Protectionist policies and diversity of entrepreneurial types

Abstract

In recent years, there have been vigorous debates on whether restricting the operations of foreign multinational enterprises (MNEs) through various protectionist policies would allow local entrepreneurs to flourish. Research suggests that, although knowledge spillovers from MNEs can provide positive impetus to local entrepreneurship, MNE may also crowd out entrepreneurial firms. In this study, we examine how policies restricting MNEs’ entry affect local entrepreneurship, especially the diversity of the new firms being created. Using an agent-based simulation of an ecosystem, we model the dynamic interplay between MNEs’ knowledge spillover and diversity of local entrepreneurship and how two protectionist policies—tariffs on imports and subsidies for local entrepreneurs—impact this dynamic. We develop a fine-grained understanding of how such policies can both enhance and constrain different types of local entrepreneurship.

Appendix. Agent-based model for local economies

The agent-based NetLogo model used in this paper can also be represented using dynamics systems approach. In the Appendix, we show an equivalent formal two-equation system dynamics approach. As a variant of Lotka-Volterra model (Takeuchi 1996), the model represents a typical dynamical system well suited for our model which has been used in prior work studying the relationship between knowledge spillovers and innovation (Parker 2010). Adopting a similar model with slight changes to account for heterogeneity in competitive interactions between various types of local entrepreneurs and MNEs, we arrive at the following dynamic structure for change in population of the four different types of local entrepreneurs. If λ_i represents the population size of local entrepreneurs for a particular type i, K denotes the knowledge spillovers from the foreign MNEs, and T and S denote tariffs and subsidies, then the proportion of local entrepreneurship varies in a manner similar to a competitive Lotka-Volterra rate of change (Takeuchi 1996)

$$ \frac{\delta \lambda i}{\delta t}={r}_i{\lambda}_i\left(1-\frac{\sum \limits_{j=1}^Ng\left(K,{\alpha}_{ij},\mu, T,S\right){\lambda}_j}{K}\right) $$

$$ \frac{\delta K}{\delta t}={p}_iK\left(1-\frac{\sum \limits_{j=1}^Nf\left(K,{\beta}_j,\mu, T,S\right){\lambda}_j}{K}\right) $$

The type of local entrepreneurship in an economy varies as a logistic growth curve (Griliches 1959) with a steady growth rate r. The effect of foreign MNEs on the population of different types of local entrepreneurs’ g (K, α_ij, u, T, S) is a function of total spillovers and the local ACAP and the level of tariffs and subsidies that can go from 0 to 1. Similarly, the effect on the amount of knowledge spillovers for a focal foreign MNE f (K, B_j, u, T, S) is a function of the relative populations of different types of local entrepreneurship, the local absorptive capacity μ, and the level of tariffs T and subsidies S. The two sets of dynamical equations represent mutual interactions between different autonomous agents. The effect of each agent on other agents in the model is captured by the matrix α_ij, and these can be negative or positive. The effect of local entrepreneurship on foreign MNEs is captured by the vector B_j. The functional forms f and g can be assumed to be linear.

1.1 NetLogo 6.0 Base Code