1、 ECONOMICS Chapter 2 MACROECONOMIC DYNAMICS Copyright 2008 AZEK/ILPIP ECONOMICS Copyright 2008, AZEK/ILPIP All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording, o
2、r otherwise, without the prior written permission of AZEK/ILPIP. Table of contents 2. Macroeconomic dynamics 1 2.1 Inflation versus unemployment: The great trade-off? 1 2.1.1 Unemployment and inflation: The Phillips curve . 1 2.1.2 The modern version of the Phillips curve . 4 2.2 Economic growth* 9
3、2.2.1 Some stylised facts* 9 2.2.2 Growth accounting* 10 2.2.3 Capital accumulation and economic growth* . 13 2.2.4 Technological progress and economic growth* 15 2.3 Business cycles* 17 2.3.1 The basics* . 17 2.3.2 The classical approach: Theory of exogenous business cycles*. 21 2.3.3 The Keynesian
4、 approach: Theory of endogenous business cycles* 23 2.3.4 Fiscal policy, monetary policy and the business cycle* 24 * final level chapter 2 / page 1 2. Macroeconomic dynamics 2.1 Inflation versus unemployment: The great trade-off? In Chapter 1, we have defined the inflation rate as the rate of incre
5、ase in the general price level. In this section we will show that there may be a close link between inflation and unemployment. This link has been discovered in a famous article by the New Zealand-born economist A.W. Phillips.1 Phillips found a negative relation between unemployment and nominal wage
6、 inflation in the UK. This relation is now known as the Phillips curve. 2.1.1 Unemployment and inflation: The Phillips curve Before venturing any further in this analysis, we need to define what is meant by “unemployment”. The most widely used measure is the unemployment rate (u). It is equal to the
7、 number of unemployed workers divided by the total labour force: . 100forceLabourunemployedofNumberu = Unemployed workers are those who are willing and able to work but cannot find a job. The labour force is not equal to the total number of people of working age. Rather, it represents the sum of all
8、 individuals who in principle are able and willing to work, whether currently employed or not. 0.0%2.0%4.0%6.0%8.0%10.0%12.0%14.0%1960 1965 1970 1975 1980 1985 1990 1995 2000USA ItalyFigure 2-1: Unemployment rates in the US and Italy, 1960-2000 1 A.W. Phillips, “The Relation Between Unemployment and
9、 the Rate of Change of Money Wage Rates in the United Kingdom, 1861-1957”, Economica, November 1958, pp. 283-299. chapter 2 / page 2 By way of an example, Figure 2-1 shows unemployment rates in the United States and Italy for the period 1960-2000. Until 1980, the unemployment rates in both countries
10、 fluctuated between 4% and 9%. Since then, the US unemployment rate began to fall steadily, while the Italian unemployment rates (along with those of most other Western European countries) rose sharply from the mid 1980s onwards. Unemployment is thus an issue of particular concern to European macroe
11、conomists. Here, we are concerned not with the causes of unemployment, but with the relationship between unemployment and inflation, which is one of the most hotly debated topics in macroeconomics. Initially, the Phillips curve received strong empirical support; in the 1960s many empirical studies d
12、ocumented a clear negative relation between unemployment and inflation for the major industrialised countries. This is confirmed by our Figure 2-2, which plots inflation against unemployment rates in the United States and Italy for the 1960s. A negative relation is apparent between the rates of infl
13、ation and unemployment in both countries. This pattern was typical for most industrialised economies during that period. Can we conclude from this that there is a stable trade-off between inflation and unemployment, as suggested by Phillipss original paper? The evidence of the 1970s and the early 19
14、80s has shown that the relation between these two macroeconomic variables is not so simple. Particularly in the 1970s, the United States and most other industrialised countries experienced periods of both higher inflation and higher unemployment. This unfortunate economic phenomenon, commonly referr
15、ed to as “stagflation”, is not consistent with the classical Phillips curve relation. (a) chapter 2 / page 3 (b) Figure 2-2: Phillips curves for the United States (a) and Italy (b), 1961-1969 (a) chapter 2 / page 4 (b) Figure 2-3: Phillips curves for the United States (a) and Italy (b), 1971-2000 Fi
16、gure 2-3 illustrates recent inflation and unemployment rates for the United States and Italy. Clearly, the data do not show a stable negative Phillips-type relation. In the US case, it looks as if there was no relation at all. 2.1.2 The modern version of the Phillips curve In addition to its empiric
17、al failure, the Phillips curve has been attacked on theoretical grounds. In particular, some leading economists began to argue in the late 1960s that instead of a simple general relation between inflation and unemployment, there only exists a relation between unexpected inflation and cyclical unempl
