Dosi y Otros 2023 Mission-Oriented Policies and The "Entrepreneurial State" at Work - An Agent-Based Exploration

Journal of Economic Dynamics & Control 151 (2023) 104650
Contents lists available at ScienceDirect
Journal of Economic Dynamics & Control

journal homepage: www.elsevier.com/locate/jedc
Mission-oriented policies and the “Entrepreneurial State” at

work: An agent-based exploration
Giovanni Dosi a, Francesco Lamperti a,b,∗, Mariana Mazzucato c,
Mauro Napoletano d,e,f, Andrea Roventini a,e
a
Institute of Economics and EMbeDS, Scuola Superiore Sant’Anna, Pisa, Italy
b
RFF-CMCC European Institute on Economics and the Environment, Milan, Italy
c
Institute for Public Purpose and Policy, University College London London, UK
d
Université Côte D’Azur, CNRS, GREDEG, France
e
Sciences Po, OFCE, France
f
Scuola Superiore Sant’Anna, Pisa, Italy
a r t i c l e i n f o a b s t r a c t
Article history: We study the impact of alternative innovation policies on the short- and long-run perfor-
Received 6 August 2021 mance of the economy, as well as on public finances, extending the Schumpeter meeting
Revised 16 December 2022
Keynes agent-based model (Dosi et al., 2010). In particular, we consider market-based in-
Accepted 26 March 2023
novation policies such as R&D subsidies to firms, tax discount on investment, and direct
Available online 30 March 2023
policies akin to the “Entrepreneurial State” (Mazzucato, 2013), involving the creation of
JEL classification: public research-oriented firms diffusing technologies along specific trajectories, and fund-
O33 ing a Public Research Lab conducting basic research to achieve radical innovations that
O38 enlarge the technological opportunities of the economy. Simulation results show that all
O31 policies improve productivity and GDP growth, but the best outcomes are achieved by ac-
O40 tive discretionary State policies, which are also able to crowd-in private investment and
C63 have positive hysteresis effects on growth dynamics. For the same size of public resources
allocated to market-based interventions, “Mission” innovation policies deliver significantly
Keywords:
Innovation policy better aggregate performance if the government is patient enough and willing to bear the
Mission-oriented R&D intrinsic risks related to innovative activities.
Entrepreneurial state
© 2023 Published by Elsevier B.V.
Agent-based modelling
1. Introduction
In this paper, we extend the Schumpeter meeting Keynes agent-based model (Dosi et al., 2010) to assess the impact of
different innovation policies on the short- and long-run performance of the economy, as well as on the public budget.
The stagnating aftermaths of the Great Recession and, more recently, of the COVID-19 pandemics, call for public policies
able to restore robust economic growth. Such crises also exacerbated the pre-existing productivity slowdown experienced
by most developed economies. This implies that government should introduce policies to influence the pace of innovation
and technological change, which are the major drivers of long-run economic growth. The Next Generation EU program
released by the European Commission goes explicitly in this direction. However, in our view, the contemporary discourse on
∗
Corresponding author at: Institute of Economics, Scuola Superiore Sant’Anna, piazza Martiri della Libertà 33, 56127 Pisa, Italy.
E-mail addresses: [email protected] (G. Dosi), [email protected] (F. Lamperti), [email protected] (M. Mazzucato),
[email protected] (M. Napoletano), [email protected] (A. Roventini).
https://doi.org/10.1016/j.jedc.2023.104650
0165-1889/© 2023 Published by Elsevier B.V.
G. Dosi, F. Lamperti, M. Mazzucato et al. Journal of Economic Dynamics & Control 151 (2023) 104650
innovation policies has been far too narrow, quite disjoint from their implications for the economic and social future of our
societies. In fact, it is remarkable that, in the past, some of the most important “innovation policies” were not called as such.
The Manhattan Project, the Apollo Program, Nixon’s “war on cancer” were not discussed, if at all, as “policies” but as major
societal objectives, well shielded from the narrow concerns of economists’ cost-benefit analyses. On the contrary, nowadays,
innovation policies - except for war-related innovations and pandemic emergencies - have to pass through the dire straits
of efficiency criteria. However, even on these narrower grounds, we shall show, innovation policies are well worth.
Innovation policies (written large, and meant to include science and technology policies) broadly refer to the design
of a variety of instruments aimed at generating new knowledge, new products and more efficient production techniques
(within an enormous literature, see from Bush et al., 1945 to Criscuolo et al., 2020; Edler and Fagerberg, 2017; Freeman
and Soete, 1997). Depending on the type and scope of the policy tools employed, innovation policy might require more or
less extensive involvement of the public sector in the economy. A broad distinction is between indirect and direct innovation
policies (Dosi, 1988; Dosi and Nelson, 2010; Mazzucato and Semieniuk, 2017). Indirect policies tend to be “market-friendly”
as they provide monetary incentives to firms to improve their innovative performance (e.g. R&D subsidies) or to speed-up
their technological renewal (e.g., investment tax discounts). In an influential debate at the OECD in the early 80s, they were
called “diffusion-oriented” policies (Ergas, 1987). Differently, direct innovation policies imply an active of role of the public
sector in shaping the rates and directions of innovative activities, which means - to paraphrase Nelson (1962) - shaping
technological landscape and search regimes, taking risks that private businesses are not willing to sustain, and pursuing
path-breaking technological developments. Direct innovation policies respond to Freeman (1987) plea for policies creating
systems and institutions able to nurture the generation and diffusion of new knowledge across the economy, the creation
of new industries and markets and - ultimately - to fuel economic growth. These policies may certainly be facilitated by
an Entrepreneurial State (Mazzucato, 2013) that takes the lead and directly invests in the search for novel technological
opportunities (possibly directed to specific missions; see also Mazzucato, 2018a and Mazzucato, 2021).
The ability of alternative innovation policies to spur innovation, crowd in private investment and deliver sustained long-
run growth is highly debated. Notwithstanding a large body of studies evaluating single policies (see Becker, 2015, for a
survey), systematic comparisons of policy designs are scarce in the literature (Grilli et al., 2018), especially from a macroe-
conomic perspective (Di Comite and Kancs, 2015). A recent review by Bloom et al. (2019) discusses pros and cons of various
instruments, suggesting a trade-off between the short run, where tax incentives and subsides are effective in stimulating
innovation, and long run outcomes, which would benefit from systemic investments in universities and education. However,
Bloom and co-authors overlook (or dismiss) direct policies, based on the argument that the effects of these policies are hard
to be identified econometrically. In addition, those policies, it is suggested, lack an economic rationale - of course in terms
of the conventional economic theory, according to which were it not for market failures and externalities, one better leave
the market and the search for innovations to itself.
In this work, we shall indeed show the robust rational of direct policies in complex evolving economies. We extend the
Schumpeter meeting Keynes (K+S) macroeconomic agent-based model (Dosi et al., 2010) to systematically compare the impact
of direct and indirect innovation policies on economic performance via their stimulus to incremental and radical innovation,
while accounting for their impact on the public budget.1 In that, the paper also contributes to the literature about modelling
of R&D, innovation activities and their impacts on the macroeconomy, integrating the representation of technological change,
its sources and consequences within an agent-based perspective (for germane contributions see Caiani et al., 2019; Dawid
et al., 2008; Dosi et al., 2019; Fagiolo et al., 2020; Gräbner and Hornykewycz, 2022; Lorentz et al., 2016; Russo et al., 2007,
the survey in Dawid, 2006 and the recent critical review by Aistleitner et al. (2021), wherein multiple modeling approaches
are discussed). Indeed, we believe that a first-order systematic comparison between “Entrepreneurial State”-like policies and
price-based R&D incentives is missing in the literature linking macroeconomic dynamics and technical change, and it would
be better carried out abstracting from choice of the particular sectors and missions to target (which is highly arbitrary and
possibly affected by political considerations), though keeping vivid the spirit of mission orientation (Mazzucato, 2013).
The K+S model is composed of two vertically-related sectors, wherein heterogeneous firms strive to develop new tech-
nologies and locally interact by exchanging capital-goods in a market with imperfect information. This is the Schumpeterian
engine of the model: new machine tools are discovered and diffuse within the economy both via imitation activities of
competing capital-good producers and via investment by consumption-good firms. Firm investment depends on firm de-
mand expectations, as well as on their financial conditions and it constitutes, together with worker consumption and public
expenditures, the Keynesian soul of the model. Aggregate demand dynamics in the model affects not only business cycles,
but also the pace of technological change (see e.g. Dosi et al., 2016). The K+S model is therefore able to go beyond the
traditional separation between “coordination” and “change” in economics (Dosi and Virgillito, 2017).
Indeed, the K+S family of models represents flexible environments which can be used as virtual laboratories for policy
experiments to investigate a variety of policy interventions and perform counter-factual analyses. We examine four stylized
innovation policy regimes and their possible combinations, namely (i) R&D subsidies to capital-good firms; (ii) tax-discounts
on consumption-good firms’ investments; (iii) the creation of a public research-oriented capital-good firm; (iv) the insti-
tution of a National Research Laboratory which tries to discover radical innovations that enlarge the set of technological
1
Agent-based models are particularly suited to evaluate different combinations of policies in frameworks characterized by deep uncertainties, technical
and structural change. More on that in Dawid and Delli Gatti (2018); Dosi and Roventini (2019); Fagiolo and Roventini (2017). We also suggest to look at
Dosi et al. (2020c) for a systematic comparison of market-based and industrial policies in fostering catching-up.
2
opportunities available in the economy. The first two experiments mimic indirect innovation policies, while the latter pair
captures key features of direct or “Entrepreneurial-State” policies, leaving aside the decision of which mission(s) to target.
Finally, we consider a benchmark scenario where the public resources is used to support private consumption instead of
innovation policies.
Simulation results show remarkable differences across innovation policy regimes. First, all innovation policies spur pro-
ductivity and GDP growth, but to different degrees, while this is not the case for transfers to households. Second, the impact
of direct innovation policies is larger vis-à-vis indirect ones and entails effects of positive hysteresis (Cerra et al., 2021; Dosi
et al., 2018) putting GDP on higher growth trajectories. However, Entrepreneurial-State policies are risky: their positive im-
pact tend to show up on longer time horizons as compared with indirect interventions, and they can fail to discover new
technologies. Nonetheless, extensive Monte Carlo analyses show that, on average, direct innovation policies deliver higher
productivity and GDP growth, while being less expensive in terms of net public resources, compared to “indirect” forms of
intervention. The impact of Entrepreneurial-State interventions is stronger when they combine the presence a public firm
with a National Research Laboratory. Conversely, indirect monetary incentives tend to be associated with some redundancy
- that is transfer of resources to firms with little effect on the intensity of search. Finally, all innovation policies we consider
crowd in private R&D investment (in line with Moretti et al., 2019 and Pallante et al., 2020), although direct interventions
provide, again, the most bang for their buck. Accordingly, our results suggest that the type of tools utilised by a mission-
oriented Entrepreneurial State (Mazzucato, 2013; 2018a; 2021) are also more effective at meeting uncontroversial innovation
policy goals of productivity and growth gains.
To sum up, our results indicate that innovation policies are highly effective. In particular, when public resources are
concentrated on clear missions and Entrepreneurial-State interventions, they appear to deliver large gains in economic per-
formance compared to policies based on monetary incentives. This should be taken into account by policy makers when
designing vast policy plans such as the Next Generation EU to jump-start growth in economies hardly hit by the COVID-19
crisis.
The rest of the paper is organized as follows. Section 2 provides a critical overview of the literature on innovation policies.
In Section 3, the K+S model is introduced. The empirical validation of the model is performed in Section 4. In Section 5, we
present the results of innovation policy experiments. Finally, Section 6 concludes the paper.
2. Innovation policies: A critical review
Economic theory identifies innovation as the most relevant driver of industrial development, specialization and long-run
economic growth. This holds true both in neoclassical (Aghion and Howitt, 1992; Romer, 1986; Solow, 1957) and evolution-
ary theories (Dosi, 1982; Dosi et al., 1994; Nelson and Winter, 1982). However, the underlying views about how knowledge
evolves, accumulates, diffuses and - ultimately - affects productivity are profoundly different across these two theoretical
paradigms (see Dosi, 1988; Dosi and Nelson, 2010, among others). Such differences also often map into opposing prescrip-
tions with respect to innovation policy.
We define innovation policies, to repeat, as the set of attempts carried out by a government to shape or influence the
generation and diffusion of new knowledge and new technologies. All this can be implemented either via monetary instru-
ments, regulations or direct interventions, often but not always with the purpose of increasing productivity and economic
growth. Some other times, they can be just be an unintended consequence of policies meant to achieve other purposes -
e.g. winning a war (Gross and Sampat, 2020; Moretti et al., 2019). But what motivates innovation policies themselves?2
The market view closely based on a neoclassical perspective justifies policies with the presence of market failures or
untraded externalities. Building on these premises, innovation policies should be designed as monetary incentives align-
ing private behaviors and delivering the socially optimal level of innovation. This implicitly shrinking the scope for a more
active role of the State in shaping the technological landscape. Indeed, assuming in a first approximation the equivalence
between technological knowledge and information, the latter has an intrinsic public-good nature, implying an endemic ten-
dency to underinvest in expensive activities of search by private profit-motivated agents (Arrow, 1951; 1962), which can be
mitigated by various forms of transfers and incentives. Another way to partially align private actors’ incentives to innovate
and social objectives - which overstretches the implications of Arrow’s argument - entails the deepening and strengthen-
ing of Intellectual Property Rights (IPR), thus supposedly increasing the equilibrium rates of allocation to R&D investments.
Though appealing as they are for their simplicity and highly influential among policy makers (EU, 2020; OECD, 2010) and
economists alike, these arguments build on weak foundations and - we argue - should not limit the innovation policy debate
to the design of tax credits, subsidies and IPRs (see, e.g., Bloom et al., 2019; Guellec and Van Pottelsberghe De La Potterie,
2003).
On the theoretical side, the focus on market-based instruments to conduct innovation policy is postulated on the role
general equilibrium effects under “complete markets”. However, there is a fundamental incompatibility between innova-
tion and general equilibrium, basically for two reasons (Cimoli et al., 2009; Stiglitz, 1996). First, if an innovation is a true
innovation, one cannot know about it ex ante, otherwise it would not be an innovation: therefore it is also impossible to at-
tribute probabilities to its occurrence, let alone having “rational expectations” about them and their mapping into expected
2
The interested reader can find critical surveys of innovation policy instruments in Bloom et al. (2019); Borrás and Edquist (2013); Edler and Fager-
berg (2017) and Mazzucato (2016).
3
costs. Thus, markets must be incomplete by definition. Second, the very presence of technological knowledge (Arrow calls it
“technical information”) implies an extreme form of increasing returns and thus ubiquitous non-convexities, multiple equi-
libria, or non- existence of equilibrium at all (see Arrow, 1996 and the comments by Arrow in Teece, 2019). Of course, with
that disappears all the welfare properties of general equilibrium, taken for granted in “market failure” evaluations. Further,
the equivalence between knowledge and information is just a first rough approximation: while information can be easy to
access, the same does not necessarily hold for knowledge. Not all knowledge can be codified: much economically useful
knowledge is tacit and heterogeneously distributed across actors and contexts (Dosi, 1988; Dosi and Nelson, 2010; Metcalfe,
2005; Nelson and Winter, 1982; Polanyi, 1944; Winter, 1998).
More generally, the empirical evidence supporting any link between incentives and the propensity to generate or acquire
knowledge is at best fuzzy. First, the available evidence backing the effectiveness of monetary subsidies and stronger IPR
regimes to stimulate private R&D spending is rather weak (Dimos and Pugh, 2016; Papageorgiadis and Sharma, 2016; Zúñiga
Vicente et al., 2014), despite the fact that these policies typically entail large fiscal costs. Indeed, firms might tend to keep
their R&D steady and simply exploit public subsidies and tax-credits to boost their profits (Marino et al., 2016; Mohnen et al.,
2017). Second, stronger IPRs might not matter significantly in firm-level decisions and can even decrease the long-run pace
of innovation (Cimoli et al., 2014; Dosi et al., 2006; Dosi and Stiglitz, 2014; Stiglitz, 2014). For example Levin et al. (1987),
Fagerberg (2017) and Cohen (2010) show that in most industries firms are not much concerned about the lack of strong IPR
as the capabilities underpinning their innovative performance cannot be copied easily (Dosi and Nelson, 2010; Edler and
Fagerberg, 2017). On the contrary, many firms have close interactions and knowledge exchanges with relevant parties (e.g.,
customers, suppliers, universities, public research institutions, etc.) which nurture the transfer of tacit knowledge during the
innovation process.
Furthermore, the market failure approach is hardly useful when radical technological change is needed (see
Mazzucato, 2016). Private businesses tend to invest in new technologies only after the high risks and uncertainty have been
absorbed by research and development activities directly funded by the public sector. In this case, mission-oriented policies
are needed to create new technologies, new sectors and new markets (Foray et al., 2012). Such innovation policies consider
the public sector as an Entrepreneurial State mostly engaged in industry creation and market shaping rather than market
fixing, actively setting new innovation directions towards significant social goals (missions). Indirectly, the scope for active,
direct innovation policies is further justified by a recent stream of studies suggesting that long-term growth is facilitated by
the development of complex products, i.e. tradable artifacts closely related to many economic activities (Hidalgo et al., 2007;
Tacchella et al., 2012), to which innovation policies might aim at, given the locally available technological competences. The
idea of market shaping and mission-orientation has began to gained acceptance in recent years in Europe where it seems to
be adopted by the European Commission - in relation to grand societal challenges such as the green transition (Mazzucato,
2018b; 2019). This finally reflects disappointment in the ability of market fixing approaches to address these challenges and
recognition that the appetite for risk, long term thinking and capacity for coordination in the private sector is inadequate
for producing a decisive shift in the direction of innovation (Mazzucato and Semieniuk, 2017; 2018).
Beyond the selection of the missions to pursue, which reflects broader societal and political objectives, the En-
trepreneurial State approach to innovation policy can be summarized across three defining features (Mazzucato, 2016). First,
public organizations should experiment, conduct research, learn and take risks. Second, policy design should create symbi-
otic private-public partnerships, overtaking the idea of de-risking the private investment and fostering a collaborative envi-
ronment, characterized by joint R&D projects to create new products and services (e.g., new vaccines; Chataway et al., 2007),
and crowding-in of private investment (see for example Engel et al., 2016; Moretti et al., 2019; Pallante et al., 2020). Finally,
it should provide a system of rewards for the public sector to ensure the long run sustainability of the high risk-taking in-
vestments described above, as well as for public accountability purpose. Along these lines, public policies should support all
phases of the innovation process, taking risks (and possible losses) that the private sector will not absorb, waiting patiently
for the rewards of innovation and coordinating activities across public and private stakeholders (Mazzucato, 2013).
Perhaps less widely acknowledged is the economic case for a mission-oriented Entrepreneurial State. The economic im-
pacts from such policies are often hard to quantify empirically (Bloom et al., 2019), being them associated with dynamic
spillovers, even when the social ones are quite obvious. A priori, we would expect Entrepreneurial State policies to have
high potential for generating growth due to the fact they target new markets, technologies and directions of discovery. This
means they have also the potential to create opportunities for advancement in productivity, consumer demand, international
competitiveness and so forth, which would not be created by the private sector alone (Mazzucato, 2013; 2018a). A recent
assessment of the US moonshot program implemented during the 1960s finds large, first order and long lasting growth ef-
fects induced by public R&D conducted through the NASA (Kantor and Whalley, 2022). However, novel opportunities come
together with additional risks, and it is not automatic that missions-oriented programs translate into success stories; to the
contrary, for every winning mission-oriented investment there are many possible failures (Mazzucato, 2016). The ultimate
result likely depends both on the uncontrollable uncertainty of the search process itself (Dosi, 1988) - which may well re-
veals unsuccessful - and on the politically controllable willingness of the government to sustain eventual costs and losses
during the research path.
The historical record provides compelling cases in support of Entrepreneurial State interventions. Governments invested
directly in the technologies that enabled the emergence of mass production and IT revolutions and undertook the bold
policies required to deploy them throughout the economy (Block and Keller, 2015; Ruttan, 2006). Many of the examples
of this relate to the pervasive impact of military and space innovation (the Manhattan project, the Apollo program and
4
ARPANET - the progenitor of the internet - are among the most famous; see Gross and Sampat, 2020; Gross and Sampat,
2021) but, more recently, successful results have been highlighted across many other technological landscapes, including
the biotechnology industry (Lazonick and Tulum, 2011), nanotechnologies (Motoyama et al., 2011), and the emerging clean-
tech sector (Mazzucato, 2015; Steffen et al., 2020). To the contrary, failed missions exist and the literature’s focus on case
studies has often tended to highlight successes and provide too few examples of failures, which may lead to overstate
the positive consequences and downplay the negative effects of mission-oriented R&D programs (Mowery, 2010). Indeed,
dramatic flops have been reported for energy, defense and aerospace related programs (Mazzucato, 2016; Mowery, 2010;
Nelson, 1982a). Even though it is empirically difficult to balance the economic significance of failures against successes,
often because of missing counterfactual cases (Mowery, 2010), this paper attempts at contributing to such assessment by
means of a simulation laboratory.
One of the implications of the “market failure approach” is that it calls for the state to intervene as little as possible in
the economy, in ways that minimize the risk of “government failures”, whatever that means in complex evolving economies.
A corollary is also the drive to outsource the innovation process from public organizations to private firms.
To the contrary, a mission-oriented Entrepreneurial State aims at shaping the direction of technological change, employ-
ing a mix of indirect instruments (schemes of incentives) and, much more important, direct interventions (e.g. through
public agencies, formal public-private collaborations, use of public banks to finance bold R&D projects), and coordinating
the governance of the whole innovation chain.3 Under this perspective, the State should not limit itself to provide funding
for basic knowledge and help protecting innovation through implementation of IPRs, as the market failure theory would
suggest, but also identify and rectify such systemic problems coordinating all levels of public administration and private
stakeholders (Edquist, 2011; Metcalfe, 1994; 1995).
3. The K+S model
We compare a variety of innovation policies and their economy-wide effects in the Schumpeter meeting Keynes model
extended to account for radical innovations and the variable cost of public debt (Dosi et al., 2013; 2010).4 Our stylized
representation of an economy is composed of a machine-producing sector composed of F1 firms, a consumption-good sector
composed of F2 firms, an ecology of consumers/workers, and a public sector. Capital-good firms invest in R&D and produce
heterogeneous machines. Consumption-good firms combine machine tools bought by capital-good firms and labour in order
to produce a final product for consumers. The public sector levies taxes on firms’ profits, pay unemployment benefits, and
implement the selected innovation policies.
3.1. Innovation and technological progress
The Schumpeterian engine of the K+S model stems from the innovation and imitation search of capital-good firms, which
produce machine-tools using labour only. The technology of the machines of vintage τ is captured by the couple of coef-
ficients (Ai,τ , Bi,τ ), where the former represents the productivity of machines employed in the consumption-good industry,
while the latter indicates the productivity of the production technique needed to manufacture the machine. Given the mon-
etary wage, w(t ), paid to workers, the unitary cost of production of capital-good firms is given by:
w(t )
cicap (t ) = . (1)
Bi,τ
Similarly, the “quality” of the machines captured by (Ai,τ ) defines the unitary production cost of consumption-good firms
(indexed by j):
w(t )
j (t ) =
ccon . (2)
Ai,τ
Capital good firms adaptively strive to increase market shares and profits trying to improve their technology via in-
novation and imitation. These processes reflect the R&D activities performed by the firm. In line with Nelson and Win-
ter (1982) and Nelson (1982b), we conceptualize R&D as a stochastic search process in which all firms in an industry face
an identical distribution of outcomes. Both innovation and imitation are costly processes: firms invest in R&D a fraction of
their past revenues in the attempt to implement incrementally new technologies, discover radically new innovations and
imitate more advanced competitors.5 Although these outcomes may differ across firms, all firms choose to undertake the
3
In this respect, various similarities are shared with the so-called “system-oriented” innovation policies (Edler and Fagerberg, 2017), which builds on
the literature on National (Freeman, 1987; Lundvall, 1988; 2010) and Sectoral (Malerba, 2002) Innovation Systems and looks at the systemic nature of the
innovation process as emerging from the interaction of a number of factors, including knowledge, skills, financial resources, demand etc. When the system
does not sufficiently provide for those factors or fails at coordinating them, a “system failure” may hamper innovation activity.
4
See also Dosi et al. (2017a) for a survey about the Schumpeter meeting Keynes family of models. Indeed, the K+S model has been extended to account
for multiple banks and fiscal-monetary policy trade-offs (Dosi et al., 2015), decentralized interactions in the labour market (Dosi et al., 2017, 2022) and the
coupled dynamics of climate climate and the economic growth (Lamperti et al., 2019; 2021; 2018a; 2020).
5
Of course, different modelling frameworks exist, for instance considering heterogeneous R&D intensities (e.g. Silverberg and Verspagen, 1994); we refer
the reader to Dawid (2006) for a broader discussion.
5
same relative amount of R&D (Nelson, 1982b). In full agreement with this assumption, Coad and Rao (2010) find that “firms
behave as if they aim for a roughly constant ratio of R&D to sales”, thereby adjusting R&D expenditures to experienced sales
growth.6 Hence,
RDi (t ) = υ Si (t − 1 ), υ ∈ {0, 1} (3)
indicates firm i’s spending in R&D, which is split into in-house (incremental) innovation (INi ) and imitation (IMi ) activities:
INi (t ) = ξ RDi (t ), IMi (t ) = (1 − ξ )RDi (t ), ξ ∈ [0, 1]. (4)
In the Appendix, we test the robustness of our findings with respect to alternative routines describing a firm’ R&D spending
decision, in particular allowing search intensity to depend on the (i) relative size of its innovation effort (Cohen and Klep-
per, 1992) and (ii) the sign of the growth previously experienced (Coad and Rao, 2010). As in Dosi et al. (2010), innovation
and imitation are depicted as two-steps processes. The first step captures firms’ search for new technologies through a draw
from a Bernoulli distribution, wherein the real amount invested in R&D (i.e. the number of hired researchers) positively
affects the likelihood of success. More precisely, the parameters controlling the likelihood of success in the Bernoulli trial
for the innovation and imitation process, θ IN (t ) and θ IM (t ) respectively, correspond to:
θiIN (t ) = 1 − e−oIN INi (t ) , oIN > 0, (5)
θiIM (t ) = 1 − e−oIM IMi (t ) , oIM > 0; (6)
where the parameters 0 < −oIN , oIM 1 capture the search capabilities of firms.
The second step differs for innovation and imitation activities. Let us consider innovation first. Successfully innovating
firms will access a new technology, whose technical coefficients are equal to:
Ai,τ +1 = Ai,τ (1 + χA,i ) (7)
Bi,τ +1 = Bi,τ (1 + χB,i ) (8)
where χA,i and χB,i are independent draws from a Beta(α , β ) distribution over the support [ξ1,i , ξ2,i ], with ξ1 < 0 and ξ2 > 0.
The support captures the technological opportunities available for the firms. Note that as χ (t ) is allowed to be negative, the
newly discovered technology may be inferior to the current one. This reflects the intrinsic trial and error process associated
to any search for new technologies.
Successful imitators have the opportunity to copy the technology (embodied in the two technical coefficients A and B) of
one of their competitors. The imitation probability negatively depends on the technological distance between each pair of
firms. More precisely, the technological space is modelled as a 2-dimensional Euclidean space (A, B ), where 2 is chosen as
the metric determining distance between couples of points:

