But the labor market in which a typical Wal-Mart is competing ISN’T hundreds of millions if not billions of people; it consists entirely of the people within commuting radius of a given store (with “commuting radius” being anything from walking distance to the maximum distance you can drive on gas that costs less than you earn). Mr. Singh in Bangalore may be perfectly willing to take that job, but he’s not in competition for it.
Moreover, if you need a gallon of milk, or some batteries in the middle of the night, then Amazon isn’t an option; Wal-Mart’s competitors will be the retail stores within a reasonable radius. There are now locations in the Great Plains where if you need a gallon of milk, you might have only a couple of choices within a 20 or 30 or even 50 mile radius (which is of course one of the penalties of living in a small town on the plains). Even in not-so-small towns, though, the consolidation of retail and the demise of mom-and-pop stores means dwindling competition (Peapod and its competitors generally don’t deliver outside of the major urban markets).
One-quarter of American workers earn less than 150% of minimum wage ($10.55/hr–cite), and they’re not all students and part-timers. A lot of these are people trying to support themselves and sometimes even families. If Wal-Mart is the major employer in their area (and Wal-Mart is now the largest private employer in the U.S.), then Wal-Mart has substantial ability to set the wage level, and at least some people will take what Wal-Mart (or other big retailers) offer, as supplemented by various welfare benefits, and be able to support themselves in some fashion. It may not be a pretty living, but it’s a living.
If the choice, however, was Wal-Mart with no welfare or finding something else, at least some of those people would choose or be compelled to find something else. That might mean moving to a different area, or getting additional training, or other steps to find a different and higher niche in the labor market. That reduces the supply of potential workers at the lower wage; companies such as Wal-Mart would then be faced with choices. With a reduced supply, they could either reduce their demand (employing fewer people) or pay more to attract sufficient numbers of desirable job candidates.
While big retailers are working to automate as much as possible, no, there are no robots yet available commercially that can stock shelves within a typical retail store. (Target, for example, has been testing a robot that can scan the shelves to find out what products are low or out-of-stock; it then sends a message to a human being to go locate the product and fill the shelves.) Nor can robots do a lot of the other jobs in retail: robots don’t fold clothes very well, or put them on hangers, for example. Even self-service checkouts have met with mixed success: some retailers are busily rolling them out, and others are busily removing them as failed experiments. Different product lines have different service requirements, and more self-service may make sense for some kinds of goods, but bricks-and-mortar retailers (and fast food restaurants, and other big employers in the low-wage economy) aren’t going to be able to get rid of having real live human beings anytime in the foreseeable future.
Lots of other low-wage jobs simply can’t be done by robots at all anytime soon: it’s going to be a person, not a drone, delivering medications and changing bedpans in a nursing home, cleaning gutters and raking leaves for a yard service, or installing windshield washers at a quickie auto repair.