i G ancar1oG .

iri
a
. . .. .. .

1t? :^,
+i
*
i -k . .. tF .1 ^.l ii^ ry i"r l^
9
r

`
Vt

L1_;;:tI;a 6

S r ,^,y bh^
.. l'r. i' %r9 >

hfDN
! ^ ^..
- 1 ^.4H '

, Y^`y. ^ '. 'ter' t^ f1,li :^ ' S! A^.T ,)4

}y}
^T ^ o.

r ,r. 3 v il
'
.'T.

i:t(a
w` !y^ ^^ l
y
i,^f .. y^ j r ,r},/Y'`tL
^ .(^ j Y

1! txl .r l %3t, ^ n. rk
, i , {^'
a
'-.{r }
^N +1^4 i?$ l 4If C'^ . .- ct

rJ t. ^i. ^ '. M1 >i,^: aaRry '. S

C _ ^ e. ^ {^
{r

g^}^,f`3tt`' ii^.
7 X^^^n.r X ^..:
4
j f ^ 1 ^ ^ - .+, ^
^ '! .i^M^,^; ^r /} it ^. .GI ^^R
t,j, ^J/S ^GGG an r^. Y ^-o ,F,Lg
>f.-^3., ^ t._`ti t 4"t ^t T L r r f',yn^('
a`
4 ^ i ^^K^.3+ ^1} S tii^ t^ :ti` i,.j ^^k ^ ^^`ra.^44^
;^,1 7. f.^iyi : +r r v
r .} ..1_ fpi ril^ C4li^,'.;t^.- 1H
.__ _ ___. _ ^,`^ .tia; j^ ^ ^"{}

^ su

UNIVERSIDADE FEDERAL DO PARAN , _ , ^y `^ yj

_____ U F P Q SISTEMA DE BIBLIOTECAS ^.

Hililifihi IIiU i VIII!J 11111 IIHflIfl ^" ef ^., s1 r#^n

00444616 ^ , ^^ ^^
! ^^r.',^,','o^ i^`+^i1^^^a
"'A.!t'?^;ri^
i :r^r^.^. t^+cx,t*X^+r*,z^;; Yt,!fir^ ^;..:^..

0f

1 ^ ,
 I Professor Dr. Giancarlo Gandolfo
University of Rome "La Sapienza"
Faculty of Economics
Via del Castro Laurenziano 9
I-00161 Roma, Italy

c 0o ' OS
{ Biblioteca de Cincias e Tecnologia-Fisica
BUSCA Inc.
Economic Dynamics: Study Edition
10/07!2008 Termo. 567._ Registro 476875
DISPENSA DE LICITA^O
U$ 94,95 R$ 162 ,00

ist ed. 1997. Corr. 2nd printing

ISBN 3-540-62760-X Springer-Verlag Berlin Heidelberg New York

The LIbrary of Congress has cataloged the 3rd cornpletely re y . and en!. ed, as follows:
Gandolfo, Giancarlo. Economic dynamics / Giancarlo Gandolfo. - 3rd, Completely rey.
and enl. ed. p. cm . Re y. ed. of: Economic dynamics, methods and models. 2nd, re y. ed.
1980, c1971. Includes bibliographical referentes and index. ISBN 3-540-60988-1 (hard-
cover: alk. paper) 1. Economics, Mathematical. 2. Economics - Mathematical models. 3.
1 Statics and dynamics (Social sciences) I. Gandolfo, Giancarlo. Economic dynamics,
methods and models. II. Title. HB135.G36 1996 330'.01'51-dc20
This work is subject to copyright. All rights are reserved, whether the whole or part of
the material is concerned, specifically the rights of translation, reprinting, reuse of il-
lustrations, recitation, broadcasting, reproduction on microfilm or in any other way,
and storage in data banks. Duplication of this publication or parts thereof is permitted
only under the provisions of the German Copyright Law of September 9, 1965, in its
current version, and permission for use must always be obtained from Springer-Verlag.
Violations are hable for prosecution under the German Copyright Law.
e Springer-Verlag Berlin Heidelberg 1997
Printed in Germany
The use of general descriptive names, registered names, trademarks, etc. in this publi-
cation does not imply, even in the absence of a speci fi c statement, that such names are
exempt from the relevant protective laws and regulations and therefore free for general
use.
Cover design: Erich Kirchner, Heidelberg
SPIN 11504658 42/3153 - 54 3 2 1 - Printed on acid-free paper
To my wife
Lucia7na
(finally!)
Preface

This is a completely rewritten and rnucli expanded version of the book pub-
lished in 1980 (tliis, in turn, was an updated edition of rny Mathematical Methods
and Models in Economic Dynamics, originally published in 1971) and reprinted
several times since then. In the last fifteen years so many new mathematical meth-
ods have been introduced in the study of dynamic economic models, that a revision
was not sufficient, but a completely new edition was called for.
This book gives a comprehensive, but simple, treatment of the mathematical
methods commonly used in dynamical economics. It also shows how they are
applied to build and analyse economic models. Accordingly, the focus is on meth-
ods, and every mathematical technique introduced is followed by its application
to selected economic models that serve as examples. The unifying principie in the
exposition of the different economic models is then seen to be the common mathe-
matical technique. This process will enable the reader not only to understand the
basic literature, but also to build and analyse her or his own models.
` Comprehensive' means that this book contains an unusually broad set of math-
ematical methods. The standard constant-coefficient linear difference and differen-
tial equations and simultaneous systems are of course thoroughly explained from
scratch. Mixed differential-difference equations are introduced by a discussion of
discrete and continuous time in economic modelling. An extensive treatment of
non-linear dynamic equations and qualitative methods (from phase diagrams to
non-linear oscillations and bifurcation theory) is given. The saddle-path proper-
ties of solutions to dynamic optimization problems and to rational expectations
models with jump variables are explained rigorously. Stability considerations are
stressed throughout, including advanced topics such as Liapunov's second method,
structural stability, conditional stability, and so on. The theory of chaotic dynam-
ics is examined at some length and appraised in relation to stochastic erraticity.
Synergetics and catastrophe theory are also (briefly) treated.
`Simple' means that the lowest possible level of mathematical prerequisites is
presupposed in the reader. He or she is assumed to have no previous familiarity
with the topics treated. Accordingly, every subject is worked out in great detail,
and no essential step in the argument is omitted. The reader is guided through a
step-by-step analysis of each topic, be it a mathematical method or an economic
model. This 'user-friendly' feature, which is also present in the exercises, will
undoubtedly be appreciated by students.
The required background for Part I consists of elementary algebra, for Part
II of the rudirnents of calculus. Advanced matrix algebra is used when necessary,
VIII