T Di, j = ( Ai − A j )2 + ( Bi − B j )2 , (9)
where the vintage of the technology employed by firms i and j is dropped to ease notation. For each imitator, competi-

tors are ranked according to their (normalized) technological distance NT Di, j = T Di, j / j T Di, j and a draw from a uniform
distribution on the unitary interval determines the firm whose technology will be imitated.
When a novel technology is developed or imitated, capital-good firms decide whether to adopt it or not by comparing
its overall costs through the following routine:
min[ phi (t ) + bccon,h ] h ∈ {in, im, τ }, (10)
where b is a payback parameter (more on that in Section 3.2), p is the price of the machine and c is the unitary production
cost a firm would incur in employing the imitated (im), newly developed (in) or available technology of vintage τ . Such
routine guarantees that capital-good firms try to improve their competitiveness by manufacturing a machine that reduces
the costs faced by their downstream clients. Once the machine to put in production is selected, capital-firms fix the price
as a constant mark-up on their unit cost of production. The capital-good market is characterized by imperfect competition:
capital-good firms advertise their product to their historical customers, as well as to a subset of potential new ones.
Beyond in-house incremental innovations and imitation, we allow for the discovery of radical innovations, which are in-
tended here as innovations that change the technological landscape and increase the technological opportunities available in
the economy. Examples of such radical innovations include electricity, energy storage and the Internet. Following Mazzucato
(2013), these innovations are rarely the outcome of a single research project within private businesses, but more likely de-
pend on a broader, systemic effort encompassing both public (from basic to applied) and private research, often carried out
through private-public collaborations and characterized by sequences of trials and errors (see also Block and Keller, 2015;
Mowery, 2010 and the discussion in Section 2). To capture these features, we model radical innovations as shifts of the
6
Fig. 1. Shift of technological opportunities implied by radical innovations.
support [ξ1 , ξ2 ] of the distribution of technological opportunities available to the firm at a given time (see also Fig. 1):
ξ1RI,i = ξ1,i + χ RI , ξ2RI,i = ξ2,i + χ RI , (11)
where ξ RI
indicate the extrema of the support after a successful radical innovation.
The probability of discovering a radical innovation depends positively on the cumulative R&D expenditures performed by

the capital-good firm (CRDi ) and by public research agencies (CRDpublic ). Private cumulative R&D, CRDi (t ) = t ∗ <s≤t RDi (s ),
∗
proxies the stock of knowledge generated by the firm over time, after the eventual discovery, at time t , of a previous
radical innovation. The probability (PiRI ) that a capital-good firm i discovers a radical innovation enlarging the technological
opportunities is then equal to:

CRDi (t ) + CRDpublic (t ) 1
PiRI (t ) = f x | x = = , (12)
GDP (t ) 1 + eη1 (η2 −x )
with η1 > 0 and η2 > 0 controlling the shape of the logistic function.7 Indeed, there is robust evidence supporting a non-
linear positive association between a sufficiently large stock of cumulated knowledge and the discovery of breakthrough
innovations (Dunlap-Hinkler et al., 2010; Kaplan and Vakili, 2015; Phene et al., 2006). Further, our formulation is reminis-
cent of radical innovation resulting from exaptation, which suggests that firms may accumulate technological knowledge
without anticipation of its subsequent uses, and a radically new technology may eventually emerge from deploying a firm’s
existing technological knowledge base into a new selection environment (see the special section edited by Andriani and
Cattani, 2016). Indeed, enlarged technological opportunities diffuse through the capital-good sector via the imitation of com-
peting firms. However, radical innovations are more difficult to copy as they increase the technological distance between the
firm mastering the new state-of-the-art technology and its competitors.
3.2. Investment and technological diffusion
Firms in the consumption-good industry produce a homogeneous good using their stock of machines and labor under
constant returns to scale. They invest to expand their capital stock and/or to replace their obsolete machines with new ones.
Note that such investments contribute to the technological diffusion of state-of-the-art technologies in the economy. As the
capital-good market is characterized by imperfect information, consumption-good firms choose their capital-good supplier
comparing price and productivity of the currently manufactured machine-tools. The model thus entails local interaction
among heterogeneous suppliers and customers.8
Let us first consider expansionary investment. Firms face a demand created by the expenditures of workers, and plan
their production according to (adaptive) expectations over such a demand, desired inventories, and their stock of invento-
ries.9 Whenever the capital stock is not sufficient to produce the desired amount, firms invest (EI j ) in order to expand their
production capacity:
EI j (t ) = K dj (t ) − K j (t ), (13)
where K dj and K j denote the desired and actual capital stock respectively.
Further, firms invest to replace current machines with more technologically advanced ones according to a payback period
routine. In a nutshell, they compare the benefits entailed by new vintages embodying state-of-the-art technology vis-á-vis
6
Empirical evidence on the manufacturing sector further suggests that, at the firm level, R&D spending is highly correlated with revenues, and that
differences in R&D intensity (given by the ratio between R&D spending and revenues) can be largely explained by industry level characteristics, while
factors as size and competition play a second-order role (e.g. Cohen et al., 1987).
7
Hence, the discovery of a radical innovation depends on the search effort exerted after another radical innovation had eventually been discovered. See
also our discussion in Section 5.1.
8
More on that in Dosi et al. (2010). Note also that machine production is a time-consuming process: consumption-good firms receive the ordered
machines at the end of the period. This is in line with a large body of literature: see, e.g., Rotemberg (2008) for details on pricing, imperfect information
and behavioural attitudes of consumers and Boca et al. (2008) for the presence of gestation lag effects in firms’ investments.
9
In the benchmark setup, expectations are myopic. The results are robust for different expectation formation mechanisms. More on that in
Dosi et al. (2020b).
7
the cost of new machines, taking into account the horizon in which they want to recover their investment. In particular,
given the set of all vintages of machines owned by firm j at time t, the machine of vintage τ is replaced with a new one
according to:
pnew
j
≤b (14)
ccon
j
(t ) − cnew
j
where pnew and cnew are the price and unitary cost of production associated to the new machine and b is a parameter cap-
turing firms’ “patience” in obtaining net returns on their investments.10 The vintages of machines that satisfies Eq. 14 con-
stitute the replacement investment of the firm, SI j (t ). Aggregate investment, in monetary terms, just sums over the value of
investments of all consumption good firms:

I (t ) = [EI j (t ) + SI j (t )] p j (t ), (15)
j
where p j (t ) is the price of a new machine available to firm j at time t.

