References and Literature
The baseline model of Schumpeterian growth presented in Section 14.1 is based on the work by Aghion and Howitt (1992). Similar models have also been developed by Segerstrom, Anant and Dinopoulos (1990), Grossman and Helpman (1991a,b).
Aghion and Howitt (1998) provide an excellent survey of many Schumpeterian models of economic growth and numerous extensions. The specific modeling assumptions made in the presentation here draw on Acemoglu (1998), which also uses the aggregate production function with proportional quality improvements and costs of production and R&D increasing proportionally with quality. The original Aghion and Howitt (1992) approach is very similar to that used in Section 14.2.Aghion and Howitt (1992) also discuss uneven growth and potential growth cycles, which were presented in Section 14.2. Uneven growth and cycles are also possible in expanding product or input variety models as shown by Matsuyama (1999, 2001). I only discussed the possibility of such cycles in the context of Schumpeterian growth, since the forces leading to such cycles are more pronounced in these models.
The effect of creative destruction on unemployment is first studied in Aghion and Howitt (1994). The implications of creative destruction for firm-specific investments are discussed in Francois and Roberts (2001) and in Martimort and Verdier (2003).
The model in Section 14.3 draws on Acemoglu (2008b) and is a first attempt to introducing productivity growth driven both by incumbents and entrants (see also Barro and Sala-i- Martin, 2004, for a model in which incumbents undertake R&D because they have cost advantage). A related paper is Klette and Kortum (2004). Klette and Kortum construct a richer model of firm and aggregate innovation dynamics based on expanding product varieties. Their key assumption is that firms with more products have an advantage in discovering more new products.
With this assumption, their model generates the same patterns of firm growth as the simple model and Section 14.3, and also matches certain additional facts about propensity to patent and the differential survival probabilities of firms by size. One disadvantage of this approach is that the firm size distribution is not driven by the dynamics of continuing plans (in fact, if new products are interpreted as new plants, the Klette-Kortum model predicts that all productivity growth will be driven by entry of new plants, though this may be an extreme interpretation, since some new products may be produced in existing plants). Lentz and Mortensen (2006) extend Klette and Kortum’s model by introducing additional sources of heterogeneity and estimate this extended model on Danish data. Klepper (1996) documents various facts about the firm size, entry and exit decisions and innovation, and provides a simple descriptive model that can account for these facts. None of these papers consider a Schumpeterian growth featuring innovation both by incumbents and entrants that can be easily mapped to decomposing the contribution of new and continuing plants (firms) to productivity growth.Step-by-step or cumulative innovations have been analyzed in Aghion, Harris and Vickers (1999) and Aghion, Harris, Howitt and Vickers (2001). The model presented here is a simplified version of Acemoglu and Akcigit (2006), which includes a detailed analysis of the implications of intellectual property rights policy and licensing in this class of models. The proof of existence of a steady-state equilibrium under a somewhat more general environment is provided in that paper. The notion of Markov Perfect Equilibrium used in Section 14.4 is a standard equilibrium concept in dynamic games and is a refinement of subgame perfect equilibrium, that restricts strategies to depend only on payoff-relevant state variables. This concept is further discussed in the Appendix Chapter C and in Fudenberg and Tirole (1994).
Blundell (1999), Nickell (1999) and Aghion, Bloom, Blundell, Griffith and Howitt (2005) provide evidence that greater competition may encourage economic growth and technological progress. The latter paper shows that industries where the technology gap between firms is smaller are typically more innovative. Aghion, Harris, Howitt and Vickers (2001) and Aghion, Bloom, Blundell, Griffith and Howitt (2005) show that in step-by-step models of innovation greater competition may increase growth. Aghion, Dewatripont and Ray (2000) provide another reason why competition may encourage growth. In their model competitive pressure improves managerial incentives and efficiency.
14.7.