Output and Inflation
Aggregate
demand determines the actual output of goods and services produced. Given a
fixed potential output for any short-term period, aggregate demand will thus
determine the unemployment rate. In the simple model used above, when aggregate
demand exceeds potential output, an inflationary gap exists and the price level
rises. However, in the real world, inflation occurs at or below full
employment.
The
inflation rate for a given time period is the per year change in price level:
INFT = (PT+1-PT)/PT. The price
level represents the overall price of all goods and services taken together.
The most commonly cited measure of the average price level is the Consumer
Price Index (CPI). This provides an index of typical consumer products
purchased by average households. However, it does not take into account the
roughly one-third of total output represented by investment expenditure. The
price level index which includes all goods and services in the economy is
called the GDP deflator.
Most firms
set their prices based on the anticipated costs of production and the
anticipated demand for the goods and services produced. These expectations are
based on past performance, economic indicators, and the thought processes of
managers. The most recent level of aggregate demand is one of the key factors
determining these expectations. The higher aggregate demand, the higher the
firm’s own recent demand is likely to have been, and the higher its
expectations of future demand. In addition, the higher the aggregate demand,
the higher the firm’s expectations about the cost of labor, materials and other
factor inputs. As a result, the higher recent aggregate demand has been, the
higher a firm is likely to set its prices. If all firms operate in this
fashion, then the rate of increase of the price level will be directly and
positively related to the level of aggregate demand.
In any short
run period, therefore, aggregate demand will influence both the unemployment
rate and the inflation rate. As a result, there will be an implied relationship
between unemployment and inflation. For each possible level of aggregate
demand, there will be corresponding rate of unemployment and of inflation. The
graph of unemployment against inflation for a varying level of aggregate demand
in the short run is called a Phillps
Curve:
In the real
world, the constraint that Y cannot exceed Q is somewhat relaxed, because of
the way we have defined full employment. Facing demand exceeding Q, some
fatories and workers can work overtime and the average frictional and
structural rates of unemployment will fall because there are so many unfilled
vacancies. Thus the economy in the short
run can ‘squeeze’ some extra production out of its resources. However, the cost
of this economic ‘boom’ is that factor prices will rise and consequently the
price level will rise at a rate higher than normal. This is represented by the
increasing slope of the Phillips Curve as unemployment goes above UF.
One might
expect inflation at full employment to be zero, because at over-full
employment, scarcity of factors of production will lead to rising demand and
thus rising prices; at under-full employment, abundance of factors of
production will lead to reduced demand and thus reduced prices; and at exact
full employment, demand and supply will be in equilibrium, resulting in stable
prices. Empirically, however, the position of the Phillips Curve has been such
that some positive rate of inflation occurs at the full emploment rate of
unemployment. This is caled the inflationary bias.
Inflationary Bias
The labor
market is quantitatively the most important factor of production, since it
accounts for roughly two-thirds of all income payments by firms. The ‘labor
market’ is in fact many different markets, as workers are specialized in many
different skills. At any given time, there will probably be some skills that
have excess demand while other skills have excess supply. All these markets
operate imperfectly, in the sense that wages do not adjust immediately to
equate supply and demand. This is particularly evident in many markets where
there is excess supply; wage rates are observed not to respond to the downward
pressure. Wage rates appear to be ‘sticky’ in the face of high unemployment.
Many reasons
are given for this observation. One is that in many labor markets, wages are
determined by collective bargaining between unions and management, sometimes
for an entire industry. The political nature of union decision-making is such
that a reduction in wages is exceedingly difficult to obtain, regardless of
economic circumstances. A reduction in wages makes everyone somewhat worse off.
However, a failure to reduce wages makes certain people (those laid off) much
worse off, to the benefit of others (those who keep their jobs). If the
economic downturn is anything short of catastrophic, less than half the workers
are likely to be laid off. If the workers have a good idea who will be axed,
then the majority of workers, voting in their own self-interest, will elect to
keep their current wages. In addition, those workers with the most seniority
are the least likely to be laid off, therefore the most likely to oppose wage
reductions—but this group of people are also likely to hold the most
influential positions within the union. Another explanation is that given that
the government will pay unemployment benefits for a while, a typical worker may
be better off accepting work at a high wage in the knowledge that there will be
occasional layoffs, than accepting work at a lower wage that continues
indefinitely.
