"US imports of goods and services from the rest of the world amount to, what, 12% of national income. Most of this trade is with the first world. Since I am too lazy to go look up exact numbers, I'll say that imports in goods and services from the third world amount to 5% of national income.
"Doesn't this place a rough upper bound of 5% in the amount of US prosperity that can be explained away as imperialist plunder?"
This question intrigues me too. The closest I get to an answer is to think about what all is counted in national income in a country like the US that would not be counted elswhere, or would make up a much smaller part. First, the US has marketized many activities that remain non-market household activities elsewhere. Laundry service is just one example. Second, the salaries of corporate executives are part of national income. What, however, do they really contribute? Third, there are many "services" here, in the way of legal, insurance, 900 numbers, etc., that don't even exist in other places but are nonetheless dutifully recorded as part of gross national product/ gross national income.
What is my point? The quantity, GDP, of which net imports are just a small percent, is a very suspect number! There is a lot of fluff in it, and there have been many efforts, currently mostly by environmentalists like Herman Daly, to come up with alternative measures. The critique environmentalists and others agree on is that GDP does not correlate well with welfare. Environmentalists add to that the interesting point (but not directly relevant to the question here) that the current measures suffer from grossly inadequate conceptions of depreciation. The idea I'm leading up to is that imports from third world countries would figure as a bigger fraction if we looked at only that economic activity that seems important to us.
However, a cautionary note is worth making here on the definition of GDP. Whether meritorious or misleading, the measure is commercially objective! That means, at least in principle--I'll leave it to someone else to explain how the totals are REALLY arrived at--any newly produced good or service that somebody is willing to pay for should get counted. What you or I think something is worth--or whether we think it has even any worth at all--is moot. The important price is market price and the important amount the amount actually transacted. (Ignore inventories for now, please!) So if I find it useful to hire Bill Gates because of his connection to Doug Henwood, and think that connection worth paying Gates $100 million a year for his services on my board of directors, there is nothing in the definition to second-guess my evaluation, and there it goes: another $100 million added to the GDP, and the service sector booms.
There is more (sorry!). GDP was conceived assuming a single, fairly simple competitive market, with no significant non-market power, whether of the monopoly or the military type. Its usual interpretation comes within the pale of making sense only when producers of similar goods and services compete on something like equal terms. It loses all meaning when the ratio of payment for the same good or service is as big as ten to one or even one hundred to one. Its meaning is distorted for our "common sense" even if the services are not the same, as when a lawyer in the US gets $100 for two hours work filling out a tax form, while a computer programmer in India has to work maybe 20 hours to get the same amount. In any case, segmented markets as well as geographically partitioned markets make GDP nearly uninterpretable.
Indeed, due to the partitions and the non-market power, the smallness of the recorded value of net imports (<5% of GDP) may reflect international relations of power and oppression more faithfully than the proportion of total welfare in the US contributed by the third world.
There is no one "right" measure; that's why it's good to have many measures; and in my opinion, the concept of GDP is ingenious in its own way; so I don't say we should just "throw it out." However, other measures should be placed alongside it. Besides those proposed by environmental economists, how about simply measuring human labor hours? What proportion of the labor contributing to life in the US is done domestically? What proportion elsewhere?
There is still more! The shift from GNP to GDP some years ago reflected a decision to measure economic activity according to the nation it took place IN rather than the nationality of the person or corporate entity the earnings went TO. Thus we have a measure of goods and services produced in the US, regardless of whether the profits go to US citizens or to Japanese citizens. Of course, it is the same elsewhere, so US citizens earning oversees contribute to the GDP of other countries, while they would have been counted as contributing to the US GNP.
All that said, I do share your suspicion that some of the "plunder" models of the world economy overstate things. After thinking through all the twists and turns of GDP, GNP and the rest of product and income accounting, however, I realise I can manage little more than a confession of ignorance, mitigated only slightly by viewing labels on products at stores and hearing anecdotal stories of what goes on in other countries. Those are the basis of my other suspicion, that that net import fraction understates things.
Michael Brun