but the main propositions can be understood without it. More mathematical
background is needed for Part III; however, this does not go inucli beyond the
knowledge acquired in any basic course of mathematics for economists. It also
should be pointed out that the organisation of Part III is different from that of
Parts I and II, where mathematical methods and illustrative economic models
are treated in separate chapters. This separation is didactically convenient when
the mathematical chapter deals with a specific and limited topic, e.g. difference
equations of a given order, linear a.nd with constant coefficients. Since each chapter
of Part III usually has a mucli broader inathematical subject matter-----bifurcation
theory, for examplethe economic illustrations are more conveniently presented
with the mathematical methods.

This book has been written for economists, but applied mathematicians inter-
ested in getting a bird's eye view of the economic applications of the mathematical
methods under consideration may fiad the economic parts of the book useful. In
fact, although the main selection criterion has been the models' suitability to il-
lustrate the mathematical point, both old classics and new research results have
been included, as vvell as both iiiicroecolioniic asid nlacroeconomic lnodeis. Also,
students in time series analysis who need a grounding in difference equations will
find Part I useful.

I am grateful to the students from all over the world who have written me over
the years to indicate unclear points and misprints, and to Flavio Casprini, Nicola
Cetorelli, Giuseppe De Arcangelis, Vivek H. Dehejia, Daniela Federici, Maria Mad-
dalena Giannetti, Michael D. Intriligator, Giovanna Paladino, Maria Luisa Petit,
Francesca Sanna Randaccio, for their advice and comments. Claude Hillinger and
Karlhans Sauernheimer discussed with me the project of this book when I was
visiting CES (Center for Economic Studies of the University of Munich) in 1992
and 1994, and gave me many useful suggestions. Finally, I would like to thank the
University of Rome `La Sapienza' for generous support in terms of research funds
and a year of sabbatical leave. None of the persons and institutions mentioned has
any responsibility for possible deficiencies that might remain.

Giancarlo Gandolfo, University of Rome `La Sapienza', December 1995

PREFACE TO THE SOFTCOVER EDITION

This student edition differs from the previous hardcover edition in one impor-
tant respect: the addition of a new section containing the answers to all exercises.
The help of Daniela Federici and Maria Maddalena Giannetti is gratefully acknowl-
edged. Also, rnisprints and errors have been corrected thanks to Serena Sordi.
In the 2005 reprint further misprints Hand errors have been corrected thanks to
Lisbeth Fajstrup and Federico Trionfetti.

University of Romc `La Sayicnza', Dcccmbcr 1996 and April 2005

Contents

PREFACE. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VII

1 Introduction 1
1.1 Definition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1 .2 Functional equations . . . . . . . . . . . . . . . . . . . . . . . 2
1 .3 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

I LINEAR DIFFERENCE EQUATIONS 5

2 Difference Equations: General Principies 7

2 .1 Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2.2 Linear difference equations with constant coefficients . . . . . 9
2.2.1 The homogeneous equation . . . . . . . . . . . . . . . 10
2.2.2 The non-homogeneous equation . . . . . . . . . . . . . 12
2.3 Determination of the arbitrary constants . . . . . . . . . . . . 13
2 .4 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14

3 First-order Difference Equations 17

3.1, Solution of the homogeneous equation . . . . . . . . . . . 17
3.2 Particular solution of the non-homogeneous equation . . . . . 21
3.2.1 g(t) is a constant . . . . . . . . . . . . . . . . . . . . . 21
3.2.2 g(t) is an exponential function . . . . . . . . . . . . . . 22
3.2.3 g(t) is a polynomial function of degree m . . . . . . . . 23
3.2.4 g(t) is a trigonometric function of the sine-cosine type . 23
3.2.5 q(t) is a combination of the previous functions . . . . : 24
3.2.6 The case when g(t) is a generic function of time. Back-
ward and forward solutions . . . . . . . . . . . . . . . 24
3.3 General solution of the non-homogeneous equation . . . . . . . 27
3.4 A digression on distributed lags and partial adjustment equa-
tions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
3 .5 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
3 .5.1 Example . . . . . . . . . . . . . . . . . . . . . . . . 31
3 .5.2 Other exercises . . . . . . . . . . . . . . . . . . . . 32
3.6 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
 Contents Contents

4 First-order Difference Equations in Economic Models 35 7.5 E)

4.1 The cobweb theorem . . . . . . . . . . . . . . . . . . . . . . . 35 71
4.1.1 The cobweb model and expectations . . . . . . . . . . 38 71
4.1.1.1 The normal price . . . . . . . . . . . . . . . . 39 R(
R
40 76
4.1.1.2 Adaptive expectations . . . . . . . . . . . . .
4.2 The dynamics of multipliers . . . . . . . . . . . . . . . . . . . 43 8 Higher
4 .2.1 The b as ic case . . . . . . . . . . . . . . . . . . . . . . . 43 8.1 In-
4.2.2 Other multipliers . . . . . . . . . . . . . . . . . . . . . 45 8.2 Di
4.2.2.1 A foreign trade multiplier . . . . . . . . . . . 46 th(
4 .2.2.2 Taxation . . . . . . . . . . . . . . . . . . . . 47 8.3 Ex
4 .3 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48 8.4 Re
4 .4 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
9 S imult:
5 Second-order Difference Equations 53 9.1 Fi
5.1 Solution of the homogeneous equation . . . . . . . . . . . . . . 53 9.1
5.1.1 Positive discriminant (0 > 0) . . . . . . . . . . . . . . 54
5.1.2 Null discriminant (A = 0) . . . . . . . . . . . . . . . . 55 9.1
5.1.3 Negative Discriminant (A <0) . . . . . . . . . . . . . 56
5.1.4 Stability conditions . . . . . . . . . . . . . . . . . . . . 58
5.2 Solution of the non-homogeneous equation. . . . . . . . . . . . 59
5.2.1 The operational method . . . . . . . . . . . . . . . . . 61
5.3 Determination of the arbitrary constants . . . . . . . . . . . . 63 9.1
5.4 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
5.4.1 Example . . . . . . . . . . . . . . . . . . . . . . . . . 65 9.2 Fir
5 .4.2 Other exercises . . . . . . . . . . . . . . . . . . . . . . 68 9.2
5 .5 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69 9.2
6 Second-order Difference Equations in Econoinic Models 71
6.1 Multiplier-accelerator interaction: the prototype model . . . . 71
6.1.1 Graphical location of the roots . . . . . . . . . . . . . . 73 ^'2'
6.2 Market adjustments and rational expectations . . . . . . . . . 75
6.3 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76 2i
6.4 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.3 Ger
80
9.3.

