There are many possible explanations of the small observed real substitution response. Perhaps not enough time elapsed after the tax changes for individuals and firms to exhibit their long-term reaction to the changed environment. It may be that the tax changes, particularly the Tax Reform Act of 1986, were so complicated that it is difficult to accurately measure how the relevant relative prices changed. Of course, another possibility is simply that the relevant elasticities of real substitution are close to zero, so that a correctly measured elasticity is also quite low.

In this paper I investigate another possible factor: that the availability of timing, financial, and accounting responses to tax changes mitigates the real response, so that even in the presence of significantly non-zero elasticities of substitution little real response occurs. To do so I construct a “general” model of behavioral response, which allows individuals to change both their labor supply and avoidance effort in response to tax changes.

The interaction between the real response and the other responses is a pervasive issue for tax analysis because it is rare that taxes are levied directly on a real activity. Taxes are not levied on the income from labor, but rather on the reported income from labor, or on the income determined by auditors. Taxes are not levied directly on future consumption, but rather on the return to some, but not all, returns to savings instruments, or on certain portfolio shifts, as in the case of capital gains taxation. Thus there are always avoidance margins as well as real margins that can potentially be adjusted. loans