Robert Reischauer and Phil Joyce's "The Federal Line Item Veto: What Is It and What Will It Do?" (Public Administration Review, March/April 1997) is a very informative and useful piece for both researchers and practitioners. However, even though the article provides a careful explanation of the federal line-item veto, it is inadequate in providing a good framework of analysis for researchers to follow in evaluating Its effects.
One problem with evaluating the line-item veto is that it isn't necessarily one procedure or rule that is always the same in its stringency or manner of operation. There are loopholes in the legislation itself, as well as strategies that can be used by Congress to limit or circumvent the new line-item power of the president (e.g., Congress can simply change the way appropriation bills are packaged). Most importantly, Congress can choose to preempt the president's item-veto power entirely by writing an exemption clause in the legislation. These loopholes imply that the line-item veto is less a limitation on the power of Congress than a statement of self-restraint that is as good as the level of control Congress is willing to put upon itself. That level may well vary.
A second problem in evaluating the effects of the line-item veto is that it Is difficult to determine whether this veto is an independent or dependent variable. In empirical analysis, in general, when an event is going to be evaluated for its effect or impact, the researcher assumes that it Is exogenous to the system of events. In a simple sense, if the change is exogenous (i.e., its value is determined outside the system or model) Its impact on the system can be determined. On the other hand, if a change is endogenous (i.e., its value is determined inside die system) it is not scientifically appropriate to evaluate the impact of this change to the system. Instead of being an agent of change, an endogenous event or process is an object of change. In other words, the first decision a researcher has to make is whether the federal line-item veto should be considered as an independent variable (cause) or a dependent variable (effect) (Cook and Campbell, 1979).
Whether a variable is independent or dependent all depends on the way a researcher sets up his or her model of analysis and the actual interaction among the events. If the line-item veto were introduced as a constitutional amendment in which there was independent interpretation from the Supreme Court and binding power on other constitutional players, especially Congress, then it would be appropriate to view the line-item veto as an exogenous change and look at its impact in the federal budgeting system. However, the current form of federal line-item veto can be redefined and modified by Congress relatively easily. It is obvious that the stringency of the federal line-item veto is determined inside the system by Congress. Therefore, it would be more appropriate to model the veto power as an endogenous change or a dependent variable.
Sorting out the exogenous and endogenous changes may change the questions one should ask in studying the subject and, most importantly, may lead to different research results and subsequent policy implications. If one treats the line-item veto power as exogenous and finds that there is a reduction in the deficit after its implementation, one may conclude that the line-item veto is a "magic bullet" for reducing the deficit. But if one views the line-item veto as endogenous (which should be the more appropriate research lens here), and there is reduction in the size of the deficit, that reduction will be interpreted as the outcome of other factors, instead of the line-item veto. Those factors may include political pressure from the public and interest groups on Congress, the political skills of the president, and changing attitutdes and power distributions inside Congress. That is, when the line-item veto is endogenous, the factors that affect it also affect the outcomes. …