Well, why not? That seems to be the central issue. And there is a classic category for the sort of reasons that would dictate a "not" if they apply: namely, that the transaction would in some way be harmful to some other person or persons. Now, if it were harmful to one of the very parties to the trade, that person would, of course, be motivated not to make the trade. In the normal case, the party does not think that, and the deal goes through. So the only relevant parties are those not parties to the contemplated transaction. The question is, when are they in fact "relevant"?

     There has been a tendency in recent decades to discuss matters of business ethics (among others) in terms of "stakeholders," these being people who have any sort of interest at stake. So defined, the notion is very broad. Much too broad, actually: if taken really literally, it would undo any rational discussion of this or any similar subjects. After all, it is a pretty safe conjecture that for any interaction whatever between any two persons whatever, we would be able to find at least one third party with an opinion, quite possibly a passionate opinion, about whether the interaction in question is a Good Thing or not – including the implication that the persons should be either allowed (because some don't mind about it), or encouraged, perhaps even required (because some are enthusiastic about it), or forbidden (because some object) to do the thing in question. Indeed, there is probably not a single namable activity undertaken by two or more persons that someone out there does not object to on the basis of some supposed "moral" principle. If we were to hold that no transaction between A and B is legitimate unless there exists no one who disapproves, all transactions whatever, we may be sure, will be illegitimate.

     In order to bring coherence and sanity back into view, we clearly need to narrow down the set of "stakeholders" in some suitable way. Do we have the resources to do this? I think that we do, and that they are the classic resources of liberal theory, plus sheer logical commonsense. Here are the necessary restrictions, as I see it.

1. The "stake" of a stakeholder cannot be ideological. That is to say, A is not a relevant stakeholder simply because A has an opinion about the issue whether persons situated in the case at hand should be treated one way or another.

2. Stakeholders must be persons whose personal well-being will be affected for better or worse by the contemplated action.

3. 'Personal' well-being in the previous point may be analyze into two sorts (by no means mutually exclusive): one, as affecting the individual concerned independently of all others; or two, as affecting that individual by virtue of his or her relation to some other person or persons. In the second case, though, we impose two further conditions. First, the involvement of those other persons must be voluntary. Otherwise we will be promoting this person’s good at the expense of someone else. Second, those other persons' involvement must, in turn, meet the other conditions spelled out here.

4. All stakeholders’ involvement in the issue before us must meet the condition of being mutually beneficial in relation to the actors (say, corporations, employees, entrepreneurs, whatever) involved.

5. All affected “stakes” must be affected significantly; trivial, unmeasurable, or effects no greater than what are entailed by fully normal activities are not allowed to count.

     These all seem to me to be minimally reasonable conditions for being concerned about putative "stakeholders." The list has an important implication: it does not allow some group of stakeholders to insist that, say, some company undertake some substantial action expensive to the company for the sake of improving the situations of that group's members on its own account. The company has to judge that it will do better by doing so. In effect, a stakeholder has a case for compulsion only when the actor in question is in fact worsening that person’s situation; a potential bettering is not enough to justify compulsion.

     Move now to the situations of people in poorer countries who might be in the international labor market. Are they stakeholders? Certainly, in the sense that there are potential benefits to them from the new employment that might result. Perhaps also in the sense that they might stand to lose, either in the sense of losing work that they now do but which will be made obsolete by the new incoming technologies, or in the sense that the new environment in which he and his fellows will be immersed once extensive foreign investment is made may be unfamiliar and even painful. Both are possible concerns, but they are quite different concerns.

     Let's begin with the first point, that foreign investment offers benefits in the way of new employment. We first need to recognize an important point about this: the new employment is certain to be at substantially higher pay levels than what prevailed before. That is because the new industries must attract these workers. The new industries typically attract relatively more capable workers, since they will need to learn new processes and routines; they may also attract the hitherto unemployed. For them the new wages will no doubt be low, but they will be higher than zero. All of this will make a considerable difference to the real income of the communities in question. We will be able to add in the side effects of increased economic activity: the newly employed with their higher wages will spend those and create more employment for others in their towns and villages.

     Where's the downside, then? Why, indeed, is increased economic activity perceived as having a 'downside'? There are two answers to this. One is that when things change, there are always some who prefer the previous ways of doing things, even if those included a lot of poverty. Change does not sit well with everybody, and that is understandable. The question is, what normative significance we should attach to this. And that in turn is going to depend largely on the question whether it can be claimed that these people are "losing" in any relevant sense.

     A main claim on behalf of capitalism generally and free trade in particular is that is that the system runs on Pareto optimality: the idea, at least, is that somebody wins and nobody loses. But how can this be if practitioners of commerce bring about unemployment for some? Losing a job is, after all, losing – isn’t it? And isn’t that bad? Doesn't commerce with these effects, then, violate the Pareto criterion – making some better off at the expense of others?

