The common structure (the common income effects) of seemingly disparate policies kept me wondering through the early 1980s about our fiscal nomenclature. How is it that we describe such similar policies so differently? And why do some certain policies that hurt young and future generations show up as budget deficits, whereas others do not? Feldstein (1974) had planted the seed for worrying about fiscal taxonomy in the 1970s in calculating unfunded social security liabilites and, thereby, implicitly comparing them with explicit liabilities. But he never clarified why it was that some liabilities that we leave for our children show up on the books, while others do not.
One night, while lying in bed on vacation, it dawned on me that the different policies we had simulated were not really different, rather we were simply using different words to describe the same policies. For example, in the case of pay-as-you-go social security we call contributions being made to the government “taxes” and the benefits received “transfer payments.” But we could just as well call the contributions “loans from us to the government” and the benefits “return of principal plus interest on those loans less an old-age tax.” In the case of the U.S., this alternative language would roughly triple the size of official government debt. I considered waking up my wife to tell her this, but thought better of it.
Instead, I wrote a series of papers on deficit delusion, and Alan and I devoted a chapter of our book on the A-К Model to the issue of deficit finance and fiscal illusion. The most important of these papers is Kotlikoff (1993) which points out that in any neoclassical model with rational economic agents and rational economic institutions (institutions that produce the same economic environment regardless of the language used to describe that environment), the “deficit” is not a well defined economic concept. Stated differently, when one uses “ the” deficit to describe a country’s fiscal policy, one is, in fact, describing its vocabulary, rather than anything about its actual policy.
A final lesson we learned was the importance of announcement effects to economic efficiency and short-run economic performance. When we orginally wrote down our model we didn’t realize that it was capable of analyzing policies that are announced in the present to take place in the future. But it was indeed. All we needed to do was to tell the model when the policy was going to take place, and the agents in the model would automatically and fully incorporate this information in their economic plans. Thus an anouncement today that the economy is switching to a consumption tax two years hence stimulates lots of consumption right now as agents seek to purchase consumption while its still relatively cheap. The short-run run on consumption produces the opposite effect of what consumption-tax advocates seek and advertise — an immediate increase in national saving and capital formation. Moreover, the distortion of short-run relative prices can be so great as to either substantially offset or fully eliminate the efficiency gains from consumption taxation.