It seems so obvious: if you don't pay employees enough to live on, you will have low morale and high turnover, lowering productivity and creating less value. The reduction in value created will always exceed the labor costs saved. So if sweatshop owners are really "greedy", why don't they see this when they look at their labor force, factory, and financial statements? Because they can't, for two reasons:
--Piece rate work does not create value and commoditizes laborers--to the factory owner's detriment.
--Archaic, irrelevant labor costing techniques developed before spreadsheets and exported overseas become dangerous agents of poor decision making in the hands of managers who don't understand the underlying math.
First, a wiki-sized history of the garment industry. Mass communications and the growth of department stores and catalogs made mass merchandising of fashion possible. However, the difficulties in predicting fashion success coupled with long lead times required to produce a clothing line made conventional financing difficult. Investors wanted a piece of the profits--but profits in this shady business too often fell victim to a bookkeeper's eraser, the back of a wise guy's truck, extravagant garment district overheads, or a foolhardy position in a fabrication with fleeting fame. So investors settled on more predictable metrics: Labor costs could be based on negotiated piece rates and materials costs on yields. Overhead could be controlled by limiting expenses to 40 or 50% of labor costs, a benchmark at the time. If overhead were higher than that the investors' share of the profits would come out of the manufacturers' pockets.
This model lasted for a while, but profits were still elusive (because piece rates produce pieces, not garments, but more on that later). The industry moved to a licensing model which just ignored profits and took fees off of sales--now Calvin Klein doesn't produce a thing and just gets a royalty check based on sales from the vertical conglomerate that manufactures the clothing.
But the piece rates and overhead multipliers survived in the costing departments like nipples on men. Factory owners and industrial engineers measured labor costs one way: with a clipboard and a stopwatch, meticulously timing production of every step multiple times for accuracy, then increasing it by some divine multiplier like 40% to account for overhead: The industrial engineering equivalent of ordering a bacon double cheeseburger with a diet coke. Garment factories shipped these metrics to Asia along with their equipment and systems of work and they exist, unquestioned, to this day.
The problem with piece rates is that they encourage workers to produce pieces--not garments. If a worker is paid by the collar they sew, you'll end up with a ton of collars, but you may not get a ton of shirts. Collars have no value--you can't sell them. The reason garment factories rarely achieved profitability in piece rate systems was that the profits ended up on the floor, in piles of work-in-process. Sure, value was created--someone would eventually make and buy a jacket, but imagine all that gratuitous effort in the name of productivity!
If a worker only sews collars, she is going to go from novice to expert in a short period of time and make the better paid veteran sitting next to her start to look expensive. This commoditizes the laborer and makes it difficult to see who has the know how and experience to create value. So when times got tough, it made perfect sense to cut labor costs by squeezing out higher paid veterans. And, because the workers were focused on producing parts rather than the whole to begin with, value creation probably didn't suffer much.
The overhead multiplier used in the garment industry became a weapon of mass destruction when Excel was introduced to the schmata trade. That overhead is 40%-50% of labor in garment manufacturing was as accepted a benchmark as the keystone markup is in retail. But using this benchmark in financial modeling and decision making without regard to its origins or the underlying math will lead to some terrible decisions.
Keep in mind that in financial statements, labor is only charged when a product is sold. All other expenses--such as the labor invested in work in process that is not made into garments--gets put into inventory. So when a spreadsheet uses 40% of labor costs to calculate overhead rather than trying to predict the actual overhead, overhead--representing "fixed" expenses--appears to fluctuate with labor costs. If you lower your labor costs, your overhead will be appear to be reduced, when in reality it stays the same and probably increases because of the need for additional management. If you increase your labor costs, or add more laborers and expand within the same facility, or you create more value and sell more units using the same workers, your overhead will appear to increase, when in reality it stays the same and actually decreases on a unit basis.
And when you tag an artificial expense onto your labor costs, you obscure your true unit profitability. In other words--YOU CANNOT SEE HOW MUCH MONEY YOUR WORKERS ARE MAKING FOR YOU, which marginalizes the impact of better workers, further commoditizing them.
One of my favorite quotes was from a one time garment magnate whose company was insolvent and we were negotiating to buy his assets. His calculations used the overhead multiplier, and mine used actual labor costs and actual overhead costs. Of course his calculations showed his company should still be in business, because they neglected the eighteen 40-yard dumpsters of obsolete work-in-process we had to remove after the deal closed. Our calculations showed much lower profitability. He said to me, "With the way you calculate profits, I can't understand how you make any money." He--and his generation of "garmentos"--would never understand that value can't be created with a calculator.
This distinction is critical to the the fight against sweatshops. There appears to be a battle within the left about whether to boycott sweatshops, police sweatshops, or just ignore them, because in many environments a job in a sweatshop is far better than the alternative. I believe in a fourth way: co-operation. There are practices in every sweatshop that hurt both workers and management, and whose elimination would help both labor and management, but management knows no alternative. They just know the clipboard, the stopwatch, and the piles of work in process showing that everyone has plenty of work. If we could just show sweatshop management how to create more value by converting that waste into wages, those factories would become more competitive, and the race to the bottom might start to reverse itself.