Crossing the line
The most fundamental shortcoming is that very little social research has the instrumental
character that is most useful to policymakers. Very little of it can say to them if you do this, then
this will happen. Very little makes any connection crossing the line that divides government and
society seen in Figure 1.
Evaluators do make that connection, and this chiefly explains the influence of the MDRC
studies. But experimental evaluations succeed only in showing that a certain program had impact.
The treatment usually comprised several services, and it is unclear which of them generated the
results. Evaluators are also reluctant to generalize across different studies to say which type of
program is best overall, the thing that policymakers most want to know. For showing which
treatment produces outcomes and for integrating results across sites, statistical studies can be
superior to experiments, even though they do not estimate impact (see further below).
Political scientists prefer to write only about political or governmental processes, eschewing
stands on policy issues. They claim no authority to second-guess the decisions of the democratic
process. Few political scientists become policy experts who attempt to advise government; they
abandon this role largely to economists.
43 Thus, few scholars of the politics of welfare reform assert
much about the substance of policy. In terms of Figure 1, few make any connection between what
policymakers do and effects on the society.