Geography Reference
In-Depth Information
purpose of which is to determine the existence and stability of equilibrium.
5
Any notion
of 'dynamics' in such models concerns only the model economy's movement towards its
equilibrium: once in that equilibrium state, the system is in stasis. Equilibrist economics
is basically antithetical to notions of history and evolution (Harris, 2004; Kaldor, 1934,
1972, 1985; Robinson, 1974; Setteri eld, 1995, 1997).
In contrast to equilibrium economic processes, in an economy characterised by path
dependence the specii c details and sequence of historical events govern the unfolding
course of development - what David calls 'historically contingent evolution'. Following
Harris (2004) and Page (2006), such a process might be expressed as:
Historical process: x
(
t
+ 1) =
F
x
(
t
)(
h
(
t
)
x
), −∞ ≤
t
≤ +∞,
h
(
t
)
x
=
x
(
t
),
x
(
t
− 1) . . .
x
(0)
where
h
(
t
)
x
is the history of past outcomes of
x
from
t
= 0 up to time
t
, and the function
F
x
(
t
) maps that history into the next outcome.
6
The outcome function can itself change
over time so it is indexed by
t
. In such a system, the present state of the economy will
depend on where it has come from, and on how it got there: this is what is generally
meant by path dependence. There is nothing inherent in such an historical process that
necessitates that it possesses or reaches a stable equilibrium state. Indeed, the concept
of path dependence can be argued to be fundamentally antagonistic to an equilibrist
methodology. According to the latter, the long-run equilibrium state can be dei ned and
reached independently of the path taken towards it, whereas with path dependence any
long- run coni guration that is reached by the economy will
depend
on the path taken
towards it. There is thus no predetermined economic outcome, no outcome independ-
ent of history or context. To argue that an economy, an economic landscape, is a path-
dependent historical process would thus seem to be incompatible with simultaneously
arguing that it is an equilibrium process.
Yet, while emphasising the need to move beyond the equilibrist methodology of main-
stream economics, Paul David and other path dependence theorists nevertheless seem
reluctant to relinquish the idea of 'equilibrium thinking' altogether.
7
In fact, recently,
David has explicitly referred to his approach to historical economics as 'path-dependent
equilibrium analysis' (David, 2005a, p. 153). His strategy to reconcile the apparent con-
tradiction in this phrase is to dei ne path dependence in terms of the dynamics associated
with particular types of non-ergodic stochastic processes and systems that possess a mul-
tiplicity of limiting distributions, that is multiple equilibria:
8
The elaboration of theories around the core concept of path dependent dynamics . . . encour-
ages and enables economists to entertain the possibility that, in place of a unique equilibrium-
seeking dynamic, they should envisage a process that is seeking an evolving and historically
contingent equilibrium. (David, 2005b, p. 2).
Under such conditions,
Small events of a random character - especially those occurring early on the path - are likely to
i gure signii cantly in 'selecting' one or other among the set of stable equilibria, or 'attractors'
(David, 2007, p. 151).
Which of these multiple equilibria is reached or 'selected', it is contended, will depend
on the initial state of the system - on initial 'random events' - and the chains of transitions
