ABSTRACT
To make the institutional relationship between model builders and policy makers viable, there have to be degrees of freedom. Policy makers may not like to be confined to discussing policy within a model framework defined by others, while model builders may like even less having to adapt to the whims of policy makers. Priorities in model-building and model improvements might thus differ. Policy makers would often be concerned with more trivial improvements like updating procedures, reconciliation of results from different (sub-)models, details of specification of no theoretical interest, etc. In the Norwegian context such needs were accommodated by the Ministry offering additional resources on a contractual basis. On the other hand, the model might in certain situations be perceived by the policy makers as being too much of a straitjacket. The MODIS model was much structure and little behaviour, and allowed the Ministry much discretionary judgement, while the MODAG model in recent years may have appeared less amenable in the confrontation between econometric evidence and the Ministry view of how the economy functioned. Generally, the Ministry seems to have stuck to a very disciplined use of the models with no fiddling with the results, but on some occasions it is known to have ‘disconnected’ certain relations while operating the model.
