CHAPTER 11. AGGREGATE DEMAND AND SUPPLY

Copyright 2011 by Richard A. Stanford


The "Keynesian Cross" analysis elaborated in Chapters 8 through 10 can go a long way toward explaining macroeconomic phenomena, but it contains a serious deficiency. It is not convenient to analyze the causes or consequences of inflation or deflation since the price level is not represented explicitly on any of the associated coordinate axes. In Keynesian theory, price level changes must be inferred from other relationships, and the consequences of price level changes can only be "talked into the analysis." This omission led during the 1960s to the development of another graphic vehicle for macroeconomic analysis that complements the Keynesian models. This newer model may be employed to examine the implications for the price level in a normally operating and growing economy.

Figure 11-1 is a short run macroeconomic analysis depicting the economy's aggregate demand (AD), its long-run aggregate supply (LRAS), and its short-run aggregate supply (SRAS) as functions of the price level (actually an index of the price level such as the CPI). For a market-driven economy like that of the United States, the normal operating capacity, N1, is achieved when its installed plant and equipment are operating at about 80 to 85 percent of rated capacity, and unemployment is no more than 5 or 6 percent of the labor force. This is the irreducible amount of unemployment attributable to structural changes and frictions of adjustment in a dynamic market economy such as that of the early-twenty-first century United States.


Figure 11-1.



Aggregate Supply

In its short-run manifestation, the "long-run aggregate supply" curve, LRAS in Figure 11-1, is fixed along a vertical path above N1. It is represented as a vertical line in conformance with the "classical dichotomy" between real and nominal causation. In the long run the real productive capacity of an economy does not vary (i.e., is perfectly inelastic) with respect to changes of the price level, a nominal quantity, but it may vary with changes of real phenomena, i.e., real matters such as resource availability, productivity, and technological changes. The long run changes of these real phenomena may be depicted in the short-run analysis as shifts or drifts of LRAS to the right with on-going growth, or as leftward shifts consequent upon supply-shocks that may have lasting adverse effects on the economy's output capacity.

In contradiction to the "classical dichotomy", modern macroeconomic theory acknowledges the possibility of a short-run relationship between real output and the price level with the SRAS curve. This curve may be coincident with the LRAS curve if input price changes match the pace of output price changes. But SRAS may diverge from LRAS and attain some degree of upward slope if input prices adjust more slowly than do output prices. Although the normal operating capacity of the economy can be exceeded temporarily, it is likely that the SRAS curve becomes ever more steeply upward sloped the farther the economy attempts to operate above its normal capacity. Also, the SRAS curve may become quite shallowly sloped if managers become resistant to cutting prices or if published price lists must be honored until new catalogs or menus can be printed.

In The General Theory of Employment, Interest, and Money, John Maynard Keynes asserted that both wages and prices tend to behave asymmetrically, i.e., they freely rise in response increases of demand or decreases of supply, but they are "sticky" in the downward direction when demand decreases or supply increases. Keynes focused upon the monopoly powers exerted by big business enterprises and labor unions as his explanations of downward stickiness of both prices and wages.

A more recent hypothesized reason for downward wage stickiness is that in some industries there appears to be an implicit contract (or at least an understanding) between management and labor to the effect that managements will not do anything overtly to cause deterioration of the purchasing power and working conditions of their labor forces. Multi-year contracts may limit the adjustability of prices or wages in some industries. Prices may be sticky in either direction because of the costs and seldomness of reprinting menus, price lists, and catalogs. Whatever the cause, to the extent that wages and prices are more inflexible in the downward direction than in the upward direction, the SRAS may tend to be more shallowly upward sloped for price level changes below the current level, but more steeply sloped for price level changes above the current level.

How do the paces of change of prices of final goods and productive inputs compare? An important concept here is that everybody's price is someone else's cost. Wages are a cost to business employers, but final goods prices figure into the costs of living of laborers. It is possible that wage rates and other input prices lag behind increases of final goods prices. This is likely because the majority of resource owners indulge in so-called default forecasting, i.e., the presumption that tomorrow will be much like today since today was a lot like yesterday. Some people may engage in adaptive forecasting when they adjust their forecasts of the near future by some proportion of their recent-past errors in forecasting the present. Both default and adaptive forecasters tend to incur systematic forecast errors because they are slow to recognize and adjust to changes in the prices of final goods.

