Now if the economy is in a steady state initially, one can use steady-state values for evaluating such derivatives because the initial (pre-policy-change) position of the economy at any point in time is the steady state position. But what if the economy is moving along a transition path? Then the derivatives of economic variables at any particular point in time must be evaluated using the values of the economy’s initial (pre-policy-change) variables at that point in time. But how does one know these values unless one actually computes the economy’s dynamic equilibrium?

In general then, one can not use the calculus to study the impact of policy changes on the economy through time unless one computes the initial (pre-policy-change) transition path. But if one can compute the initial transition path (the transition path under the initial policy), one can also compute the new transition path (the transition path under the new policy) and compare them. So the calculus is really of no use. Furthermore, restricting oneself to evaluating derivatives for economies that are initially in a steady state is only valid for steady states that can be computed independent of the transition path.

But some steady states can not be so computed, specifically steady states the value of whose long-run policy variables are set as functions of what happens during the transition. As an example, take an economy in a steady state with no social security. Now suppose the government announces that it is going to establish a pay-as-you-go social security system starting in five years by setting, on a permanent basis, social security’s benefit at 40 percent of the average wage prevailing in five years. In order to figure out what that benefit level and its associate payroll tax are, one has to calculate the transition and, among other things, determine the level of average wages in year 5. fast payday loans

The complaint about the model being a “black box” is puzzling given that there are only a limited number of elements in the model — a standard, time-separable CES utility function, a standard CES production function, standard quadratic capital adjustment costs, a familiar fiscal structure, a few other quite straightforward elements, and no uncertainty. If this most basic neoclassical life-cycle growth model constructed along a realistic time dimension is viewed as a “black box,” it’s certainly high time to study that box. The “black box” pejorative is also inappropriate given that the results of the policy simulations are readily explained in terms of economic theory, invariably coincide with one’s economic intuition, and are easy to double check.