He is assuming that there will be at some point a foreigner who will want to build a factory and you will refuse – because you take your “social contracts and customs” argument seriously.

And this one incident will make your “social contracts and customs” argument harmful; never mind all other foreigners who only want to own, not build. Just a single foreigner who does want to build makes James’ case.

Buying a factory may be better, or worse, than the current situation. If you buy a factory and run it less well than the previous owners, then the economic activity of the country has suffered. Your argument seems to be that the company may be run better – to which the obvious counter charge is, and it may be run worse.

So to summarise. If a foreigner builds a new factory, this is a good thing. If a foreigner buys an existing factory, this may be a good thing, or a bad thing, depending upon the personal qualities of the new owner. You seem to be assuming that a foreign owner will necessarily have better personal qualities than the native owners; which seems implausible.

Daniel and Cian have responded by saying there is a clear difference in economic terms between a foreigner buying a factory and a foreigner building a factory. I’m not disagreeing with that. I’m making a different point: the explanations of why it’s dangerous to allow foreign capital in apply equally well to a foreign investor building a factory as to one buying a factory. In Daniel’s original piece, the reasons he cited why a country might want to keep foreign investors out were: being able to govern foreign companies via domestic laws, more control over tax base, access to shipping and accounting records, desire not to be taken to WTO tribunal.

My point is that all of these concerns are just as applicable to a foreign company that’s building, rather than just buying, a factory.

Now, the argument then becomes, I suppose, a cost-benefit one: letting foreigners build a factory is bad (for reasons Daniel cites), but it’s good because it spurs economic activity, while letting foreigners buy a factory is bad (for reasons Daniel cites), and is of negligible value when it comes to economic activity. I don’t agree with this analysis—I think the negative effects of allowing in long-term foreign capital are negligible at best, and that the positive effects of foreign acquisitions can be meaningful—but my point in the last post was simply that when you’re talking about the dangers of allowing foreign investors in, you necessarily are talking about foreigners who want to build as well as buy.

I don’t see this at all; a new factory has to bid away labour and other resources from the rest of the economy, so it has to work on the basis of marginal opportunity costs. Buying an existing intramarginal unit leaves you with a very different production function. The only case in which James is right here is the typical economics 101 assumption one of homogeneous capital, and thus I think this is a good example of how blackboard models cause people to make serious mistakes.

Restriction on ownership can only be used, it seems, to prevent foreign capital from entering the country.

So, Daniel is right – these are two different concepts, in fact they are exactly the opposite.

However, contrary to Daniel’s argument, the former seems much easier to justify than the latter; you probably do need a fair amount of sophistry and demagoguery in most cases to justify your refusal to allow foreign capital to flow into your country. You certainly want more capital, not less.

Does it make sense what I’m saying?

Um, no they don’t. The worst that can happen if a foreigner builds a factory is that it will fail. However given that there was no factory before, and economic activity will have occured, this is still preferable to no investment.

In contrast, if a foreigner buys a company and causes it to fail (or do worse than it did before) – then a reduction in economic activity has occurred as a result.

Clear distinction.

I also, of course, disagree that it’s hard to see the positive benefits. The Gourinchas study is about the direct effects of increased foreign capital flows into developing countries, and demonstrates convincingly that these are quite small. But it also suggests that there may be indirect effects of free-capital markets that have nothing to do with capital deepening, and this—it seems to me—is where most of the gains from free-capital markets come from. (This was the case I tried to lay out in the Guardian piece.) To take only the example you cite, having Japanese auto factories here didn’t just get around tariffs. It also dramatically improved the productivity of workers in the US auto industry, and the US auto-parts industry, too. That’s a small improvement, but a valuable one.

Maybe, but the case for allowing FDI in developed countries is also rather weak. That is, it’s hard to see the positive benefits that come. One case where I think the results were positive was where Japanese auto companies got around tariffs by building plants here—but those benefits were related to the implicit lowering of tariffs.

If Surowiecki really can’t see the difference between financial speculation (writing somebody a check for something to transfer ownership) and real economic investment then he needs to switch careers. Of course, the heavily ideological pop economics of the 1990s (which probably influenced Surowiecki, since that’s when he broke in) often depended on blurring precisely that difference.

The appeal to Krugmanite authority doesn’t particularly help, since who knows the context of those quotes. If Krugman meant that foreign purchase of a company is not portfolio investment he is just clearly wrong. But my guess would be that he was driving toward a different point: that when you actually buy a majority interest in a company you are therefore more likely to use your control of that company to make real investments in the country through building new plants for the company, etc. So the two are not the same but they are correlated. Presumably a large ownership share of company stock means that your investment is less liquid, implying a longer term committment, so perhaps you will stick around and actually do something in the country. Well, maybe, but maybe not. Maybe you will sell off the company’s assets and then go invest the proceeds in some other country. One can argue this makes some contribution to the “destruction” part of “creative destruction”, and 90s-style free market ideology would probably say it is therefore productive labor, but it isn’t economic investment.

Clearly, it depends on the circumstances, on a case by case basis. Clearly a state with well-established industry has a certain kind of advantage and a politically stable state with low-wage labor and weak environmental laws has a different kind of advantage; so how can this be so simple?

So, should we have tariffs on the exploitation of such new technology? So as to avoid the harm to, as you correctly note, those US Steel workers?

Stating that competition can only be meaningful when all operate under the same labor laws is a little extreme. Do you seriously mean that the sweat shop worker should not get $40 a month, that she should go back to the paddy field and get only $10 a month because employers won’t or can’t buy her the same $10,000 a year’s worth of medical insurance that an American worker gets?

The point about mercantilism, protection and so on is not that they do not favour certain interests within a nation state. They do, clearly. Which is in fact precisely the argument. They favour certain narrow groups within that State, rather than the wider public. Which is why “tariffs are bad” is an impeccably liberal position to hold. Those who benefit from them are not the great unwashed and it is that very same great unwashed that we’re supposed to be aiding.

(Yes, I know, I’m a classical liberal, not a liberal, but in this instance the two philosophies should, at least, be urging the same behaviour.)