A consequence of these systematic forecast errors is that during a period of inflation, the wages of labor and the prices of non-labor inputs tend to rise more slowly than do final-goods prices. With input costs rising more slowly than output prices, the profit margins of producers increase and they attempt to hire more inputs in response. Because input owners may be slow to perceive the increases in their costs of living, they acquiesce in offering larger quantities of their inputs to the employers than they might if they were fully cognizant of the increases in their costs of living. The net effect is to cause a divergence between the short-run aggregate supply curve and the long-run aggregate supply curve. Because of the slowness of response of input owners relative to output producers in adjusting their respective prices, the SRAS curve attains some positive slope (diverging from the vertical) even though the LRAS is vertical.

Economists of the Rational Expectations school of thought theorize that if the world were populated exclusively by people who form expectations rationally, there would be no divergence of the SRAS from the LRAS. But in the imperfect world in which we live, only the few people who form expectations rationally are able to adjust their prices apace with producer prices. It is likely that only a small proportion of any population forms expectations rationally, primarily because only a few possess the requisite intelligence, knowledge, and perceptiveness. Those who do possess these requisites attempt to use all available information in forming expectations of future states. Their forecasts tend to be more accurate and timely than are default and adaptive forecasts, and forecast errors tend to be random (rather than systematic). To the extent that their forecasts are more accurate, rational expectations decision makers can protect themselves from negative events or take advantage of entrepreneurial opportunities. In so doing, they tend to adjust their own prices apace with the prices of final goods. For rational expectations decision makers, the SRAS would be vertical and coincident with the LRAS. Even so, to the extent that other members of the population indulge in default and adaptive forecasting behavior, the SRAS curve diverges from LRAS to exhibit an upward slope.

The SRAS curve probably changes shape over time. If inflation has not been a problem recently and is not expected in the near future, when it does occur the SRAS may have a very shallow upward slope which diverges only slightly from the horizontal. This is the same as to say that the majority of the population is surprised by the rising prices. But as inflation becomes an on-going problem which is recognized and expected by even those who form expectations adaptively or by default, the SRAS will become ever more steeply upward-sloped and approach the locus of the vertical LRAS. We might also hypothesize that if more people were to begin to form expectations rationally, the SRAS would become more steeply upward sloped and eventually converge with the vertical position of LRAS.


Aggregate Demand

The AD function probably has a graphic appearance similar to microeconomic demand curves. However, while it is possible to aggregate individual consumer demands for a particular item into a market demand curve or a demand curve faced by a single seller of the item, it is not feasible to extend the aggregation process to the whole economy because of the so-called "adding-up" problem. In such an aggregation it would be necessary to add unlike units of different goods. Therefore, the vertical axis for the AD function must be (an index of) the general price level rather than the price of any single good or service. The horizontal axis must be the level of real output rather than the quantity of any single good or service which is part of real output.

We should approach the AD curve from the perspective of its upward slope from right to left. This is because we have had so very little experience with price deflation in recent decades. From the "Great Depression" forward our experience mostly has been with price inflation. There are several possible reasons for the upward slope of the AD function from right to left. One is the so-called real wealth effect, i.e., at higher price levels people tend to feel a sense of impoverishment since their real wealth has less purchasing power; vice versa, if price levels were to fall, they might feel more affluent due to the fact that their real wealth has greater purchasing power. People probably adjust their purchases with respect to these feelings of affluence and impoverishment.

Another explanation for the right-to-left upward slope of AD lies in the Fisher effect, the tendency for nominal interest rates to rise as prices rise in order to compensate lenders for the loss of purchasing power which they would experience over the term of the loans which they extend to borrowers. Interest sensitive purchases (e.g., houses, automobiles) tend to vary inversely with the interest rate, so the aggregate demand for the economy's output tends to diminish with higher prices and consequently higher nominal interest rates. Yet another reason for the slope of the AD function lies in the trade balance effect, the likelihood that at higher domestic prices relative to foreign prices, exports of goods produced in the country will decrease at the same time that imports of foreign goods increase.

The classical Quantity Theory of Money can be represented by the identity, MV=PY, where M is the money stock, V is the income velocity of money, P is an index of the price level, and Y is real output. If this identity is solved for Y=MV/P, aggregate demand AD can be substituted for Y so that AD=MV/P. Then, if the money stock and income velocity both remain constant, it can be seen that AD varies inversely with the price level, and a panel of it would follow a unitarily elastic downward sloping path when plotted in the coordinate space of Figure 11-1. However, the modern version of the quantity theory of money recognizes that V probably does vary directly with the price level because a changing price level affects the nominal rate of interest which affects velocity. The effect of the changing velocity is to render the AD curve somewhat more inelastic than it would have been in a constant-velocity world.


