A Stitch in Haste

It's tempting for capitalist-libertarians to cite the recent announcement by Marriott International that all its U.S. and Canadian hotels will go completely smoke-free as a demonstration of the viability of private markets to allocate, and re-allocate, resources based on consumer preferences as they evolve over time (i.e., that demand creates its own supply) and that government mandates in private markets, such as smoking bans, are unnecessary.

But I wonder if there isn't another factor at work here.

If one state (e.g., California) mandates higher emission standards for automobiles, then which is more likely: that auto makers will produce two different kinds of cars or that they will simply mass-produce the more compliant vehicle? The presence of economies of scale strongly encourages the former.

So too, perhaps, with Marriott:

Currently more than 90 percent of Marriott guest rooms are already non-smoking and smoking is prohibited in many public spaces due to local laws.

Maybe the company's management simply tired of worrying about complying with a mish-mosh of local laws, the violation of any one of which could lead to both fines and negative publicity (who wants an army of Eliot Spitzer acolytes pursuing them over smoking ban violations?). Perhaps it was simply more prudent to ban all smoking everywhere and be done with it.

The real test of the evolution of the markets for smoke-free environments will be, not hotels, but casinos -- which, incidentally, enjoy almost universal and almost absolute exemptions to smoking bans, thanks of course to the Politics of Pull.

So let's start a pool: How long until we see a major casino in a major market go completely smoke-free? I'm guessing it will be at least another ten years.

And let's not forget that none of this economic analysis changes the inarguable impropriety of smoking bans as property rights violations.

For Discussion: Marriott International does not actually own its branded hotels -- it is, for the most part, an operator and franchiser only. Does this change the analysis? If so, how?