7 Highe -order Difference Equations 83 9.3.

7.1 Solution of the homogeneous equation . . . . . . . . . . . . . . 83
7.2 Particular solution of the non-homogeneous equation . . . . . 84
7.2.1 The operational method . . . . . . . . . . . . . . . . . 85
7.3 Determination of the arbitrary constants . . . . . . .. . . 87
7.4 Stability conditions . . . . . . . . . . . . . . . . . . . . . . . . 87
7.4.1 Necessary and sufficient stability conditions 9.4 Exe
(Samuelson's form) . . . . . . . . . . . . . . . . . . . . 88 9.4.
7.4.2 Necessary and sufficient stability conditions 9.4.
(Cohn-Schur form) . . . . . . . . . . . . . . . . . . . . 89 9.5 Ref
Contents XI

7 .5 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
7 .5.1 Example . . . . . . . . . . . . . . . . . . . . . . . . . . 91
7.5.2 Other exercises . . . . . . . . . . . . . . . . . . . . . . 92
7.6 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92

8 Higher-order Difference Equations in Economic Models 95

8.1 Inventory cycles . . . . . . . . . . . . . . . . . . . . . . . . . . 95
8.2 Distributed lags and interaction between the multiplier and
theaccelerator . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
8.3 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
8.4 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102

9 Simultaneous Systems of Difference Equations 103

9.1 First-order 2x2 systems in normal form . . . . . . . . . . . . 103
9.1.1 General solution of the homogeneous system: first
method. . . . . . . . . . . . . . . . . . . . . . . . . . 103
9.1.2 General soliltion of the homogeneous system:
second (or direct) method . . . . . . . . . . . . . . . . 106
9.1.2.1 Unequal real roots . . . . . . . . . . . . . . . 106
9.1.2.2 Equal real roots . . . . . . . . . . . . . . . . . 108
9.1.2.3 Complex roots . . . . . . . . . . . . . . . . . 109
9.1.3 Particular solution. Determination of the arbitrary
constants. . . . . . . . . . . . . . . . . . . . . . . . . 110
9.2 First order n x n systems in normal forro . . . . . . . . . . . . 111
9.2.1 Direct matrix solution . . . . . . . . . . . . . . . . . . 114
9.2.2 Stability conditions . . . . . . . . . . . . . . . . . . . . 115
9.2.2.1 A digression on not-wholly-unstable systems . 118
9.2.2.2 Proof of the stability conditions . . . . . . . . 120
9.2.3 Particular solution . . . . . . . . . . . . . . . . . . . . 121
9.2.3.1 The operational method . . . . . . . . . . . . 122
9.2.4 Deterrnination of the arbitrary constants . . . . . . . . 125
9 .3 General systems . . . . . . . . . . . . . . . . . . . . . . . . . . 126
9.3.1 First-order systems not in normal form . . . . . . . . . 126
9.3.2 Higher-order systems . . . . . . . . . . . . . . . . . . . 127
9.3.2.1 An example . . . . . . . . . . . . . . . . . . . 127
9.3.2.2 The general case . . . . . . . . . . . . . . . . 129
9.3.2.3 Transformation of a liigher-order system into
a first-order system in normal form . . . . . . 130
9.3.2.4 Stability conditions for higher-order systems . 132
9.4 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132
9.4.1 Example . . . . . . . . . . . . . . . . . . . . . . . . . . 132
9.4.2 Other exercises . . . . . . . . . . . . . . . . . . . . . . 133
9.5 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134
XII Contents

10 Simultaneous Difference Systems in Economic Models 137

10 .1 Cournot oligopoly . . . . . . . . . . . . . . . . . . . . . . . . . 137
10.2 Multiplier effects in an open economy . . . . . . . . . . . . . . 140
10.3 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143
10 .4 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143

II LINEAR DIFFERENTIAL EQUATIONS 145

11 Differential Equations: General Principies 147
11.1 Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147
11.2 Linear differential equations with constaiit coeflicients . . . . . 148
11.2.1 The homogeneous equation . . . . . . . . . . . . . . . 149
11.2.2 The non-homogeneous equation . . . . . . . . . . . . . 150
11.3 Determination of the arbitrary constants . . . . . . . . . . . . 152
11.4 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154

12 First-order Differential Equations 155

12.1 Solution of the homogeneous equation . . . . . . . . . . . . . . 155
12.2 Particular solution of the non-homogeneous equation . . . . . 158
12.2.1 g(t) is a constant . . . . . . . . . . . . . . . . . . . 158
12.2.2 g(t) is an exponential function . . . . . . . . . . . . . . 159
12.2.3 g(t) is a polynomial function of degree rn . . . . . . . . 1 59
12.2.4 g(t) is a trigonometric function of the sine-cosine type . 160
12.2.5 g(t) is a combiiiation of the previous functions . . . . . 160
12.2.6 g(t) is a generic function of time. The method of vari-
ationof parameters . . . . . . . . . . . . . . . . . . . . 161
12.3 General solution of the non-homogeneous equation . . . . . . . 162
12.4 Continuously distributed lags and partial adjustment equations 163
12.5 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165
12.5.1 Example . . . . . . . . . . . . . . . . . . . . . . . . . . 165
12:5.2 Other exercises . . . . . . . . . . . . . . . . . . . . . . 167
12.6 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167

