production, Alchian (1950) and Friedman (1953) took an instrumentalist stance towards economics as a
science. Their argumentation holds that, shortly stated, the assumptions of a theory need not be realistic as
long as the hypotheses derived from the theory are helpful in predicting the future.
Ironically, Friedman’s use of the theory of natural selection to defend neoclassical economics has
probably accelerated to development of evolutionary economics as an alternative to neoclassical
economics. Not long after Friedman’s (1953) essay, scholars argued if economic ‘natural selection’
underlies the economy, economic theory should explicitly model the economy as an evolutionary process
rather than implicitly assuming this to be the case (Koopmans 1957; Winter 1964). This position is in line
with realism, as opposed to instrumentalism, in that evolutionary models try to unravel the precise
mechanisms driving change. Put differently, scholars adhering to realism are primarily interested in
understanding the underlying mechanism in society, and only secondarily in prediction (which is
inherently difficult due to the sheer complexity of social and economic life). In this light, the most
important contribution to evolutionary economics has been the seminal work by Nelson and Winter
(1982). They showed that an important part of neoclassical theory can indeed be formally reformulated as
an evolutionary theory of economic change. A core assumption in their work has been that firms are
bounded rational (Simon 1955), and react to their economic environment according to routines rather than
through maximisation of profits. From this, it follows that the competitiveness of firms depends on their
capability to improve their routines through innovation over time. Put differently, where neoclassical
economics deals with price competition and static efficiency, evolutionary economics deals with
Schumpeterian competition and dynamic efficiency (Schumpeter 1942). This explains why evolutionary
economics is also called neo- or post-Schumpeterian economics (Andersen 1994), and since have focused
on Schumpeterian themes as technological development, industrial organisation, and long-term
macroeconomic growth.
3.2 The mathematics debate
The second debate unites evolutionary scholars with neoclassicals, and differentiates them from
institutional economists. The debate centres around the role of mathematical modelling in economics, or
perhaps more accurately, in social science in general. Where evolutionary and neoclassical economists
frequently formalise ideas in abstract modelling, the majority of institutionalists rejects the use of
modelling, because it does not capture the contextual nature of economic life, and life in general
(Hodgson, 1998; Martin, 2000). This rejection of the use of formal mathematical models is not true for all
proponents of institutional economics. Hodgson (1998) summarised the attitude of institutional
economists toward mathematical modelling as follows: “ ... mathematical and statistical techniques are
recognized as the servants of, rather than the essence of economic theory” (p. 173). In doing so, they
criticise the neo-classical attitude to exclude those factors from their analysis that might be relevant (such
as institutional factors), but cannot be formalised and expressed in mathematical forms within a
maximization-equilibrium framework.
Evolutionary economics, though critical towards the exact assumptions used in neoclassical models,
do not reject the use of mathematical modelling as such. Mistakenly, the use of mathematics has been
confused with an anti-realist modelling. As argued by Maki (1992) on various occasions, abstracting from
specifics by formalising particular economic mechanism does not necessarily imply an instrumentalist
position on models as tools for prediction. Mathematical modelling can also serve a realist program aimed
10