A Detailed Treatment of Aggregate Demand

Keynesians turned macroeconomic theory and policy completely around as they shifted the focus from the supply-side to the demand-side of the economy.   The secret to hitting an income target was the effective management of aggregate demand.  The logic of the Keynesian position is apparent in the AS-AD diagram below.  Once a target rate of output/income has been established (Y*), the policy maker needed only to design a policy that would move aggregate demand to the appropriate level. To successfully hit these targets, policy makers needed to have a complete understanding of the individual components of aggregate demand. 

Keynesian AS-AD Model

KeyASAD.gif (2291 bytes)

Fortunately, the theorists devoted a considerable amount of time and effort to searching for the secrets of the five components of demand [Consumption, Investment, eXports, iMports, and Government expenditures].  Their energies were focused on the first two of these and in this section we will follow that lead and examine in some more detail Consumption and Investment spending.


Consumption and Saving have always been at the center of discussions of macroeconomics, although the focus has shifted between the two. In the pre Keynesian era the focus was on Saving, with Consumption viewed as simply a residual. Given a level of income determined in the labor market, analysis of household behavior focused on the decision to save.

This changed, however, with the Keynesian emphasis on aggregate demand.  Consumption spending is the largest component of aggregate demand and therefore any effort to understand aggregate demand would need to start with consumption. The revival of conservative theory in the 1980s prompted a rediscovery of saving, but this discussion will need to wait until we get to the 1980s. At this time we will look a bit more closely at consumption spending.

Consumption is by far the largest component of aggregate demand, accounting for approximately two-thirds of total spending - a share that has been increasing since the mid 1960s.  Consumption spending consists of the three categories durables (ex. automobiles, appliances), nondurables (ex. clothing, food) , and services (ex. education, health care).

As for the separate elements of consumption, a faster growth in durable and service  expenditures has decreased nondurables' share of consumption spending from approximately 1/2 to 1/3.  The share accounted for by durables and services, meanwhile, has increased from 1/9 to 1/6 and from 2/5 to 1/2. Throughout the entire period we also find that expenditures on durables tend to be quite volatile while expenditures on services and non durables tend to be quite stable.

ConsGDP.gif (3249 bytes)Cons_parts.gif (3434 bytes)

Theoretical analysis of consumption began with Keynes who took the eminently plausible position that the single most important determinant of consumption was income.  An increase in income could be expected to increase consumption by some fraction of the increase in income.  As Keynes saw it, the low level of demand could be traced back to the low level of income - people without jobs and income would not be buying. This certainly seemed to be born out by the data as evident in the scatter diagram of personal consumption expenditures and GNP data from the US economy, although remember Keynes did not have access to these data.

Cons_Income.gif (4455 bytes)

The scatter diagram based on historical post WWII data provides strong evidence of a relationship between the income and consumption expenditures - in fact it is a VERY good relationship that we could capture in a linear equation.  This is why you see a consumption function of the form C = a + b*Y become a standard feature of macroeconomic models.  If you look at the scatter diagram and think about drawing a line that passes through the points, a would be the intercept and b would be the slope.  Of these two parameters, it was the slope which Keynesians focused on since this is the marginal propensity to consume (MPC) that was at the center of the multiplier concept.

As we saw earlier in our discussion of the 1930s, the implications of this observation for macro policy can not be overestimated since it produced a complete reversal of the perception of government spending's impact on the economy. Where we once viewed an increase in government spending as crowding-out private spending, we now saw increases in government spending acting as a catalyst to economic expansion.  But there were some potential problems. First, there was the fear of stagnation that seemed to be the logical conclusion to be drawn from the model. To understand the nature of the perceived problem, let us simply divide the consumption equation above by Y to obtain the equation specifying consumption's share of total income.

C/Y = a/Y + b

As the economy expands, consumption's share of total income will fall because the term a/Y declines as Y increases. If this happens, then one of the other components of aggregate demand would need to increase faster than the economy to keep the economy from stagnating.  Stated a bit differently, since the MPC was less than 1, as income increase consumption would increase more slowly.  If the economy was to continue to expand, there would need to be above average increases in the other components of aggregate demand or growth would be slowed.  The logical choice would have been the government, but this triggered concerns about the growth in the size and scope of government involvement in the economy.

There was also the issue of the marginal propensity to consume (MPC) which has a key role in the Keynesian model since the MPC shows up in the formula for the multiplier. The higher the value for the MPC, the larger the value of the multiplier and therefore the greater the effectiveness of fiscal policies. 

One "solution" to the problem was offered by Milton Friedman who developed the Permanent Income Hypothesis that focused attention on lifetime earnings rather than current income as the primary determinant of consumption.  At its center was the view that a medical student would consume more than a carpenter because of the higher expected lifetime income of the student. Similarly, if there was a slowdown in the economy during a recession, individuals would be unlikely to substantially adjust downward their estimates of lifetime income and therefore they would tend to adjust to the income loss with lower saving rather than lower consumption spending - a low value for the MPC.