13 First-order Differential Equations in Economic Models 169

13.1 Stability of supply and demand equilibrium . . . . . . . . . . 169
13.2 The neoclassical growth model . . . . . . . . . . . . . . . . . . 175
13.2.1 Existente of a growth equilibrium . . . . . . . . . . . . 176
13.2.2 Stability of growth equilibrium . . . . . . . . . . . . . 178
13.2.3 Refinements . . . . . . . . . . . . . . . . . . . . . . . . 181
13.2.3.1 Depreciation and technical progress . . . . . . 181
13.2.3.2 Golden rule . .. . . . . . . . . . . . . . . . . . 183
1
13.2.4 Further developments . . . . . . . . . . . . . . . . . . . 184
13.2.4.1 Adjustment time or, how long is the long rlm? 184
Contents XIII

13.2.4.2 (3-convergente, e--convergence, and all that . . 187

13.2.4.3 Endogenous growth . . . . . . . . . . . . . . . 189
13 .3 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189
13.4 References . . . . . ^ ^
191

14 Second-order Differential Equations 193

14.1 Solution of the homogeneous equation . . . . . . . . . . . . . . 193
14.1.1 Positive discriminant (A > 0) . . . . . . . . . . . . . . 194
14.1.2 Null discriminant (A -= 0) . . . . . . . . . . . . . . . . 195
14.1.3 Negative discrirninant (0 <0) . . . . . . . . . . . . . . 196
14.1.4 Stability conditions . . . . . . . . . . . . . . . . . . . . 198
14.2 Particular solution of the non-homogeneous equation . . . . . 199
14.2.1 Variation of parameters . . . . . . . . . . . . . . . . . 200
14.3 General solution of the non-homogeneous equation . . . . . . . 202
14.4 Determination of the arbitrary constants . . . . . . . . . . . . 203
14.5 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203
14.5.1 Examples . . . . . . . . . . . . . . . . . . . . . . . . . 203
14.5.2 Other exercises . . . . . . . . . . . . . . . . . . . . . . 205
14.6 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 5

15 Second-order Differential Equations in Economic Models 207

15.1 The second-order accelerator . . . . . . . . . . . . . . . . . . . 207
15.2 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 210
15.3 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 212

16 Higher-order Differential Equations 213

16.1 Solution of the homogencous equation . . . . . . . . . . . . . . 213
16.2 Solution of the non-llomogeneous equation . . . . . . . . . . . 215
16.2.1 Variation of parameters . . . . . . . . . . . . . . . . . 215
16.3 Determination of the arbitrary constants . . . . . . . . . . . . 218
16.4 Stability conditions . . . . . . . . . . . . . . . . . . . . . . . . 219
16.4.1 Necessary and sufficient stability conditions
( Routli-Flurwitz) . . . . . . . . . . . . . . . . . . . . . 221
16.4.2 Necessary and sufficient stability conditions
(Linard-Chipart) . . . . . . . . . . . . . . . . . . . . . 223
16 .5 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223
16.5.1 Example . . . . . . . . . . . . . . . . . . . . . . . . . . 223
16 .5.2 Other exercises . . . . . . . . . . . . . . . . . . . . . . 225
16.6 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 225

17 Higher-order Differential Equations in Economic Models 227

17.1 Feedback control and stabilisation policies . . . . . . . . . . . 227
17.1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . 227
17.1.2 Three types of stabilisation policy . . . . . . . . . . . . 228
XIV Cont.ents Contents

17.1.2.1 Proportional stabilisation policy . . . . . . . 231 19 Differer

17.1.2.2 Mixed proportional-derivative stabilisation po1- 19.1 Sta
icy. . . . . . . . . . . . . . . . . . . . . . . . 232 19.1
17.1.2.3 Integral stabilisation policy . . . . . . . . . . 233 19.1
17.2 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 234 19.2 Hin
17.3 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 235 19.3 A d
19.4 BaL
18 Simultancous Systems of Differcntial Equations 237 19.5 Exe
18.1 First-order 2x2 systems in normal form . . . . . . . . . . . . 237 19.6 RefE
18.1.1 General solution of the hornogeneous system: first
method. . . . . . . . . . . . . . . . . . . . . . . . . . 238
18.1.2 General solution of the homogeneous system: second
III AD)
(or direct) method . . . . . . . . . . . . . . . . . . . . 240 20 Compar
18.1.2.1 Unequal real roots . . . . . . . . . . . . . . . 241 20.1 Intrc
18.1.2.2 Equal real roots . . . . . . . . . . . . . . . . . 243 20.2 The
18.1.2.3 Complex roots . . . . . . . . . . . . . . . . . 244 20.2
18.1.3 Particular solution. Determination of tlie arbitrary 20.2
constants. . . . . . . . . . . . . . . . . . . . . . . . . 245 20.3 Con
18.2 First order n x n systems in normal forro . . . . . . . . . . . . 245 20.4 Con
18.2.1 Solution of the homogeneous system . . . . . ... . . . 247 `cori
18.2.1.1 The matrix exponencial . . . . . . . . . . . . 249 20.4
18.2.2 Stability conditions . 251 20.5 Exti
18.2.2.1 D-stability, and stabilisation of matrices . . . 254 20.5
18.2.2.2 Sensitivity analysis . . . . . . . . . . . . . . . 20.6G Ele r.
255
20.7
18.2.2.3 A digression on not-wholly-unstable systems . 259
IS-L
18.2.2.4 Proof of the stability conditions . . . . . . . . 262
20.8 Exe]
18.2.3 Particular solution . . . . . . . . . . . . . . . . . . . . 264
20.9 Refe
18.2.3.1 Variation of parameters . . . . . . . . . . . . 264
18.2.4 Determination of the arbitrary constants . . . . . . . . 265 21 Stability
18.3 General systems . . . . . . . . . . . . . . . . . . . . . . . . . . 266 21.1 Intrn
18.3.1 First-order systems not in normal form . . . . . . . . . 26 i 21.2 Basi
18.3.2 Higher-order systems . . . . . . . . . . . . . . . . . . . 261 21.2
18.3.2.1 An example . . . . . . . . . . . . . . . . . . . 26 21.2
18.3.2.2 The general case . . . . . . . . . . . . . . . 27( 21.2
21.3 Qua
18.3.2.3 Transformation of a higher-order system into
a first-order system in normal form . . . . . . 27: 21.3
21.3
18.3.2.4 Stability conditions for higher-order systems . 27;
18.4 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . 27:
18.4.1 Example . . . . . . . . . . . . . . . . . . . . . . . . . . 27
18.4.2 Other exercises . . . . . . . . . . . . . . . . . . . . . . 27
18.5 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 21.4 Quz
ts Contents Xv

