Costs and Consequences: Despite Reforms, Telecom Regulations Are Still Costing Consumers Billions of Dollars
Ellig, Jerry, Regulation
THE FEDERAL COMMUNICATIONS COMMISSION spent approximately $1.2 billion on telecommunications and broadband regulation in fiscal 2004. Yet the actual cost is more than 60 times that amount.
In a Mercatus Center working paper released earlier this year, I assembled estimates from scholarly literature suggesting that federal telecommunications and broadband regulation generates $75 billion in wealth transfers annually, reduces consumer welfare by $25 billion annually, and reduces social welfare by $41 billion annually.
With the exception of monthly local phone service, most telecommunications services have relatively elastic demand. Because regulation acts as a tax on those services, it generates substantial reductions in consumption and production. For most regulations, the resulting "excess burden percentage"--the value of lost output divided by the amount of wealth transferred--usually exceeds the excess burden associated with general federal taxation. Economic research usually finds that the excess burden of general taxation is between 25 and 40 percent; that is, output falls by between 25 and 40 cents when taxes raise a dollar of revenue. The average excess burden of telecommunications regulation almost always exceeds 40 percent and sometimes exceeds 60 percent.
Regulatory outcomes have not been researched as extensively as costs. Existing research suggests that the most costly telecommunications regulations have negligible effects on the outcomes they are supposed to accomplish. For many other regulations, there is little or no peer-reviewed research or independent program evaluation assessing whether the regulations have accomplished their intended outcomes.
Tables 1 and 2 summarize the current state of research on the costs and consequences of federal telecommunications and broadband regulation. Although this lengthy menu of regulations may seem intimidating, the regulations can be divided into four categories based on the economic effects that create costs: entry barriers, taxes and subsidies, mandated services, and network sharing requirements.
Entry barriers limit competition, thereby raising prices and reducing the amount of service consumed. The two principal entry barriers are spectrum allocation and satellite regulation. The latter has been studied little; the former, much.
The FCC ultimately decides how much spectrum will be devoted to which private uses, and it then licenses equipment to use that spectrum. Broadcasters, for example, can only use their assigned spectrum for broadcasting. Spectrum allocation has gotten more flexible in recent years; wireless communications spectrum, for example, can be used for any wireless communications service. Nevertheless, it is likely that the government has under-allocated spectrum to wireless services.
COSTS A study by my George Mason University colleague Thomas Hazlett and coauthors estimates that the price of wireless service would fall by 50 percent if another 200 MHz of spectrum were devoted to wireless service. The price reduction would save consumers $54 billion annually and generate another $23.4 billion annually in consumer surplus from expanded use of wireless services.
Hazlett's numbers are based on international data from wireless phone service, but many observers believe the most valuable wireless services facilitated by an increase in wireless spectrum would be data services, not voice. In addition, his study assesses only the impact of allocating an additional 200 MHz, not the impact of moving to a more competitive market in spectrum across the board. Therefore, his figures should be interpreted as only a rough approximation of the true cost of spectrum allocation policy. They suggest that we pay a huge cost for having government allocate spectrum among various uses.
The costs of current spectrum allocation policy can be expected to fall sometime after 2006, if the FCC carries through on its plan to auction an additional 90 MHz of spectrum that year. …