Average wages and average consumers: a mistake
This is the fourth in a series of articles: sweatshops (intro), sweatshop definition, and sweatshop pay. The previous post on the importance of the marginal concept in price theory is also relevant; I would have used a better article, but the wiki entry on marginalism is uncharacteristically bad.
In the essay, "Defending Sweatshops: Too Much Logic, Too Little Evidence," Rothstein fails to understand the same principles of economics as O'Reilly in my marginal post, especially the importance of the marginal cost and subjective value (marginal utility). He and other writers on the sweatshop front make frequent statements such as, "wouldn't you willingly pay a little more for a t-shirt so that sweatshop workers can be paid a 'decent' wage?" You and I may, but that completely misses the mark.
Let's go back to the marginal cost of labor (the cost of one more unit, be it man-hour, piece, whatever). What we need to see is that margin, not the average. The factory is humming along, selling everything they make. Now, you raise wages. On the margin, you have to decide if the selling price for that t-shirt is going to equal or exceed the cost of labor to make it. At some point, as you raise wages, the answer is no, the cost of one more unit of work exceeds the benefit the factory can gain from it. This is the margin.
Downstream, there are similar considerations. Consumers have other uses of their money, including mortgage, fuel, and groceries. The value they can gain from one more pair of jeans has to compete with those other uses of their money. When the cost of those jeans exceeds the benefit of them (measured in terms of rent, groceries, or fuel foregone), then you stop buying jeans. This is the margin. What consumers spend more time "at the margin"? Those with less disposable income, aka, the poor. The wealthy and the middle class aren't forced to confront tradeoffs as often. Of course, if we are talking about $100 designer jeans, that's not a decision the poor should be facing (though I'd appreciate it if someone can explain why high-dollar Nikes are so popular among urban utes).
Marginal analysis determines whether one more pair of jeans will be made or sold at a given price. If jeans go unsold, the price is too high (storage costs, too, so there is a time limit). If people line up to wait for deliveries, the price is too low (waiting in line costs, too, so there is a time limit). The analysis of whether people will buy at a higher price, and therefore whether increases in labor costs can be passed on, depends on people right at the margin of deciding to buy or not.
You are a factory, selling all the jeans you can make for $10/each. Your competitor sells them for $10.05. Your customer buys all they can from you, then all they can from your competitor, and then maybe a few more from another, more expensive competitor ($10.10). Their marginal cost is $10.10, but their average costs are between $10 and $10.10 (depending on the mix they buy). If your costs drift upwards to $10.10 and higher, you may not be able to sell all you can make with all these well-paid workers unless the prices downstream also go up, correspondingly.
There are competing theories on how significant the impact of raising wages by mandate is on the labor pool.
As to the impact on lower skilled workers, the key to understanding how much a firm is willing to pay for a worker is productivity: if the worker creates $X worth of marginal value, then pay him $X.
I once saw this exchange on usenet:
#1) I own a restaurant. I'm not stingy or rich; my waitresses make more than I do. I pay busboys minimum. I can't afford a raise in the minimum.
#2) Then get rid of the waitress.
That doesn't make sense. If you get rid of waitress, someone else has to do that work, probably the owner. Now someone has to do his work (cook?), probably someone not as skilled. The result: lower quality of food, lower quality of service at given price; or the same quality, but lower price (because you have to wait longer); or same quality, same price, but fewer customers served (because you have to wait longer). The benefit to owner is down, benefit to customers is unchanged or down, benefit to the waitress is down. The ostensible benefit to the busboy (assumes no resulting change in busboy working conditions, up to and including layoff) is up. Net benefit to society: down. The problem here is that the owner has saved the cost of the waitress, but the average productivity in his restaurant has plummeted. The change in the wage did nothing to help it, either.
Try this pop quiz: wich of the following result in higher productivity?
Free riding doesn't necessarily mean that firm 2 is immoral, cheating, or suffering from some other dire malady. Consider the example of the collection plate (from Tom Tietenberg's excellent Environmental Economics and Policy). You may firmly believe that your church needs $25/week to operate (it is worth $25 to you). You shouldn't overspend, since after all you also need to give to Habitat for Humanity, Nature Conservancy, etc. Your ethics require you to maximize the benefit you get from every investment, including (or perhaps, especially) your charity. So the collection plate comes to you, you see that there is already $20 in it (from someone else who believes as you do, but can only afford $20), so you put in $5. When you add up both utilities, you see that the aggregate utility of church is $45 (or perhaps more), but the total contributed is $25. Church is going to be underproduced, even though everyone believes they are doing their best to fund it.