-1 19 Differential Equation Systems in Economic Models 279

19.1 Stability of Walrasian general equilibrium of exchange . . . . . 279
2 19.1.1 Static stability . . . . . . . . . . . . . . . . . . . . . . 280
3 19 .1.2 Dynamic stability . . . . . . . . . . . . . . . . . . . . . 283
4 19.2 Human capital in a grwtlh model . . . . . . . . . . . . . . . . 286
5 19.3 A digression on 'arrow diagrams' . . . . . . . . . . . . . . . . 291
19.4 Balanced growth in a multi-sector economy . . . . . . . . . . . 293
7 19 .5 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 298
.7 19 .6 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 301

8
III ADVANCED TOPICS 303
o 20 Comparative Statics and the Correspondence Principie 305
1 20.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . 305
3 20.2 The method of comparative statics . . . . . . . . . . . . . . . 306
4 20.2.1 Purely qualitatively comparative statics . . . . . . . . 310
20.2.2 The inverse comparative statics problem . . . . . . . . 310
20.3 Comparative statics and optimizing behaviour . . . . . . . . . 311
20.4 Comparative statics and dynamic stability of equilibrium: the
T `correspondence principie' . . . . . . . . . . . . . . . . . . . . 314
20.4.1 Criticism and qualifications . . . . . . . . . . . . . . . 316
1 20.5 Extrema and dynamic stability . . . . . . . . . . . . . . . . . 318
20.5.1 An application to the theory of the firm . . . . . . . . 323
1
20.6 Elements of comparative dynamics . . . . . . . . . . . . . . . 324
20.7 An illustrative application of the correspondence principie: the
IS-LM model . . . . . . . . . . . . . . . . . . . . . . . . . . . 325
20.8 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 328
20 .9 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 329

21 Stability of Equilibrium: A General Treatment 331

21.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . 331
21.2 Basic concepts and definitions . . . . . . . . . . . . . . . . . . 332
21.2.1 Stability . . . . . . . . . . . . . . . . . . . . . . . . . . 333
21.2.2 Ftirther definitions . . . . . . . . . . . . . . . . . . . . 337
21.2.3 Structural stability . . . . . . . . . . . . . . . . . . . . 338
21.3 Qualitative methods: phase diagrams . . . . . . . . . . . . . . 341
21.3.1 Single equations . . . . . . . . . . . . . . . . . . . . . . 342
21.3.2 Tuvo-equation simultaneous systems . . . . . . . . . . . 346
21.3.2.1 Introduction: phase plane and phase path . . 346
21.3.2.2 Singular points . . . . . . . . . . . . . . . . . 347
21.3.2.3Graphical construction of the trajectories . . 350
21.3.2.4Linear systems . . . . . . . . . . . . . . . . . 356
21.4 Qtuantitative methods . . . . . . . . . . . . . . . . . . . . . . . 360
XVI Cont.ents Contents

21.4.1 Linearisation . . . . . . . . . . . . . . . . . . . . . . . 360 24Introdue

21.5 Elements of the qualitative theory of difference equations . . . 363 24.1 Prel
21.5.1 Single difference equations . . . . . . . . . . . . . . . . 363 24.2 Som
21.5.2 Two simultaneous difference equations . . . . . . . . . 368 24.2.
21.6 Econornic applications . . . . . . . . . . . . . . . . . . . . . . 368 24.2.
21.7 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 369
21.8 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 370 24.2.
24.3 Limi
22 Saddle Points and Economic Dynarnics 373 24.3.
22.1 Saddle points in optimal control problems . . . . . . . . . . . 374 24.3.
22.1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . 374 24.3.
22.1.2 The maximum principie . . . . . . . . . . . . . . . . . 375
22.2 Optimal ecoriomic growth . . . . . . . . . . . . . . . . . . . . 377
22.2.1 Optimal growth: traditional . . . . . . . . . . . . . . . 378
24.4 The
22.2.1.1 The setting of the problem . . . . . . . . . . . 378
24.4.
22.2.1.2 The optimality conditions in the basic neo-
24.4.
classical inodel . . . . . . . . . . . . . . . . . 381
22.2.1.3 Saddle-point transitional dynarnics in the ba- 24.4.
sic neoclassical model . . . . . . . . . . . . . 384
22.2.2 Optimal growth: endogenous . . . . . . . . . . . . . . 386
22.2.2.1 A model of optimal endogenous growth . . . . 386 24.5 Exer
22.2.2.2 The conditions for optimal endogenous growth 388 24.6 Refe
22.2.2.3 Optimal endogenous growth: saddle-point tran-
sitional dynamics . . . . . . . . . . . . . . . . 390 25 Bifurcati
22.3 Rational expectations and saddle points . . . . . . . . . . . . 394 25.1 Intrc
22.3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . 394 25.2 Bifui
25.2.
22.3.2 Rational expectations, saddle points, and overshooting 397
25.2.
22.3.3 Rational expectations and saddle points: the general
25.2.
case. . . . . . . . . . . . . . . . . . . . . . . . . . . . 401
25.2.
22.4 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 403
25.2.
22.5 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 405