This
rigidity does not occur in the upwards direction. Workers are always generally
happy to accept more money. As discussed previously, full employment does not
mean zero unemployment. It is still possible (even likely) that under full
employment, some labor markets will have excess demand while others will have
excess supply. In markets with excess demand, wages can be expected to rise
relatively quickly, but in markets with excess supply, wages will only fall
slowly, if at all. Firms which face excess demand for labor will expect their
costs to rise and will therefore set higher prices. However, firms which face
excess supply of labor will not have a reasonable expectation of falling costs,
and will therefore leave prices unchanged. This will result in an increase in
the average price level.
If
unemployment rates are high enough, the downward pressure on wages will be
sufficient to overcome downward wage rigidity and wages and prices will fall.
There have been very few occasions where this has occurred; the most striking
example is the Great Depression in the 1930s, where, in the face of extremely
high unemployment rates, the inflation rate was negative for several years on
many countries.
Properties of the Phillips Curve
The Phillips
Curve has another important property: It is not linear. Its curvature suggests
that the nature of the trade-off between inflation and unemployment depends on
where the economy currently falls on the curve. At high rates of unemployment,
the curve is relatively flat: It takes a large increase in unemployment to
effect a small increase in inflation. At lower rates of unemployment, a small
change in unemployment will result in a much larger change in inflation.
This can be
explained as follows: If the initial condition is high unemployment, then most
labor markets will be characterized by excess supply and very few by excess
demand. If unemployment increases, the excess demand in those few markets will
be reduced, so those few firms will still increase their prices but not by as
much as they would have otherwise. However, the downward rigidity of the labor
markets already experiencing excess supply will be such that the firms
operating in those markets will not change their prices. As a result, the
change in the rate of inflation will be small. However, if the initial
condition is very low unemployment, most labor markets will be characterized by
excess demand and very few by excess supply. If unemployment increases, the
upward pressure on wages will be decreased, perhaps sharply in those cases
where the initial excess demand was severe. Since very few labor markets were
in excess supply conditions, most firms will still expect an increase in labor
costs, but less (possibly much less) than previously. As a result, price
increases for the majority of firms will be less than they would have been, and
there will be some firms which might otherwise have set very sharp price
increases who no longer need to do so. The reduction in the rate of incrase of
the average price level will be very noticeable.
The
existence of a Phillips Curve causes a problem for government policymakers. A
choice must be made between the evils of unemployment and of inflation. Policy
tools that affect aggregate demand cannot be used to fight inflation and
unemployment at the same time. If aggregate demand is controlled to achieve
full employment, some inflation will generally result. If aggregate demand is
controlled to eliminate inflation, high unemployment will generally result.
Employment vs. Inflation in the Long
Run
The Phillips
Curve is a short run relationship. In the long run, the Phillips Curve can
shift its position and change its curvature. Thus, the rate of inflation in the
long run depends not only on where on the Phillips Curve the economy is
operating, but also where the Phillips Curve is positioned. The fact that the
Phillips Curve can shift over time causes some difficulty in interpreting
historical data, because it is hard to know whether any given change reflects a
shift in or a movement along the Phillips Curve. A moving Phillips Curve can
result in a situation where unemployment and inflation move in the same
direction. This has caused some to conclude that there is no Phillips Curve. It
is important to remember that the Phillips Curve is a short run relationship.
However,
short-run decisions should not be made in the absence of consideration of their
long-run impacts. It might be possible, in normal situations, that high
unemployment and low inflation today might permit preferred combinations of
unemployment and inflation that would not have been possible otherwise. In such
situations, the appropriate sort-term policy goal might be to choose an
aggregate demand target below potential output.
