#546 1985 · US Federal Communications Commission (Michael J. Marcus) · Telecommunications regulation
The FCC couldn't police interference in three junk radio bands, so it let anyone transmit there and promised nothing
the problem
New wireless devices needed spectrum but licensing every one would take years
background
Radio spectrum regulation since Marconi's era ran on one principle: interference is prevented by exclusivity. The FCC assigned each licensee its own frequency and enforced against anyone who transmitted on someone else's, and getting a new use authorized meant a rulemaking process that could take years per applicant. By the early 1980s a new generation of low-power devices — cordless phones, wireless data terminals, inventory scanners — had no realistic path to spectrum on that timeline, and the FCC itself lacked the staff to arbitrate a growing swarm of small, individually low-stakes requests one exclusive license at a time.
The bands the FCC could plausibly open quickly were the ones nobody wanted: the Industrial, Scientific and Medical (ISM) bands at 902-928, 2400-2483.5 and 5725-5850 MHz, reserved for microwave ovens and diathermy machines that radiate RF energy with no power limit at all — engineers who worked in the industry called them the "garbage bands." Granting exclusive licenses there was pointless, since no licensee could ever be protected from a microwave oven's own unlimited emissions; the FCC's own conventional tool, exclusivity, simply did not apply.
what everyone would do
The available fix, and the only one the FCC's own regulatory toolkit really offered, was more licensing: open a rulemaking, define a new class of exclusive-use license for the new low-power devices, and let each applicant fight through the years-long approval process for its own protected slice of spectrum. That path could not keep pace with a wave of new low-power devices arriving faster than the Commission could arbitrate exclusive rights one applicant at a time, and it offered nothing at all for the ISM bands, where no exclusivity was even physically enforceable against a microwave oven's unlimited emissions.
what they saw
Marcus saw that the FCC's job was not actually to prevent interference — it was to prevent interference from mattering, and those are different problems with different fixes. Spread-spectrum modulation, spreading a signal across many frequencies so a receiver can pick its own transmitter's pattern out of what looks like noise to everyone else, meant devices could survive simultaneous, uncoordinated use by design. If the technology itself absorbed the interference problem, the regulator didn't need to coordinate access at all — it only needed to open the door and let engineering do what licensing used to do.
the move
FCC engineer Michael J. Marcus proposed, and the Commission adopted on May 9, 1985 (Docket 81-413), a rule permitting anyone to operate spread-spectrum transmitters in the ISM bands without a license, provided they used frequency-hopping or direct-sequence modulation — a wartime signal-guidance technique whose expired patents made it freely available — and accepted, in the FCC's own words at a later industry seminar, that they were "operating on a sufferance basis": no interference protection, no guaranteed channel, no recourse if another device stepped on their signal.
why it works
Because entrants got no guarantee of a clear channel, only permission to try, the FCC never had to arbitrate whose use took priority — it pushed that entire cost onto the device makers, who had every commercial incentive to build modulation and error-correction robust enough to work anyway, since a product that couldn't survive the shared band simply wouldn't sell. That incentive, not central coordination, is what actually kept the bands usable as device density climbed through the late 1980s and 1990s: manufacturers who under-engineered lost customers to ones who didn't, and the survivors' technology got steadily better at coexisting, all without a single interference ruling from the Commission.
the payoff
Adoption was slow at first — the first equipment authorization under the new rule wasn't granted until 1988, four years later, to a Canadian firm, Telesystems SLW — but by the early 1990s, as the FCC widened the allowed bandwidth, a wireless LAN industry (Apple among the earliest entrants) had adapted its equipment to survive in what one 1994 industry paper called the bands' 'feral territory,' engineering around interference rather than waiting for a regulator to prevent it. The IEEE 802.11 committee, chartered that same year, built the Wi-Fi standard directly on top of the 2.4 GHz ISM allocation Marcus had opened nine years earlier.
where it breaks
This only works when the underlying technology can genuinely be engineered to tolerate uncoordinated, simultaneous use — a resource where sharing degrades gracefully rather than catastrophically. It also requires participants able to bear the engineering cost of building for a no-guarantees environment, which is why the earliest ISM-band devices were built by well-resourced firms like Apple and IBM rather than small independents, and it needs a low enough initial value in the resource that incumbents don't fight the giveaway to a standstill, as the telecommunications industry's 'overwhelming opposition' to the original 1985 rule shows was a real risk even in bands nobody else wanted.
what came after
Every Wi-Fi router, Bluetooth device, cordless phone and garage-door opener sold today still runs on the exact regulatory carve-out the FCC issued in 1985; a 2023 Indiana Law Journal analysis by Professor John Blevins argues the outcome was not, as often assumed, a happy accident but a deliberately sustained regulatory strategy the FCC kept nurturing through repeated rule revisions into the 2000s. The 'commons' approach — open entry, no guarantees, let engineering do the coordinating — was later extended to wider swaths of spectrum as its results in the original junk bands proved out.
references
- [1]Darwinism and the ISM Bands (IEEE P802.11-94/91)IEEE 802.11 Working Group document archive, 1994grouper.ieee.org
- [2]The Regulatory Triumph of Wi-FiThe Regulatory Review, University of Pennsylvania Carey Law School, 2023theregreview.org