According to amendments to the Clean Air Act of 1970, new coal power plants have to install scrubbers to reduce the carbon and ash of their emissions. Environmental groups called this amendment a great victory for clean air. The scrubbers, which are about as large as the power plant itself, consume a great deal of power (10% the plant generates) and are very expensive to operate. There are two basic types of coal in the United States that could be mined for such power plants: “dirty” coal (which has a high carbon and ash content, mined in the east) and “clean” coal (which has a low carbon and ash content, mined in the west). The latter is slightly more expensive, but does not need to be scrubbed (and is in fact cleaner than scrubbed emissions from dirty coal).The punchline to all of this is that power plants buy dirty coal instead of clean coal since scrubbers have to be installed regardless. In the end, we get dirtier air (scrubbed dirty coal is dirtier than unscrubbed clean coal), the opposite of what the Clean Air Act was suppose to do. In a podcast about this topic, Bruce Yandle notes that environmentalists, scrubber makers, dirty coal miners, and railroad companies (who specialized in that kind of coal transport) celebrated at the regulation. Strange bedfellows, indeed.
A student notes that the rule should be that all power plants purchase clean coal. It's certainly a step in the right direction, but not likely to be a good, lasting solution. Whenever you discover a law encourages people to do X when Y is more efficient, the proper response is not to require people to do Y. Just because it's specific, doesn't mean it's going to be smart. Institutional and technological change might make X better later, or a third option, Z, better than Y. The goal is not to force people down a particular road but to encourage them to the road that's most efficient at any given time. In other words, taxing the emissions (with all that calculation problems that brings along) is a much smarter solution. It not only deters the essence of what we dislike, it encourages new ways to solve the problem. Striving for specificity, no matter how smart it might seem in the short run, is ultimately a recipe for centralization and encourages the delusion that "just the right static requirements" are better than the competing efforts of countless millions.