Likewise with any public good: it is likely to be underproduced, and that fact doesn't change even if you fund it through taxation. Education has both a private and public component; we as a society benefit from literate citizens, but each individual benefits from that part of their education that increases their productivity and therefore pay. If Rothstein really believes that improving productivity through education is a good thing, he needs to find a better way of funding education than forcing employers to raise wages. To me, that is just like saying that we need to fund NASA on the hopes that they will invent something useful to us, like Tang. Without any prompting from us, factories will determine the proper amount of training necessary because they will adequately fund the training that benefits them, and underfund education that benefits the public, including their competitors. Conceptually, I usually differentiate between education (general knowledge like reading and writing) and training (specific knowledge/skills like operating a particular machine).
I am ready to concede that there is some slop in the cracks and that management may not be as efficient as we sometimes assume, at least in the footwear, textiles, and apparel (FTA) industry where actual sweatshops may exist (but not in the way and not to the degree that many activists claim). As I have observed before, I believe that the historical trends that brought about the present state of FTA (low tech, family ownership, generational transfer) have made them insular, resistant to change, and impervious to modern management methods. I'm not the only one. By "modern management", I mean actual management, not using "offshoring" as a crutch. By resorting to this one-tool toolbox, the MBAs have reduced apparel to a boring commodity using the same technique as every one of their competitors, yet they still can't figure out why they aren't making money (actually, several are highly profitable). I also believe that Kathleen is correct when she says that they will never invent a way to handle material except by human hands, and that this is the reason it is so hard to improve productivity in the industry by automation, the technique that has revolutionized so many other industries (though I would say, "not for a very, very long time" instead of "never"). So what is my counter-proposal for improving productivity and pay in this industry? See my next post on the subject.
Sweatshop
In the essay, "Defending Sweatshops: Too Much Logic, Too Little Evidence," Rothstein fails to understand the same principles of economics as O'Reilly in my marginal post, especially the importance of the marginal cost and subjective value (marginal utility). He and other writers on the sweatshop front make frequent statements such as, "wouldn't you willingly pay a little more for a t-shirt so that sweatshop workers can be paid a 'decent' wage?" You and I may, but that completely misses the mark.
Let's go back to the marginal cost of labor (the cost of one more unit, be it man-hour, piece, whatever). What we need to see is that margin, not the average. The factory is humming along, selling everything they make. Now, you raise wages. On the margin, you have to decide if the selling price for that t-shirt is going to equal or exceed the cost of labor to make it. At some point, as you raise wages, the answer is no, the cost of one more unit of work exceeds the benefit the factory can gain from it. This is the margin.
Downstream, there are similar considerations. Consumers have other uses of their money, including mortgage, fuel, and groceries. The value they can gain from one more pair of jeans has to compete with those other uses of their money. When the cost of those jeans exceeds the benefit of them (measured in terms of rent, groceries, or fuel foregone), then you stop buying jeans. This is the margin. What consumers spend more time "at the margin"? Those with less disposable income, aka, the poor. The wealthy and the middle class aren't forced to confront tradeoffs as often. Of course, if we are talking about $100 designer jeans, that's not a decision the poor should be facing (though I'd appreciate it if someone can explain why high-dollar Nikes are so popular among urban utes).
Marginal analysis determines whether one more pair of jeans will be made or sold at a given price. If jeans go unsold, the price is too high (storage costs, too, so there is a time limit). If people line up to wait for deliveries, the price is too low (waiting in line costs, too, so there is a time limit). The analysis of whether people will buy at a higher price, and therefore whether increases in labor costs can be passed on, depends on people right at the margin of deciding to buy or not.
You are a factory, selling all the jeans you can make for $10/each. Your competitor sells them for $10.05. Your customer buys all they can from you, then all they can from your competitor, and then maybe a few more from another, more expensive competitor ($10.10). Their marginal cost is $10.10, but their average costs are between $10 and $10.10 (depending on the mix they buy). If your costs drift upwards to $10.10 and higher, you may not be able to sell all you can make with all these well-paid workers unless the prices downstream also go up, correspondingly.
There are competing theories on how significant the impact of raising wages by mandate is on the labor pool.
- Standard answer - raise prices of labor, decrease quantity of labor bought
- No impact - This was the finding of David Card and Alan Krueger's 1994 empirical study, “Minimum Wages and Employment: A Case Study of the Fast-Food
Industry in New Jersey and Pennsylvania.” American Economic Review, Vol. 84, No. 4, pp. 772-93. Unfortunately, I couldn't find a free version, so see Bryan Caplan's post for an explanation and discussion, especially Dsquared's exahnge with James. - Differentiated impact
- A higher minimum wage encourages more dropouts because it makes not going to school a better use of their time
- A higher minimum wage favors people with more skills, i.e. against dropouts and minorities (note the contradiction with the previous point)
- Gordon Tullock (via Tyler at MR): "the government can make an employer raise nominal money wages, but can't stop him from turning off the air conditioner."