Increases of Aggregate Demand and Supply

In the long run, the economy's normal operating capacity, N, tends to increase with economic growth, that is, as population grows and productivity increases with technological advance and capital investment. Recent experience in Western market economies indicates that N increases on average between 2 and 3 percent per annum, but varies significantly with cyclical or irregular macroeconomic fluctuations. Long-run growth and short-run fluctuations may be depicted by rightward or leftward shifts of the vertical LRAS curve.

The LRAS curve also shifts to the right in the long run as N increases, carrying with it the SRAS curve. Economists known as supply-siders have hypothesized that in the United States the rightward progress of LRAS may have been retarded in the post World War II period due to the disincentive effect of progressive taxation and the costs to the business community of governmental regulation. Some supply-siders have even suggested the possibility that the LRAS has drifted leftward and the SRAS curve upward and to the left at an almost imperceptible pace in response to these forces. But the SRAS curve may also shift in the short run in response to supply shocks and changing perceptions of costs by producers when inflation or deflation occurs.

Both the output level (Y) and the price level (P) of the economy tend to change in response to shifts of the aggregate demand and aggregate supply curves. In Figure 11-2, an increase of aggregate supply curve from SRAS1 to SRAS2 results in an excess of output relative to demand expenditures at the current price level, with the consequence of inventory accumulation of E'-E1. Businesses incur both explicit and implicit costs when inventories accumulate above planned or intended levels. The explicit costs result from the additional storage and security facilities required to contain the additional stocks. The implicit costs are the opportunity costs of the firm's working capital which is tied up in inventory and thus not available to meet other expenses (payroll, energy bills, etc.) until inventory can be disposed of. The effort by firms to "move" the accumulating stocks will tend to put downward pressure upon prices which may be manifested in reduced list prices, increased discounts off of list prices, or sales.


Figure 11-2.


The macroeconomic adjustment to the rightward shift of the SRAS curve is a movement downward along the AD curve from E1 to E2 as output increases to Y2 and the price level falls. If the rightward-shift of SRAS is attributable to a growth phenomenon, then the normal operating capacity of the economy will increase to N2 and the LRAS will become vertical above it. The managerial implications of a rightward shift of the LRAS curve are clear: productive capacity can be expanded and output increased, although prices may have to be lowered.

In Figure 11-3, a rightward-shift of aggregate demand from AD1 to AD2 produces an excess of demand relative to output at the current price level, and results in inventory depletion of E'-E1. If inventories have become swollen in the past, the depletion may be desirable and occasions no immediate response. However, eventually firms must respond to the depleting inventories or "stock-out" and be "out of business" in those items.


Figure 11-3.


One adjustment to an inventory draw-down is to take the occasion of the strong market to raise prices, although in many cases it may not be possible to implement desired price increases until new menus or catalogs can be distributed. Another means of stemming the inventory depletion is to increase output rates by accelerating assembly processes and using more labor and material inputs. Increased labor usage can be accomplished by extending work hours with overtime, adding shifts, calling back to work employees who have been laid off, or hiring new employees.

Demand pull inflation is attributable to a rightward shift of the aggregate demand curve.  Graphically, the adjustment proceeds along the SRAS curve from E1 to E2 with consequent demand-pull inflation from P1 to P2.  If the new output exceeds the normal operating capacity of the economy, N1, the availability of additional materials inputs may be limited by supply bottlenecks which will eventually result in rising materials costs. Also, this output expansion will tighten the labor market and bid wage rates upward. When these upward price pressures eventually become translated into rising input costs, managers likely will revise production targets and schedules downward, with a consequent leftward shift of short-run aggregate supply until it reaches SRAS2. The price level increase from P2 toward P3 may be understood as cost-push inflation which is attributable to a leftward or upward shift of the aggregate supply curve.  The cost-push inflation comes to an end only when the output level has returned to the economy's normal operating capacity which is sustainable in the long run. The aggregate demand increase achieved no lasting increase of real output, but resulted in price inflation from P1 to P3.

What is the managerial implication of an increase of AD which takes the economy beyond its normal operating capacity? Increasing costs may lead to price increases, but it would be inappropriate to consider expanding plant capacity. By way of comparison, the AD increase illustrated in Figure 10-3 resulted in inflation, but no permanent output increase, while the SRAS increase illustrated in Figure 10-2 followed from a permanent increase in the normal operating capacity from N1 to N2, and yielded some price deflation.


