To show the actual choices, figure 12 shows the path of inflation and unemployment under these assumptions: path A illustrates the most conservative policies, with unemployment allowed to rise to 7 percent, and to be maintained at that deep recession level until near price stability has been restored in 1974. Path B shows maintenance of the 6 percent unemployment rate until price stability is achieved in 1975. Path C assumes the unemployment path of a standard economic forecast showing unemployment falling to 5.2 percent by the end of 1972 and falling to 4.7 percent by the end of 1973, and staying at that level thereafter. Price stability is never restored and the curve returns to the neighborhood of the explosive area by 1974. Path D shows a quick return to 4.5 percent unemployment by the summer of 1972 and a drop to 3.5 percent thereafter. The wage-price explosion would quickly resume.

It can be seen that a continuation of the previous policies, relying exclusively on unemployment to bring the wage-price system back to stability, would have been a very costly and drawn out process. Indeed, the unemployment cost of the classical route to price stability would obviously have been prohibitive.

If the goals had been more moderate, looking only to an ultimate target of a 3 percent inflation rate, the path would still have been difficult. The trouble is that the three percent price goal leaves the economy in a condition where inflationary expectations remain lively. The expectational inflation severity factor stays in force, and only a sufficiently high rate of unemployment could neutralize the inherently explosive character of the system under that circumstance. The fact is that once the inflationary expectations have become fully aroused,

intermediate policies aiming only to bring the inflationary rate down to "acceptable" levels require so sensitive a management of demand to keep the economy from relapsing into an explosive wage-price pattern that it is probably impractical.

D. STUDIES OF PHASE II AND BEYOND

The projected Administration monetary and fiscal policies combined with probable consumer and investor responses produce the gradual recovery depicted earlier as Path C. To analyze the potential of Phase II, an additional series of simulations were run. Table 9 presents the baseline results and the potential benefits and problems of policies intended to achieve the wage and price targets of 5.5 percent and 2.5 percent respectively. Phase II controls are assumed to be in effect through the third quarter of 1973.

In all three studies, the average annual unemployment rates for 1971 to 1975 are projected to follow the pattern: 6.0 percent, 5.7 percent, 5.0 percent, 4.5 percent, 4.5 percent. Without wage-price controls inflation would not drop below 3 percent in this interval.

Part II of the table analyzes the economy if only wage controls are successfully pursued. Such a program would reduce the first-year wage inflation by 0.8 percent. If labor accepted this reduction of income, the system would settle down on its own to price and wage behavior only a few tenths above the Phase II goals. By 1975 price and wage levels could be 3 percent lower as a result of the better path.

The difference between the controlled wage path and the free path measures the current stress on the wage controls created by the uncontrolled price path. The accumulated stress shows how this pressure builds up over time from current and past differences. It is certainly doubtful whether the one-sided program would be accepted by labor.

If only prices can be successfully controlled, the stress on the controls will be greater. Large wage increases in 1971 and 1972 will create strong cost pressures in addition to the problems created by the expanding unfilled orders of a recovering economy. With prices limited to a 2.5 percent rate, wages decelerate gradually to a 5 percent rate but the process will be slow, too slow for business to be expected to comply. The accumulated price stress will reach almost 2 percent by the assumed end of the two-year price control program.

In summary, one-sided programs will create great strains. On the other hand, if both elements of Phase II succeed, the system settles down near the Phase II targets and would be near its long-run Phillips Curve by the second half of 1973.

E. THE LONG-RUN PHILLIPS CURVE

Since the economy, in fact, faces a relatively horizontal Phillips Curve in the short run whose location in the unemployment-inflation space is largely predetermined by the price history, the definition of a long run Phillips Curve has to be a somewhat synthetic construct. Further, demand policy in the U.S. is largely oriented towards short run goals, so the short run Phillips Curve considerations dominate. Nonetheless, policy should also consider the longer run choices, the implications of any particular short run policy on the subsequent path of the system. The long run Phillips Curve also can show the true structural characteristics of the economy with regard to unemployment and inflation. What does the Phillips Curve look like when the present is neither punished by a poor past history of inflation, nor rewarded by an exceptionally good price history bought with high unemployment? We define the long run Phillips Curve as follows: assume that a given unemployment rate is maintained so long that the initial conditions and the lag structures of the equations no longer have an effect.

