Getting started
Inflation refers to a sustained and general increase in prices (decrease in the value of money), not a one-shot price increase nor a price increase specific to one or a limited set of goods and services. Such price increases may arise from either increases in aggregate demand or decreases in aggregate supply.
We distinguish two distinct mechanisms by which inflation may be initiated. A demand-pull inflation refers to an inflationary process that begins with an expansion of aggregate demand. Any factor contributing to increased aggregate demand --such as increases in the money supply, in gov't expenditures, or in net exports -- can initiate a demand-pull inflation. The most notable such inflation in recent U.S. history was that which began in the late 1960s with Lyndon Johnson's "guns-and-butter" policy (sharply increased government expenditures for the war in Vietnam, while social spending also was increasing).
Increased aggregate demand tends to pull up prices as it also increases real GDP, but in the long run the resulting overheating of the economy contributes to wage increases. This in turn shifts the SAS curve to the left and hence reduces real GDP while further increasing the price level.
Note, however, that the process just described corresponds to a one- (or at most two-) shot increase in the price level. For sustained increases in prices to occur, it is necessary for there to be continuing increases in aggregate demand. As Parkin notes, this will occur only if there are continuing increases in the quantity of money.
The second variety of inflation is referred to as cost-push inflation. Here, the initial impulse promoting higher prices is an increase in costs of production (wages or raw materials).
Such an increase shifts the SAS curve to the left, and thereby tends to push prices up while at the same time reducing the level of real GDP (stagflation).
As before, this represents but a one-shot change. In order for a cost-push supply shock to result in an inflation spiral, it is necessary for there to be a stimulative response on the part of monetary authorities (undertaken to stimulate the economy and increase real GDP) that pushes out aggregate demand, and that is followed by a subsequent cost-push increase and increased aggregate demand. This description characterizes experience in the U.S. following the first oil shock in the 1970s.
Expectations and inflation
Irrespective of the source of inflation, economic agents factor in expectations of inflation to their behavior (note the endogeneity of economic outcomes with respect to expectations; contrast with expectations about the weather). In particular, nominal wage increases sought by workers and agreed to by employers are determined in part by the rate of inflation expected by the respective parties (reflects efforts to maintain real wage rates). Similarly, nominal interest rates will be viewed by lenders and borrowers through the lens of anticipated inflation (the higher the expected inflation rate, the lower the real interest rate, cet. par.; recall that the real interest rate equals the nominal interest rate minus the rate of inflation).
If expectations regarding inflation are incorrect (what Parkin calls unanticipated inflation), this entails both redistributive consequences and costs for the economy. For example, if inflation is underestimated (actual inflation turns out to exceed expected inflation), there will be a redistribution of income away from workers and lenders and toward firms and borrowers (negotiated wage increases will not maintain real wages, and nominal interest rates will yield relatively low real interest rates). Conversely, if inflation is less than expected, workers gain at the expense of firms, and lenders gain at the expense of borrowers.
In addition to these redistributive consequences of incorrect expectations of inflation, there are also some real costs. In the labor market, incorrect expectations result in real wage rates that are either too high or too low. In the former case, firms end up laying off some workers, producing lower output, and earning less profit; in the latter case, firms seek to hire additional labor, but workers are more likely to change jobs in search of higher real wages -- and this increased labor turnover entails higher costs for workers (job search) and firms (overtime, maintenance costs).
In the capital market, incorrect expectations of inflation result in a real interest rate that is too low or too high. This, in turn, means that the amount of lending and borrowing that actually occurs will be different from what lenders and borrowers would have chosen with perfect foresight (cf., ex ante vs. ex post).
Clearly, then, economic agents have a stake in accurately predicting levels of inflation. Economists assume that people make use of all available relevant information in their efforts to anticipate the future. In particular, forecasts based on all available information are characterized as rational expectations. For purposes of predicting inflation, relevant information here means the information that economic theory predicts to be relevant to determination of the price level.
Information relevant to determining the price level is reflected in expectations concerning aggregate supply and aggregate demand. As before, we must distinguish between the short run and the long run. In the short run, expected inflation will be determined by expectations regarding aggregate demand (and especially, growth of the quantity of money) and aggregate supply (and especially, the level of money wages).
In the long run, by contrast, wages are flexible and eventually adjust to the price level so as to achieve full employment. Hence, expected aggregate demand and expected long-run aggregate supply determine the expected price level and hence expected inflation.
Now let's consider the role of expectations in influencing the process of inflation. In particular, if aggregate demand is expected to increase and thereby raise prices, there will be corresponding upward pressure on wages so as to keep the real wage rate steady. If the expectations of increased demand are correct (and wages are adjusted accordingly), the economy will experience inflation as anticipated, with no change in real GDP.
However, if the actual growth of aggregate demand differs from expected growth, then actual inflation will be different from its expected level. The resulting incorrectly anticipated inflation will lead to deviations of real GDP from potential GDP. Thus, we see that what is expected can influence the outcome -- i.e., outcomes are endogenous with respect to expectations.
More specifically, if actual growth in AD exceeds expected growth, there is unanticipated inflation that looks like demand-pull inflation: real GDP exceeds potential GDP, placing additional upward pressure on wages (because prices have increased by more than wages) and setting the stage for a demand-pull inflation.
Conversely, if actual growth in AD falls short of expected growth, there is a good chance that the "too-large" wage adjustments will initiate a process of cost-push inflation. That is, at the outset there will be a recessionary gap, with unemployment above the natural rate. If policy-makers stimulate aggregate demand to reduce unemployment, cost-push inflation may result.
Clearly, then, there are costs of unanticipated inflation, in the form of deviations of real GDP from its full-employment level. As Parkin notes, however, there are also costs of anticipated inflation. These costs include transactions costs ("bootleather" or search costs plus efforts to avoid using the depreciating currency), decreases in potential GDP (increased transactions costs divert resources from producing goods and services; cf., the flat tax and tax accountants/lawyers), and declines in long-term growth (rising effective tax rates, increased uncertainty about long-run inflation slowing investment, resources devoted to avoiding the costs of or profiting from inflation rather than producing innovations). These costs are small when inflation is low, but they are likely to be substantial in high-inflation environments.