Keynesians initially explained the existence of the Phillips curve as caused by "bottlenecks" encountered as an economy moved ever closer to full employment. After it became apparent that trade-offs between unemployment and inflation might be unstable, a "structural-shock" theory was developed to show why Phillips curves might shift. We describe both approaches in this section.
The Keynesian Structuralist Approach
An early Keynesian explanation for the Phillips curves---the structuralist approach---goes like this: As unemployment falls and full employment is approached, it becomes increasingly costly to produce extra output. One reason is that as more and more industries operate nearer their capacities, firms increasingly encounter "bottlenecks" that boost costs, which are then passed on to consumers as higher prices.
Major structural bottlenecks occur in labor markets. Competition among firms for the best workers becomes increasingly vigorous as unemployment rates fall. Most of the best workers retain jobs even when unemployment is high. The new Keynesian dual labor market analysis presented in the last chapter suggests that when workers are displaced from primary labor markets (due, for example, to down-sizing and restructuring), they may be forced into less attractive secondary markets. Workers remaining unemployed (or underemployed) in prosperous times often lack the skills required for jobs during hard times; other workers' skills may have become rusty during extended periods of joblessness. Consequently, employers begin offering higher wages to "pirate" qualified workers away from other firms as unemployment rates fall. This competition drives up costs, prices, and so on.
Structural Shocks and Wage/Price Stickiness
More sophisticated Keynesian explanations for Phillips curves' shapes hinge on two key ideas: (a) Shocks to market demands and supplies continually bombard all economies, and (b) wages and prices are assumed downwardly "sticky." That is, wages and prices rise more easily than they fall. At any moment in time the structure of demands---both private and public---is in flux. These are shocks to the economy. As new products are developed or national priorities change, the demands for some goods rise while those for others fall.
Output growth in sectors where demand grows is limited in the short run by the capital and trained labor readily available. Thus, wage and price hikes are the primary adjustments to stronger demand. In sectors where demand shrinks, the norm is production cutbacks and layoffs---not price reductions. Workers accept wage cuts only reluctantly, so firms find it easier to cut wage costs through layoffs. Some temporary price cuts (e.g., automobile rebates) may be used to liquidate excess inventories, but these price cuts will be accompanied by declining output and rising unemployment. This is a major difference between traditional Keynesian and Classical analysis; different conclusions about how rapidly wages and prices adjust to bring about equilibrium and whether adjustment occurs in output or prices.
Suppose Aggregate Demand is stable while the structure of demand changes. Prices in growing sectors rise faster than outputs, but prices fall slowly if at all in declining sectors, while outputs plummet. Net result? A higher price level, more unemployment, and less total output.
Structural shocks cause Aggregate Supply to fall temporarily because of friction encountered in moving resources from declining to growing sectors.
How this Keynesian analysis is consistent with Phillips curves is shown in Figure 5.
Figure 5 Full Employment Output, Shocks, and the Rate of Inflation: The Modern Keynesian Approach
A modern Keynesian approach suggests that recurring shocks to large numbers of individual markets may create excessive unemployment that can only be overcome by expansionary policies, lending a natural bias for inflation to an economy based on the market system.
At our initial equilibrium at both points a, the price level is stable at 100. Assume that Aggregate Demand remains at AD0 but that its structure changes, jarring resource markets. Aggregate Supply shrinks from AS0 to AS1, and equilibrium moves from point a to point b. Given downward wage-price stickiness in declining markets, structural changes raise the price level. If policymakers try to maintain output at Qf and unemployment at Uf, expansionary policies will increase Aggregate Demand from AD0 to AD2; the economy shifts from point b to point c in both panels of Figure 5. Thus, adopting policies to hold unemployment at Uf yields annual inflation of 3 percent if structural changes of the same magnitudes and frequencies continually recur.
This result suggests that continuous changes in the structure of economic activity create a "natural" inflationary bias in the economy. Policymakers might slash Aggregate Demand to AD1 (point d in Panel A) to suppress "natural" inflation, but unemployment would then rise to U1 in Panel B. Thus, policymakers can reduce either the unemployment or the inflation caused by structural economic changes, but not both. Less inflation means fewer jobs, and vice versa.
A Keynesian Theory of Stagflation
Classical theory discounts the possibility of lasting Phillips curves and emphasizes the stabilizing qualities of automatic market adjustments. Erratic government policy is cited as a major source of macroeconomic shocks. On the other hand, Keynesian structuralist and "shock" theories help rationalize the existence of Phillips curves. But, respectively, how well do these competing approaches illuminate the rocky path the U.S. economy has followed in the past three decades?
Keynesians considered multiple disruptions as important in disrupting Aggregate Supplies and Phillips curves in the 1970s, classifying shocks under five broad headings: (a) inflationary expectations, (b) external shocks, (c) labor market disturbances, (d) structural changes in product markets, and (e) disruptions emerging in the public sector. Let us examine each, remembering that Aggregate Supply and the Phillips relationship can shift in either direction.
Adaptive Inflationary Expectations
Any event that reduces Aggregate Supply induces supply-side inflation and worsens the trade-off between unemployment and inflation (shifts the Phillips curve to the right). Keynesians recognize that inflationary expectations negatively affect Aggregate Supply. (In fact, A. W. Phillips' 1958 analysis of trade-offs between unemployment and inflation included statistical adjustments for past inflation and several other factors that he viewed as explaining placement of the curve.) The theory of adaptive expectations identifies recent inflation as the major determinant of expectations about future inflation.