Consumption-good firms sets the price of their final good applying a variable mark-up on their unit cost of production.
In line with the evolutionary literature and a variety of “customer market” models (Phelps and Winter, 1970), the mark-up
changes over time according to the evolution of firm’s market shares: firms increase prices if their market share is ris-
ing and decrease it when the market share falls. Consumers have imperfect information regarding the final product (see
Rotemberg, 2008 for a survey on consumers’ imperfect price knowledge) which prevents them from instantaneously switch-
ing to the most competitive producers. For this reason, market competition is captured via a replicator dynamics: the market
share of firms more competitive than the industry average increases, while that of less competitive ones shrinks over time.
Firms’ competitiveness depends on their price and on their capacity to satisfy demand in the past.11
At the end of each period, consumption-good and capital-good firms compute their profits and update their stock of
liquid assets. As in Dosi et al. (2010), firms finance production using internal funds and, if the latter are not sufficient - they
borrow from credit lines, over which they pay a constant interest rate, up to a maximum debt to revenues ratio. Firms with
zero market shares or negative net assets die and a new breed of firms enters the market. Overall, the number of firms is
fixed, hence any dead firm is replaced by a new one. In line with the empirical literature on firm entry (Bartelsman et al.,
2005), we assume that entrants are on average smaller than incumbents, with the stock of capital of new consumption-good
firms and the stock of liquid assets of entrants in both sectors being a fraction of the average stocks of the incumbents.
Concerning the technology of entrants, new consumption-good firms select amongst the newest vintages of machines, while
the technology of new capital-good firms is on average worse than incumbents’ ones.
3.3. The public sector and the macroeconomic framework
Workers-consumers have a marginal propensity equal to one in the model. Accordingly, aggregate consumption (C) is
computed by summing up over the income of both employed and unemployed workers:
C (t ) = w(t )LD (t ) + wU [LS (t ) − LD (t )], (16)
where w represent wages, wU the unemployment subsidy and LD and LS labour demand and labour supply respectively. Ag-
gregate labor demand is computed summing up the labor demands of capital- and consumption-good firms; labor supply is
exogenous and inelastic, as in Dosi et al. (2010). Wages are linked to the dynamics of productivity, prices and unemployment
rate by the following wage equation:

¯ (t )
ABcpi(t ) U (t )
w(t ) = w(t − 1 ) 1 + ψ1 + ψ2 + ψ3 , (17)
¯ (t − 1 )
AB cpi(t − 1 ) U (t − 1 )
where AB¯ indicates the average productivity in the economy, cpi is the consumer price index and U stands for unemploy-
ment rate. The labor market does not necessarily clear and both involuntary unemployment and labor rationing can occur.
The unemployment subsidies - a fraction of the current market wage - are paid by the public sector (G indicates such
spending), which also levies taxes on firm profits. Taxes and subsidies are the fiscal leverages that contribute to the aggre-
gate demand management regimes. Further, the government can run innovation policy incurring in additional spending as
indicated by IP (more on that in Section 5). The deficit is then equal to:
De f (t ) = G(t ) − T axes(t ) + CD(t ) + IP (t ), (18)
10
Our assumptions are in line with a large body of empirical literature showing that replacement investment is typically not proportional to the capital
stock, but a crucial strategic decision of firms (see e.g. Eisner, 1972; Feldstein and Foot, 1971; Goolsbee, 1998).
11
Unfilled demand is due to the difference between expected and actual demand. Firms set their production according to the expected demand. If a
firms is not able to satisfy the actual demand, its competitiveness is accordingly reduced. On the contrary, if expected demand is higher than actual one,
inventories accumulate.
8
where CD indicates the cost of public debt (i.e. interests on previous debt) and satisfies CD(t ) = r pd (t )P D(t − 1 ), with P D
denoting the stock of public debt and r pd (t ) the interest rate. Differently from Dosi et al. (2010), the interest rate on gov-
ernment bonds changes over time according to the ratio between the public debt and GDP:
PD PD
r pd (t + 1 ) − r pd (t ) = (t + 1 ) − (t ) , (19)
GDP GDP
with > 0. Indeed, public debt is implicitly placed within a external international financial market charging higher interest
rates in presence of larger stock of debt. The above assumption allows one to capture the long-run cost of innovation policies
and the possible emergence of vicious debt cycles triggered by excessive public expenditures.12
Finally, the model satisfies the standard national account identities: the sum of value added of capital- and consumption
goods firms equals aggregate production (in our simplified economy there are no intermediate goods). In turns, the value of
total output coincides with the sum of aggregate consumption, investment and change in inventories:

Qi (t ) pi (t ) + Q j (t ) pcj (t ) ≡ Y (t ) ≡ C (t ) + I (t ) + N (t ), (20)
i j
where Qi and Q j represent the production of capital and consumption good firms, respectively, pi and pcj the prices charged
by capital and consumption good firms, while C, I and N stand for consumption, investment and the variation of invento-
ries, all measured in nominal terms. The real stance of each aggregate variable can be obtained dividing by the appropriate
deflator (CPI for aggregate consumption and inventories; PPI for investments), and the real GDP simply corresponds to their
sum.13
4. Simulation set-up and empirical validation
The foregoing model does not allow for analytical, closed-form solutions. This is a distinctive feature of many ABMs
that stems from the non-linearities present in agent decision rules and their interaction patterns, and it implies running
computer simulations to analyze the properties of the stochastic processes governing the coevolution of micro and macro
variables (more on that in Fagiolo et al., 2019; Fagiolo and Roventini, 2017; Windrum et al., 2007). In what follows, we
therefore perform extensive Monte-Carlo analyses to appropriately account for cross-simulation variability. More precisely,
all results are presented either as single simulation runs, to show the behaviour of our artificial economy along an hypo-
thetical scenario, or as averages across two-hundreds independent simulations to identify robust emerging properties and
to perform statistical testing across scenarios and policy experiments.
Before running policy experiments, the model has undergone an indirect calibration exercise. Then we “empirically vali-
dated” the model, i.e. we studied its capability to account for a large ensemble of macro and micro stylized facts (see Fagiolo
et al., 2019; Windrum et al., 2007)14
In particular, the parameter space has been extensively explored (through random sampling and in absence of innovation
policy) in search of the three properties listed below. Then, a best candidate vector of parameters has been retained (see the
Appendix for the list of parameters) and - finally - a series of validation tests based on stylized facts (i.e. known empirical
regularities) replication have been performed. The three properties we looked for are: (i) long-run growth and business
cycles punctuated by infrequent yet possibly deep crises, (ii) a sustainable pattern of deficits reflecting a balanced fiscal
policy and (iii) a vivid process of firm competition sustained by innovation and imitation with very rare yet possible radical
innovations. Such an approach guarantees a good degree of empirical validity to our simulation experiments and finds in
line with the prevailing practices in the agent-based economic literature (Fagiolo et al., 2019).15 That said, we stress that
our exercises should be taken much more as thought experiments aimed at unveiling mechanisms, comparing policies and
establishing robust rankings, rather than delivering quantitative predictions. Table 1 shows the stylized facts that the model
replicates.16
Fig. 2 shows the dynamics of the model in the “no innovation policy” baseline relying on a single model run, while
Table 2 reports a series of summary statistics over the Monte Carlo ensemble. The model robustly generates endogenous
self-sustained growth patterns characterized by the presence of persistent fluctuations and rare crises. The positive trend in
productivity and aggregate output is driven by the innovation activity of capital-good firms and the processes of technolog-
ical diffusion occurring horizontally via the imitation activity of competitors, as well as vertically through the investment
choices of consumption-good firms.
12
For more experiments on the short- and long-run impact of fiscal policies on public debt as well as economic dynamics, see Dosi et al. (2015). Further,
we check that debt to GDP ratios remain under control (below 150%) in all simulations.
13
Thanks to the micro-founded nature of the model we are able to collect the prices charged and the quantities produced by each firm in each sector,
which - in turns - give the consumption price index (CPI) and the producer price index (PPI). The prices charged by consumption good firms ( pcj ) correspond
to a markup over production costs, see Dosi et al. (2010) for details.
14
See also Lamperti (2018a,b), Guerini and Moneta (2017) and Lamperti et al. (2018c). By indirect calibration we refer to an algorithmic procedure wherein
a set of empirical properties is targeted, a parameter vector satisfying them is obtained and the replication of a larger set of stylized facts is examined.
15
See Reissl et al. (2021) for a recent application of an indirect calibration approach to a quantitative oriented macro-economic input-output agent-based
model.
16
We point the reader to Dosi et al. (2017a) for a more detailed overview of these facts and to the Laboratory for Simulation Development website for
the code of the K+S model (without innovation policies), which can be used to generate the data and inspect the stylized facts.
9
Fig. 2. Model behaviour under the “no innovation policy” baseline. Selected indicators are reported for a single model run. GDP, Labour productivity, Real
wage are in logs. Negative public deficit indicates a surplus.
Table 1
Main empirical stylized facts replicated by the DSK model. .
Stylized facts Empirical studies (among others)
Macroeconomic stylized facts

SF1 Endogenous self-sustained growth Burns and Mitchell (1946); Kuznets and Murphy (1966)
with persistent fluctuations Stock and Watson (1999); Zarnowitz (1985)
SF2 Fat-tailed GDP growth-rate distribution Castaldi and Dosi (2009); Fagiolo et al. (2008)
Lamperti and Mattei (2018)
SF3 Recession duration exponentially distributed Ausloos et al. (2004); Wright (2005)
SF4 Relative volatility of GDP, consumption and investments Napoletano et al. (2006); Stock and Watson (1999)
SF5 Cross-correlations of macro variables Napoletano et al. (2006); Stock and Watson (1999)
SF6 Pro-cyclical aggregate R&D investment Wälde and Woitek (2004)
Microeconomic stylized facts
SF7 Firm (log) size distribution is right-skewed Dosi (2007)
SF8 Fat-tailed firm growth-rate distribution Bottazzi and Secchi (2003, 2006)
SF9 Productivity heterogeneity across firms Bartelsman and Doms (2000); Dosi (2007)
SF10 Persistent productivity differential across firms Bartelsman and Doms (2000); Dosi (2007)
SF11 Lumpy investment rates at firm-level Doms and Dunne (1998)
SF12 Firm bankruptcies are counter-cyclical Jaimovich and Floetotto (2008)
10
Table 2
Summary statistics for selected indicators in the “no innovation policy” baseline; 200 runs are used. HHI
stands for Hirschman-Herfindahl Index; Cap-Good indicates the Capital Good sector and Cons-Good the
Consumption good sector; Lik. stands for Likelihood; incr. and rad. for incremental and radical respec-
tively. Crises are defined as events where either GDP drops by more than 3% in a single period or four
consecutive periods of negative growth are observed.
Variable Mean St. Dev Variable Mean St. Dev
GDP growth 0.0268 0.0012 Unemployment 0.0610 0.0376

GDP volatility 0.0819 0.0005 Productivity growth 0.0258 0.0012
Deficit on GDP 0.0434 0.0553 HHI Cap-Good sector 0.6691 0.0601
Lik. of crises 0.0462 0.0399 HHI Cons-Good sector 0.0032 0.0001
Lik. of (incr.) innovation 0.571 0.0424 Lik. of imitation 0.6012 0.0502
Lik. of (incr.) inno. & imit. 0.218 0.0370 Lik. of (rad.) innov. 1.25 10−05 0.0000
Table 3
Results from Experiment 1 (R&D subsidies). Rows reports the average relative performance of each experiment
with respect to the “no innovation policy” baseline (Baseline) over 200 Monte Carlo runs; for example 1.2
indicates that the experiment has produced an average value of the relevant statistic that is 20% higher than
in the baseline. Symbol ∗ indicates a statistical significant difference between the experiment and the baseline
at 5% as resulting from a t-test on the means. GDP vol. stands for GDP volatility as proxied by the standard
deviation of the growth process; Unempl. stands for unemployment and empl. for employment; Deficit and
Fiscal cost are expressed as relative to GDP.
GDP growth GDP vol. Unempl. Periods full empl. Deficit Fiscal cost
Baseline 2.68% 0.08 6.10% 16% 4.34% 0.00
Size of the subsidy

