We review the theoretical links between growth and agglomeration. Growth, in the form of innovation, can be at the origin of catastrophic spatial agglomeration in a cumulative process a la Myrdal. One of the surprising features of the Krugman [Journal of Political Economy 99 (1991) 483-499] model, was that the introduction of partial labor mobility in a standard "new trade model" with trade costs could lead to catastrophic agglomeration. The growth analog to this result is that the introduction of endogenous growth in the same type of "new trade model" can lead to the same result. A difference with the labour mobility version is that the results are easier to derive from the analytical point of view in the endogenous growth version. We show that the relation between growth and agglomeration depends crucially on capital (human or physical) mobility between regions. The absence of capital mobility is at the heart of the possibility of spatial agglomeration with catastrophe. In addition, growth alters the process of location even without catastrophe. In particular, and contrary to the fundamentally static models of the New Economic Geography, spatial concentration of economic activities may be consistent with a process of delocation of firms towards poor regions. Finally, the presence of localized technology spillovers implies that spatial agglomeration is conducive to growth.
