Government interference in private markets for exhaustible resources is a ubiquitous phenomenon. Each level of government — federal, state or provincial, and even local — has its own tax policy. Tax-policy instruments include royalties, depletion allowances, severance taxes, price controls and supports, environmental regulations, freight subsidies, profit taxes, and property taxes, to name only a few.

These taxes may be levied at every stage of production. For example, property and some ad valorem taxes are levied on ores in the ground, severance taxes are paid when the ore leaves the ground, royalties and depletion allowances may be collected at the first sale (after some degree of processing), and price controls may be placed on finished products.

Not only does the government intervene in private resource markets with great regularity but the size of tax effects is not small. For example, depletion allowances in the United States have been as large as thirty three and one third percent of revenues. And price controls in some periods have kept petroleum prices at less than half world prices.

A great deal of evidence for the U.S. economy suggests that after changes in aggregate nominal demand (1) prices move «sluggishly» toward their eventual long-term values, that (2) output responds directly to these changes in demand, and (3) these output effects linger «peristently» over time. Most econometric evidence suggests that something of the order of four to six years must pass before the ultimate effects on prices and output will be reached. As commonly interpreted, this lag has profoundly disturbing implications about the functioning of markets and about the rate of agents’ expectations in macroeconomics.

The conventional interpretation of the output-price adjustment lag -- to borrow the label suggested recently by R.J. Gordon (1982) -- comes unders the « Natural Rate Hypothesis and Gradual Adjustment of Prices » (NRH-GAP) approach. After a change in nominal demand, according to NRH-GAP, some combination of forecast inefficiency and institutional constraints on price setting prevents agents for a time from choosing the set of prices that would clear output markets at the « natural » level. In the meantime and because of this price stickiness aggregate output exhibits persistent deviations from its natural path.

It is always a risky enterprise to attempt to project « stylized facts » on to any complex actual economic situation; just when one feels that a recognisable pattern has clearly emerged the situation will change and invalidate ones hoped for empirical regularity. Yet, having said this, it appears that it is possible over the period since the second world war to broadly distinguish between groups of nations on the basis of their balance of payments behaviour. First, there are what have been called the strong currency countries like Germany and Switzerland. These countries tend to run habitual balance of trade surpluses, their economic policies are never constrained by the necessity of avoiding trade account deficits and their most characteristic trading problem has been the existence of occasionally massive capital inflows. Second, there are a larger group of what might be called normal countries like Belgium and Sweden. In the recent difficult times the performance of these countries has deteriorated slightly, but broadly, over the past thirty-five years they have managed to achieve rough long-run trade balance equilibrium without suffering excessive unemployment or unduly restricting the growth rate of their economies.

In an innovative and important paper Hall (1978) demonstrated that the assumption of rational expectations when applied to the life cycle or permanent income hypothesis would, given a number of further assumptions such as constancy of the real rate of interest, imply that real consumption follows a martingale. Thus in logarithms

log C - log C- 1 = α + ∊

where C is real consumer expenditure, α is a constant which allows for real growth in the economy and ∊ is a white noise error term.

Hall presented empirical evidence for the U.S. economy which was broadly consistent with the hypothesis. However, subsequent work by a number of authors (eg Muellbauer (1983) and Hendry and von Ungern-Sternberg (1980)) appears to be inconsistent with the hypothesis. Past values of economic variables, such as consumption or income, are significant determinants of the first difference of the logarithm of consumption which violates the hypothesis.