23 Liapunov's Second Method 407

23.1 General concepts . . . . . . . . . . . . . . . . . . .. . . . . . . 407
25.2.
23.2 The fundamental theorems . . . . . . . . . . . . . . . . . . . . 408
25.3 Bifi.0
23.3 Some economic applications . . . . . . . . . . . . . . . . . . . 413
25.3.
23.3.1 Global stability of Walrasian general equilibrium . . . . 413
25.3.
23.3.2 Rules of thumb in business management . . . . . . . . 420 25.3.
23.3.3 Price adjustment and gligopoly under product differ- 25.3.
entiation. . . . . . . . . . . . . . . . . . . . . . . . . . 421
23.4 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 425
23.5 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 426 25.4 Exer
tts Contents XVII

i 0 24 Introduction to Nonlinear Dynamics 429

;
3 24.1 Preliniinary rernarks . . . . . . . . . . . . . . . . . . . . . . . 429
1
3 24.2 Some integrable differential equations . . . . . . . . . . . . . . 431
8 24.2.1 First-order and first-degree exact equations . . . . . . . 431
8 24.2.2 Linear egliations of the first order with variable coeffi-
9 cients . . . . . . . . . . . . . . . . . . . . . . . . . . . 434
24.2.3 The Bernoulli equation . . . . . . . . . . . . . . . . . 436
24.3 Limit cycles and relaxation oscillations . . . . . . . . . . . . . 437
24.3.1 Limit cycles: the general theory . . . . . . . . . . . . . 437
24.3.2 Limit cycles: relaxation oscillations . . . . . . . . . . . 439
24.3.3 Kaldor's non-linear cyclical model . . . . . . . . . . . . 441
24.3.3.1 The model . . . . . . . . . . . . . . . . . . . 441
24.3.3.2 Kaldor via relaxation oscillations . . . . . . . 445
24.3.3.3 Kaldor via Poincar's limit cycle . . . . . . . 447
24.4 The Lotka-Volterra equations . . . . . . . . . . . . . . . . . . 449
24.4.1 Construction of the integral curves . . . . . . . . . . . 454
24.4.2 Conservative and dissipative systems, and
irreversibility. . . . . . . . . . . . . . . . . . . . . . . 456
24.4.3 Goodwin's growth cycle . . . . . . . . . . . . . . . . . 458
24.4.3.1 The model . . . . . . . . . . . . . . . . . . . 458
24.4.3.2 The phase diagram of the model . . . . . . . 461
24.5 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 464
24.6 Referentes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 466

25 Bifurcation Theory 469

25.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . 469
25.2 Bifurcations in continuous time systems . . . . . . . . . . . . . 469
25.2.1 Codimension-one bifurcations . . . . . . . . . . . . . . 471
25.2.2 The Hopf bifurcation . . . . . . . . . . . . . . . . . . . 475
25.2.3 Sensitivity analysis and bifurcations: a reminder . . . . 479
25.2.4 Kaldor's non-linear cyclical model again . . . . . . . . 480
25.2.5 Oscillations in optimal growth models . . . . . . . . . . 481
25.2.5.1 The model . . . . . . . . . . . . . . . . . . . 481
25.2.5.2 The optimality conditions . . . . . . . . . . . 483
25.2.5.3 Emergente of a I lopf bifurcation . . . . . . . 484
25.2.6 Cycles in an IS-LM model with pure money financing . 486
25.3 Bifurcations in discrete time systems . . . . . . . . . . . . . . 488
25.3.1 Codimension-one bifurcations . . . . . . . . . . . . . . 489
25.3.2 The Hopf bifurcation in discrete time . . . . . . . . . . 491
25.3.3 Kaldor's cyclical model in discrete time . . . . . . . . . 492
25.3.4 Liquidity costs in the firm . . . . . . . . . . . . . . . . 494
25.3.4.1 The model . . . . . . . . . . . . . . . . . . . 494
25.3.4.2 The dynamics . . . . . . . . . . . . . . . . . . 497
25.4 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 499
 XVIII Contents Contents
fi
25 .5 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 501 27.8 Exer
27.9 Refe
26 Complex Dynamics 503
26.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . 503 Bibliogri
26.2 Discrete time systems and chaos . . . . . . . . . . . . . . . . . 505
26.2.1 The logistic map . . . . . . . . . . . . . . . . . . . . . 505 Name Ira
26.2.2 Intermittency . . . . . . . . . . . . . . . . . . . . . . . 512
26.2.3 The basic theorems . . . . . . . . . . . . . . . . . . . . 512 Subject
26.2.4 Discrete time chaos in economics . . . . . . . . . . . . 515
Answers
26.2.4.1 Chaos in growth theory . . . . . . . . . . . . 515
26.2.4.2 Exchange rate dynamics and chaos . . . . . . 516
26.3 Continuous time systems and chaos . . . . . . . . . . . . . . . 518
26.3.1 The Lorenz equations, strange attractors, and chaos . . 518
26.3.2 Other routes to continuous time chaos . . . . . . . . . 521
26.3.2.1 The Rdssler attractor . . . . . . . . . . . . . 521
26.3.2.2 The Shil'nikov scenario . . . . . . . . . . . . . 521
26.3.2.3 The forced oscillator . . . . . . . . . . . . . . 522
26.3.2.4 The coupled oscillator . . . . . . . . . . . . . 522
26.3.3 International trade as the source of chaos. . . . . . . . 526
26.3.4 A chaotic growth cycle . . . . . . . . . . . . . . . . . . 527
26.4 Significante and detection of chaos: Stochastic dynamics or
chaos? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 528
26.5 Other approaches . . . . . . . . . . . . . . . . . . . . . . . . . 532
26.5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . 532
26.5.2 Fast and slow, and syllerget,ics . . . . . . . . . . . . . . 533
26.5.3 Catastrophe theory . . . . . . . . . . . . . . . . . . . . 536
26.6 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 538
26.7 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 540

27 Mixed Differential-Difference Equations 545

27.1 General concepts . . . . . . . . . . . . . . . . . . . . . . . . . 545
27.2 Continuous vs discrete tiene in economic models . . . . . . . . 545
27.3 Linear mixed equations . . . . . . . . . . . . . . . . . . . . . . 550
27.4 The method of solution . . . . . . . . . . . . . . . . . . . . . . 550
27.5 Stability conditions . . . . . . . . . . . . . . . . . . . . . . . . 555
27.6 Delay differential equations and chaos . . . . . . . . . . . . . . 556
27.7 Some economic applications . . . . . . . . . . . . . . . . . . . 556
27.7.1 Kalecki's business cycle model . . . . . . . . . . . . . . 557
27.7.1.1 The model . . . . . . . . . . . . . . . . . . . 557
27.7.1.2 The dynaniics . . . . . . . . . . . . . . . . . . 559
27.7.2 A formalization of the classical price-specie-flow mech-
anisln of balance of payI ients adjust,nient, . . . . . . . . 563
27.7.2.1 The model . . . . . . . . . . . . . . . . . . . 563
27.7.2.2 Stability . . . . . . . . . . . . . . . . . . . . . 565
itent., Contents XIX