18、oyment. This modern version of the inflation-unemployment trade-off has become known as the expectations-augmented Phillips Curve. The first step in an exposition of this theory is to define various types of unemployment that are distinguished in the models. The natural rate of unemployment (u*) is
19、defined as the rate that will prevail if the labour market is in equilibrium. Thus, somehow paradoxically, the natural rate of unemployment corresponds to full employment. This reflects the reality that there will always be some unemployment arising from the fact that workers who change jobs or ente
20、r the labour force mostly need some time to find suitable employment. Unemployment that arises from time spent on job searching is referred to as frictional unemployment. A related concept is that of structural unemployment, which is defined as unemployment which results from a mismatch of job vacan
21、cies with the supply of labour available, caused by shifts in the structure of the economy. Frictional and structural unemployment together make up “natural” unemployment. chapter 2 / page 5 The difference between the actual unemployment rate and the natural rate is called the cyclical unemployment
22、rate.2 If cyclical unemployment exists, the labour market is not in equilibrium, since the demand for labour is insufficient to absorb the supply of labour. Suppose that the average wage in the economy is determined as follows: , )uF(uE(P)W *= where W average nominal wage E(P) expected price level.
23、The term (u-u*) represents cyclical unemployment. F is a negative function, since the wage relates negatively to the rate of cyclical unemployment. This expression thus implies the assumption that unemployment is a disciplining device in the sense that when unemployment is high, the average wage lev
24、el will tend to fall. Also note that wages are expressed as a function of the expected price level. This reflects the realistic assumption that workers, in negotiating future wages, will take account not just of the current price level but of their forecast of inflation over the period of the wage c
25、ontract. This wage-setting relation is very similar to the one we introduced in section 1.5.1. The only difference is that cyclical unemployment is explicitly taken into account. Some algebraic manipulations now allow us to derive the modern version of the Phillips curve. Suppose that we have a simp
26、le form of the function F: ( ) . uu1E(P)W *= If firms set their prices as a constant mark-up on the nominal wage, the price level in the economy will be: ( ) .W 1P += Note that this equation corresponds to the price-setting relation of section 1.5.2. (where A = 1) Substituting for the average wage l
27、eads to: ( ) ( ) . uu1E(P)1P *+= After introducing time subscripts, this can be rewritten as: ( ) ( ) . uu1)E(P1P *tttt += Dividing both sides of the above equation by Pt-1, we obtain: ( ) ( ) . uu1P )E(P1PP *tt1tt1tt +=As previously shown, the inflation rate is 2 Sometimes this type of unemployment
28、 is also referred to as “Keynesian” unemployment. chapter 2 / page 6 . 1PPP PP1tt1t1tt =tpi Hence, . 1PP1tttpi+=Substituting this expression in the above formula, and noting that , )E(1P )E(P1tttpi+=leads to: ( ) . uu1)E(1()1()1( *tttt pi+=pi+ Taking logs from both sides, we obtain3: . )u(u)E( *tttt
29、 +pi=pi This equation describes the expectations-augmented Phillips curve, which states that there is a negative relation between unexpected inflation, (pit E(pit), and cyclical unemployment, ( *tt uu ). It is common practice to plot inflation against unemployment, rewriting the above equation in th
30、e following form: . uu)E( t*ttt += pipi Figure 2-4 illustrates this linear version of the expectations-augmented Phillips curve, which is also known as the short-run Phillips curve. We find the traditional negative relationship between unemployment and inflation. Remember, however, that we are deali
31、ng here with cyclical unemployment and with unanticipated inflation. The negative relation only holds if the other variables on the right-hand-side of the last equation remain unchanged. An increase in expected inflation E(pi) or in the natural rate of unemployment, u*, increases the constant term o
32、f the Phillips curve and the curve moves up and to the right, from PC1 to PC2. This is illustrated in Figure 2-4 by a move from M to M, which could represent, for instance, an increase in expected inflation. This suggests that the trade-off between inflation and unemployment is not stable. Whenever
33、expected inflation or the natural rate of unemployment changes, the Phillips curve will shift. 3 Remember that ln(1+x) x for x small enough. chapter 2 / page 7 Figure 2-4: The short-run Phillips curve Now, suppose that agents correctly anticipate inflation, in the sense that the expected inflation i
