References and Literature

The neoclassical growth model goes back to Frank Ramsey’s (1928) classic treatment and for that reason is often referred to as the “Ramsey model”. Ramsey’s model is very similar to standard neoclassical growth model, except that it did not feature discounting.

All growth and macroeconomic textbooks cover the neoclassical growth model. Sargent and Ljungqvist (2004, Chapter 14) provides an introductory treatment of the neoclassical growth model in discrete time. Barro and Sala-i-Martin (2004, Chapter 2) provides a de­tailed treatment focusing on the continuous-time economy. Blanchard and Fisher (1989, Chapter 2) and Romer (2006, Chapter 2) also present the continuous-time version of the neoclassical growth model. These books use the necessary conditions implied by the Maxi­mum Principle, including the strong version of the transversality condition, and characterize utility-maximizing consumption behavior. The implicit assumption is that any other paths of consumption and assets are either less desirable or violate the no-Ponzi condition. As discussed in the previous chapter, some more care is necessary in applying the Maximum Principle in the presence of the no-Ponzi condition and also it is important to verify that the necessary conditions in fact characterize a global optimum. The difficulty is that these necessary conditions do not apply at the boundaries, thus the argument for ruling out paths that reach zero consumption is not fully rigorous. Instead of focusing on necessary conditions, I used the sufficient conditions in Theorem 7.14 from the previous chapter.

Ricardian Equivalence discussed in Exercise 8.33 was first proposed by Barro (1974). It is further discussed in Chapter 9.

A systematic quantitative evaluation of the effects of policy differences is provided in Chari, Kehoe and McGrattan (1997). These authors follow Jones (1995) in emphasizing differences in the relative prices of investment goods (compared to consumption goods) in the Penn Worlds tables and interpret these as due to taxes and other distortions. This interpretation is not without any problems. In particular, in the presence of international 348

trade, these relative price differences will reflect other technological factors or possible factor proportion differences (see Chapter 19, and also Acemoglu and Ventura, 2002, and Hsieh and Klenow, 2006). Parente and Prescott (1994) use an extended version of the neoclassical growth model (where the “stock of technology,” which is costly to adopt from the world frontier, is interpreted as a capital good) to perform similar quantitative exercises. Other authors have introduced yet other accumulable factors in order to increase the elasticity of output to distortions (that is, to reduce the α parameter above). Pete Klenow has dubbed these various accumulable factors introduced in the models to increase this elasticity the “mystery capital” to emphasize the fact that while they may help the quantitative match of the neoclassical-type models, they are not directly observable in the data.

8.14.