Why so little demand for protectionism?

Paul Krugman asks a very good question, namely why the political pressures for protectionism in the midst of recessions and depressions have been so weak.  While I do not disagree with his points (which cite institutions such as the WTO and EU), I am surprised by what he leaves out.  Here is a summary of Spence and Hlatshwayo on U.S. labor markets:

Looking back on the period from 1990 to 2008, the co-authors found that 97 percent of the 27.3 million U.S. jobs created were in the non-tradable sector. (The five largest non-tradable sectors, mentioned above, contributed 65 percent of the 1990-2008 jobs growth.) “The employment creation occurred mostly in non-tradable sectors — where we don’t have international competition,” Spence said.

In other words, with more jobs in the service sector, we are practicing increased “protectionism by any other name,” often with the law and with regulation but in many cases cultural barriers and lack of trade networks will suffice.  Trade costs for many services are in any case high and thus the constituency for protectionism or further protectionism is not quite there.  The workers who might have supported tariff-based or quota-based protectionism thirty-five years ago already have lost their jobs to foreign trade and they or their descendents have moved to more heavily protected service sectors.  As we should recall from the literature on the gravity equation, explicit tariffs are only a small part of the actual barriers to trade.

A second issue is the where the actual burden of foreign competition is falling, given a much higher degree of globalization.  The Mexicans are worried about Chinese competition, but they are not mainly worried about Chinese competition pulling Mexican consumers away from Mexican products (chili peppers are one exception here).  Mostly they are worried that Chinese competition has taken away many of Mexico’s export markets elsewhere, and putting tariffs on Chinese goods coming into Mexico won’t stop that.