5% 1.04 1.01 0.98 1.04 1.25 0.9% ∗
∗ ∗
10% 1.08 1.02 0.98 1.08 1.39 2.2% ∗
∗ ∗ ∗
15% 1.10 0.97 0.96 1.17 1.14 2.6% ∗
∗ ∗
30% 1.18 0.99 0.95 1.37 0.94 6.4% ∗
Simulation results also show the presence of fierce Schumpeterian competition taking place at the microeconomic level.
For instance, on average, slightly more than half of capital-good firms successfully introduce an innovation or copy the
technology of a competitor in every simulation step, while just one-fifth perform both activities. The likelihood of radical
innovations is remarkably low, and only a private firm is able to obtain one in a single run.
Government deficit averages around reasonable levels (4-6% of GDP) for a developed economy, while displaying large
spikes during crises, characterized by surges in unemployment. Beyond such rare crises, whose likelihood is relatively low
(around 5%), public finances often register a surplus, guaranteeing the long run sustainability of debt.
The bottom panels in Fig. 2 show the cyclical components of the GDP, consumption and investment time-series generated
by our model. They show the presence of vibrant fluctuations in all series, punctuated by deep downturns. Such fluctuations
are genuinely endogenous, as no aggregate exogenous shock is present in the model. In addition, consumption and invest-
ments are, respectively, more and less volatile than output, in tune with empirical evidence (Napoletano et al., 2006; Stock
and Watson, 1999).
As mentioned above (see Table 1), the baseline configuration of the K+S model is able to account for a wide set of microe-
conomic empirical regularities concerning e.g. firm size and growth-rate distributions, productivity dynamics, investment
patterns; see also Dosi et al. (2017a) for additional details. This reflects the strong explanatory capabilities of agent-based
models as discussed in Haldane and Turrell (2019) and Dosi and Roventini (2019).
Overall, our “no innovation policy” baseline reflects an economy where decentralized interactions give rise to stable
properties at the macroeconomic level (all standard deviations in Table 2 are relatively low compared to the averages), with
sustained growth and healthy public finances. Against such background we test a series of policy regimes aimed at further
stimulating innovation, productivity and long-run growth, while maintaining public deficit and debt under control.
5. Innovation policy experiments
As emphasized in Section 2, innovation policy encompasses a variety of instruments, ranging from monetary incentives
such as R&D subsidies and tax credits (indirect interventions) to direct spending in public research activities (for example,
in the US, funding basic research through the National Sciences Foundation as well as public organizations like DARPA of the
US Department of Defense). In this Section we rely on controlled simulation experiments to investigate the macroeconomic
effects of different policy instruments: Section 5.1 first describes the different policy interventions, while simulation results
are spelled out in Section 5.2. A sensitivity analysis of the main results can be found in the Appendix (Table 9), together
with a set of additional experiments (see Tables 10, 11 and 12) where we test alternative designs of the policy interventions
described below, as well as alternative behavioural assumptions.
11
5.1. A “menu” of innovation policies
We consider five different types of innovations policies and we also experiment with ensembles of different interventions.
Experiments 1 and 2 consider indirect policy interventions typical of the market failure approach, whereas Experiments 4 and
5 explore direct Government interventions and are akin to the Entrepreneurial State framework (see Mazzucato, 2013 and
the discussion in Castelnovo and Florio, 2020).17 As an additional benchmark, we consider a scenario (Experiment 3) where
public expenditures sustain only private consumption and hence cannot have a direct influence on productivity growth.
Experiment 1. (R&D subsidies) The Government provides a R&D subsidy to firms in order to increase their research efforts.
Larger R&D investments may increase the chances of discovering novel machines, more efficient production techniques or,
finally, they may speed up horizontal technological diffusion via imitation of competitors. We assume that public subsidies
qRD > 0 are proportional to firm’s past spending in research and innovation (RDi ):
RDi (t ) = υ Si (t − 1 ) + qRD RDi (t − 1 ). (21)
Experiment 2. (investment tax discount) Under this intervention, consumption-good firms receive a government-financed
discount on their investments in novel capital goods, whose size - relative to the price of the new machine - amounts to
dT D . This policy is supposed to speed up technological diffusion vertically, as consumption-good firms firms pay a lower
prices whenever they replace current machines with new ones embedding state-of-the-art technologies. Under this policy,
the pay-back period routine (cf. Eq. 14) becomes:
pnew (1 − dT D )
≤ b. (22)
ccon
j
− cnew
Experiment 3. (public expenditures directed to private consumption) This experiment mimics a scenario where public
transfers boost household consumption expenditures. Of course, in this framework, they do not directly affect the innova-
tion and investment decisions of firms, but they might increase productivity growth via more sustained levels of aggregate
demand. In the model, consumption positively affects demand expectations and thus expansionary investment. This exper-
iment may thus have, via this channel, a positive effect on R&D in the capital good sector, which depends on past sales.
Nevertheless, its impact is expected to be lower compared to R&D subsidies and direct government innovation policies.
Experiment 4. (a public capital-good firm) In an Entrepreneurial State framework, new public entities are created to shape
the innovation landscape by engaging and coordinating research in given fields and diffusing the relevant knowledge to
facilitate technological progress (see Sections 1 and 2). In this experiment, the government creates and fund a public firm
in the capital-good sector. Similarly to privately owned firms, the new public firm satisfies the demand of machines coming
from consumption-good firms and performs innovation and imitation activities. However, four key differences apply: i) the
public firm allocates all its profits (pf ) to R&D; ii) it is bailed out by the government in case of failure (negative net liquid
assets); iii) it can receive additional funds from the government (IP ) to perform extra research activities; and iv) it fosters
the diffusion of its technology to its competitors which can freely imitate it if their cumulated knowledge is sufficiently
large. In particular, the R&D expenditure of the public firm ( p f ) amounts to:
RDpf (t ) = υ S p f (t − 1 ) + pf (t − 1 ) + IP (t ). (23)
Any capital-good firm i can freely imitate the public firm if its (normalized) technological distance - which stems from
the history and direction of its innovations - from the public firm (NT D p f, j , cf. Eq. 9) is smaller than a fixed threshold
φ ∈ (0, 1 ). In other words, the public firm discloses the blueprints of its technology to private firms that are sufficiently
close from a technological perspective. In turn, private firms can decide to use the public firm’s blueprints, if convenient
(i.e. using the same routine described by equation (10)). However, a more technologically distant firm may still imitate the
public firm according to the process described in Section 3.1. In general, we design Experiment 4 to account for the role
that public firms cover within Entrepreneurial State-like programs, in which they both contribute to the search process for
novel technologies and further facilitate their diffusion (see Chiang, 1991; Mazzucato, 2013; Nelson, 1982a).18 Fig. 3 shows
a stylized representation of such a “local” process of knowledge diffusion. Obviously, private firms will decide whether to
adopt the technology of the public firm only if it is convenient on the basis of the routine expressed by Eq. (10).
Experiment 5. (a national research laboratory) The last experiment captures another essential feature of an Entrepreneurial
State, i.e. the creation and funding of public institutions that discover radical innovations enlarging technological opportu-
nities in the economy (as for national research laboratories and the Internet, see Section 2), while bearing the risks and the
costs of such ventures. In particular, we introduce a national research lab (NRL) that (i) performs basic research but does not
produce; (ii) takes stock of all the knowledge developed in the economy, (iii) tries to enlarge the set of technological op-
portunities available for capital-good firms through the discovery of radical innovations (see Section 3.1). At each time step,
17
In the Appendix, we test the robustness of such policies to various alternative assumptions.
18
Consider, for example, the experiences of the Italian IRI and the French government-controlled electricity company EDF. See also Castelnovo and Flo-
rio (2020).
12
Fig. 3. Experiment 4 : knowledge diffusion by the public firm.
the NRL receives public funding form the government to perform its research activities. Further, as it is a purely research-
oriented organization, it is able to exploit the entire body of knowledge available in the economy to perform its research.
Hence, the discovery of a radical innovation by the NRL is assumed to depend on its cumulative search efforts (CRD public ),
as well as on those performed by capital-good firms (CRDi ):

i CRDi (t ) + CRDpublic (t ) 1
RI
PNRL (t ) = f x | x = = . (24)
GDP (t ) 1 + eη1 (x−η2 )
Equations (24) and (12) reveal a clear asymmetry between the NRL and private firms. In particular, in the search for radical
innovations, the NRL leverages the knowledge stock developed by the whole economy; differently, capital-good firms build
on own and public cumulative R&D, but not on the knowledge stock developed by other private firms. Such an assumption is
meant to reflect the difference between a research-oriented public organization and a profit-driven private firm (Mazzucato,
2013; Nelson, 1982a), and is grounded on the historical evidence that public agencies and laboratories have typically engaged
in projects characterized by large technological breadth, merging pieces of technical knowledge developed across time by a
number of public and private organizations (consider, for example, projects developed at the NASA and ARPA-E, as well as
the experiences of mixed Bell labs or the Xerox Park; see also Mazzucato, 2013 and Kantor and Whalley, 2022).19 However,
the ability of the NRL to exploit the existing body of knowledge is not perfect, in the sense that additional R&D activities
just raise the probability of finding a radical innovation, but they never guarantee such a discovery.
Further, a NRL that discovers a radical innovation also provides free access to the new technological opportunities it
involves, de facto moving the distribution of innovative possibilities for the whole economy.20 This is another difference
between the NRL and private capital-good firms (see Section 3.1).
Fig. 4 exemplifies how cumulative R&D affects the discovery of a radical innovation by the NRL across multiple model
runs. Indeed, as the economy-wide knowledge stock accumulates the probability of radical innovation increases logistically,
according to Eq. 24. Contrarily, when cumulative R&D is relatively low, the likelihood of finding an innovation approaches
zero which - in other words - implies that the NRL is not able to exploit the knowledge of the economy to enlarge the
technological opportunities. When a radical innovation is discovered, the probability of finding a new one drops (panel B)
and it re-starts increasing as long as additional R&D is performed, either publicly or privately.21
5.2. Simulation results
To ensure the comparability of results across the different policy experiments, we keep constant the fiscal cost of the
innovation policies in the various regimes. In particular, we first perform Experiment 1 (R&D subsidy) by setting the size
of the subsidy (qRD ∈ {5%, 10%, 15%, 30%}). Then, we inspect the results of the model (see Table 3) and select a reference
scenario whose fiscal cost — expressed in terms of average expenditure for the innovation policy relative to GDP — is
imposed to all other experiments. In particular we use qRD = 15% as our reference scenario, where the average cost of the
innovation policy amounts to 2.6% of GDP. When running all other experiments, the size of the policy intervention is then
equal to IP (t ) = 0.026 · GDP (t ).
Single innovation policies. Figs. 5 and 6 show the patterns of GDP (and public deficit) for a single run of the five inno-
vation policy experiments. First, all innovation policies have a positive effect on the long-run output trend of the economy
(although to different extent). This is not the case for transfers supporting private consumption (Exp. III), which do not have
19
For example, the recent Cancer Moonshot program, largely coordinated by the NIH, collects research initiatives across a wide spectrum of scientific and
technical fields, integrating and levering on knowledge stocks produced by several private and public firms, as well as universities and laboratories.
20
In the current set-up, we cannot study mission-oriented innovation policies directed to specific missions, as the model does not allow for multiple
industries. Hence, we cannot study how such policies trigger the direction of technical change through the emergence of new sectors and markets. We
leave such developments to future research (see also our discussion in Section 6).
21
This also indirectly reflects the idea that when missions are achieved, organizations often go through a phase of change wherein part of the knowledge
stock is lost and needs being rebuilt towards new missions. For example, DeLong et al. (2004) reports “[... ] to go to the moon again, we’ll be starting
from scratch. [... ] in the 1990s NASA lost the knowledge it had developed to send astronauts to the moon. In an era of cost-cutting and downsizing,
the engineers who designed the huge Saturn 5 rocket used to launch the lunar landing craft were encouraged to take early retirement from the space
program”.
13
Fig. 4. Dynamic behavior of the cumulative R&D intensity between two successive radical innovations (x in Eq. 24; panel A), probability that the NRL
RI
discovers a radical innovation (PNRL ; panel B) and occurrence of a radical innovation (panel C) in Exp. V. Each line corresponds to a different model run..
Table 4
Comparison of different innovation policy experiments and their combinations. Rows report the average rel-
ative performance of each experiment with respect to the “no innovation policy” baseline over 200 Monte
Carlo runs. Symbol ∗ indicates a statistical significant difference between the experiment and the baseline
at 5% as resulting from a t-test on the means. GDP vol. stands for GDP volatility proxied by the standard
deviation of the growth process; unempl. stands for unemployment and empl. for employment; deficit is
expressed as relative to GDP.
Policy GDP growth GDP vol. Unempl. Periods full empl. Deficit
Baseline 2.68% 0.08 6.10% 16% 4.34%

