Yes indeed, economic analysis does consist of ceteris paribus
exercises--but such analyses can be made much more convincing if the
effects of other factors (such as ongoing rationalization) affecting the
target variable one is interested in controlling (such as unemployment) are
somehow simultaneously accounted for, instead of simply being ignored--at
least as they appear to be in discussion thus far. This is a standard of
inquiry which is applicable to all policy analysis, and indeed is applied
in a broad range of policy evaluation studies using multivariate
methods--not just "unconventional" proposals to cut unemployment through
tax incentives.
The effects of any policy need to be assessed in relation to an
alternative scenario in which the proposed new policy is not implemented.
If the projected changes in the target variable which can be attributed to
the new policy do not stand in any reasonable proportion to likely changes
in the target variable in the absence of a policy, it can be questioned as
to whether there are good grounds for implementing it. Any such
projections are of course fundamentally uncertain, and there are always
considerable costs to "doing nothing"--they need to be included in the cost
evaluations too. If the relative merits of a WST can be argued on the
basis of more extensive benefit-cost analyses of this kind (perhaps
available in the paper which Geoff Beacon refers to), I'm naturally willing
to be convinced.
all best
EC
-------------------------------------------------------------
Posted to uk-policy, a service of Nexus. http://www.netnexus.org/
Hosting and email provided by new media consultants On-Line Publishing