These parameters appear in the state decision schedule because they influence state' expectations of how future (AER) will effect capital. Further, notice that all of variables in appear in the decision of war schedule, since via equation  all of these variables help agents forecast future (AER). ( Compare this with common econometric practice of using only current and lagged values of the (AER) as proxies for expected future (AER).)
The fact that  and  share a common set of parameters ( the matrix) reflects the principle that actors' optimal decision rule for accumulating gains , described as a function of current and lagged state and information variables , will depend on the ...view middle of the document...
Returning to the more general decision of war example, the usual econometric practice has been roughly as follow. First, a model is typically specified as estimated of the form
where is a vector of free parameters of dimension conformable with the vector .Second , holding the parameter fixed, equation (6) is used to predict the implications of alternative paths for the (AER) .This procedure is equivalent to estimating equation from historical data when and then using this same equation to predict the consequence for capital accumulation of instituting an on-again, off-again (AER) policy of the form . Doing this assumes that a decision of war schedule of the form can be found with a single parameter vector that will remain fixed regardless of the rules for administering the (AER).(17)
The fact that equations  and  share a common set of parameters implies that the search for such a regime-independent decision schedule is misdirected and bound to fail. This theoretical presumption is backed up by the distressing variety of instances in which estimated econometric models have failed tests for stability of coefficients when new data are added. This problem can not be overcome by adopting more sophisticated and more general lag distributions for the vector as perhaps was hoped in the 1960s.
The decision of war example illustrates the general presumption that the systematic behavior of state and the random behavior of International system both will change whenever state's constraints change, as when power policy or other parts of the environment change. to make reliable statements about policy interventions, we need dynamic models and econometric procedures which we consistent with this general presumption. Foremost, we need a new and stricter definition of the class of parameters that can be regarded as structural. The body of doctrine associated with the simultaneous equations model in econometrics properly directs the attention of the researcher beyond reduced form parameters to the parameters of structural equations which are meant to describe those aspects of other actor's behavior that remain constant across a range of hypothetical environments. Although such structural parameters are needed to analyze an interesting class of policy interventions, most often included among them have been parameters of equations describing the rules of expectations for private agents. Going to peace, and make decision of war, are all examples of such rules of expectations. In dynamic settings, regarding the parameters of these rules of expectations as structural or invariant under policy interventions violates the principle that optimal decision rules depend on the environment in which agents believe they are operating.
If parameters of decision rules can not be regarded as structural or invariant under policy interventions, deeper objects that can must be sought. The best that can be hoped for is hat parameters...