I - R&D subsidies 1.10 ∗ 0.97 0.96 1.17 ∗ 1.14 ∗
II - Investment tax discount 1.08 1.22 ∗ 0.97 1.38 ∗ 1.34 ∗
III - Transfers to consumption 0.96∗ 0.98 0.92 ∗ 1.08 ∗ 1.45∗
IV - Public firm 1.27∗ 1.53∗ 0.88∗ 1.28∗ 1.19∗
V - National Research Lab 1.55∗ 2.01∗ 0.88∗ 1.52∗ 0.78∗
I + II 1.01 0.96 1.13∗ 1.22∗ 1.35∗
IV + I 1.16∗ 1.12∗ 0.94∗ 1.32∗ 1.27∗
IV + II 1.12∗ 1.40∗ 0.96 1.49∗ 1.34∗
V+I 1.37∗ 1.99∗ 0.74∗ 1.36∗ 0.90∗
V + II 1.22∗ 1.35∗ 0.87∗ 1.39∗ 0.95
IV + V 1.67∗ 2.60∗ 0.77∗ 1.61∗ 0.77∗
significant effects compared to the baseline scenario. Furthermore, a stark contrast emerges between indirect (Exps. I and
II) and direct (Exps. IV and V) innovation policies: while R&D subsidies and tax incentives produce a permanent upward
shift in the GDP level compared to the baseline (with subsidies being much more effective than tax-credits, see also Figs. 7
and 8), Entrepreneurial State interventions, either in the form of research-oriented public capital good firms or as a national
research laboratory, produce robust GDP growth accelerations (see panels A and C of Fig. 6).
Further, direct intervention policies are more effective than indirect ones as far as public finances are concerned. Indirect
policies generate public deficit-to-GDP ratios that tend to be constant yet higher than in the baseline scenario (see panel
B of Fig. 8 and Table 4). Entrepreneurial State interventions generate instead deficits-to-GDP ratios that are decreasing over
time and that, in the case of experiment V, are lower than in the baseline (see again Table 4).22 Decreasing deficits-to-GDP
22
The highest deficit is recorded when public transfers finance private consumption (Exp. III). However, in all policy scenarios the ratio between public
debt and GDP does not increase over time.
14
Fig. 5. Dynamics of GDP and public deficit across experiments for indirect innovation policies. Each row of panels corresponds to an experiment: panels
A and B to Experiment 1 (R&D subsidies), panels C and D to experiment II (Investment tax discount), panels E and F to experiment III (Transfers to
consumption). Each plot shows a single model run under the experiment and the “no innovation policy” baseline.
ratios are result of the growth accelerations induced by direct innovation policies as the fiscal cost is constant across policy
scenarios.
The superior performance of direct innovation policies vis-à-vis indirect ones is confirmed by the summary statistics re-
ported in Table 4. The battery of Monte Carlo statistics shows in particular that Experiment 5 is the best innovation policy to
implement as it solves the growth-deficit trade-off (with respect to the baseline) that characterizes instead all other policy
regimes and it guarantees a superior trajectory for the economy characterized by higher average growth, lower unemploy-
ment output, and the lowest impact on public finances (the higher volatility is due to the jump in technological opportu-
nities). Experiment 4 ranks second as it improves the performance of the economy. However, its lower (positive) impact on
growth is not enough to improve the average deficit to GDP ratio with respect to the “no innovation policy” baseline. Indi-
15
Fig. 6. Dynamics of GDP and public deficit across experiments for direct innovation policies. Each row of panels corresponds to an experiment: panels A
and B to Experiment 4 (Public firm) and panels C and D to experiment V (National Research Lab). Each plot shows a single model run under the experiment
and the “no innovation policy” baseline.
Fig. 7. Dynamics of GDP (panel A) and public deficit (panel B) across experiments. Averages over 200 Monte Carlo runs. Exp. I: R&D subsidies; Exp. II:
Investment tax discount; Exp. III: Transfers to consumption; Exp. IV: Public firm; Exp. V: National Research Lab.
16
Fig. 8. Distribution of GDP growth (panel A) and public deficit (panel B) values across experiments. Pooling of averages over 200 Monte Carlo runs, each
observation corresponds to the Monte Carlo average in a given simulation step. Exp. I: R&D subsidies; Exp. II: Investment tax discount; Exp. III: Transfers
to consumption; Exp. IV: Public firm; Exp. V: National Research Lab.
rect innovation policies (Exps. I and II) are more effective to stimulate productivity and GDP growth in the short-run (Fig. 7),
but they are overtaken by Entrepreneurial-State interventions in the long-run, and they worsen public finances across the
whole simulation span. More precisely, tax discounts does not significantly improve neither output growth nor the employ-
ment rate with respect to the baseline, while R&D subsidies do both. However, we show in the Appendix (see Table 10)
that tax credits on R&D - and not on physical investments - prove superior to Exp. I under the assumption that firms an-
ticipate the policy and raise their search efforts accordingly. Such effect is driven by R&D credits guaranteeing higher liquid
resources to firms, which cushion against downswings of the cycle. As expected, public transfers to consumption ranks last
with a negative impact on GDP growth and public deficit (but lower average unemployment rate). A number of additional
tests are collected in the Appendix. Tables 9 and 12 show the robustness of Exp. V’s superior performance against a number
of alternative specifications. Interestingly, we run a scenario in which the discovery of radical innovations depends solely
on the R&D conducted within the organization. In this setting the NRL is way less effective at stimulating growth than in
the standard Exp. V, which points to the importance of building public laboratories that effectively leverage on private R&D.
In particular, the NRL achieves - on average - less than one radical innovation per run (0.96; Table 12), which should be
compared to the 2.15 radical innovations per run of the standard experiment (see also Table 9). Intuitively, this translates
in a trajectory characterized by sustained yet lower growth, reduced volatility and slightly larger deficit. Nevertheless, even
in this modified experiment, the NRL policy emerges as comparatively superior - on average - to other innovation policies
(Table 4). Finally, Table 11 shows that our ranking of innovation policies is robust to alternative specifications of the routine
used by capital-good firms to invest in R&D, though some quantitative differences emerge.
Combinations of innovation policies. We also consider different pairs of innovation policies by equally splitting the
public resources across the two interventions, thus guaranteeing comparability with previous (stand-alone) experiments.
Simulation results reveals interesting synergies and redundancies across policies (see Table 4). First, the joint implemen-
tation of R&D subsidies with Entrepreneurial-State policies (IV+I and V+I) delivers higher output growth and employment
levels while shrinking deficits with respect to Experiment 1 alone. However, such a combination is outperformed by both
stand-alone “public firm” and “NRL” interventions (Exps. IV and V). Splitting resources across research subsidies and tax
discounts (Exp. I + II) worsen the dynamics of GDP and public finances relative to the Experiments 1 and 2 alone, showing
a mutually defeating effect of incentives to private firms in fostering innovation and growth. Indeed, by reducing public sup-
port both to technology discover (Exp. I) and downstream diffusion (Exp. II) such coupled policy turns out to be unsuccessful
across both dimensions: reduced investment by consumption good firms both retard the penetration of novel technologies
and decrease sales in the capital good sector, which in turn decreases future R&D spending. Finally, the best results are
obtained when the synergies between Entrepreneurial-State policies (Exp. V + IV) are fully exploited. Indeed, such a policy
combination improves the performance of the economy and reduces the deficit-to-GDP ratio vis-á-vis the two interventions
in isolation. The faster technological diffusion guaranteed by the presence of the public firm stimulates productivity and -
hence - demand growth, which raise R&D spending (see also Tables 5 and 7 below) in the economy finally reducing the
risks of failure of the NRL.
17
Table 5
Experiment 4 : imitation of the public firm. Values represent
the average number of times the public firm is imitated by a
private firm in each simulation span and over 200 Monte Carlo
runs (capital good sector is composed of 50 firms).
Per-period imitations of the public firm
Mean Max Min St. Dev.
Simulation span
[1-100] 0.7 3 0 1.8
[101-200] 2.2 7 0 2.5
[201-300] 4.6 8 0 3.1
[301-400] 1.8 4 0 2.2
Fig. 9. Technological embeddedness of the public firm and aggregate productivity growth in the economy under Experiment 4 (Public firm). Each point
represents the average over a Monte Carlo run; 200 runs are used.
Positive hysteresis. Accelerations in either GDP or productivity growth, which underlie the superior performance of di-
rect innovation policies, are the result of positive hysteresis, i.e. a permanent increase of the growth possibilities of the
economy.23 For instance, in Exp. IV the public firm induces a rapid and temporary process of knowledge accumulation and
diffusion that has positive permanent effects on the level of output. In Exp. V we observe instead super hysteresis, i.e. a per-
manent surge of GDP growth rate. This is explained by the fact that a NRL has relatively higher chances to introduce radical
innovations, which shifts to the right the entire distribution of technological, and thus growth, opportunities, with respect
to private firms. Table 9 (in the Appendix) reports the share of runs with at least one radical innovation, which is the ulti-
mate responsible for the change of paradigm in the growth dynamics. Thus, while almost all hysteresis literature focuses on
the long-lasting impact of recessionary shocks on employment and GDP (see e.g., Cerra et al., 2021; Dosi et al., 2018), our
results show that Entrepreneurial State innovation policies can positively affect the growing possibility of the economy.
Table 5 allows one to better understand the microeconomic drivers of hysteresis in our simulation experiments. During
the initial stages of the simulation, i.e. when the innovation policy has still to exert its effects, the public firm is rarely
imitated by its private competitors. However, as time goes by, the higher R&D propensity of the public firm maps into more
innovations, which move its technology towards the frontier and thus increase the imitation rates of its private counterparts.
In turns, the sustained imitation process spurs the diffusion of state-of-the-art technologies in the economy and triggers the
temporary GDP growth accelerations shown in panel C of Fig. 6. However, this process eventually stops (see panel C of
Fig. 6 and panel B of Fig. 7) and the aggregate growth rate of the economy falls back to previous levels, for two reasons.
First, the public firm extracts productivity gains from a constant technological opportunity landscape. This sets an upper
bound on the productivity gains it can diffuse to the rest of the economy. Second, most firms are able to catch up the
technology of the public firm over time. The latter is therefore less and less imitated over time (cf. the lower imitation rates
in the last part of the simulation in Table 5), which introduces a further slow-down on the overall growth process.
The ability of the public firm to trigger a diffusion process stimulating productivity and output growth correlates robustly
to its degree of technological embeddedness (Fig. 9), defined as the average technological distance between the public firm and
its private competitors (see Eq. (9) in Section 3.1). Simulation runs wherein the private firms are able to quickly catch-up
the public one display - ceteris paribus - higher productivity growth (Fig. 9). These results deliver two policy implications:
23
In macroeconomics, hysteresis is defined as a situation where a shocks permanently affect the path of the economy.
18
Fig. 10. Dynamics of GDP (panel A) and public deficit (panel B) in Experiment 5 (National Research Lab), multiple runs in different shades of color.
(i) Entrepreneurial-State-like policies may need time to display their positive results, especially at the macroeconomic level;
(ii) the position of public firms in the technological space can play a significant role in boosting the growth rate of the
economy.
A NRL-based direct innovation policy thus delivers a superior performance - on average- compared to indirect policies.
At the same time, it may also imply some risks, which are associated to the ability of this policy to effectively trigger
technological breakthroughs that enlarge the set of technological opportunities. Fig. 10 shows the dynamics of GDP and of
the deficit-to-GDP ratio in five selected runs, which capture two qualitatively opposite patterns associated with that policy.
In the first one, output growth exhibits a positive structural break and super hysteresis emerges. This virtuous dynamics is
triggered by the discovery of radical innovations by the NRL and its subsequent diffusion in the economy. On the contrary,
in the second pattern shown in the figure, the R&D activity by the National Research Lab is not able to deliver a major
technological breakthrough. In this case, the innovation policy does not spur GDP growth, but it raises the public deficit
and the ratio between public debt and output (cf. Fig. 10), resembling those displayed in Experiment 3 (i.e. unproductive
spending; see panel F in Fig. 6). While experiments from I to III display a rather homogeneous behavior across runs (see the
distributions of Fig. 8), Exp. V and VI induce a trade-off between superior average growth performance and higher risks of
policy failures. Indeed, simulation results clearly reveal the perils of Entrepreneurial State policies wherein for every winning
investment there are many possible failures (Mazzucato, 2016).24 Nonetheless, the likelihood of these failed trajectories
remains limited. The distributions of the average deficit and debt-to-GDP ratios emerging from the Monte Carlo exercise
suggest that in Exp. V, the public R&D investment, which, to repeat, is comparable to that of other ones, lead most of the
time to the discovery of a radical innovation that keep public finance under control or even in surplus (see also Fig. 11 and
Table 9 in the Appendix).
The status of public finances differs sharply across experiments (see, e.g. Fig. 7). Indirect innovation policies (Exp. I and II)
and transfers to consumption increase the deficit to GDP ratio of the economy with respect to the baseline, as the additional
growth they eventually generate does not fully compensate for the cost of the policy itself. However, the government incurs
in a stable series of deficits both over time and across runs (see Fig. 8, signaling relatively low risks from these policy
scenarios. Differently, direct innovation policies (Exp. IV and, especially, V) induce an initial phase of deficit, which gradually
improves over time as the knowledge stock of the economy increases and the policies get efficacy boosting output growth
and fiscal revenues. However, failures of the Entrepreneurial State could impede such a shift in the stream of deficits, making
Exp. IV and V comparatively riskier.
Crowding-in effects. Finally, we investigate whether public innovation policies crowd out or crowd in private R&D ex-
penditures. In particular, in line with Moretti et al. (2019), we study the possible additionality of innovation policies relative
to firms’ R&D investment, by performing OLS regressions on the artificial data generated by different policy experiments:25
log RDi,s,t = β1 log IPs,t−1 + β2 log GDPs,t−1 + λi + μs + νt + εi,s,t (25)
24
For example, the US Department of Energy provided large-scale guaranteed loans to two green-tech companies: Solyndra ($500 million) and Tesla
Motors ($465 million). While the latter is regarded as a success story, the former went bankrupt with a loss for the public agency.
25
Our artificial economy offers a convenient setting to estimate Eq. (25) across different experiments: multiple model runs are independent by con-
struction, while offering across-run variability ensured by the stochastic nature of the model; the size of the innovation policy is comparable both across
experiments and time and individual-level fixed effects absorb firm-specific shocks that differentiate capital-good businesses in our economy.
19
Fig. 11. Status of public finances under Experiment 5 (National Research Lab); panel A reports the distribution of simulation-average deficits and panel B
the distribution of simulation-average debt; 200 runs are used. The blue line indicates the mean while the dashed red line crosses the x-axis at zero. (For
interpretation of the references to colours in this figure legend, the reader is referred to the web version of this article.)
Table 6
Crowding-in of private investments in R&D. Each column reports the estimates of Equation (25) using data relative to different
experiments; 200 Monte Carlo runs are employed. Exp. I: R&D subsidies; Exp. II: Investment tax discount; Exp. III: Transfers to
consumption; Exp. IV: Public firm; Exp. V: National Research Lab.
Dependent variable: log firm R&D(t)
(baseline) (Exp. I) (Exp. II) (Exp. III) (Exp. IV) (Exp. V)

∗∗∗ ∗∗∗ ∗ ∗∗∗
log public R&D(t-1) 0.000 0.643 0.066 - 0.031 0.594 0.511∗∗
(-) (0.005) (0.009) (0.018) (0.011) (0.241)
log GDP(t-1) 0.784∗∗∗ 0.533∗∗∗ 0.660∗∗∗ 0.659∗∗∗ 0.572∗∗∗ 0.589∗∗∗
(0.003) (0.003) (0.004) (0.004) (0.004) (0.004)
Individual-level FE Yes Yes Yes Yes Yes Yes
Period-level FE Yes Yes Yes Yes Yes Yes
Run-level FE Yes Yes Yes Yes Yes Yes
Observations 1,960,000 1,960,000 1,960,000 1,960,000 1,960,000 1,960,000
Adjusted R2 0.4600 0.4732 0.4287 0.4101 0.5654 0.4932
F Statistic 243,119,189∗∗∗ 288,637,079∗∗∗ 65,888,745∗∗∗ 57,034,106∗∗∗ 57,034,106∗∗∗ 57,034,106∗∗∗
Note: ∗ p<0.1; ∗∗
p<0.05; ∗∗∗
p<0.01.
Table 7
Crowding-in of private investments in R&D. Each column reports the estimates of Equation (25) using data relative to different experi-
ments; 200 Monte Carlo runs are employed. Exp. I: R&D subsidies; Exp. II: Investment tax discount; Exp. III: Transfers to consumption;
Exp. IV: Public firm; Exp. V: National Research Lab.
Dependent variable: log firm R&D(t)
(baseline) (Exp. V+I) (Exp. V+II) (Exp. IV+I) (Exp. IV+II) (Exp. IV+V)
∗∗∗ ∗∗∗ ∗∗∗ ∗∗∗
log public R&D(t-1) 0.00 0.631 0.460 0.931 0.531 1.330∗∗∗
(-) (0.005) (0.005) (0.005) (0.005) (0.005)
log GDP(t-1) 0.784∗∗∗ 0.560∗∗∗ 0.580∗∗∗ 0.560∗∗∗ 0.560∗∗∗ 0.540∗∗∗
(0.003) (0.003) (0.003) (0.003) (0.003) (0.003)
Individual-level FE Yes Yes Yes Yes Yes Yes
Period-level FE Yes Yes Yes Yes Yes Yes
Run-level FE Yes Yes Yes Yes Yes Yes
Observations 1,960,000 1,960,000 1,960,000 1,960,000 1,960,000 1,960,000
Adjusted R2 0.4802 0.4758 0.5804 0.5698 0.5554 0.5960
F Statistic 243,119,189∗∗∗ 297,909,025∗∗∗ 255,425,660∗∗∗ 408,800,073∗∗∗ 264,841,402∗∗∗ 565,446,088∗∗∗
Note: ∗ p<0.1; ∗∗
p<0.05; ∗∗∗
p<0.01.
where RD refers to private R&D, IP indicates the monetary size of the innovation policy and λi , μs and νt are individual-
level, model-run level, and period-level fixed effects. Econometric results show that innovation policies produce sig-
nificant crowding-in of private R&D expenditures across all experiments (Tables 6 and 7). However, stark differences
emerge in the impact of different policies. The estimated elasticity of private R&D to public research-related spending
ranges from 0.07 (Exp. II) to 1.3 (Exp. IV + V), with the elasticity of R&D subsidies (Exp. I) being at an intermediate
level between such boundaries yet delivering a positive significant effect (which is consistent with recent evidences, see
Santoleri et al., 2020, and references therein). Remarkably, these results are qualitatively and quantitatively comparable to
20
those of Moretti et al. (2019) on OECD countries (who report an elasticity of private to public R&D of about 0.6%) and of
Pallante et al. (2020) on the US (who find that private R&D increases by more than 0.1% for every additional percentage
point of spending in public mission-oriented research). Table 7 further confirm the synergies between direct policy inter-
ventions (with the elasticity of private R&D corresponding to 1.3 when Exp. V is combined with exp. IV, against 0.6 for Exp.
I and 0.5 for Exp. II), which maximize the crowding in of private investments.
6. Conclusions
If and how innovation policies should be designed is one of the major challenges facing policy makers and societies at
large. This work contributes to the ongoing debate extending the Schumpeter meeting Keynes agent-based model (Dosi et al.,
2010) to assess the impact of different public innovation interventions on the short- and long-run performance of the econ-
omy, as well as on the public budget. More precisely, we have considered indirect innovation policies supporting the R&D
activity and capital-good investment of private firms and direct intervention encompassing a public firm developing new
technologies and freely diffusing them into the economy, as well as a National Research Laboratory (NRL) engaged in fron-
tier research to discover radical innovations. The last two policies are akin to the interventions implemented by an En-
trepreneurial State (Mazzucato, 2013).
Our results show that the most effective innovation policies involve the creation of public research bodies, which we
label National Research Labs. Such a policy facilitate the discovery of radical innovations that enlarge the set of techno-
logical opportunities available to private firms, and trigger the emergence of positive hysteresis dynamics. The outcome is
a higher growth potential of the economy and a lower unemployment rate while the public deficit is kept under con-
trol. Positive synergies can be activated combining the previous policy with the creation of a public firm developing new
technologies and easing technological diffusion of state-of-the-art capital goods. Indirect innovation policies also increase
economic growth while keeping the public budget under control. However, their impact is lower than the one of direct poli-
cies. Entrepreneurial-State policies comes with the risk of deteriorating public finances in those cases where the publicly-
discovered technologies do not diffuse enough or large-scale and high-risk research projects seeking radical innovations fail.
However, for the same amount of public resources allocated to market-based interventions, Entrepreneurial-State innovation
policies deliver significantly better aggregate performances if the government is willing to patiently bear the intrinsic risks
related to innovation. Finally, in line with the empirical evidence (Moretti et al., 2019; Pallante et al., 2020), we find that
innovation policies crowd-in private R&D investment, and such a result is stronger for direct innovation policies.
Overall, this paper supports the idea that public policies aimed at stimulating basic research improve the economic
performance (see e.g., Akcigit et al., 2020; Kantor and Whalley, 2022) and stimulate the private search for innovations (e.g.
Rosenberg and Nelson, 1994). In contrast with Bloom et al. (2019), we show a clear economic rationale for mission-oriented
research programs which strengthens both recent empirical evidence (Pallante et al., 2020) and the historical analysis of
large government-led research programs (Gross and Sampat, 2020; 2021; Kantor and Whalley, 2022). While mission-oriented
policies are often criticized for the alleged inability of the government to select challenges and technologies, we find that
innovation policies inspired by an Entrepreneurial State approach are valuable independently of the specific mission to be
targeted. Though risky, they are way more effective than a number of alternatives at enlarging the pool of knowledge and
the technological opportunities available to the economy as well as at facilitating their diffusion. This is beneficial and
socially desirable per-se. Hence, given a societal challenge, our results hint that genuine Entrepreneurial State-like policies
should patiently promote the discovery of radical innovations through broad and ambitious research-oriented programs,
sustain the costs of eventual early failures, and ease knowledge diffusion through public-private interactions. Such a genuine
Entrepreneurial State approach will further solve the apparent short term trade-offs between innovation policies and other
forms of public spending.
This work can be extended along several directions. One natural avenue of further research would formalize mechanisms
of interactions and research partnerships between private firms, public firms and the NRL, thereby linking the present anal-
ysis to the role of R&D networks. Further, one could study the impact of innovation policies targeting workers’ skills, which
would constrain the discovery and diffusion of new technologies. This could be done starting from the labour-augmented
K+S model (see Dosi et al., 2020, 2022, and references therein). Third, one could consider the possible interactions between
innovation policies and the financial sector, and the possible introduction of a development bank extending the framework
in Dosi et al. (2015). Fourth, one could study mission-oriented innovation policies triggering clear missions, such as the
fight to climate change and the orderly decarbonization of the economy. Such interventions could be studied in an extended
version of the Dystopian Keynes meet Schumpeter model (Lamperti et al., 2019; 2021; 2018b; 2020), which builds on the
shoulders of the baseline model analyzed here. Relatedly, a promising new research avenue we plan undertaking will enrich
a multi-sector model (e.g. Dosi et al., 2022) with lobbying dynamics (e.g. Isley et al., 2015), enabling the analysis of various
innovation policies targeting endogenously defined and possibly competing missions.
Acknowledgments
The authors are indebted with Herbert Dawid, Mattia Guerini, Pierre Mohnen, Gianluca Pallante, Emanuele Russo, Danilo
Spinola, Tania Treibich and three anonymous reviewers for helpful comments and discussions. Further, the authors thank all
21
participants to EMAEE 2017, WEHIA 2018, WEHIA 2019, EAEPE 2018, EAEPE 2019, CEF 2018 and the First European Confer-
ence “Moving the Frontier of the Macroeconomic Modelling of Research & Innovation”. The authors acknowledge financial
support from the H2020 project GROWINPRO (GA No 822781).
Appendix.
Parameters’ table
Table 8
Model’s main parameters and initial conditions.
Description Symbol Value
Monte Carlo replications MC 200

