This does not mean, however, that it was somehow inefficient for
the Route 128 region to have organized around integrated systems firms, which possessed con- siderable advantage in the era of the minicomputer, the region’s most important high-technology product (Robertson 1995).
458. Klepper (2016).
459. In both cases, these transaction costs are often what I like to call dynamic transaction costs (Langlois 1992b).
460. Klepper (2016, pp. 112–28).
461. Lécuyer (2006, pp. 200–207).
462. Nicholas (2019, pp. 195–96).
463. Lécuyer (2006, pp. 259–60).
464. Sporck (2001, p. 139).
465. Lécuyer (2006, p. 257).
466. Klepper (2016, p. 123).
467. Noyce and Hoff (1981).
468. Hitzlik (1999, p. xxi).
469. Freiberger and Swaine (2000, pp. 36–53). This account of the early history of the mi-
crocomputer industry generally follows Langlois (1992a).
470. Indeed, the 100-wire bus of the Altair, the S-100 bus, would eventually be enshrined as the
IEEE 696 bus by the Institute of Electrical and Electronics Engineers (Noyce and Hoff 1981, p. 16). 471. Levering, Katz, and Moskowitz (1984, p. 351).
472. Langlois (1992a); Langlois and Robertson (1992).
473. Freiberger and Swaine (2000, pp. 111–24).
474. Isaacson (2014, p. 266).
Notes to Chapter 9 661
475. Moritz (1984, p. 191).
476. L. R. Shannon, “A Decade’s Progress,” New York Times, September 8, 1987, p. C7. The TRS-80 was built around the Zilog Z80, essentially a clone of the Intel 8080. Zilog was a spinoff from Intel, founded by Federico Fagin, a onetime Fairchild employee who had headed develop- ment of the 4004 and the 8080.
477. The machine was designed by former Motorola employee Chuck Peddle, using the MOS Technology 6502 microprocessor, a clone of a Motorola chip, which Peddle had also designed. Commodore had absorbed MOS Technology, Peddle’s spinoff from Motorola.
478. Isaacson (2011, p. 57).
479. Moritz (1984, p. 29).
480. Freiberger and Swaine (2000, pp. 261–70); Moritz (1984).
481. Moritz (1984, p. 126).
482. Data from Apple Computer, cited in “John Sculley at Apple Computer (B),” Harvard
Business School Case no. 9-486-002, May 1987, p. 26. Baldwin (2019, p. 19) cites IDC data that put Apple’s 1983 revenues at $1.1 billion.
483. Isaacson (2011, p. 71).
484. Moritz (1984, pp. 200–201).
485. Scott Mace, “Assembling Micros: They Will Sell No Apple before Its Time,” Infoworld,
March 8, 1982, p. 16.
486. Levering, Katz, and Moskowitz (1984, p. 132).
487. Bresnahan, Greenstein, and Henderson (2012, p. 218).
488. Chposky and Leonsis (1988, p. 9).
489. “How the PC Project Changed the Way IBM Thinks,” Business Week, October 3, 1983, p. 86. 490. Baldwin (2019, p. 2).
491. Chposky and Leonsis (1988, pp. 43–53); Freiberger and Swaine (2000, pp. 330–37). 492. In the end, three operating systems were available for the IBM PC, including a 16-bit
version of CP/M; but PC-DOS was the only operating system initially available, and IBM priced it at only $60, one-third the price of the cheapest alternative. Scott Mace, “IBM Releases CP/M- 86 for the Personal Computer after Delay,” InfoWorld, April 26, 1982, p. 8.
493. Gates (1995, p. 48).
494. Chposky and Leonsis (1988, pp. 88, 68).
495. “Personal Computers: And the Winner Is IBM,” Business Week, October 3, 1983, p. 78. 496. Chposky and Leonsis (1988, p. 24).
497. IDC, cited in Baldwin (2019, p. 19).
498. Langlois and Robertson (1992).
499. Baldwin (2018, p. 3).
500. Gerstner (2002, pp. 110–20).
501. Baldwin (2019, p. 20).
502. US Bureau of the Census, Current Population Survey, various years. Today far more
than 90 percent own a computer, although that term has been defined—as it should be—to include smartphones and tablets.
503. “Computer Use at Work in 2003,” Bureau of Labor Statistics, US Department of Labor, The Economics Daily, https://www.bls.gov/opub/ted/2005/aug/wk1/art03.htm (accessed June 8, 2021).
662 Notes to Chapter 9
504. Computer and Business Equipment Manufacturers Association, 1992 Information Tech- nology Industry Data Book, p. 94.
505. Matt Hamblen, “Update: PC Market Declines in 2001; Slow Turnaround Expected,” Computerworld, January 18, 2002.
506. Merchant semiconductor firms are those that sell into the market rather than producing for their own consumption. In fact, some 30 percent of American production in 1986 was captive, most of that internal production by IBM, by far the largest producer of semiconductors in the world at the time. If captive production is counted, US production was some 30 percent higher than Japanese production in 1986 (Langlois, Pugel, Haklisch, Nelson, and Egelhoff 1988, p. 27).
507. Borrus (1988); Ferguson (1988); Prestowitz (1988).
508. MIT Commission (1989, p. 20).
509. Semiconductor Industry Association 2021 Factbook, April 21, 2022, https://www
.semiconductors.org/resources/factbook/, p. 3.
510. Lécuyer (2006, p. 282).
511. Burgelman (1994, p. 33).
512. Irwin (1996, p. 28). Both the US and Japan had tariffs on imported semiconductors.
These were reduced in 1978 and ultimately eliminated in 1985. 513. Moore and Davis (2004, p. 20).
514. Burgelman (1994, pp. 32–34).
515. Pillai (2020).
516. Borrus (1988, p. 144).
517. Prestowitz (1988, pp. 146, 135–36).
518. Dataquest, cited in Methé (1991, p. 69).
519. Irwin (1996, p. 15); Irwin and Klenow (1994).
520. Irwin (1996, p. 42).
521. Japan would indeed be successful in developing a computer industry to rival IBM—just
in time for the “competitive crash” of the mainframe and minicomputer industries that, as we will soon see, would throw IBM itself into crisis.
522. Callon (1995, p. 37).
523. Fransman (1990, p. 80).
524. Sigurdson (1986, p. 53).
525. Callon (1995, p. 119); Sigurdson (1986, p. 83).
526. Prestowitz (1988, pp. 148–61).
527. Flamm (1996, pp. 159–226); Irwin (1996).
528. Flamm (1996, p. 240).
529. Irwin (1996, pp. 38–39). DOD was generally liberal on trade issues because its client
defense contractors tended to be export oriented.
530. Grindley, Mowery, and Silverman