The theory of adaptive expectations suggests that expectations about future inflation are typically a weighted average of past rates of inflation.
Suppose, for example, that people expected inflation during the next year to equal (1) half the inflation of the past year, plus (2) one-third of the inflation experienced a year ago, plus (3) one-sixth of the inflation from two years back. If inflation was 14 percent in the past 12 months, 9 percent for the year earlier, and 18 percent three years ago, then, because 14/2 + 9/3 + 18/6 = 13, people would expect roughly 13 percent inflation in the coming year. Different people might assign different weights to different years. This approach suggests that the more severe the recent history of inflation, the faster Aggregate Supply recedes to the left and the more severe will be the tradeoff between unemployment and inflation.
Shocks to the System
During 1974--1981, recurrent oil price hikes boosted firms' energy costs, reducing Aggregate Supply and eroding the Phillips curve. This is shown in Figure 5. Wars and bad weather or other natural disasters have similar effects. On the other hand, discoveries of new resources, technological advances, prolonged periods of economic stability, or reductions in international tensions may improve the Phillips trade-off.
Labor Market Changes
Incentive structures that affect work effort also help explain the Phillips curve's location. Unemployment compensation, for example, affects many workers' labor-leisure choices. Higher unemployment compensation payments will encourage greater unemployment, intensifying inflationary pressure. Unemployment compensation (a) encourages temporary layoffs by firms because it eases hardships on firms' workers when orders for output drop, and (b) extends the average duration of unemployment. Social Security currently encourages early retirement. Increased payments under most welfare programs may also worsen the trade-off between unemployment and inflation.
Some recent reforms may reverse these effects. For example, welfare recipients who work can use subsidized day-care for their children, government now supports job training for the "hard-core" unemployed, and the penalties exacted from Social Security recipients who earn income have been reduced.
In 1994, President Clinton proposed revamping the unemployment compensation system to help reduce the tradeoff between unemployment and inflation (shifts the Phillips curve to the left). The Reemployment Act consolidates into a single comprehensive system six government programs that currently serve dislocated workers. This consolidation will, at one location, permit workers to apply for jobless benefits, get counseling and access their skills, and, if needed, apply to enter retraining programs. The plan is designed to streamline the current process and prepare displaced workers for a changing economy that demands that new employees be flexible and useful.
All these reforms may soften trade-offs between inflation and unemployment by increasing work effort and Aggregate Supply. Policies to foster mobility and match idle workers and job openings also improves the Phillips trade-off, as will programs to reduce discrimination so that productivity becomes the primary criterion for employment.
These changes in labor policy are directed at counteracting worker displacement caused by global competition and corporate restructuring. Dual labor market analysis suggests that competition for primary sector jobs paying higher efficiency wages will intensify. Enhancing and individual's productivity will increase their chances of becoming reemployed and retaining that position in the long-run.
Demographic changes also affect how unemployment and inflation are related. Post-- World War II baby boomers flooded labor markets during 1962--1980. But younger workers tend to bounce from job to job and are unemployed more often, worsening the Phillips relationship. As this group ages, the maturation of much of the work force should reduce the severity of this trade-off. Women's labor force participation rates grew markedly from the 1950s onward. Even though this influx of labor bolstered Aggregate Supply, it also may to have worsened the Phillips curve because women's unemployment rates are typically higher than those of males.
Collective bargaining may also affect the Phillips curve. If growth of unions' power enables them to obtain contracts with inflationary wage hikes, or if strike activities intensify, the trade-off between inflation and unemployment will deteriorate. As trade union membership as a percent of the labor force has declined, the effect of unions on unemployment and inflation has noticeably weakened.
Structural Changes in Product Markets
One explanation for the shape and existence of Phillips curves is drawn from continuous changes in the structure of demands for goods, combined with wage and price stickiness and a policy of trying to maintain full employment. The menu of choices between unemployment and inflation worsens if external shocks become stronger or if structural changes occur more rapidly. If consumers' preferences or investors' perceptions of the economic outlook change markedly, the structure of output may also change drastically. Of course, this trade-off is more favorable the slower or less extreme these changes are during a given interval.
Changes in laws governing foreign trade become more important as our economy becomes more internationally oriented. Hikes in tariffs or cuts in import quotas raise the prices of imported goods and can certainly worsen the dilemma posed by the Phillips curve. Competition is also reduced. Similarly, growth of domestic monopoly power and the resulting drives for greater profits will worsen the inflation/unemployment trade-off.
Public Sector Changes
Just as shocks to the structure of private demands can shift the Phillips curve, so can disruptions to the composition of public sector demands for goods and resources. Changes in tax structures, subsidies, and transfer payments also affect the trade-off. Major revisions of such regulations as those policed by the Occupational Safety and Health Administration, minimum-wage laws, environmental protection, leasing policies for mineral exploration on public lands, or changes in property rights structures may all shift the Phillips curve relationship. The direction of the shift depends on whether a particular regulatory change enhances or encumbers economic efficiency.
For modern Keynesians, an explanation for the stagflation of the 1970s goes something like this: Rising inflationary expectations and disruptions associated with the Vietnam War caused the first shift in the Phillips curve in the early 1970s. The increase in world oil prices in 1974 caused the second shift. More oil price hikes in 1979 and 1981, coupled with worldwide crop failures, caused further shifts of the Phillips curve in the early 1980s. And relative cessation of shocks to the economy in the mid-1980s ultimately improved the Phillips curve trade-off.