Figures 11 and 12 (see above) indicate this curve as the heavy line in the center of the short run curves. It consists of two parts: a nearly vertical element which defines the critical unemployment rate-the lower bound of sustainable long run performance, and a second segment which is a conventional Phillips Curve. Even if a good past history allows the economy to temporarily reach a combination of unemployment and prices to the left of the critical rate of unemployment, the subsequent path of the system is essentially vertical and explosive. Thus the critical rate of unemployment is defined to be that rate which produces a rate of wage increase sufficiently above productivity to push price inflation above the critical 2.5 percent rate which brings the expectational inflation severity factor into play.

If unemployment is higher than the critical rate, the system will ultimately show declining inflation until it settles down on the appropriate point on the long run Phillips Curve. This segment of the curve is stable: if unemployment were constant, wages and prices would settle to steady and mutually consistent rates of increase.

The long run Phillips Curve is a measure of the structural element in the inflation problem. That the natural rate of unemployment is as high as 4 percent is disappointing from the social point of view. In the 1960's the Council of Economic Advisers defined an interim unemployment goal of 4 percent. Despite the manpower and equal opportunity programs designed to improve the functioning of the labor market plus whatever small steps may have been taken to improve the performance of product markets, the interim goal of 1962 continues to escape our grasp. The few years of unemployment below the interim target produced a wage-price explosion. In the absence of major efforts to improve the structure of the economy, future governments would be well advised not to exceed the old interim targets, and perhaps not even to reach them.

This seems to be a very high rate of unemployment to be necessary to hold the aggregate wage rate within 2 to 3 percent of productivity growth. Apparently, product markets permit full transmission of cost increases over a wide range of demand conditions. Labor markets produce wage increases above productivity even when there is considerable unemployment. It is the totality of our product and labor market structures asymmetries, barriers to entry, other monopolistic elements, information imperfections, protection from foreign competition, discrimination in employment and other factors--which limits our economic performance to the inadequate possibilities sketched by the long run Phillips Curve.

1. Present Results and the Accelerationist Position

As the preceding analysis has demonstrated, the weights which wage and price inflation receive as determinants of each other have important implications for the stability of the economy and hence for the options available to policy makers. If prices and wages fully respond to each other, no permanent tradeoff between unemployment and inflation exists. This is the position of the "accelerationist" school. Expectations are assumed to be consistent and to derive primarily from past experience, recognizing the average as well as the trend of the recent past. Wage inflation is asserted to fully respond to anticipated price inflation to maintain an expected real wage level.

Prices increase proportionately with unit labor costs. In the long run anticipated price inflation is identical to actual price inflation and equals wage inflation adjusted for productivity gains. The "natural" rate of unemployment is determined by the structure of the labor market. The mobility and flexibility of the work force, the job-search behavior of participants, the adaptability of productive processes and the efficiency of information collection and dissemination procedures are prime determinants. According to the monetarist version of the theory, the price levels are determined solely by the increase in money demand, which in turn is determined by the rise in the money supply. The key difference between this view and that which emerges from the present study is the existence of a permanent tradeoff in periods of unemployment above the "natural" rate. The price equations of the two theories are nearly the same: both indicate that, aside from deviations due to productivity and unfilled order irregularities, prices will rise with unit labor cost. However, we place greater emphasis on the role of monopoly elements in the determination of the critical rate of unemployment. In our view, complete transmission of cost increases should not occur when capacity significantly exceeds demand. A major difference comes on the wage side, where our equation implies that the system has a considerable range of stability.

If the accelerationist's view is accurate, it is only the destabilizing actions of government, and perhaps capital purchasers, which preclude a world of no inflation and 4 percent unemployment. The only roles for policy are to stabilize demand to eliminate this stimulus to cycles and to make the labor market more efficient in order to reduce the "natural" rate of unemployment. No tradeoff exists because only one problem need exist: unemployment.

F. AN ANALYTICAL VIEW OF THE MODEL OF THIS STUDY

Various analytical versions of the wage-price mechanism can be utilized to demonstrate the logic of the long and short run relationships. To simplify the analysis, unfilled orders and tax rates are assumed constant and output prices are set equal to consumer prices. Let:

The price response to current wage increases is not complete due to rigidities of price-setting and to an adjustment mechanism. If previous unit labor costs did not equal their anticipated values, current prices would be adjusted to take this discrepancy as well as expected future unit labor cost inflation into account. This is analogous