US environmental policy is largely based on regulatory standards.  At least on the surface, this approach appears to be consistent with the popular conception of environmental quality as a collective good to which all citizens are entitled.  Since polluters have no formal entitlement to pollute, the government can regulate their emissions as strictly as it likes.  Regulation is the obvious, direct approach.  If a firm is emitting too much, the government can tell it exactly what pollution abatement equipment to install (a technology standard), or specify a maximum amount it is allowed to emit (an emissions standard), or specify the maximum allowable concentration of pollutants in the environment at various receptor points (an ambient standard).  This overall approach is often referred to as "command and control."

In reality, the public's demand for environmental quality must be balanced against its need for jobs, incomes and consumer products, of which pollution is a byproduct.  Except in "Captain Planet" cartoons, it's never a black-and-white, good-guy/bad-guy issue.  The truth is that firms do have, and always have had, some presumptive rights to pollute.  The question is: how much?

Designing an efficient regulatory program is rarely a straightforward proposition.  The ultimate policy objective is typically some ambient level of environmental quality, which can be monitored at various receptor points.   Receptors may have differing degrees of vulnerability to pollution damage; for example, a toxic spill near an elementary school is taken more seriously than the same spill occurring in a barren, isolated area.  Consequently it may make sense to define differing ambient standards at different receptor points.  Since the environment tends to commingle pollutants from various sources in somewhat unpredictable ways, it may be difficult to quantify the contributions of various sources to pollutant concentrations at different receptor points.  Thus an ambient standard provides little guidance on how much each source should reduce its emissions so that ambient standards are met at all receptor points.

A second-best approach is to improve ambient quality by setting emissions standards at the various point sources.  The problem here is that heterogeneous sources may have widely varying costs of pollution abatement, so any uniform emissions standard--all firms reduce emissions to X tons each, or all firms reduce emissions by Y percent--is very unlikely to minimize aggregate abatement costs across all firms.  Emissions standards are one remove from ambient standards.  It is uncertain what ambient quality improvements are actually obtained from any set of emissions standards.

The relationship between firms and the regulatory agency is inherently confrontational, and firms have clear incentives to exaggerate their compliance burdens, exploiting an information asymmetry: the firms have far more information about their own production processes and pollution control costs than the regulatory agency.  Thus it is often difficult for the regulators to challenge firms' claims of unreasonable economic hardship caused by the emissions standards.

Emissions standards require reliable monitoring of firms' emissions.  If the regulatory agency lacks the budget for thorough monitoring, it may be forced to rely on firms' self-monitoring and self-reporting of emissions violations.  (The EPA's Toxic Release Inventory program is almost entirely based on firms' self-monitoring, with occasional agency audits of some firms to encourage accurate reporting.)  Obviously standards compliance is likely to be lower under self-monitoring than agency monitoring.

In reality, most firms have multiple emissions points, with an emissions standard for each point.  This brings up various practical complications.  If an expanding firm needs to add more emissions points, should the agency simply issue new permits for these points, using the same per-point standards?  If the expanding firm is to be held at or near its existing aggregate emissions standard, should it be allowed to shift emissions allowances between different points, or trade off one type of emission for another?  Negotiating such trade-offs can involve high transactions costs and bureaucratic red tape.

A third-best approach is for the regulatory agency to mandate specific pollution abatement technologies for each firm or emission point.  Since this ignores the natural information asymmetry mentioned above, it often results in inefficient "one size fits all" technologies being mandated for heterogeneous firms.  It is uncertain how much emissions reduction will actually result from implementation of a technology standard.  Thus technology standards are one remove from emissions standards, and two removes from ambient standards.  For many firms the mandated technology won't be the most cost-efficient means of pollution abatement.  And once a firm has reluctantly installed the standard technology, it won't necessarily maintain or operate it efficiently without additional prodding from the regulatory agency.

How much can the regulatory agency actually make firms pay for their pollution abatement?  Congress's legislative ambition can far surpass the practical reach of any regulatory program.  For example, the Clean Water Act (1972) envisioned a complete cessation of discharges into navigable waterways as "best practicable" (reasonable cost) technology standards would be supplanted within five years by "best available" technologies (regardless of cost).  The history of the Clean Water Act illustrates most of the practical regulatory pitfalls noted above.