501 27 .8 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 67
27.9 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 568
503
503 Bibliography 571
505
505 Name Index 593
512
512 Subject Index 599
515
Answers to Exercises 611
515
)16
518
18
21
21
21
22
22
26

8
;2
2
3
6
3
 XXII

21.9 Multil,
21.10 Natur(
21.11 Graph
21.12 Arrow
21.13 Limit
L ist of Figures 21.14 Phase
gent n
21.15 Phase
monot
3.1 (a) Monotonic and clivergent; (b) Oscill atory aIld convergcnt; 21.16 Phase
(c) Monotonie ancl clivenceIlt; (d) Oscillatory 11 Id civergcIIt; oscill a
(e) Constant; (f) Oscillatory with constant amplitude. . . . . 19
3.2 Lo a dif
Joining the points of t, he solut ion ference equatioil . . 22.1 Equili
20
22.2 The n
4.1 The cobweb theorem . . . . . . . . . . . . . . . . . . . . . . . 36 22.3 An oF
4.2 The cobweb theorem: abnormal cases . . . . . . . . . . . . . 38 22.4 Ratio:
4.3 The simple dynamic multiplier . . . . . . . . . . . . . . . . . 45 shooti
22.5 Ratio:
5.1 Second-order difference equations: real roots . . . . . . . . . . 60
5.2 An apparently divergent, solution . . . . . . . . . . . . . . . . 64 23.1 Perfec
5.3 A constant-amplitude oscillation . . . . . . . . . . . . . . . . G8 24.1 Kaldc
6.1 Samuelson's multiplier-accelerator diagram . . . . . . . . . . 74 functi
6.2 Roots in Hicks' second-order rnodel . . . . . . . . . . . . . . . 79 24.2 The c
24.3 Kaldc
8.1 Metzler's inventory cycle: stability regions . . . . . . . . . . . 97 24.4 Lotka
24.5
24. Lotk^
12.1 Solutions to a first-order differential equation . . . . . . . . . 157
24.5 LotkiP
13.1 a) excess clemand, b) e xcess demand prive . . . . . . . . . . . 171 24.7 The i
13.2 The neoclassical aggregate growth model . . . . . . . . . . . . 178 24.8 Good
24.9 Path
17.1 Proportional stabilisation policy in the Pllillips model . . . . 232 24.10 Good

19.1 Stability of growth equilibrium with human capital . . . . . . 288 25.1 Sadd
19.2 Construction of an arrow diagram . . . . . . . . . . . . . . . 292 25.2 Tr ae
25.3 Pite
21 .1 Stability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 334
The
25.4
21 .2 Asymptotic stability . . . . . . . . . . . . . . . . . . . . . . . 335
25. 5 The
21.3 First-order non-linear differential equation: phase diagram . . 343
25.6 The
21.4 First-order non-linear differential equation: spiral-like phase
diagram. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 344 26.1 The
21.5 Singular points of 2x2 systems: the nodo . . . . . . . . . . . 348 26.2 Flip
21.6 Singular points of 2x2 systems: the saddle point . . . . . . . 349 26.3 Styli
21.7 Singular points of 2x2 systems: the focus . . . . . . . . . . . 350 26.4 Winq
21.8 Singular points of 2x2 systems: the centre . . . . . . . . . . . 351 26.5 Cha(
 XXII List of Figures

21.9 Multiple singular points of 2x2 systems . . . . . . . . . . . . 352

21.10 Nature of the points not lying on cp ' = 0 . . . . . . . . . . . . 353
21.11 Graphical determination of the nature of a singular point . . 354
21.12 Arrows and centres . .. . . . . . . . . . . . . . . . . . . . . . . 356
21.13 Limit cycles . . . . . . . . . . . . . . . . . . . . . . . . . . . . 357
21.14 Phase diagram for a first-order difference equation: conver-
entmonotonic
g movement . . . . . . . . . . . . . . . . . . . . 364
21.15 Phase diagram for a first-order difference equation: divergent
monotonicmovement . . . . . . . . . . . . . . . . . . . . . . 365
21.16 Phase diagram for a first-order difference equation: various
oscillatorymovernents . . . . . . . . . . . . . . . . . . . . . . 367
19
22.1 Equilibrium point; of the neoclassical optimal growth model . 383
22.2 Tlie neoclassical optirnal growth inodel: transitional dynamics 384
16 22.3 An optimal endogenolis growt,li model: transitional dynamics 392
:8 22.4 Rational expectations and saddle path dynamics: the over-
5 shooting phenomenon . . . . . . . . . . . . . . . . . . . . . . 396
22.5 Rational expectations and exchange-rate overshooting . . . . 400
D
23.1 Perfect complementarity . . . . . . . . . . . . . . . . . . . . 419

24.1 Kaldor's business cycle model: the investment and saving

functions . . 442
24.2 The dynamics of Kaldor's cycle . . . . . . . . . . . . . . . . . 444
24.3 Kaldor's model va Poincar-Bendixson . . . . . . . . . . . . 448
24.4 Lotka-Volterra equations, diagram 1 . . . . . . . . . . . . . . 452
24.5 Lotka-Volterra equations, diagram 2 . . . . . . . . . . . . . . 453
24.6 Lotka-Volterra equations, construction of the orbits . . . . . . 455
24.7 The real wage rate and ernployment . . . . . . . . . . . . . . 460
24.8 Goodwin's growth cycle . . . . . . . . . . . . . . . . . . . . . 462
24.9 Path of u and v in Goodwin's growth cycle . . . . . . . . . . 463
24.10 Goodwin's elementary non-linear cyclical model . . . . . . . . 465

