The Brazilian market reserve (Cobra, Scopus, Itautec)

**The reasoning was explicit and it was not stupid.** Brazil had a balance-of-payments crisis and an import-substitution strategy. The argument was that protected for a decade, domestic firms would build genuine capability and could then compete openly - and Japan and the United States were both cited as precedents for protectionism that had worked. Academics, industry, the press, the government and the opposition all supported it.

**The case against it is substantial and should be stated first, because it is the one usually left out of nostalgic accounts.** Freed of foreign competition, several manufacturers did not pioneer; they cloned. Machines closely modelled on Apple, Tandy, Atari and Commodore products appeared under Brazilian names. Scopus copied Microsoft's MS-DOS and shipped it as SISNE, drawing a piracy threat from Microsoft. In 1985 the United States formally accused Brazil of unfair trade practices and applied retaliation. And between 1989 and 1992 every semiconductor plant in the country closed or left. Smuggling was widespread enough that a police operation seizing contraband machines produced public anger rather than approval.

**And the case for it also has numbers, which is why the argument has never been settled.** Between 1984 and 1987 the Brazilian microcomputer market grew at about 7.4 per cent a year, the highest rate in the capitalist world at the time. Domestic firms went from 23 per cent of the market in 1979 to 40 per cent by 1982. Installed computers went from 1,200 units in 1974 to 23,200 by 1982.

**The figure that matters most for this site is a different one. Employment in informatics and automation went from about 42,000 people in 1984 to about 74,000 by 1989.** Whatever those companies were building, they were hiring, training and pulling technology through the economy - and a generation of Brazilian engineers learned the trade inside them.

**That is the part that outlasted the policy.** The law expired in 1992 as scheduled and most of the manufacturers did not survive contact with open competition. But the people did, and they went into the distributors, integrators, carriers and banks that appear elsewhere on this timeline. **A protectionist industrial policy failed at building a computer industry and succeeded at building a workforce, which was not its stated aim.**

**Where the companies went is the second finding, and it is not a coincidence.** Cobra was acquired by Banco do Brasil. Scopus was bought by Bradesco, and later passed to IBM Brasil. Itautec became a division of Itaú. Microsiga renamed itself Totvs and went into enterprise software. Three of the largest surviving names were absorbed by banks and turned to banking automation - **which is a substantial part of why Brazilian banking technology became unusually advanced.** A protected computer industry, unable to compete on hardware, was reabsorbed by its largest customers and became very good at exactly one thing.

**Read against the rest of this timeline, the pattern is familiar and uncomfortable.** Every argument about domestic technology capability - who builds it, who is allowed to buy it, what happens to the people trained under it - is a re-run of this one, with different vocabulary. The honest summary is that the policy achieved something real and did not achieve the thing it was for, and that both halves are usually reported separately by people who prefer one of them.