![Figure 3: Average growth rate and technology level with different values of thresholds [γ1, γ2].](/figures/figure-3-average-growth-rate-and-technology-level-with-d74tul60.webp)
Q2. What is the definition of cluster volatility?
The cluster volatility is a well known phenomenon in the financial market literature (see Cont, 2007), and implies that large changes in variable values occur preferably at neighboring times, reflecting the tendency for markets to move from stable to more turbulent periods.
Q3. What is the reason for the increase in the average technology level?
In particular, when the State funds all the innovating firms or jointly the group of isolate and collaborative firms, the average technology level increases more than linearly with σ.
Q4. What is the probability of a firm becoming a standalone innovator?
Firms with a probability higher than Z̄t = (1−γ1)ZMaxi,t would try to become standalone innovators, those with a probability lower than Zt = (1− γ2)ZMaxi,t are willing to become imitators (with 0<γ1, γ2<1 being parameters), while are collaborative innovators all the other ones (i.e., those with Zt < Zi,t < Z̄t).
Q5. What is the evidence of a decay in the time series of the aggregate output?
The share of failures is quite constant during the simulation, consequently, a decay in the time series of the aggregate output can be interpreted as caused by the simultaneous failure of relatively large firms (see Fig. 5, right side).
Q6. How many simulations of the model are there?
The following results reproduce the outcome of 100 simulations of the model with increasing levels of the R&D expenditure parameter σ, starting from 0% to 50% with steps of 0.5%.
Q7. What is the effect of the power law on the average technology level of the economy?
the average technology level of the economy can register a large fall because of the lost of the reached technology knowledge, dissipated with the failing firm.
Q8. What is the evidence of the logarithm of firm growth rates?
in line with other empirical works (Amaral et al., 1997; Bottazzi et al., 2001;10/21Fagiolo and Luzzi, 2006), the authors show that the probability distribution of the logarithm of firm growth rates is tent-shaped and can be fitted by an asymmetric Laplace distribution (double exponential), whose tails decay much slower than in a Gaussian distribution (see Fig. 4 (right side)).
Q9. How can a multiplicity of heterogeneous agents generate economic regularities?
following the agent-based computational economic approach (Colander et al., 2008; Tesfatsion and Judd, 2005), the authors have proved that a multiplicity of interacting heterogeneous agents, whose decisions are determined by evolving decision rules, can generate economic regularities without resorting to any full rationality of a Bayesian representative-agent.16/21
Q10. What are the main characteristics of the empirical regularities?
The empirical regularities, as a relatively stable skewed firm size distribution (Axtell, 2001; Gaffeo et al., 2003), the Laplace distribution of firms’ growth rates (Stanley et al., 1996; Bottazzi and Secchi, 2003), the firms’ heterogeneity with respect to employed technology (Silverberg and Verspagen, 2005) and others important growth type stylized facts (Kaldor, 1961; Audretsch, 1997) are, instead, well reproduced by agent-based models.
Q11. How many people are in the group of isolated innovators?
12/21a probability of belonging to the group of isolated innovators, 8% to the group of collaborative innovators and 90% to the imitators.