Time steps in economic system T 400
Number of firms in capital-good industry F1 50
Number of firms in consumption-good industry F2 200
Capital-good firms’ mark-up μ1 0.02
Consumption-good firm initial mark-up μ̄0 0.3
Uniform distribution supports [ ϕ1 , ϕ2 ] [0.10, 0.90]
Wage setting AB ¯ weight ψ1 1
Wage setting cpi weight ψ2 0
Wage setting U weight ψ3 0
R&D investment propensity ν 0.02
R&D allocation to innovative search ξ 0.5
Sensitivity of innovation success to R&D oIN 0.3
Sensitivity of innovation success to R&D oIM 0.3
Beta distribution parameters (innovation) (α , β ) ( 3, 3 )
Beta distribution support (innovation) [ξ1 , ξ2 ] [−0.10, 0.10]
Radical innovation shift χ RI 0.025
New customer sample parameter ω̄ 0.5
Desired inventories l 0.1
Payback period parameter b 120
Logistic parameter for curve’s steepness η1 1.5
Logistic value for sigmoid’s midpoint η2 6
Sensitivity of interest on public bonds on debt to GDP ratio 0.01
Inflation adjustment parameter γπ 1.10
Unemployment adjustment parameter γU 1.10
Income tax rate taxi 0.1
Profit tax rate tax p 0.1
Unemployment subsidy rate wU 0.5
Sensitivity analysis
Here we perform a sensitivity analysis of the model’s behaviour under direct innovation policies (Exp. IV and Exp. V)
where we let vary one parameter at the time. All other parameters are set to their benchmark configuration (Table 8). Over-
all, we find that results are robust to changes in the value of parameters governing the probability of discovering a radical
innovation, the size of the shift in technological opportunities and the cost of public debt. This reinforces our conclusion
about the dynamics of growth and the health of public finances in the scenarios where Entrepreneurial State policies are
implemented. In addition, we notice that (i) when the probability of a radical innovation is more sensitive to the cumulative
stock of R&D, Exp. IV sometimes triggers a radical innovation by either a private or public firm, which did not happen in
the benchmark configuration; (ii), if the shift in technological opportunity due to a radical innovation is sufficiently large,
the average deficit over the simulation turns to positive thanks to the super hysteric impact on growth.
Additional policy and behavioural experiments
In what follows, we perform a battery of simulation exercises wherein we (i) study indirect innovation policies under
alternative designs, (ii) test all policies under different R&D investment routines, and (iii) modify the equation governing
the probability of a radical innovation. Overall, our results are robust to such alternative configurations which nonetheless
provide some novel insights.
Indirect innovation policies under alternative designs

Here we modify the design of indirect innovation policies (Exp. I and II). First, we consider a variant of Exp. I wherein we
introduce a R&D tax credit; second, we test different values of the pay-back parameter b, which modifies the consumption-
good firm’s patience in renovating their capital stock. In particular, the first variant of Exp. I introduces a tax credit that
22
Table 9
Sensitivity analysis to key parameters. Values refer to averages across Monte Carlo experiments of size 50. # rad. inn.
indicates the average number of successful radical innovations per run; runs RI indicate the share of runs with at
least one radical innovation. GDP gr., Une. and Def. stands for GDP growth unemployment and deficit to GDP ratio,
respectively..
Exp. IV Exp. V
# RI runs RI GDP gr. Une. Def. # RI runs RI GDP gr. Une. Def.
benchmark 0 0% 3.40% 5.38% 5.17% 2.15 86% 4.15% 5.37% 3.39%

η1 =1 0.02 1% 3.41% 5.30% 5.10% 3.22 92% 5.22% 3.82% 2.16%
η1 =1.2 0 0% 3.40% 5.38% 5.17% 2.46 88% 4.33% 4.45% 3.10%
η1 =2 0 0% 3.40% 5.38% 5.17% 0.21 18% 2.76% 6.83% 6.13%
η2 =4.5 0.16 14% 3.47% 4.93% 5.02% 4.31 100% 6.23% 3.67% –2.47%
η2 =5 0.03 3% 3.41% 5.35% 5.15% 3.58 98% 5.72% 3.80% –1.03%
η2 =7 0 0% 3.40% 5.38% 5.17% 0.65 60% 2.90% 6.24% 5.89%
=0.05 0 0% 3.40% 5.38% 8.66% 2.15 86% 4.15% 5.37% 4.71%
=0.1 0 0% 3.40% 5.38% 10.34% 2.15 86% 4.15% 5.37% 8.49%
=0.005 0 0% 3.40% 5.38% 4.51% 2.15 86% 4.15% 5.37% 3.08%
χ RI = 0.0125 0 0% 3.40% 5.38% 5.17% 2.15 86% 3.61% 6.01% 4.32%
χ RI = 0.05 0 0% 3.40% 5.38% 5.17% 2.15 86% 6.73% 2.93% –2.88%
Table 10
Comparison of different experiments on alternative designs of indirect innovation policies. Symbol ∗ indicates
a statistical significant difference between the experiment and the baseline at 5% as resulting from a t-test
on the means. GDP vol. stands for GDP volatility proxied by the standard deviation of the growth process;
unempl. stands for unemployment and empl. for employment; deficit is expressed as relative to GDP..
Policy GDP growth GDP vol. Unempl. Periods full empl. Deficit
Baseline 2.68% 0.08 6.10% 16% 4.34%

I - R&D subsidies 1.10 ∗ 0.97 0.96 1.17 ∗ 1.14 ∗
Ib - R&D tax credits, no anticip. 1.03 0.95∗ 0.96 1.02 1.09∗
Ic - R&D tax credits, anticipation 1.11∗ 0.90∗ 0.95∗ 1.20 ∗ 1.18 ∗
II - Invest. tax disc. (b=120) 1.08 1.22 ∗ 0.97 1.38 ∗ 1.34 ∗
II - Invest. tax disc. (b=90) 1.15∗ 1.42 ∗ 1.05 1.42 ∗ 1.21 ∗
II - Invest. tax disc. (b=160) 1.02 1.10∗ 1.08∗ 0.93∗ 1.39∗
capital-good firms receive in the period after having perfomed R&D (Exp. Ib); the fiscal discount amounts to a percentage
dTRD
D
of its previous expenditures in R&D:
T D,i (t ) = dT D RDi (t − 1 ).
DRD RD
(26)
Such discount boost firms profits, but it does not immediately translate into different R&D expenditures. Hence, we further
extend the experiment by assuming that capital good firms believe the policy regime, anticipate the effect of the fiscal
discount and increase their R&D by an equal amount (Exp. Ic). Finally, we experiment with different values of the pay-
back parameter b (see Eqs. 10 and 14), which proxies consumptio good firms’ “patience” in obtaining net returns on their
investments and, hence, could affect the effectiveness of tax discounts on physical investments.
The results in Table 10 highlight that discounts on physical investments (Exp. II), aimed at fostering technological dif-
fusion, deliver different dynamics with respect to R&D tax credits. Indeed, R&D credits prove more effective that physical
investment discounts at stimulating growth, yet only if we assume that firms anticipate the fiscal relief and destinate an
equal amount to R&D (Exp. Ic). In this scenario, the economy-wide performance is superior to the case of subsidies (Exp.
I), as lower taxation improves firms’ balance-sheet and reduce the volatility of growth, cushioning downward phases of the
cycle. However, we stress that the empirical literature reports mixed evidence about the crowding-in effect of R&D tax cred-
its, as some studies suggest that firms might just use the policy opportunistically to boost their profits (Marino et al., 2016;
Mohnen et al., 2017) while keeping search efforts unaltered. Indeed, Exp. Ib clearly shows that when firms do not anticipate
the policy and fail to increase their R&D expenditures the public intervention turns largely ineffective. Finally, we find that
the economy-wide impact of physical investments’ discounts depends heavily on firms’ “patience” in repaying their invest-
ments in capital goods: lower values of the pay-back parameter b boost investment and growth, though this comes at the
cost of larger fluctuations.
Innovation policies under different R&D investment routines

So far we have assumed that capital-good firms’ R&D spending correspond to a constant fraction of their past revenues.
Though there is empirical evidence supporting our modelling choice (Coad and Rao, 2010), some heterogeneity in R&D
intensity exist and might relate to firms’ size and their growth dynamics (e.g. Coad and Rao, 2010; Cohen and Klepper,
1992). In particular, Coad and Rao (2010) further robustly show that manufacturing firms are relative less willing to decrease
23
Table 11
Comparison of innnovation policies under various specifications of R&D investment routines. Rows
reports the average relative performance of each experiment with respect to the “no innovation
policy” baseline (Baseline) over 200 Monte Carlo runs. Symbol ∗ indicates a statistical significant
difference between the experiment and the baseline at 5% as resulting from a t-test on the means.
GDP vol. Symmetric reaction to growth corresponds to θ1 = θ2 = −0.5 and θ3 = 0; asymmetric re-
action to growth corresponds to θ1 = −0.7, θ2 = −0.3 and θ3 = 0; dependence on size corresponds
to θ1 = θ2 = 0 and θ3 = 0.1.
symmetric reaction to growth asymmetric reaction to growth
GDP growth Growth vol. Deficit GDP growth Growth vol. Deficit
Baseline 2.55% 0.10 4.11% 2.67% 0.13 2.17%

Exp. I 1.13∗ 0.92∗ 1.13∗ 1.18∗ 0.87∗ 1.05
Exp. II 1.08 1.20∗ 1.32∗ 1.08 1.24∗ 1.33∗
Exp. IV 1.33∗ 1.47∗ 1.17∗ 1.35∗ 1.5∗ 1.14∗
Exp. V 1.49∗ 1.87∗ 0.85∗ 1.52∗ 1.83∗ 0.81∗
dependence on size
GDP growth Growth vol. Deficit
Baseline 2.75% 0.09 4.02%
Exp. I 1.12∗ 0.98 1.12∗
Exp. II 1.09 1.18∗ 1.35∗
Exp. IV 1.45∗ 1.67∗ 1.07∗
Exp. V 1.50∗ 1.84∗ 0.90∗
R&D after a negative growth episode than they are in increasing it after a positive growth shock. Moreover, Cohen and
Klepper (1992) argue that - within industries - R&D intensity proportionally depends on the relative size of the search
intensity effort. In line with this empirical evidence, we compare our innovation policies under different rules determining
R&D spending, wherein we include dependence of the search intensity on experienced growth shocks and heterogenous
R&D intensity:
RDi (t ) = υ Si (t − 1 ) ∗ [1 + θ1 gSi (t − 1 )1[gSi < 0] + θ2 gSi (t − 1 )1[gSi > 0] + θ3 fRD,i (t − 1 )] (27)
where θ1 < 0, θ2 < 0, θ3 > 0, gS indicates sales growth, fRD the R&D expenditure relative to the overall R&D effort of capital-
good firms and 1[·] is and indicator function.
Table 11 collects the results. We find that when R&D expenditures adjusts to growth shocks, either symmetrically or
asymmetrically, R&D subsidies become comparatively more appealing than in the benchmark case where R&D solely de-
pends on past revenues. Indeed, they counterbalance the dependence of R&D spending from sales fluctuations at the firm
level. Intuitively, when R&D reacts asymmetrically to past sales growth (and more pronouncedly to negative growth shocks),
subsidies reinforce the countercyclical effect of R&D that characterizes this scenario. Differently, the impact of all other ex-
periments is similar to the analysis reported in the main text. Further, when we allow R&D investment to depend on its
size, our results are broadly confirmed. The only notable difference concerns Experiment 4 (Public firm). Indeed, under this
scenario, the larger tendency to spend in R&D of the public firm (see Eq. 23) vis-á-vis other capital-good firms kickstarts
a virtuous cycle triggering higher search for new technologies and lower risk of lagging behind in the technological space.
This reinforces the diffusion process that lies at the core of the growth stimulus brought about by a public firm. Indeed, in
the set-up where R&D adjusts to size, Experiment 4 and Experiment 5 turned out being almost comparable with respect to
their economy-wide effects.
Alternative assumptions on radical innovation occurrence

We test the robustness of our results to an alternative assumption concerning the occurrence of a radical innovation.
Specifically, we design a scenario - labeled as Exp. V (only own R&D) - where the probability of discovering a radical inno-
vation (see Eqs. 12 and 24) depends only on the cumulative R&D performed within the organization. This holds true both
for the NRL and other capital good firms.
Table 12 collects the results and contrast them with the baseline and the Exp. V discussed the main text (standard). In the
baseline scenario, we find no differences between the current setup of the model with respect to the benchmark one (see
Table 4). Indeed, in both cases, the cumulative stock of knowledge developed within a single capital good firm turned out
being insufficient to the discovery of a radical innovation during our simulation runs (see also Table 9). Similarly, the NRL is
way less effective at stimulating growth than in the benchmark Exp. V, wherein it leverages on the knowledge stock of the
entire economy. It achieves - on average - less radical innovations, which obviously reflects on the average dynamics of the
model. Indeed, the economy shows significantly lower growth, relatively deficits, as well as markedly reduced volatility than
in the standard experiment. Unemployment is low in both scenarios and statistically significant difference are not found. On
average, this setting of experiment V is still superior to other innovation policies (Table 4).
24
Table 12
Baseline and two alternative specifications of Experiment 5: as in the main text (standard)
and with radical innovation depending on own cumulative R&D. The second and third rows
report the average relative performance of with respect to the “no innovation policy” baseline
(Baseline) over 200 Monte Carlo runs, but for the number of radical innovations (# rad. inn).
Symbol ∗ indicates a statistically significant difference with respect to the baseline, at 5% and
resulting from a t-test on the means. Similarly, symbol † indicates a significant difference
with respect to the Exp. V of the main text (standard). GDP vol. stands for GDP volatility
proxied by the standard deviation of the growth process; unempl. stands for unemployment;
deficit is expressed as relative to GDP.
Policy # rad. inn. GDP growth GDP vol. Unempl. Deficit
Baseline 0 2.68% 0.08 6.10% 4.34%