Decreases of AD and AS

As illustrated in Figure 11-4, a collapse of aggregate demand which shifts AD1 to AD2 results in inventory accumulation and induces an adjustment downward along the SRAS curve from E1 toward E2. Unemployment worsens as production levels are cut in response to the increasing inventories, and prices begin to soften. The price level decrease from P1 to P2 may be regarded as demand pull deflation since it resulted from a decrease of aggregate demand.  However, as we have noted above, prices may be somewhat sticky in the downward direction. If prices cannot change to absorb some of the AD collapse, the brunt of the adjustment must be borne by falling output, possibly as far as E'.


Figure 11-4.


If some price deflation does occur, the lower prices eventually are translated into lower production costs. Employers and production planners may alter their plans to increase output, especially after inventories have begun to deplete. This will have the effect of increasing aggregate supply to SRAS2, with adjustment from E2 toward E3 until there is a return to N1, the normal operating capacity of the economy. In such a soft or weakening economy, managers should expect to slow production rates and orders of materials, and may have to cut prices. This second stage of price level decline might be thought of as cost push deflation since it followed from the increase of short-run aggregate supply.  However, since output tends to return to its level before the demand collapse, it may be more appropriate for managers to try to hold production rates constant while letting inventory variation absorb the effects of the demand collapse.

Figure 11-5 illustrates an aggregate supply decrease from SRAS1 to SRAS2 due to a supply shock (the causes of supply shocks are explored in Chapter 14). Product shortages emerge as output falls, and inventories begin to shrink. In such a market environment, prices become firmer and managers may be tempted to take the occasion to announce price increases. As the higher prices become translated into increased production costs, cost-push inflation from P1 to P2 ensues in the adjustment from E1 toward E2. With worsening unemployment which lowers spendable income, aggregate demand can be expected to fall toward AD2, tending to bring prices back to original level before the shock. The economy adjusts along the path from E2 toward E3.


Figure 11-5.


When the falling prices become translated into decreasing production costs, managers may revise production plans and increase output (to replenish depleted inventories). Aggregate supply will shift back toward SRAS1. With the fall of prices and the recovery of the economy, aggregate demand will increase toward its original position, AD1, and output will return to N1, the normal operating capacity of the economy. The adjustment path may pass near to E4 before returning E1. Market and inventory realities may force managers to make the adjustments in production rates and prices just described, but since both aggregate output and prices will tend to return to their pre-shock levels, a better strategy (as in the case of a demand collapse) may be to try to weather the storm by holding constant both prices and production levels, while letting inventories serve as the shock absorber.

Supply-siders have hypothesized that a continuing leftward or upward drift of the SRAS curve (possibly attributable to the disincentive effects of progressive income taxation and to the costs of governmental regulation of business) results in on-going cost-push inflation. The growth of real output may slow, or output may decrease absolutely. Economists have referred to the combination of on-going inflation combined with slow output growth as stagflation. There appears to be no automatic mechanism which will tend to return the economy to its normal operating capacity when SRAS continues an upward-leftward drift. Some hypothesize that the on-going inflation will create an inflation psychology in which managers begin to raise their prices in anticipation of future cost increases. We shall explore in Chapter 14 the role of the government in enabling and in treating such a phenomenon.


Aggregate Expenditure and Aggregate Demand

The aggregate expenditure and aggregate demand schedules do not occupy the same two-dimensional graphic plane even though they share a common axis. Assuming a linear consumption function and that the short run aggregate supply curve does not diverge from the long run aggregate supply curve, the relationship between AE and AD may be illustrated in the three-dimensional panel (a) of Figure 11-6. In this depiction, the left-hand floor axis is the price level, P. The right-hand floor axis is real output, Y. The vertical dimension is expenditures, E. Aggregate demand is represented by the (blue) curved surface ABCD, with functional specification

(1)      AD = g ( P | Y, r, ... ).

In this specification AD is the dependent variable and P is the principal independent variable. Output, Y, the real interest rate, r, and all other independent variables are assumed constant. The long run aggregate supply surface, Y1ZMN, intersects the aggregate demand surface along path JJ' which corresponds to price level P1. The path FJ' above P1J locates the two dimensional aggregate expenditure function at the intersection of aggregate demand ABCD with long run aggregate supply Y1ZMN. P1J' is the E=Y equilibrium path which rises at an angle of 45 degrees from the floor. When panel (a) of Figure 11-6 is viewed from the perspective of the 0Y axis so as to collapse it into two dimensions, it appears as depicted in panel (b), which corresponds to Figure 8-1. When panel (a) of Figure 11-6 is viewed from above, it appears as depicted in panel (c), which corresponds to Figure 11-1.


