The main hypothesis of the RET can best be illustrated through an example using the Breland family. Suppose the world consists of two periods—today and tomorrow—and assume that the government spends a fixed amount each period. The government has many ways it can finance its expenditures. It can a) balance its budget each period by levying taxes both today and tomorrow which generate revenue equal to its spending each period. Or it can b) run a deficit today by issuing bonds, and raise taxes tomorrow. Or it can c) run a surplus today and then cut taxes tomorrow. Notice that regardless of whether the government uses method a), b), or c) to finance its expenditures, the total lifetime tax liability of the Breland family is the same. While the timing of that tax liability is different in each case, the total value of the Breland's tax liability is fixed. If the Brelands seek to smooth their consumption over time, as discussed in last month's article, then what is relevant to them is the total value of their tax liability, not its timing. Hence, the timing of taxes and deficits used by the government to finance its spending is irrelevant to the Breland family's consumption choices.<Seater, p. 143, asserts that the Ricardian Equivalence Theorem is simply "a generalization of the Life Cycle-Permanent Income hypothesis (LCPIH)" in which the consumption-based LCPIH is extended to include government purchases, taxation, and debt. See the January 1998 edition of Fraser Forum for a simple discussion of the LCPIH.>

Let us make this example more concrete. Assume that the government decides to cut taxes in the present period with no change in expenditures (operate in deficit today) and that this tax cut generates $500 in extra disposable income for the Breland family in the current period. Assume also that the lifetime tax liability of the Brelands is $100,000, based on the level of government spending during their lifetime. If the Brelands are forward-looking, they will realize that lower taxes today imply higher taxes sometime in the future. In order to smooth their consumption over time, the Brelands will save the extra $500 generated by the tax cut and use these savings to purchase $500 in government bonds. When the bonds mature, the Brelands will use the proceeds from the bonds (principle and interest) to compensate for the higher taxes they will have to pay in the future. The Brelands choose to do this because they realize that with no compensatory decrease in spending, the government has simply shifted their tax liability to the future. The RET, therefore, predicts that a tax cut, with government spending held constant, will have no effect on consumption or capital accumulation.

The RET is often discounted in public policy discussions regarding the government budget. This is because many scholars believe that the assumptions built into the RET are not generally applicable in the "real world." One major criticism of the RET is that, applied in a dynamic setting, it requires individuals to have an infinite planning horizon. This is obviously implausible given that nobody lives forever! Indeed, Paul Samuelson and Peter Diamond, two prominent Keynesian economists, have formalized this criticism and shown that, if individuals only care about their own consumption, then the RET is invalid. <See Peter A. Diamond (1965), "National Debt in a Neoclassical Growth Model," American Economic Review, 55: 1126-1150; Paul A. Samuelson (1958), "An Exact Consumption-Loan Model of Interest With or Without the Social Contrivance of Money," Journal of Political Economy, 66: 467-82.> In such a world, it is possible for individuals to make themselves better off following a tax cut by forcing future generations to bear the burden of the future tax increase.

In a famous 1974 paper published in the Journal of Political Economy, Robert J. Barro re-asserted the applicability of the RET. <Robert J. Barro (1974), "Are Government Bonds Net Wealth?" Journal of Political Economy, 82: 1095-1117.> According to Barro, individuals with finite lives will behave as if they have an infinite planning horizon if they care about the welfare of their children. In other words, if parents are "inter-generationally altruistic," children (and future generations) are incorporated into the planning decisions of parents. Altruism of this sort allows the "treatment of another's utility as an extension of one's own." <Seater, p. 147.> In such a world, the RET holds, since the parents behave as if they expect to live forever!

Both the Samuelson-Diamond critique and Barro's re-assertion of the RET are again illustrated using the Breland family. Suppose the government implements a 10 percent tax cut which is financed by government bonds. Assume also that the Breland parents will not live to see their taxes increase. If the Brelands behave according to the Samuelson and Diamond model (i.e. they only care about their own welfare), then they will use the proceeds from the tax cut to increase their consumption and thereby force their children, John and Mary, to bear the burden of a future tax increase. The RET obviously fails to hold in this scenario since the tax cut induced the Breland parents to increase their current consumption. However, if the Brelands behave as Barro posits (i.e. they care about their welfare and their children's welfare), then the fact that Bob and Jillian may not live to see their taxes rise is irrelevant. If Bob and Jillian are inter-generationally altruistic, they will save the tax cut and pass on their saving to their children as a bequest in order to offset their children's higher future tax liability. In this scenario, RET holds.

Many critics of deficit financing often decry the fact that higher deficits today will leave future generations worse off. Steven Landsburg, an economics professor at the University of Rochester, makes an interesting counter-argument. In his new book entitled Fair Play, Landsburg argues that current deficits will only leave your children worse off if you want them to. <Steven Landsburg (1997), Fair Play: What Your Child Can Teach You About Economics, Values, and the Meaning of Life, New York: The Free Press.> As we noted above, RET fails to hold if parents do not care about the welfare of their children. Since there is nothing stopping parents from leaving altruistic bequests to future generations, children are only worse off if their parents want them to be.

Given the restrictive assumptions which must hold in order for the RET to be true, it is difficult to accept that the RET is strictly applicable to the real world. Nonetheless, there is common agreement among macroeconomists about the approximate applicability of the RET. <In particular, Seater concludes that the RET is a "close approximation of reality."> Indeed, it is interesting to note that statistical tests of the RET have, in general, been inconclusive. In a recent paper by Emmanuela Cardia, it was shown that most of the statistical tests used to determine whether or not the RET holds are invalid. <Emmanuela Cardia (1997), "Replicating Ricardian Equivalence Tests with Simulated Series," American Economic Review, 87: 65-79.> In this paper, Cardia constructed two artificial economies—one in which the RET holds and one in which it does not—and then used these artificial economies to generate simulated data. Cardia then applied a number of standard statistical tests to the simulated data and found that the tests were unable to distinguish the Ricardian economy from the non-Ricardian one. Clearly the jury is still out on the Ricardian Equivalence Theorem.

There is a very important public policy implication that can be derived from the RET which substantially alters the terms of the current debate in Canada about "how to spend the fiscal dividend." If one believes that the RET is a reasonable approximation of reality, then the debate on how to spend the fiscal dividend is really a red herring. The fundamental insight of the RET is that while the timing of taxes and deficits has no impact on real economic activity, the amount the government spends does. This is because greater levels of government spending raise an individual's total tax liability, but changes in the timing of deficits and taxes (holding government spending constant) do not. Viewed from this perspective, the central policy issue is how much government should spend, and not how it finances that spending. Policy discussion should therefore focus on the optimal size of government (i.e., how much should government spend?) rather than on the mix of debt and taxes used to finance those expenditures. <For further information see the forthcoming Fraser Institute publication on the optimal size of government in Canada.>