34、s equal to the actual inflation (E(pit) = pit). This is a realistic assumption in the long run since economic agents cannot permanently underestimate or overestimate inflation. This enables us to simplify the Phillips curve equation as follows: , )u(u)E( *tt += pipi tt , )u(u0 *tt = . uu *tt += The
35、last equation shows clearly that, when workers can perfectly predict inflation, the rate of unemployment is completely unaffected by the inflation rate. Since the / term is likely to be small (especially in competitive markets), we can neglect it and rewrite the above relation in the form: . uu *tt
36、The above equation describes the long-term Phillips curve. Hence, in the long run, when there are no inflation surprises, the actual rate of unemployment is equal to the natural rate, regardless of the inflation rate. This is illustrated in Figure 2-5. Unemployment, uPC112u1PC2MMchapter 2 / page 8 F
37、igure 2-5: The long-run Phillips curve According to the modern version of the Phillips curve, therefore, policy makers do not have a general choice between fighting inflation or fighting unemployment. They cannot reduce the unemployment rate by allowing inflation to increase, except for very short p
38、eriods of time. The problem with trying to reduce unemployment by generating “surprise inflation” is that economic agents will consider this years surprise as next-years expected inflation rate and build this into their wage agreements. Hence, once there has been an unanticipated surge in inflation
39、in this model it will be difficult and costly to reduce the inflation rate again. A short-term inflationary boost may well yield a short-term drop in cyclical unemployment, but in the long term unemployment will return to its natural rate, but inflation could be stuck at a permanently higher level.
40、This is another way of formalising the monetarist view, according to which monetary policy (which affects inflation) has no power to boost real output in the long run. This reasoning also underpins the view that monetary policy should be conducted by an independent central bank. Elected governments
41、might be more short-termist in their policy goals than central bank officials with a fixed-term appointment. In particular, it has been argued that politicians might create surprise inflation ahead of elections, so as to reduce cyclical unemployment on election day. In the long run, such a policy wo
42、uld have no permanent effect on unemployment, but it would saddle the economy with high inflation. Unemployment, u12u*Long-term PCchapter 2 / page 9 2.2 Economic growth* 2.2.1 Some stylised facts* Economic growth is the expansion of a countrys aggregate output. Countries that enjoy faster economic g
43、rowth will have higher future income and will provide more consumption and higher living standards for their citizens. Growth is therefore the principal focus of long-run macroeconomic analysis. Before introducing the main theories of economic growth, it is enlightening to give some stylised facts a
44、bout growth, by looking at some cross-country evidence. Figure 2-6 shows real GDP per capita for several countries for the period 1950-1992. 020004000600080001000012000140001600018000200001950 1955 1960 1965 1970 1975 1980 1985 1990JapanVenezuelaIndiaUSACHGermany1985 USDFigure 2-6: Real GDP per capi
45、ta for a sample of countries, 1950-1992 (Source: Heston, A. and Summers, R., Penn World Tables 5.6) Since aggregate income is distributed among a countrys citizens, GDP per capita is a more illustrative indicator of living standards than total GDP. GDP per capita is simply GDP divided by the populat
46、ion. Moreover, it is difficult to compare growth rates of output in different countries when output is reported in local currency and in nominal terms. For this reason, we have reported data series in constant US dollars (with reference year 1985). Some salient features are evident from Figure 2-6:
47、Per-capita incomes are on a long-term increasing trend in most countries. Most of our sample countries are substantially richer in 1992 than they were in 1950. The example of Venezuela, however, shows that protracted periods of negative growth are also possible. chapter 2 / page 10 The income gap be
48、tween rich and poor countries is large and persistent. For example, US per-capita income was about 14 times that of India in 1992 as well as in 1950. There are large differences in growth rates of some countries. Thus, in 1950, Venezuela enjoyed a higher GDP per capita than West Germany. Over the su
49、bsequent 40 years, the situation was reversed, and in 1992 Germany was twice as rich as Venezuela. Economic growth is thus neither universal nor irreversible. In addition, simple algebra shows that over a long period of time even an apparently small difference in the rate of economic growth can translate into a large difference in the income of the average household. Nobody has a complete understanding of why economies grow, and nobody has an infallible policy formula f