Figure 11-6.



Income and Expenditure Dynamics

In panel (a) of Figure 11-6, any shift of the aggregate demand surface or the long run aggregate supply surface will change the intersection of the two and the corresponding price level, thereby shifting the aggregate expenditure curve. Figure 11-7 extends Figure 11-6 to add aggregate expenditure surface RSTU in panel (a). Its functional specification is

(2)      AE1 = f ( Y, P | r, ...),

which has two principal independent variables, Y and P, all others assumed constant. The RSTU surface is represented as being linear, but it could in reality be concave downward if the consumption function is concave downward as depicted in Figure 10-2, or if a progressive income tax is imposed (discussed in Chapter 9). Aggregate expenditure surface RSTU slopes upward along the 0Y axis because increasing income enables greater expenditures, but it slopes downward along the 0P axis since, as the price level increases relative to any particular income level, real expenditures must fall.

It can now be seen in panel (a) of Figure 11-7 that aggregate expenditure path FJ' (red) is a vertical slice through the aggregate expenditure surface parallel to the 0Y axis at price level P1. Suppose that long run aggregate supply increases from Y1 to Y2. The new intersection of aggregate supply with aggregate demand is along path KK' (blue). The ensuing adjustment process (described below) causes the price level to fall to P2, and a new slice through the aggregate expenditure surface along path GK'. When viewed from the perspective of axis 0Y as illustrated in panel (b) of Figure 11-7, it is apparent that AE2 lies above AE1. The conclusion is that any increase of aggregate supply which decreases the price level will cause the AE schedule to shift upward. Likewise, a decrease of aggregate supply which increases the price level will cause the AE schedule to shift downward.


Figure 11-7.


Figure 11-8 illustrates the effect of an increase of aggregate demand. In panel (a) the aggregate demand surface shifts outward from the origin to new locus A'B'C'D'. The intersection of the new aggregate demand surface with the unchanged aggregate supply is along path LL'. The ensuing adjustment process (described below) causes the price level to rise to P3, and another new slice through the aggregate expenditure surface along path HL'. This time, in panel (b) the new aggregate expenditure schedule, AE3, lies below the original aggregate expenditure schedule at AE1. The conclusion is that an increase of aggregate demand which causes the price level to rise will cause the AE schedule to shift downward. It also follows that a decrease of aggregate demand which decreases the price level will cause the AE schedule to shift upward.


Figure 11-8.



Conclusions

Aggregate demand and supply analysis coupled to Keynesian Income and Expenditure analysis point to several important macroeconomic conclusions:

1.  Any initiating shift of aggregate demand in a market economy is likely to induce a subsequent opposite-direction shift of short-run aggregate supply as people "wake up" to what is happening to their incomes and costs of living.  A comparable conclusion may be stated with respect to an initiating shift of short-run aggregate supply.

2.  A first-stage of demand-pull inflation (or deflation) in a market economy is likely to elicit a second-stage of cost-push inflation (or deflation).  A comparable conclusion may be stated with respect to a first-stage of cost-push inflation (or deflation).

3.  A market economy may be induced by an aggregate demand increase to produce temporarily at a rate of output exceeding its normal operating capacity, but subsequent adjustment forces will tend to return the output rate to its normal operating capacity.

4.  A collapse of aggregate demand (as may have occurred in the Great Depression of the 1930s) will result in falling output and price level, but subsequent adjustment forces will tend to return both the output rate and the price level to their former levels.  However, the time required for recovery may be excessive relative to political realities.

5.  A supply shock in a market economy will lead to a temporary decrease of the rate of output and corresponding increase of the price level, but subsequent adjustment forces will tend to return both the output rate and the price level to their former levels.  No such conclusion follows from progressive decreases of aggregate supply attributable to excessive regulation or progressive taxation.

6.  Although the market economy's output rate tends to gravitate toward its normal operating capacity, increases of aggregate demand or aggregate supply may lead to permanent changes of the price level.


What's Ahead

We now have in place complementary macroeconomic models which are sufficient to many of the needs of identifying the managerial implications of macroeconomic change. In subsequent chapters we may apply the models to matters of growth, inflation, shocks, and policy.