As to the impact on lower skilled workers, the key to understanding how much a firm is willing to pay for a worker is productivity: if the worker creates $X worth of marginal value, then pay him $X.
I once saw this exchange on usenet:
#1) I own a restaurant. I'm not stingy or rich; my waitresses make more than I do. I pay busboys minimum. I can't afford a raise in the minimum.
#2) Then get rid of the waitress.
That doesn't make sense. If you get rid of waitress, someone else has to do that work, probably the owner. Now someone has to do his work (cook?), probably someone not as skilled. The result: lower quality of food, lower quality of service at given price; or the same quality, but lower price (because you have to wait longer); or same quality, same price, but fewer customers served (because you have to wait longer). The benefit to owner is down, benefit to customers is unchanged or down, benefit to the waitress is down. The ostensible benefit to the busboy (assumes no resulting change in busboy working conditions, up to and including layoff) is up. Net benefit to society: down. The problem here is that the owner has saved the cost of the waitress, but the average productivity in his restaurant has plummeted. The change in the wage did nothing to help it, either.
Try this pop quiz: wich of the following result in higher productivity?
- Educate workers
- Replace workers with capital (automation)
- Turn down the air conditioner and cut back on other compensation
- In a high-turnover environment, allow workers to leave through attrition and replace with more experienced workers
- Lay off less experienced or less capable workers
- Educating them makes them more productive, but education costs. There is a break-even point.
- Introducing machinery makes them more productive, but machinery costs. There is a break-even point.
- Cutting costs in other areas increases total (mulifactor) productivity, but there is a break-even point. Hot, sweaty, dissatisfied workers aren't as productive.
- Replacing lower productivity workers with higher productivity workers raises the average workforce productivity. It doesn't help the workers who leave.
- Removing lower productivity workers raises the average workforce productivity.
Free riding doesn't necessarily mean that firm 2 is immoral, cheating, or suffering from some other dire malady. Consider the example of the collection plate (from Tom Tietenberg's excellent Environmental Economics and Policy). You may firmly believe that your church needs $25/week to operate (it is worth $25 to you). You shouldn't overspend, since after all you also need to give to Habitat for Humanity, Nature Conservancy, etc. Your ethics require you to maximize the benefit you get from every investment, including (or perhaps, especially) your charity. So the collection plate comes to you, you see that there is already $20 in it (from someone else who believes as you do, but can only afford $20), so you put in $5. When you add up both utilities, you see that the aggregate utility of church is $45 (or perhaps more), but the total contributed is $25. Church is going to be underproduced, even though everyone believes they are doing their best to fund it.
Likewise with any public good: it is likely to be underproduced, and that fact doesn't change even if you fund it through taxation. Education has both a private and public component; we as a society benefit from literate citizens, but each individual benefits from that part of their education that increases their productivity and therefore pay. If Rothstein really believes that improving productivity through education is a good thing, he needs to find a better way of funding education than forcing employers to raise wages. To me, that is just like saying that we need to fund NASA on the hopes that they will invent something useful to us, like Tang. Without any prompting from us, factories will determine the proper amount of training necessary because they will adequately fund the training that benefits them, and underfund education that benefits the public, including their competitors. Conceptually, I usually differentiate between education (general knowledge like reading and writing) and training (specific knowledge/skills like operating a particular machine).
I am ready to concede that there is some slop in the cracks and that management may not be as efficient as we sometimes assume, at least in the footwear, textiles, and apparel (FTA) industry where actual sweatshops may exist (but not in the way and not to the degree that many activists claim). As I have observed before, I believe that the historical trends that brought about the present state of FTA (low tech, family ownership, generational transfer) have made them insular, resistant to change, and impervious to modern management methods. I'm not the only one. By "modern management", I mean actual management, not using "offshoring" as a crutch. By resorting to this one-tool toolbox, the MBAs have reduced apparel to a boring commodity using the same technique as every one of their competitors, yet they still can't figure out why they aren't making money (actually, several are highly profitable). I also believe that Kathleen is correct when she says that they will never invent a way to handle material except by human hands, and that this is the reason it is so hard to improve productivity in the industry by automation, the technique that has revolutionized so many other industries (though I would say, "not for a very, very long time" instead of "never"). So what is my counter-proposal for improving productivity and pay in this industry? See my next post on the subject.
Sweatshop
<< Home