A different slant on the problem was provided by the Life-Cycle Model of consumption built on the assumption that lifetime consumption was linked to lifetime income through an 'intertemporal' budget constraint - what you spend today depends on what you expect to earn in future years and spending decisions are based on a preference for avoiding large changes in spending.  The implication of the model is similar to that of the Permanent Income Model.  In the short run we may find income changes producing less than proportional changes in consumption spending as individuals attempt to smooth their consumption patterns over their lifetimes.  Both theories highlighted the importance of consumer expectations as a determinant of demand and the differences between temporary and permanent income changes.  What is added by the Life-Cycle model is age structure as an influence on aggregate spending.  This latter effect helps explain the difficulties Japanese officials experienced in the late 1990s when they attempted to get aging Japanese consumers to spend money to stimulate aggregate demand. The Japanese, worried about retirement, continued to save despite the government's efforts to get them spending.  Finally, the model provides a basis for understanding how a greater fear of nuclear war translates into higher consumption spending. If you expect nuclear war, you expect to die earlier and therefore you will save less for your retirement.   These models also suggested that wealth should be a determinant of consumption spending, which explains how the performance of the stock market can affect spending.  

Once we accept the intertemporal nature of consumption decisions, another factor exerting a potential impact on consumption spending is the structure of the capital market.  If it is easy to borrow funds, you would expect that consumption spending would tend to be higher.  If you think of social security as a policy designed to make it very easy for individuals to live beyond their earned income after retirement, then you would expect that social security systems would lower savings and increase consumption.  Similarly, the greater down payments that Asian households must pay for their homes reduces current consumption spending and raises savings.


The second component of aggregate demand is investment spending which is a relatively small component of aggregate demand. It's importance is derived from its volatility, its supply - side effect, and its sensitivity to interest rates. As you can see in the diagram below, the swings in the level of investment spending are substantially larger than the swings in consumption spending, although the trends appear to be similar.  The most graphic indication of the volatility of investment spending came in an earlier time period, in the Great Depression, when investment spending fell 90 percent.

Invest1.gif (4376 bytes)

Just as consumption spending can be decomposed into three separate components, so too can investment spending be decomposed.  The two broadest categories are fixed investment and change in business inventories. Under the category of fixed investment there are two components - residential and nonresidential investment, the latter being further decomposed into structures and equipment.  

As you can see in the diagram below, of the components of investment spending, nonresidential structures is the most stable and producer durables (equipment) is the fastest growing.  In 1995, investment spending in producer durables was 8 times higher than the level in 1959, while nonresidential structures was only twice the earlier level. 

Invest_parts.gif (6614 bytes)

Inventory investment is important for two reasons. First, it is the key link between the macro identities outlined earlier and the equilibrium conditions that are at the heart of the macroeconomic models. By definition, any mismatch between production and consumption shows up as a change in inventories. Included here would be desired changes in inventory, such as inventory reductions reflecting the move toward just-in-time production, and undesired changes, such as automobiles that accumulate in GM lots due to a drop in demand. Only when there is no undesired changes in inventories, will we find ourselves in equilibrium.

Inventory investment is also important because it is widely recognized as a leading indicator. If, for example, there is an undesired build-up of inventories, then this is likely to send a signal to producers to cut back on production to bring their inventories back to desired levels. The build-up in inventories will 'lead' the eventual reductions in output and employment and the rising rate of unemployment. For this reason, inventory investment is one of the components in the index of leading indicators published monthly by the Conference Board. The major weakness of this inventory investment as a leading indicator is that there is no way to decompose the published statistics on inventory investment into the desired and undesired components.

Residential construction is also a very volatile component of aggregate demand.  The range of values for year-to-year changes in residential construction is twice that for investment in plant and equipment. More importantly, the turning points in residential construction investment tends to precede the turning points in GNP. To understand the business cycles, therefore, one needs to have some understanding of the determinants of this component of spending, something which we will examine in detail at a later time. At this time it is important only to know that residential construction is also important because it is generally recognized as being the component of aggregate demand that is most sensitive to macro policy changes. For example, the Fed's decision in the early 1990s to drive down interest rates was designed to quickly increase the demand for housing, which would in turn generate increased construction expenditures. Finally, residential construction is important for psychological reasons. There is probably no better sign of economic growth then the construction of new homes. A construction boom will be accompanied by a growing optimism, reflected in measures of consumer sentiment, and firming of housing prices which will increase household wealth, a determinant of consumption expenditures.

Investment spending on plant and equipment (structures and producer durables) is considered important to the welfare of the economy because it has a supply-side effect in addition to its demand-side effect. The increased expenditures on plant and equipment will generate demand initially, but in the future the expansion and modernization of the capital stock can be expected to translate into higher productive capacity. It is for this reason that so much attention in recent years has been focused on tax policies such as investment tax credits and accelerated depreciation, that have been aimed at increasing investment expenditures.

The final issue regarding investment expenditure is its sensitivity to interest rates.  The importance of this feature of investment spending can be traced to the importance of interest rates in our later discussion of monetary policy.  An example of this would be the decision by Alan Greenspan to reduce interest rates in late 1998, or to raise rates in early 2000.  The decision in 1998 was made in an effort to stimulate spending to offset the declines in demand associated with the Asian crises, while the 2000 decision was designed to slow down an overheated economy.  One way to see the relationship between interest rates and investment spending would be to use one of the on-line financial calculators to determine the influence of interest rates on monthly mortgage payments.  What you will find is that an increase in the interest rate will raise the monthly payment which should reduce the number of individuals who buy a house.

As for the other two components of demand, we will examine them in some detail at later times.  We will look at the international components in our discussion of the 1970s while the government will be the subject of our discussions of the 1980s.