocked bearing fits a thousand installations instead of thirty; stocking depth treated as a strategic quantity (parts measured in years, not in weeks of lead time); and the IPRCs and technical schools aligned to the actual machine census of the country rather than to curricula imported alongside the machines. The district metalworkers already do the work; the state's task is to credential them, capitalise them, and buy from them, because a country that can repair everything it operates has, quietly, acquired half the skills needed to make it.

Three honest amendments from the record. First, skills are the binding constraint, not capital: the factories that stall in East Africa stall for want of maintenance engineers and line supervisors, not machines, which is why the repair economy belongs in the first tier; it is the country's apprenticeship system wearing overalls. Second, energy prices discipline everything: no fabrication competes while power costs what it costs in Kigali, which welds this chapter to chapter 33's grid programme. Third, the regional tier is the fragile one: the EAC that must carry shared manufacturing is the same EAC whose members close borders at each ordinary quarrel. The Gatuna crossing with Uganda stayed shut from 2019 to 2022; Burundi's frontier has opened and closed with its politics; and the community admitted the DRC in 2022 even as relations between Kigali and Kinshasa sank to their worst in a generation. Every one of those facts is a line in the manufacturing plan, because it makes Rwanda's neighbourhood diplomacy a manufacturing input as surely as electricity. Domestication's hardest components are not technical.

The resistance to expect

Chapter 31 predicted that as stress rises, the core's instruments shed their politeness in order. Applied to manufacturing, the sequence is foreseeable because each step has a present-day precursor: first, trade-preference discipline of the AGOA type against protection; then intellectual-property enforcement against copying: the TRIPS regime exists to criminalise the exact toolkit above, and the 2021–22 fight over waiving it for vaccines showed the core will defend it even at the price of a pandemic prolonged; then finance conditionality against state procurement preference; and finally, for minerals and inputs, the direct methods of which the eastern Congo is the standing exhibit. A Rwanda that domesticates seriously should expect to be called protectionist by the tariff-built, a pirate by the patent-holding copiers, and a security risk by the militarised, and should read each accusation, as this chronicle has learned to, as a progress report.

The movement, closed

Four chapters, one argument. The system that this chronicle spent twenty-seven chapters dissecting will, on the stated assumption, break in the 2040s, politically before physically, and at the periphery before the core (chapter 31). The regional forecast is specific enough to engineer against (chapter 32). The survival consumables (calories, water, watts, medicine) can be domestically secured by a country that starts now (chapter 33). And the machines that make all of it can be owned, at survival grade, by 2040, provided the country accepts the penalties that the trading order imposes on those who leave their assigned role, penalties Rwanda has already sampled and already survived (this chapter).

The deeper continuity is the one the chronicle began with. Chapter 1 argued that the moral systems which survived the deep past were those fierce enough at the boundary and cooperative enough inside it. A country preparing for rupture is running that same selection problem at national scale: hard to starve, hard to bully, hard to recolonise, and worth living in, or the hardness has no point. The record assembled across thirty-one chapters says no one else will run it on Rwanda's behalf. It also says (and this is the note the chronicle chooses to end its forward movement on) that the country has done a harder thing than this before, from a worse starting point, inside living memory.

Footnotes

1. The canonical account of the toolkit and its later prohibition: Ha-Joon Chang, Kicking Away the Ladder: Development Strategy in Historical Perspective (London: Anthem, 2002); Alexander Hamilton, Report on the Subject of Manufactures (1791); Friedrich List, The National System of Political Economy (1841). ↩
2. EAC Heads of State communiqué, Arusha, March 2016 (phase-out by 2019); SMART petition to USTR, March 2017; USTR out-of-cycle review of Kenya, Tanzania, Uganda and Rwanda announced June 2017; presidential determination of March 2018; Office of the United States Trade Representative, "President Trump Suspends Duty-Free Apparel Benefits for Rwanda under AGOA" (30 July 2018). On the second-hand trade's effect on African textile sectors, Garth Frazer, "Used-Clothing Donations and Apparel Production in Africa," Economic Journal 118 (2008), pp. 1764–1784. ↩
3. CIMERWA PLC disclosures and PPC Ltd annual reports: PPC acquired a 51 per cent stake in 2013, and the new Rusizi plant, commissioned in the mid-2010s, carries a nameplate capacity of about 600,000 tonnes a year. ↩
4. On the import dependency: Africa CDC / African Union, Partnerships for African Vaccine Manufacturing framework (2021), which records that Africa consumed about a quarter of the world's vaccines while producing around one per cent of them, and sets the 60%-by-2040 target. BioNTech SE press materials and Government of Rwanda statements: ground-breaking June 2022, Kigali BioNTainer inauguration 18 December 2023. On the TRIPS waiver fight of 2020–22 and dose hoarding, WHO and WTO records and the extensive contemporary reporting. The African Medicines Agency treaty entered into force in November 2021; Rwanda was subsequently selected as host country. ↩
5. Volkswagen Group and Government of Rwanda announcements on the Kigali assembly operation (inaugurated June 2018, semi-knocked-down kits paired with a mobility-services venture); Mara Group launch materials, Kigali, October 2019, and subsequent business-press reporting on the venture's commercial difficulties. ↩
6. Beside Chang: Alice H. Amsden, Asia's Next Giant: South Korea and Late Industrialization (Oxford: OUP, 1989); Robert C. Allen, Global Economic History: A Very Short Introduction (Oxford: OUP, 2011), on the standard model of catch-up; Doron S. Ben-Atar, Trade Secrets: Intellectual Piracy and the Origins of American Industrial Power (New Haven: Yale, 2004). On the policy instrument: Republic of Rwanda, Ministry of Trade and Industry, Made in Rwanda Policy (2017). ↩