Professor Bainbridge has recently been musing on so-called "living wage theory" and its relationship to Wal-Mart, et al. It is an interesting question, with a number of interesting assumptions (such as that the total wage pool is zero-sum). It also sheds some light on a problem that authors have—one of even greater dimension than that for "standard economy" workers.
Although the data are extremely sparse upon the ground, the page, or wherever they may be found, the data that I do have indicate that income disparities are even higher among authors than they are between Wal-Mart workers and senior executives. On an hourly basis, the typical author earns less than does a Wal-Mart worker, and his/her income is quite irregular if a freelancer. On the other hand, we have the riches of Joanne Rowling, Stephen King, et al. (I would argue that Rowling, King, and other financially successful authors have almost always done far more to "deserve" high recompense than many, or even most, corporate executives—but that is a personal-values based argument that will not change any preconceived notions.) The key similarity is that the market inefficiencies of the markets themselves so vastly outweight any internal efficiencies achieved via conglomeration and commodification that asking about "living wages" may be entirely the wrong inquiry.
In abstract economic theory, wages are paid for exactly the value of the work performed. Of course, abstract economic theory always assumes the existence of a can opener to open the cans of worms it creates. The whole argument over "living wages" necessarily assumes that bargaining positions between workers and employers are—or at least should be—exactly equal, from both a preexisting power perspective and an access to relevant information perspective. Given that the last such bargaining session took place in northern Idaho in 1876, it seems to me that the "living wage" argument's conflation of the strictly economic value of the work performed and the moral value of the standard of living created by those various wages is precisely what a good scientist should avoid: the merger of independent variables into a single measured variable. This is particularly true when the measured variable is in effect a binary value (any y < x is an unacceptable result, whereas any y ≥ x is an acceptable result).
This argument becomes even more convoluted when one distinguishes between the purported necessities available at Wal-Mart (although I find it hard to believe that Spider-Man underwear is a "necessity") and the purported fripperies of the arts. I would argue that the arts are at least as important as underwear; after all, how much more charity expenditure goes to the arts than to clothing the homeless, even when there is no tax advantage? So, basically, my final position is that the whole "living wage" argument is asking the wrong question, or at least a premature one. The question that must be asked is "what is the minimally acceptable standard of living, and what are the acceptable mechanisms for providing that?" Only after one concludes that the only acceptable mechanism for providing a minimal standard of living is through wages for specific work does the "living wage" argument gain any relevance. And that is not a self-evident proposition; consider, for example, what it says about the proper way to handle systemic unemployment.
Eventually, I'll get to the point on how this relates to the arts. But only after I finish engaging in tryptophantasies over the next couple of days. I am not a connoisseur of wines like Professor Bainbridge; I am pretty damned good in the kitchen, though. (It's only the cleanup afterward that's discouraging.)