Exp V (standard) 2.15 1.55∗ 2.01∗ 0.88∗ 0.78∗
Exp. V (only own R&D) 0.96 1.23∗ ,† 1.47∗ ,† 0.92∗ 0.89∗ ,†
References
Aghion, P., Howitt, P., 1992. A model of growth through creative destruction. Econometrica 60 (2), 323–351. http://www.jstor.org/stable/2951599.
Aistleitner, M., Gräbner, C., Hornykewycz, A., 2021. Theory and empirics of capability accumulation: implications for macroeconomic modeling. Res Policy
50 (6), 104258.
Akcigit, U., Hanley, D., Serrano-Velarde, N., 2020. Back to basics: basic research spillovers, innovation policy and growth. Rev Econ Stud 88 (1), 1–43.
doi:10.1093/restud/rdaa061.
Andriani, P., Cattani, G., 2016. Exaptation as source of creativity, innovation, and diversity: introduction to the special section. Industrial and Corporate
Change 25 (1), 115–131.
Arrow, K.J., 1951. An extension of the basic theorems of classical welfare economics. Technical Report. Proceedings of the second berkeley symposium on
mathematical statistics and probability.
Arrow, K.J., 1962. Economic welfare and the allocation of resources for invention. In: Nelson, R.R. (Ed.), The Rate and Direction of Inventive Activity. Prince-
tion University Press, pp. 609–625.
Arrow, K.J., 1996. Technical information and industrial structure. Industrial and corporate change 5 (2), 645–652.
Ausloos, M., Miśkiewicz, J., Sanglier, M., 2004. The durations of recession and prosperity: does their distribution follow a power or an exponential law?
Physica A 339 (3–4), 548–558. doi:10.1016/j.physa.20 04.03.0 05.
Bartelsman, E., Scarpetta, S., Schivardi, F., 2005. Comparative analysis of firm demographics and survival: evidence from micro-level sources in OECD coun-
tries. Industrial and Corporate Change 14 (3), 365–391.
Bartelsman, E.J., Doms, M., 20 0 0. Understanding productivity: lessons from longitudinal microdata. J Econ Lit 38 (3), 569–594.
Becker, B., 2015. Public r&d policies and private r&d investment: asurvey of the empirical evidence. J Econ Surv 29 (5), 917–942.
Block, F.L., Keller, M.R., 2015. State of Innovation: The US Government’s Role in Technology Development. Routledge.
Bloom, N., Van Reenen, J., Williams, H., 2019. A toolkit of policies to promote innovation. Journal of Economic Perspectives 33 (3), 163–184.
Boca, A., Galeotti, M., Himmelberg, C.P., Rota, P., 2008. Investment and time to plan and build: a comparison of structures vs. equipment in a panel of Italian
firms. J Eur Econ Assoc 6 (4), 864–889.
Borrás, S., Edquist, C., 2013. The choice of innovation policy instruments. Technol Forecast Soc Change 80 (8), 1513–1522.
Bottazzi, G., Secchi, A., 2003. Common properties and sectoral specificities in the dynamics of US manufacturing companies. Review of Industrial Organiza-
tion 23 (3–4), 217–232.
Bottazzi, G., Secchi, A., 2006. Explaining the distribution of firm growth rates. Rand J Econ 37 (2), 235–256.
Burns, A.F., Mitchell, W.C., 1946. Measuring business cycles. National Bureau of Economic Research. http://EconPapers.repec.org/RePEc:nbr:nberbk:burn46-1.
Bush, V., et al., 1945. Science, the endless frontier. Technical Report. US Office of Scientific Research and Development.
Caiani, A., Russo, A., Gallegati, M., 2019. Does inequality hamper innovation and growth? an AB-SFC analysis. Journal of Evolutionary Economics 29 (1),
177–228.
Castaldi, C., Dosi, G., 2009. The patterns of output growth of firms and countries: scale invariances and scale specificities. Empir Econ 37 (3), 475–495.
http://EconPapers.repec.org/RePEc:spr:empeco:v:37:y:2009:i:3:p:475-495.
Castelnovo, P., Florio, M., 2020. Mission-oriented public organizations for knowledge creation. In: The Routledge Handbook of State-Owned Enterprises.
Routledge, pp. 587–604.
Cerra, V., Fatás, A., Saxena, S., 2021. Hysteresis and business cycles forthcoming.
Chataway, J., Brusoni, S., Cacciatori, E., Hanlin, R., Orsenigo, L., 2007. The international AIDS vaccine initiative (IAVI) in a changing landscape of vaccine
development: a public/private partnership as knowledge broker and integrator. The European Journal of Development Research 19 (1), 100–117.
Chiang, J.-T., 1991. From ’mission-oriented’to ’diffusion-oriented’paradigm: the new trend of US industrial technology policy. Technovation 11 (6), 339–356.
Cimoli, M., Dosi, G., Maskus, K.E., Okediji, R.L., Reichman, J.H., Stiglitz, J.E., 2014. Intellectual Property Rights: Legal and Economic Challenges for Develop-
ment. Oxford University Press.
Cimoli, M., Dosi, G., Stiglitz, J., 2009. Industrial Policy and Development: The Political Economy of Capabilities Accumulation. Oxford University Press.
Coad, A., Rao, R., 2010. Firm growth and R&D expenditure. Economics of Innovation and New Technology 19 (2), 127–145.
Cohen, W.M., 2010. Fifty years of empirical studies of innovative activity and performance. In: Handbook of the Economics of Innovation, Vol. 1. Elsevier,
pp. 129–213.
Cohen, W.M., Klepper, S., 1992. The anatomy of industry R&D intensity distributions. Am Econ Rev 773–799.
Cohen, W.M., Levin, R.C., Mowery, D.C., 1987. Firm size and r & d intensity: a re-examination. J Ind Econ 543–565.
Criscuolo, C., Gonne, N., Kitazawa, K., Lalanne, G., 2020. A fresh look at industrial policies: old debates, new perspectives. Technical Report. OECD, DSTI-CIIE.
Dawid, H., 2006. Agent-based models of innovation and technological change. Handbook of computational economics 2, 1235–1272.
Dawid, H., Delli Gatti, D., 2018. Agent-based macroeconomics. Handbook of Computational Economics 4, 63–156.
Dawid, H., Gemkow, S., Harting, P., Kabus, K., Neugart, M., Wersching, K., 2008. Skills, innovation, and growth: an agent-based policy analysis. Jahrbücher
für Nationalökonomie und Statistik 228 (2–3), 251–275.
DeLong, D.W., Storey, J., et al., 2004. Lost Knowledge: Confronting the Threat of an Aging Workforce. Oxford University Press.
Di Comite, F., Kancs, d., 2015. Macro-economic models for R&D and innovation policies. Technical Report. IPTS Working Papers on Corporate R&D and
Innovation.
Dimos, C., Pugh, G., 2016. The effectiveness of R&D subsidies: ameta-regression analysis of the evaluation literature. Res Policy 45 (4), 797–815.
Doms, M., Dunne, T., 1998. Capital adjustment patterns in manufacturing plants. Rev Econ Dyn 1 (2), 409–429.
Dosi, G., 1982. Technological paradigms and technological trajectories: a suggested interpretation of the determinants and directions of technical change.
Res Policy 11 (3), 147–162.
25
Dosi, G., 1988. Sources, procedures, and microeconomic effects of innovation. J Econ Lit 26 (3), 1120–1171. http://www.jstor.org/stable/2726526.
Dosi, G., 2007. Statistical regularities in the evolution of industries. a guide through some evidence and challenges for the theory. In: Malerba, F., Brusoni, S.
(Eds.), Perspectives on innovation. Cambridge University Press, pp. 1110–1121.
Dosi, G., Fagiolo, G., Napoletano, M., Roventini, A., 2013. Income distribution, credit and fiscal policies in an agent-based Keynesian model. Journal of
Economic Dynamics and Control 37 (8), 1598–1625. doi:10.1016/j.jedc.2012.11.008.
Dosi, G., Fagiolo, G., Napoletano, M., Roventini, A., Treibich, T., 2015. Fiscal and monetary policies in complex evolving economies. Journal of Economic
Dynamics and Control 52, 166–189. doi:10.1016/j.jedc.2014.11.014.
Dosi, G., Fagiolo, G., Roventini, A., 2010. Schumpeter meeting Keynes: a policy-friendly model of endogenous growth and business cycles. Journal of Eco-
nomic Dynamics and Control 34 (9), 1748–1767.
Dosi, G., Freeman, C., Fabiani, S., 1994. The process of economic development: introducing some stylized facts and theories on technologies, firms and
institutions. Industrial and corporate change 3 (1), 1–45.
Dosi, G., Freeman, R.B., Pereira, M.C., Roventini, A., Virgillito, M.E., 2020. The impact of deunionization on the growth and dispersion of productivity and
pay. Technical Report. National Bureau of Economic Research.
Dosi, G., Marengo, L., Pasquali, C., 2006. How much should society fuel the greed of innovators?: on the relations between appropriability, opportunities
and rates of innovation. Res Policy 35 (8), 1110–1121.
Dosi, G., Napoletano, M., Roventini, A., Stiglitz, J.E., Treibich, T., 2020. Rational heuristics? expectations and behaviors in evolving economies with heteroge-
neous interacting agents. Econ Inq 58 (3), 1487–1516.
Dosi, G., Napoletano, M., Roventini, A., Treibich, T., 2016. The short- and long-run damages of fiscal austerity: Keynes beyond schumpeter. In: Stiglitz, J.,
Guzman, M. (Eds.), Contemporary Issues in Macroeconomics. Palgrave Macmillan UK.
Dosi, G., Napoletano, M., Roventini, A., Treibich, T., 2017. Micro and macro policies in the keynes+ schumpeter evolutionary models. Journal of Evolutionary
Economics 27 (1), 63–90.
Dosi, G., Nelson, R.R., 2010. Technological change and industrial dynamics as evolutionary processes. In: Hall, B.H., Rosenberg, N. (Eds.), Handbook of the
Economics of Innovation. Elsevier: Amsterdam, pp. 51–127. chapter 4
Dosi, G., Pereira, M.C., Roventini, A., Virgillito, M.E., 2017. When more flexibility yields more fragility: the microfoundations of Keynesian aggregate unem-
ployment. Journal of Economic Dynamics and Control 81, 162–186.
Dosi, G., Pereira, M.C., Roventini, A., Virgillito, M.E., 2018. Causes and consequences of hysteresis: aggregate demand, productivity, and employment. Indus-
trial and Corporate Change 27 (6), 1015–1044.
Dosi, G., Pereira, M.C., Roventini, A., Virgillito, M.E., 2022. Technological paradigms, labour creation and destruction in a multi-sector agent-based model.
Research Policy 51 (10), 104565.
Dosi, G., Roventini, A., 2019. More is different... and complex!: the case for agent-based macroeconomics. Journal of Evolutionary Economics 29, 1–37.
doi:10.10 07/s0 0191-019- 00609- y.
Dosi, G., Roventini, A., Russo, E., 2019. Endogenous growth and global divergence in a multi-country agent-based model. Journal of Economic Dynamics and
Control 101, 101–129.
Dosi, G., Roventini, A., Russo, E., et al., 2020. Public policies and the art of catching up: matching the historical evidence with a multicountry agent-based
model. Industrial and Corporate Change doi:10.1093/icc/dtaa057.
Dosi, G., Stiglitz, J.E., 2014. The role of intellectual property rights in the development process, with some lessons from developed countries: an introduction.
In: Cimoli, M., Dosi, G., Stiglitz, J.E. (Eds.), Intellectual Property Rights: Legal and Economic Challenges for Development. Oxford University Press Oxford.
Dosi, G., Virgillito, M.E., 2017. In order to stand up you must keep cycling: change and coordination in complex evolving economies. Structural Change and
Economic Dynamics 56, 353–364.
Dunlap-Hinkler, D., Kotabe, M., Mudambi, R., 2010. A story of breakthrough versus incremental innovation: corporate entrepreneurship in the global phar-
maceutical industry. Strategic Entrepreneurship Journal 4 (2), 106–127.
Edler, J., Fagerberg, J., 2017. Innovation policy: what, why, and how. Oxford Review of Economic Policy 33 (1), 2–23.
Edquist, C., 2011. Design of innovation policy through diagnostic analysis: identification of systemic problems (or failures). Industrial and corporate change
20 (6), 1725–1753.
Eisner, R., 1972. Components of capital expenditures: replacement and modernization versus expansion. Rev Econ Stat 54 (3), 297–305.
Engel, D., Rothgang, M., Eckl, V., 2016. Systemic aspects of r&d policy subsidies for r&d collaborations and their effects on private r&d. Industry and
Innovation 23 (2), 206–222.
Ergas, H., 1987. Does technology policy matter? In: Guile, B.R., Brooks, H. (Eds.), Technology and Global Industry: Companies and Nations in the World
Economy. National Academy Press Washington, DC.
EU, 2020. Moving the frontier of macroeconomic modelling of research and innovation policy. Technical Report. Directorate General for Research and
Innovation (European Commission) doi:10.2777/34199s.
Fagerberg, J., 2017. Innovation policy: rationales, lessons and challenges. J Econ Surv 31 (2), 497–512.
Fagiolo, G., Giachini, D., Roventini, A., 2020. Innovation, finance, and economic growth: an agent-based approach. Journal of Economic Interaction and
Coordination 15 (3), 703–736.
Fagiolo, G., Guerini, M., Lamperti, F., Moneta, A., Roventini, A., 2019. Validation of agent-based models in economics and finance. In: Computer Simulation
Validation. Springer, pp. 763–787.
Fagiolo, G., Napoletano, M., Roventini, A., 2008. Are output growth-rate distributions fat-tailed? some evidence from OECD countries. Journal of Applied
Econometrics 23 (5), 639–669.
Fagiolo, G., Roventini, A., 2017. Macroeconomic policy in DSGE and agent-based models redux: new developments and challenges ahead. Journal of Artificial
Societies and Social Simulation 20 (1).
Feldstein, M.S., Foot, D.K., 1971. The other half of gross investment: replacement and modernization expenditures. The Review of Economics Statistic 53 (1),
49–58.
Foray, D., Mowery, D.C., Nelson, R.R., 2012. Public r&d and social challenges: what lessons from mission r&d programs? Res Policy 41 (10), 1697–1702.
Freeman, C., 1987. Technology, Policy, and Economic Performance: Lessons from Japan. Pinter Pub Ltd.
Freeman, C., Soete, L., 1997. The Economics of Industrial Innovation. Psychology Press.
Goolsbee, A., 1998. The business cycle, financial performance, and the retirement of capital goods. Rev Econ Dyn 1 (2), 474–496.
Gräbner, C., Hornykewycz, A., 2022. Capability accumulation and product innovation: an agent-based perspective. Journal of Evolutionary Economics 32 (1),
87–121.
Grilli, L., Mazzucato, M., Meoli, M., Scellato, G., 2018. Sowing the seeds of the future: policies for financing tomorrow’s innovations. Technol Forecast Soc
Change 127, 1–7.
Gross, D.P., Sampat, B.N., 2020. Inventing the endless frontier: The effects of the World War II research effort on post-war innovation. Technical Report.
National Bureau of Economic Research.
Gross, D.P., Sampat, B.N., 2021. The Economics of Crisis Innovation Policy: A Historical Perspective. Technical Report. National Bureau of Economic Research.
Guellec, D., Van Pottelsberghe De La Potterie, B., 2003. The impact of public R&D expenditure on business R&D. Economics of innovation and new technol-
ogy 12 (3), 225–243.
Guerini, M., Moneta, A., 2017. A method for agent-based models validation. Journal of Economic Dynamics and Control 82, 125–141.
Haldane, A.G., Turrell, A.E., 2019. Drawing on different disciplines: macroeconomic agent-based models. Journal of Evolutionary Economics 29, 39–66.
doi:10.10 07/s0 0191-018-0557-5.
Hidalgo, C.A., Klinger, B., Barabási, A.-L., Hausmann, R., 2007. The product space conditions the development of nations. Science 317 (5837), 482–487.
26
Isley, S.C., Lempert, R.J., Popper, S.W., Vardavas, R., 2015. The effect of near-term policy choices on long-term greenhouse gas transformation pathways.
Global Environ. Change 34, 147–158.
Jaimovich, N., Floetotto, M., 2008. Firm dynamics, markup variations, and the business cycle. J Monet Econ 55 (7), 1238–1252.
Kantor, S., Whalley, A., 2022. Moonshot: Public R&D and growth.
Kaplan, S., Vakili, K., 2015. The double-edged sword of recombination in breakthrough innovation. Strategic Management Journal 36 (10), 1435–1457.
Kuznets, S.S., Murphy, J.T., 1966. Modern Economic Growth: Rate, Structure, and Spread, Vol. 2. Yale University Press New Haven.
Lamperti, F., 2018. Empirical validation of simulated models through the GSL-div: an illustrative application. Journal of Economic Interaction and Coordina-
tion 13 (1), 143–171.
Lamperti, F., 2018. An information theoretic criterion for empirical validation of simulation models. Econometrics and Statistics 5, 83–106.
Lamperti, F., Bosetti, V., Roventini, A., Tavoni, M., 2019. The public costs of climate-induced financial instability. Nat Clim Chang 9 (11), 829–833.
Lamperti, F., Bosetti, V., Roventini, A., Tavoni, M., Treibich, T., 2021. Three green financial policies to address climate risks. Journal of Financial Stability 54,
100875. doi:10.1016/j.jfs.2021.100875.
Lamperti, F., Dosi, G., Napoletano, M., Roventini, A., Sapio, A., 2018. Faraway, so close: coupled climate and economic dynamics in an agent-based integrated
assessment model. Ecol. Econ. 150, 315–339.
Lamperti, F., Dosi, G., Napoletano, M., Roventini, A., Sapio, A., 2018. Faraway, so close: coupled climate and economic dynamics in an agent-based integrated
assessment model. Ecol. Econ. 150, 315–339. doi:10.1016/j.ecolecon.2018.03.023.
Lamperti, F., Dosi, G., Napoletano, M., Roventini, A., Sapio, A., 2020. Climate change and green transitions in an agent-based integrated assessment model.
Technol Forecast Soc Change 153, 119806.
Lamperti, F., Mattei, C.E., 2018. Going up and down: rethinking the empirics of growth in the developing and newly industrialized world. Journal of Evolu-
tionary Economics 28 (4), 749–784.
Lamperti, F., Roventini, A., Sani, A., 2018. Agent-based model calibration using machine learning surrogates. Journal of Economic Dynamics and Control 90,
366–389.
Lazonick, W., Tulum, O., 2011. Us biopharmaceutical finance and the sustainability of the biotech business model. Res Policy 40 (9), 1170–1187.
Levin, R.C., Klevorick, A.K., Nelson, R.R., Winter, S.G., Gilbert, R., Griliches, Z., 1987. Appropriating the returns from industrial research and development.
Brookings Pap Econ Act 1987 (3), 783–831.
Lorentz, A., Ciarli, T., Savona, M., Valente, M., 2016. The effect of demand-driven structural transformations on growth and technological change. Journal of
Evolutionary Economics 26 (1), 219–246.
Lundvall, B.-A., 1988. Innovation as an interactive process: from user-producer interaction to the national system of innovation. In: Dosi, G. (Ed.), Technical
Change and Economic Theory. London: Pinter, pp. 349–369.
Lundvall, B.-A., 2010. National systems of innovation: Toward a theory of innovation and interactive learning. London: Pinter.
Malerba, F., 2002. Sectoral systems of innovation and production. Res Policy 31 (2), 247–264.
Marino, M., Lhuillery, S., Parrotta, P., Sala, D., 2016. Additionality or crowding-out? an overall evaluation of public r&d subsidy on private r&d expenditure.
Res Policy 45 (9), 1715–1730.
Mazzucato, M., 2013. The Entrepreneurial State: Debunking Public vs. Private Sector Myths. Anthem Press, London.
Mazzucato, M., 2015. The green entrepreneurial state. In: Scoones, I., Leach, M., Newell, P. (Eds.), The Politics of Green Transformations; Scoones, I., Leach,
M., Newell, P., Eds. Routledge London, pp. 134–152.
Mazzucato, M., 2016. From market fixing to market-creating: a new framework for innovation policy. Industry and Innovation 23 (2), 140–156.
Mazzucato, M., 2018. Mission-oriented innovation policies: challenges and opportunities. Industrial and Corporate Change 27 (5), 803–815.
Mazzucato, M., 2018. Mission-oriented research & innovation in the European Union. Technical Report. Directorate-General for Research and Innovation
(European Commission).
Mazzucato, M., 2019. Governing missions in the European Union. Technical Report. Directorate-General for Research and Innovation (European Commission).
Mazzucato, M., 2021. Mission Economy. Penguin, Allen Lane-Penguin.
Mazzucato, M., Semieniuk, G., 2017. Public financing of innovation: new questions. Oxford Review of Economic Policy 33 (1), 24–48.
Mazzucato, M., Semieniuk, G., 2018. Financing renewable energy: who is financing what and why it matters. Technol Forecast Soc Change 127, 8–22.
Metcalfe, J.S., 1994. Evolutionary economics and technology policy. The economic journal 104 (425), 931–944.
Metcalfe, J.S., 1995. Technology systems and technology policy in an evolutionary framework. Cambridge J Econ 19 (1), 25–46.
Metcalfe, J.S., 2005. Systems failure and the case for innovation policy. In: Innovation Policy in a Knowledge-based Economy. Springer, pp. 47–74.
Mohnen, P., Vankan, A., Verspagen, B., 2017. Evaluating the innovation box tax policy instrument in the netherlands, 2007–13. Oxford Review of Economic
Policy 33 (1), 141–156.
Moretti, E., Steinwender, C., Van Reenen, J., 2019. The Intellectual Spoils of War? Defense R&D, Productivity and International Spillovers. Technical Report.
National Bureau of Economic Research.
Motoyama, Y., Appelbaum, R., Parker, R., 2011. The national nanotechnology initiative: federal support for science and technology, or hidden industrial
policy? Technol Soc 33 (1–2), 109–118.
Mowery, D.C., 2010. Military R&D and Innovation. In: Handbook of the Economics of Innovation, Vol. 2. Elsevier, pp. 1219–1256.
Napoletano, M., Roventini, A., Sapio, S., 2006. Are business cycles all alike? a bandpass filter analysis of Italian and US cycles. Rivista Italiana degli Economisti
1, 87–118.
Nelson, R. (Ed.), 1982. Government and Technical Progress: A Cross-industry Analysis. Pergamon Press, New York; Toronto.
Nelson, R., Winter, S., 1982. An evolutionary theory of economic change. Harvard Business School Press, Cambridge.
Nelson, R.R., 1962. Introduction. In: Nelson, R.R. (Ed.), The Rate and Direction of Inventive Activity. Princetion University Press, pp. 609–625.
Nelson, R.R., 1982. The role of knowledge in R&D efficiency. Q J Econ 97 (3), 453–470.
OECD, 2010. The OECD innovation strategy: Getting a head start on tomorrow.
Pallante, G., Russo, E., Roventini, A., et al., 2020. Does mission-oriented funding stimulate private R&D? Evidence from military R&D for US states. Technical
Report 2020/32. LEM Working Papers, Sant’Anna School of Advanced Studies.
Papageorgiadis, N., Sharma, A., 2016. Intellectual property rights and innovation: a panel analysis. Econ Lett 141, 70–72.
Phelps, E.S., Winter, S.G., 1970. Optimal price policy under atomistic competition. In: Microeconomic Foundations of Employment and Inflation Theory. WW
Norton New York, pp. 309–337.
Phene, A., Fladmoe-Lindquist, K., Marsh, L., 2006. Breakthrough innovations in the US biotechnology industry: the effects of technological space and geo-
graphic origin. Strategic management journal 27 (4), 369–388.
Polanyi, K., 1944. The Great Transformation: The Political and Economic Origins of Our Time. Boston: Beacon.
Reissl, S., Caiani, A., Lamperti, F., Guerini, M., Vanni, F., Fagiolo, G., Ferraresi, T., Ghezzi, L., Napoletano, M., Roventini, A., et al., 2021. Assessing the economic
effects of lockdowns in Italy: a dynamic Input-Output approach. Technical Report 2021/03. LEM Working Papers, Sant’Anna School of Advanced Studies.
Romer, P.M., 1986. Increasing returns and long-run growth. Journal of political economy 94 (5), 1002–1037.
Rosenberg, N., Nelson, R.R., 1994. American universities and technical advance in industry. Res Policy 23 (3), 323–348.
Rotemberg, J.J., 2008. Behavioral aspects of price setting, and their policy implications. Technical Report. National Bureau of Economic Research.
Russo, A., Catalano, M., Gaffeo, E., Gallegati, M., Napoletano, M., 2007. Industrial dynamics, fiscal policy and r&d: evidence from a computational experiment.
Journal of Economic Behavior & Organization 64 (3–4), 426–447.
Ruttan, V.W., 2006. Is War Necessary for Economic Growth?: Military Procurement and Technology Development. Oxford University Press.
Santoleri, P., Mina, A., Minin, A.D., Martelli, I., 2020. The causal effects of R&D grants: evidence from a regression discontinuity. Technical Report. LEM
Working Papers, Sant’Anna School of Advanced Studies.
27
Silverberg, G., Verspagen, B., 1994. Learning, innovation and economic growth: a long-run model of industrial dynamics. Industrial and Corporate Change 3
(1), 199–223.
Solow, R.M., 1957. Technical change and the aggregate production function. Rev Econ Stat 39 (3), 312–320.
Steffen, B., Egli, F., Schmidt, T.S., et al., 2020. The role of public banks in catalyzing private renewable energy finance. In: Donovan, C.W. (Ed.), Renewable
Energy Finance: Funding The Future Of Energy. World Scientific, pp. 197–216.
Stiglitz, J.E., 1996. Whither socialism?. MIT press.
Stiglitz, J.E., 2014. Intellectual property rights, the pool of knowledge, and innovation. Technical Report. National Bureau of Economic Research.
Stock, J.H., Watson, M.W., 1999. Business cycle fluctuations in US macroeconomic time series. In: Handbook of Macroeconomics, Vol. 1, Part A. Elsevier,
pp. 3–64. doi:10.1016/S1574-0048(99)01004-6.
Tacchella, A., Cristelli, M., Caldarelli, G., Gabrielli, A., Pietronero, L., 2012. A new metrics for countries’ fitness and products’ complexity. Sci Rep 2 (1), 1–7.
Teece, D.J., 2019. Roundtable discussion nathan rosenberg memorial issue. Industrial and Corporate Change 28 (2), 409–417.
Wälde, K., Woitek, U., 2004. R&D expenditure in G7 countries and the implications for endogenous fluctuations and growth. Econ Lett 82 (1), 91–97.
Windrum, P., Fagiolo, G., Moneta, A., 2007. Empirical validation of agent-based models: alternatives and prospects. Journal of Artificial Societies and Social
Simulation 10 (2), 8.
Winter, S., 1998. Knowledge and competence as strategic assets. In: Klein, D.A. (Ed.), The Strategic Management of Intellectual Capital. Butterworth-Heine-
mann.
Wright, I., 2005. The duration of recessions follows an exponential not a power law. Physica A 345 (3), 608–610.
Zarnowitz, V., 1985. Recent work on business cycles in historical perspective: a review of theories and evidence. J Econ Lit 23 (2), 523–580. http://www.
jstor.org/stable/2725624.
Zúñiga Vicente, J.A., Alonso-Borrego, C., Forcadell, F.J., Galán, J.I., 2014. Assessing the effect of public subsidies on firm R&D investment: a survey. J Econ
Surv 28 (1), 36–67.
28

