It's claimed as one of the great victories for enlightened (sorry) regulation, the way that the EU and US have both banned the incandescent light bulb through bureaucratic action. The ban came about by raising the efficiency standards required: this meant that the traditional bulb could no longer be sold.

The argument in favour of doing things this way was, in public at least, that everyone's too stupid (or, in a more polite manner, subject to hyperbolic discounting) to realise that the new bulbs will actually save them money in the long term by consuming less electricity. There are also the more cynical in the industry who insist that it's actually a case of regulatory capture. The light bulb manufacturing companies managing to get us all away from using cheap as spit bulbs and onto something with a decent margin on it.

But there's an interesting new paper that puts that first explanation to rest:

It is often suggested that consumers are imperfectly informed about or inattentive to energy costs of durable goods such as cars, air conditioners, and lightbulbs. We study two randomized control experiments that provide information on energy costs and product lifetimes for energy efficient compact fluorescent lightbulbs (CFLs) vs. traditional incandescent bulbs. We then propose a general model of consumer bias in choices between energy-using durables, derive sufficient statistics for quantifying the welfare implications of such bias, and evaluate energy efficiency subsidies and standards as second best corrective policies if powerful information disclosure is infeasible. In the context of our theoretical model, the empirical results suggest that moderate CFL subsidies may be optimal, but imperfect information and inattention do not appear to justify a ban on traditional incandescent lightbulbs in the absence of other inefficiencies.

To translate that for you: people aren't as dumb as the bureaucrats think. We're all perfectly capable of working out whether the energy savings will make up for the higher initial cost.

This has a number of implications in hte larger world as well: for example, it means that bureaucratic regulation on car mileages (like CAFE in hte US) is contra-indicated. A simple tax on petrol will drive up average mpg because we're not all as thick as bricks. Assuming that climate change really is a problem that must be dealt with then a carbon tax is going to do the job. For we're not all so dim that we cannot work out the utility of using fossil fuels or not given the change in prices.

That is, we don't need to be regulated into behaviour, we can be influenced into it through hte price system. Something that really shouldn't be all that much of a surprise to us market liberals: for we're the people who already insist that people do indeed respond to price incentives in markets.

Now all we've got to do is convince our rulers of the righteousness of this point: so they can stop writing those darn regulations.