Policy and Complex Systems
The methodological work to be done is huge , now we move only from ( A ) to ( B ) and to ( C ), using the recipes tool we have ( C ) emerging from ( B ). The crucial point in the model is that the network is emerging , not engineered ad hoc .
We stress this last sentence : we have interesting works on ABM and networks ( De Caux et al ., 2014 ; Hamill & Gilbert , 2009 ) where the agents act to generate the network ; in our context , agents are activated , following their internal rules and capabilities , by the events , and the network emerges as a side effect , as in the real world . We can apply to our ( C ) structures exactly the same range ( palette ) of algorithms that we can apply to ( D ).
As illustrated by Table 1 , that compares traditional NA with ABM – NA , the process that underlies the two kinds of modeling is radically different .
The same marked difference can be found in the output . ABM – NA produces a model that can be used to simulate the longterm effect of policy when individuals and their network adapt to it or introduce new behavior .
5 . The Broader Policymaking Perspective
Taking a broad complexity view on policymaking takes us very far from what has been done so far in the field . Even after disenchantment with the beneficial and efficient properties of the selforganization of market , economists have trusted their abilities to control the economy . The idea of intervening in the economy when it fails to adjust spontaneously or when there is the need to steer it toward a ( politically ) given direction has dominated economics in the last century independently of the prevailing theoretical background . In fact , the various schools of economic thought differ mainly in the prescribed control tools ( e . g ., monetary vs . fiscal policy ), sharing unfaltering confidence in the idea that economies work as machines and that “ equilibrium ” is the key to their functioning .
By “ control ” we mean the possibility of adjusting , according to the prescriptions of the various economic theories , some given variables such as public expenditure or the
Table 1
Steps Network Analysis Agent-based Network Analysis
I
II
III
IV
Take data from real world
Observe regularities
Generate theoretical networks with the desired properties
Measure the distance between theoretical and actual networks by means of network statistics
Take data from real world ( on microbehavior of network )
Build an ABM of the phenomenon of interest
Observe emerging networks ( if any )
Study the dynamic properties of the emerging network
87