Dosi y Otros 2023 Mission-Oriented Policies and The "Entrepreneurial State" at Work - An Agent-Based Exploration

Uploaded by

Copyright:

Available Formats

Dosi y Otros 2023 Mission-Oriented Policies and The "Entrepreneurial State" at Work - An Agent-Based Exploration

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Dosi y Otros 2023 Mission-Oriented Policies and The "Entrepreneurial State" at Work - An Agent-Based Exploration

Uploaded by

Copyright:

Available Formats

Journal of Economic Dynamics & Control 151 (2023) 104650

Contents lists available at ScienceDirect

Journal of Economic Dynamics & Control

Mission-oriented policies and the “Entrepreneurial State” at

2. Innovation policies: A critical review

3. The K+S model

3.1. Innovation and technological progress

RDi (t ) = υ Si (t − 1 ), υ ∈ {0, 1} (3)

INi (t ) = ξ RDi (t ), IMi (t ) = (1 − ξ )RDi (t ), ξ ∈ [0, 1]. (4)

θiIN (t ) = 1 − e−oIN INi (t ) , oIN > 0, (5)

θiIM (t ) = 1 − e−oIM IMi (t ) , oIM > 0; (6)

Ai,τ +1 = Ai,τ (1 + χA,i ) (7)

Bi,τ +1 = Bi,τ (1 + χB,i ) (8)

min[ phi (t ) + bccon,h ] h ∈ {in, im, τ }, (10)

Fig. 1. Shift of technological opportunities implied by radical innovations.

3.2. Investment and technological diffusion

where p j (t ) is the price of a new machine available to ﬁrm j at time t.

3.3. The public sector and the macroeconomic framework

C (t ) = w(t )LD (t ) + wU [LS (t ) − LD (t )], (16)

De f (t ) = G(t ) − T axes(t ) + CD(t ) + IP (t ), (18)

4. Simulation set-up and empirical validation

Stylized facts Empirical studies (among others)

Macroeconomic stylized facts

Variable Mean St. Dev Variable Mean St. Dev

GDP growth 0.0268 0.0012 Unemployment 0.0610 0.0376

Size of the subsidy

5. Innovation policy experiments

5.1. A “menu” of innovation policies

Fig. 3. Experiment 4 : knowledge diffusion by the public ﬁrm.

5.2. Simulation results

Baseline 2.68% 0.08 6.10% 16% 4.34%

Per-period imitations of the public ﬁrm

Mean Max Min St. Dev.

log RDi,s,t = β1 log IPs,t−1 + β2 log GDPs,t−1 + λi + μs + νt + εi,s,t (25)

Dependent variable: log ﬁrm R&D(t)

(baseline) (Exp. I) (Exp. II) (Exp. III) (Exp. IV) (Exp. V)

Dependent variable: log ﬁrm R&D(t)

Description Symbol Value

Monte Carlo replications MC 200

Additional policy and behavioural experiments

Indirect innovation policies under alternative designs

benchmark 0 0% 3.40% 5.38% 5.17% 2.15 86% 4.15% 5.37% 3.39%

Baseline 2.68% 0.08 6.10% 16% 4.34%

Innovation policies under different R&D investment routines

symmetric reaction to growth asymmetric reaction to growth

Baseline 2.55% 0.10 4.11% 2.67% 0.13 2.17%

RDi (t ) = υ Si (t − 1 ) ∗ [1 + θ1 gSi (t − 1 )1[gSi < 0] + θ2 gSi (t − 1 )1[gSi > 0] + θ3 fRD,i (t − 1 )] (27)

Alternative assumptions on radical innovation occurrence

Policy # rad. inn. GDP growth GDP vol. Unempl. Deﬁcit

Baseline 0 2.68% 0.08 6.10% 4.34%

You might also like