There has not been a lot of papers on this topic, but lately there seems to develop a small literature. Charlot et al. (2004) find that it is not easy to find out what is Pareto-optimal in the Core Periphery model. Well, the problem is rooted deep, I would say. Let’s give the word to Stiglitz (Reflections on the state of theory of monopolistic competition, 2004, p. 146):
This, while half a century ago, some may have marvelled at the wonders of the market, believing that some how we had thebest of all possible worlds, tody we are cognisant of the limitations of our world – but unsure how, or in fact whether, it can be improved upon. Making matters more complicated is the fact that individuals do differ in their preferences, and that a major source of product differentiation is the response to these differences; but accordingly, policies which affect the extent of diversity of product differentiation, can have markerly different effects on different individuals.
So atop of these problems in the NEG you have individuals that might differ in preferences for location. This has been recognized by Tiebout as a problem of voting with one’s feet, and it reappears in the Core Periphery model. Applying it to European structural policy, the whole concept of convergence can be put into question. If people are not willing to move, some laggards in the periphery will probably stay behind forever. Or, if some people move away, we have adverse selection and only the losers stay. For more on this topic see Brakman et al. (2004).