(1) Starting to export involves a major one time start up cost.
(2) High productivity domestic firms are more likely to start to export.
(3) The entry of high productivity firms into the export market tends to put low productivity firms serving only the domestic market out of business.
(4) In high skill economies, lowering the start up costs associated with exporting in high skill industries can greatly increase the proportion of firms that export.
Older models look mostly at comparative advantage in terms of labor costs per unit produced.
[M]ost firms do not engage in trade, but those that do are on average more productive and pay higher wages. . . . [W]hen a country opens to trade, more productive firms grow relative to less productive firms . . . shifting labor and other resources to the better organized firms and increasing overall productivity. Even if workers do not switch industries, they move from firms that are either poorly managed or that use less advanced technology and production processes toward the more productive firms. Thus, firm-level evidence demonstrates that trade allows not only economy-wide advances through resource allocation, but also allows within-industry productivity advances through reallocation of resources across firms. This shift has clear welfare-enhancing impacts[.]
Where does all this lead politically?
First, it lend to support for Obama administration proposals to reduce one export business start up costs, if you care about total economic welfare, because it makes aid to companies trying to get started in exporting look less like corporate welfare, and more like a one time investment in alleviating government created costs that will give rise to an economic engine that will create jobs in the long term.
Compared to the long term revenue costs associated with pervasive incentives to favor unearned income (i.e. "to invest") in the tax code, helping companies beat the red tape involved in getting started in exporting looks like chump change with much bigger rewards.
Second, "a commitment to an open trading system requires more than 'just say no' to protectionist measures. It also requires establishing the policy bases for mitigating protectionist pressures. This in turn . . . requires ensuring that the benefits of higher productivity and lower prices are also shared by those that experience the effects of competition from foreign producers."
In other words, if you want to be able to politically sell free trade over the increasingly popular notion of a mechantalist/protectionist approach for international trade, you have to be able to mute the objections of the less productive domestic market firms that get hammered by it.
Low productivity domestic businesses can be very sympathetic. Look at the staying power that liquor stores have had in continuing to exclude grocery stores and convenience stores in Colorado from selling full strength beer, wine and liquor. Bit by bit this domestic protectionist wall is starting to show chinks. But, everyone involved in the debate assumes that grocery stores and convenience stores would decimate liquor stores if all of them were allowed to sell all alcoholic beverages on an equal basis. In this debate, the presumably original arguments for liquor stores which is that grocery stores or convenience stores would be less effective at policing alcohol control laws have almost no currency. The liquor store lobby's argument is pretty much purely the concern large numbers of small, neighborhood, often family run businesses would be put out of business if grocery stores and convenience stores were allowed to compete with them on a level playing field.
The blog post I link at Econbrower also points out that "productivity is not the be all or end all" without really getting what that means in the way, for example, that Jared Diamond did in a second edition epilogue to his groundbreaking book "Guns, Germs and Steel."
A market with lots of domestic oriented low productivity firms (and I think that the empirical evidence would show that scale is as important a factor as good business management in the gains that exports make when they start exporting), often have higher quality and product diversity that confers intangible benefits that a narrow productivity measure would suggest.
A good example of a domestic oriented, low productivity, small scale industry is German beer making. There are large numbers of small brewpubs, and nothing approximating national high productivity, large volume products like Budweiser in this market. This is good for people who want to have high quality craft brews available to them. It is bad for people who want consistent, low cost beer.
Family restaurants are domestic oriented low productivity businesses. McDonald's is a high productivity business that operates on a large scale.
Quaint local main street shops are low productivity businesses. Wal-mart is a high productivity business.
Favoring high productivity, export oriented businesses does not mean favoring the "best" businesses in the marketplace, because productivity isn't as universally a good thing as economists assume.
Globalist leaning economists tend to have a productivity is everything approach and think that opponents to globalism are simply irrational. But, once you realize that undo attention to favoring productivity itself comes with serious costs of its own, and usually hurts local, small scale businesses that have other benefits that they bring to the table, the vigorous anti-globalist movement looks far less irrationally reactionary and purely selfish.