     There is an answer to this that is more than just credible – it's actually right! The answer has two parts. The first is that 'worse off' requires a point of comparison, a base line. One is worse off now that one is unemployed, we will suppose, than before, when one was employed. But the subtle implication of talking as if this were the relevant point of comparison is that people own their jobs, that they are entitled to them. And they are not, of course. If I work for you, then I have an agreement to work: to do various tasks for you, at such-and-such a level of pay from you, yes – but for how long? In taking me on, you did not, after all marry me. All you did was employ me, and that perhaps for a certain specific period of time. The relevant point of comparison is not "my" job, but my contract. That agreement tells me, in a general way, for how long I work for you. Though it is too bad for me if, at the end of my contractual period, you do not renew my engagement, it is wrong to say that you have inflicted a damage or loss on me in so doing. A better way to look at it is that in the previous period, while I was working for you, you were providing me with a benefit. Now that the contract has run out, you no longer provide me with that benefit – but that is all. You do not, for example, shoot me or steal something from me. You only cease extending to me a benefit that you previously did provide. The correct baseline of comparison is how you are in relation to me, without that particular job. If you let me go, I am now where I was before you hired me – not worse off. (Sometimes this won't be true, or not exactly true. Working many years for you might meanwhile make me incapable of doing many other things that I used to be able to do. My re-employability might be eroded by the character of my work for you. That's important, of course. But it's also rather obvious, and something one should take into account in going into a particular line of work. In some cases, it would suggest the advisability of insurance, or of writing a contract which took this into account).

     Answering this in the way that I have just done runs the risk of being accused of splitting hairs. But to a libertarian, this does not look like a hair. Nor, I should think, to anyone else. Your employer cannot shoot you or beat you, as he could back in the era of slavery. All he can do is let you go, and even then only in accordance with the terms of your employment contract. In the case where you are working for what, by western standards, is a pittance, you may not have a written contract and that, of course, is importan too. We would need to ask what should be on the unwritten contract or understanding between us; as indeed we will do, below. We should also appreciate that if you are earning 16˘ per hour, and are doing all right, it would be pretty surprising if your employer wanted to let you go with an airy wave of the hand. At that wage, you are a bargain, and your employer is quite unlikely not to have noticed that. (Here we note also the flip side of the preceding point about possibly eroded future employability: for as you work at this job, you become increasingly desirable for sustained employment as compared with newcomers who have no experience. You can, simply, do it better than your competitors.)

     The question is whether we can be content with this observation. And the trouble is that perhaps "we" will, but there's a great likelihood that the workers who lose their jobs as a result of outsourcing or removal to a foreign country are not going to be. So the question is whether they have a point.

     Dani Rodrik, a Harvard economist who has done much thinking about this problem, writes:

"Compensation of losers could take care of the problem, as the larger economic pie resulting from trade in principle allows the losers to be compensated in full while leaving the beneficiaries ... still better off. And this is indeed the first line of defense in the classroom and in most policy debates when the economist presents the case for gains from trade... However, compensation rarely takes place in practice and never in full. There are good theoretical reasons ... why this is so." ²

     Rodrik goes on to suggest that there are further and better reasons for international trade. In particular, its effect is exactly like that of technological improvement, which people generally admire and approve of. Still, he holds that this defense of free trade works only if the gains from trade are “consistent with the prevailing norms and rules of 'fair play' at home." (31) And of course, if these norms and rules are well taken, then he's right. Now imagine that the outsourcing leading to local unemployment saves money by using 12-year-olds working in sweatshop conditions. There's a good chance that this would lead to uproar back home. And yet, Libertarians may approve anyway – so this brings us to the crunch. In Honduras, say (hypothetical location of the outsourced manufacturing in Rodrik's example), it may well be the custom for families to encourage, or even to compel, their children to work when we would not dream of doing so. Interestingly, however, there is evidence that the "we" who would not dream this are mostly academics, pundits, intellectuals, and in general middle and upper class bourgeoisie. Low-income area workers understand about such things and so may not tend, for example, to support restrictive legislation that would prohibit import of goods made by child labor (or other sorts of labor that we don’t like). (33)

     Still, though, they and economists as well are leery about relaxing standards for migrants workers. Suppose those Honduran families come to work in Ohio, sweatshop child workers and all? They would soon be descended on by the authorities. And, it appears, most economists would approve. But what is the difference in fact between sweatshop labor in place X and sweatshop labor in place Y? Is there more sense in tolerating it across a border than within it?

     There is some kind of answer to something like this last question. As it stands, the answer is No. But you and I might very well not like to live in a neighborhood where people needed to resort to this. There are things we could do about it locally: set up a charitable organization to assist the families in question, for example. Doing that wouldn't commit us to doing it anywhere else, nor, especially, would it allow us to compel the families in question to refrain from practices we don't like.

     Rodrik also points out that American governmental practice on such matters has been riddled with inconsistency. The government objects to other countries doing things that it itself does regularly. It is pro free trade when that is to its supposed advantage, and anti when it's the other way around.

     In fact, however, the idea that trade restrictions are "to country X's advantage" is flawed in any case. If some company can do better by engaging labor in another country, the result will be lower prices in the home country, more work or at better wages for people in the other country, and an expanding market in the other country for the home country's products. Our right to work and trade is not a right to do so at other people's expense. But peaceable work and trade is not actually done at others' expense, so long as their options are not closed off by ill-considered constraints imposed by governments.