پایداری بدهی های عمومی و درآمد درون زا در برزیل: شواهد از سری های زمانی از سال 1947

In this paper, we investigate three central issues in public finance. First, was the path of public debt sustainable during 1947–1992? Second, how has the government balanced the budget after shocks to either revenues or expenditures were observed? Third, are expenditures exogenous? The results show that (i) debt is sustainable in econometric tests, with the budget being balanced almost entirely through changes in taxes, regardless of the cause of the initial imbalance. Expenditures are weakly exogenous; (ii) the behavior of a “rational” Brazilian consumer may be consistent with Ricardian Equivalence; (iii) seigniorage revenues are critical in restoring intertemporal budget equilibrium.

At least since the end of WWII, the Brazilian economy has been plagued with chronic public deficits and relatively high inflation — a typical example of seigniorage-financed deficits. However, there have been very few episodes when a sharp increase in public debt was observed. Using national account data for the revenue-GDP and the expenditure-GDP ratios from 1947 to 1992, this paper studies three central issues in public finance. First, was the path of public debt sustainable during this period? Second, if debt is sustainable, how has the government historically balanced the budget after shocks to either revenues or expenditures were observed? For example, given an unpredictable increase in expenditures, there are two polar forms of balancing the budget. One is to increase the present value of taxes and the other is to decrease the present value of expenditures. From the point of view of a rational Brazilian taxpayer, it is important to learn as to what extent these two forms of balancing the budget have occurred. Third, what are the results of exogeneity tests for government expenditures and tax revenues? This last issue is motivated by the fact that models of seigniorage-financed deficits such as Bruno and Fischer (1990), and the extension in Ruge-Murcia (1999), assume, respectively, that the fiscal deficit and that government expenditures are exogenous; the latter is also present in the tax-smoothing model of Barro (1979). Moreover, in a more flexible framework than Barro (1974), Bohn (1992) shows that the exogeneity of expenditures is a necessary condition for Ricardian Equivalence. Given that post-war Brazil seems to fit the seigniorage-financed deficit model, and that it is desirable to discuss Ricardian Equivalence here, we find it useful to apply exogeneity tests to Brazilian fiscal figures. As far as we know, this paper is the first work to do it in this context. Following Hamilton and Flavin (1986) and Bohn (1991), the first two issues are investigated using unit-root tests, co-integration tests, and calculating “unconventional impulse–response functions.” The latter are based on Vector Error-Correction Model (VECM) estimates, where a balanced-budget restriction is imposed after appropriate testing. The search for a sensible test of exogeneity has lead us to use the definitions of weak-, strong- and super-exogeneity in Engle et al. (1983). They propose likelihood-based definitions that can be tested, showing why the previous concepts (pre-determinedness and strict exogeneity) were incomplete or misleading. Exogeneity tests were conducted using the framework proposed in Johansen, 1992 and Johansen, 1995. This paper has three main findings. First, debt is sustainable, with the budget in Brazil being balanced almost entirely through changes in taxes, regardless of how the initial imbalance was generated. Expenditures are weakly exogenous in the sense of Engle et al. (1983). Second, the behavior of a rational Brazilian consumer with Ricardian preferences is consistent with the Ricardian Equivalence result (Barro (1974)). Given, for example, a tax break today, since his/her best forecast for restoring fiscal balance is that the current imbalance will be fully reversed by future tax increases, consumption and welfare will be unchanged.1 Third, we show that seigniorage revenues are critical to restore intertemporal budget equilibrium, since, when we exclude them from total revenues, debt is not sustainable in econometric tests. These results match the conventional wisdom about Brazilian public finance and are broadly consistent with the theoretical model of optimal seigniorage of Bruno and Fischer (1990), its extension in Ruge-Murcia (1999), and with the empirical findings of Pastore (1995), Ruge-Murcia (1999), and Rocha (1997). The paper is organized as follows: in Section 2, the methodology is presented; in Section 3, the data set is discussed; in Section 4, the empirical results are presented and in Section 5. we conclude.

This paper presents tests on the sustainability of Brazilian public debt for the post-war period — 1947–1992. They show that debt is sustainable only if seigniorage is included as a government revenue. In this case, exogeneity tests suggest that expenditures are exogenous. Unconventional impulse–response results show that, regardless of how an initial fiscal imbalance is originated (shocks to expenditures or revenues), it is eliminated through a change in taxes. This last result is consistent with Ricardian-Equivalence behavior for Brazilian consumers with proper preferences. Based on the evidence, we find that Brazil has followed closely a spend-and-tax policy in the post-war era, with seigniorage having a crucial role in balancing the budget. As a consequence, Brazilian inflation was consistently high during this period. A brief reflection on the status of the Real Plan is appropriate. Since the beginning of the plan in July 1994, seigniorage revenues have decreased sharply. Expenditures have increased considerably, generating a persistent deficit that runs (at the time of writing — June, 1998) at about 7% of GDP. Public debt have increased almost threefold since then. For exogenous expenditures, as verified in the sample 1947–1992, there are two polar forms of restoring long-run equilibrium: (i) increase taxes, excluding seigniorage, or (ii) increase seigniorage revenues. In the first case, Brazilians will be the most heavily taxed citizens in Latin America. However, there are very few services as a counterpart of these taxes-education, infrastructure, legal system, etc. In the second case, inflation will increase again, a price Brazilians may not be willing to pay for fiscal balance. Of course, expenditures could also be cut: let us just hope that they have ceased to be “exogenous.”