Guatemala auctioned spectrum on demand and let buyers pick the use — prices halved

Almost every spectrum regulator allocates the same way: the state decides which service and which technology live in each band, and licensees cannot change either without approval. Economists since Coase had argued this is backwards — the allocator has no reliable information about what a band is worth and sits under constant political pressure — but in 1996 one small Central American country actually wrote Coase's critique into law. Guatemala's General Telecommunications Law made spectrum something users could call into existence rather than something the state doled out.

Under Article 54, any person or company could request an unallocated band, triggering a public tender for a Título de Usufructo de Frecuencias. The winner chose the service and technology; the title restricted only frequency range, power, geography and duration. Titles could be leased, sold or subdivided, so spectrum flowed to whoever valued it most without regulator sign-off. The Superintendencia de Telecomunicaciones adjudicated interference — nothing else. Before the reform, a state agency controlled frequencies below 800 MHz and a state company everything above.

The results showed up fast. Between 1997 and 2001 Guatemalan mobile penetration compounded at 71.7% a year against a Latin American average of 46.8%, and by 2002 average retail mobile prices were roughly half the regional average. A later econometric study across sixteen Latin American countries (2000–2004) attributed the price falls to spectrum abundance: more MHz in the market meant more competition and lower prices, and concluded that liberalization had achieved its policy goals in Guatemala — and in El Salvador, which copied the law.

The design's quiet radicalism is the direction of permission. Spectrum does not wait for a national plan; a user with a business plan can force the allocation question, and the auction answers it. The state keeps only the veto physics demands — power limits and interference protection — and gives up the industrial-policy lever of choosing technologies and champions.

Anyone may request a band; auction it; winner picks the use

Demand-driven allocation ends the regulator's information problem: the people who know a band's value are the ones asking for it, and the tender extracts that value as a price.

Removing service and technology mandates lets equipment and business models change without renegotiating a license — the constraint that made legacy licenses brittle.

Tradable titles create a working secondary market, so spectrum reallocates continuously instead of at relicensing intervals.

The regulator's narrowed role (interference referee, not planner) shrinks the surface for capture and political favoritism that Coase warned about.

Abundant, contestable spectrum let mobile operators compete on price; the measured outcome was penetration growth of 71.7% a year and prices half the regional norm.

The demand side knows a resource's value better than its allocator. Let users summon the auction, price it in a tender, and restrict only what physics requires — interference, not purpose.

A 2022 World Bank review still describes Guatemala's usufruct titles as unique in the region — freely tradable with only a notification, and grantable only on an interested party's request, so the regulator cannot launch allocations on its own. But the limits showed. The law never created a process to move bands from reserved 'public' categories into the marketable regime: an April 2015 resolution to open 1,670–1,859 and 2,110–2,290 MHz was suspended a month later under court provisions. Mexico, Honduras and the Dominican Republic have since allocated more mobile spectrum.