25.1 Saddle-node (or fold) bifurcation, y' = a - y 2 . . . . . . . . . 470

25.2 T anscritical bifurcation, y' = ay - y 2 . . . . . . . . . . . . . 472
25.3 Pitchfork bifurcation, y' = ay - y 3 . . . . . . . . . . . . . . . 474
m
25.4 The Hopf bifurcation . . . . . . . . . . . . . . . . . . . . . . . 478
25.5 The fold bifurcation in discrete time, y t+l = a - y . . . . . . 489
25.6 The supercritical flip bifurcation, y t+l = ay t - ayt . . . . . . 490
26.1 The logistic curve y t+l = ay t (1 - y t ) . . . . . . . . . . . . . . 506
26.2 Flip bifurcation in the logistic equation . . . . . . . . . . . . 507
26.3 Stylized period doubling in the logistic equation . . . . . . . . 508
26.4 Windows in the logistic equation . . . . . . . . . . . . . . . . 509
26.5 Chaos in the logistic difference equation . . . . . . . . . . . . 510
List of Figures XXIII

26.6 An example of intermittency . . . . . . . . . . . . . . . . . . 511

26.7 The Loreriz egtiations (Guckenlicirner asid I-colmes) . . . . . . 519
26.8 A two-dimensional torus in three-dimensional space . . . . . . 524
26.9 A catastrophe in inflation (Woodcock and Davis) . . . . . . . 537
27.1 Roots of Kalecki's model . . . . . . . . . . . . . . . . . . . . 562
List of

3.1 Numeri
ence eq

4.1 Numeri

5.1 Consta,

6.1 Roots e
6.2 Stabilit

21.1 Singul a
a
2i y 1 +
I)

List of Tables

3.1 Numerical computation of the solution to a first-order differ-

ericeequation . . . . . . . . . . . . . . . . . . . . . . . . . . . 32

4.1 Numerical exarnple of t.lie inultiplier . . . . . . . . . . . . . . 50

5.1 Coristant-amplitude trigonometric oscillation . . . . . . . . . 67

6.1 Roots of Hicks' model . . . . . . . . . . . . . . . . . . . . . . 78

6.2 Stability condition in Goodwin's model . . . . . . . . . . . . 80

21.1 Singular points of the linear system y = a

li y l + a
12 y 2, y
z =
a21y 1 + a
22 y
2 . . . . . . . . . . . . . . . . . . . . . . . . . . . 358
 2

1.2 Fui
Let us now ti
f unctional eS
the scope of

Chapter 1 which are sul]

The basic
where the unk
means to find
equation. Nc
Introduction function whic
Two poin
function we i
(for example,
1.1 Definition when explaini
solution of a 1
tion, and the
`A system is dynamical if its behavior over time is deterrnined by functional conditions.
equations in which variables at different points of time are involved in am
The seconc
essential way'.
tion we have 1
This definition, due to Frisch (1936) and Samuelson (1947), is based on value of the ii
the formal characteristics common to all problems studied by economic simple examp
dynamics, also called dynamical economics (these two expressions will be Let lis con
lised interchangeably). Other definitions, based on the economic substance a specific func
of those problems, are possible: e.g., economic dynamics is concerned with the stated eqi
growth, or business cycles, or stability, or economic change. Such definitions (x), the value
have in fact been suggested: seo, for example Baumol (1970), Day (1994), It is easy to c
and the instructive survey by Machlup (1959). But they are inevitably par- calculus, y'(x)
tial, and a complete definition of this type would reduce to a cumbersome which gives y'
list of problems, with the danger of omitting some of them and of not includ- However, for
ing newly arising problerns. The Frisch-Samuelson formal definition, on the from a: there
contrary, is precise and general, and we shall adopt it. equation ident
Before commenting on it, let us recall that, according to another formal omit `identica.
definition (Hicks, 1939), economic dynamics is identified with those parts of ' means to satiE
economic theory where every quantity must be dated, whereas in economic In general,
statics we need not trocable about dating. But in this way we would include obvious remar
in economic dynamics rnany non-dynamic phenomena. For example, think functional equ
of a case in which all quantities have the same date. This may mean that case outside e,
a certain phenomenon has taken place at a certain point of time (and this function knov^
may be important, but is not dynamics), or that a variable at time t depends on economic
on another variable at the same time t (and this too may be important ti me, and use
e.g. consumption is assumed to depend on current income and not on lagged part of the de:
income but, again, it is not dynamics). The definition based on `dating', In fact, y'(;
then, is too vague and cannot be accepted. of y at any pc
 2 C}zapter 1. Introduction

1 12 Functional equations
Let us now tu 11 back to the initial definition and intuitively explain what a
functional equation is. The general theory of functional equations is outside
the scope of tlie present book, and we shall give only some basic notions,
which are sufficient for our purposes.
The basic concept is the following: a functional equation is an equation
where the unknown is a function. Everybody knows that to solve an equation
means to find that value (or those values) of the unknown which satisfy the
equation. Now, to solve a functional equation means to find an unknown
function which satisfies the functional equation identically.
Two points must be stressed in this definition. The first is that by
f; function we mean the f orrri of the finiction, aparta from arbitrary constants
(for exarnple, y = Ae x , where A is an arbitrary constant). As we shall see
when explaining the various functional equations appearing in this book, the
solution of a functional equation determines the forro of the unknown func-
tion, and the determination of the arbitrary constant(s) requires additional
conditions.
The second point is that by `to satisfy identically' we mean that the func-
tion we have to find rnust satisfy the functional equation for any admissible
value of the independent variable appearing in the function. The following
simple example may clarify this point.
Let us consider the functional equation y'(x) y(x) = 0. We must find
a specific function (in one independent variable) which identically satisfies
the stated eqilation, i.e. a function sucll tllat, for any value of its argument
(x), the value of the function and the value of its first derivative are equal.
It is easy to check that this function is y(x) = Ae', since, from elemeritary
calculas, y'(x) = Ac x = y(x) for any x. Now consider the function y = ax+b,
which gives y' = a; if we let x = (a b) /R, we have also y = a, i.e. y' = y.
However, for any other value of x the value of the function will be different
from a: therefore, the function y = ax b does not satisfy our functional
equation identically. As a matter of terminology, from now on we shall usually
omit `identically', it being understood that to `satisfy' a functional equation
means to satisfy it identically.
In general, the symbol x can stand for any variable, not only time. This
obviouis rerrlark is useful to avoid the mistake of believing that, in economics,
functional equations are used only in dynamical problems (an example of a
case outside econoinic dynamics is the classic problem of obtaining a utility
function knowing the marginal rato of substitution). Since this is a book
on economic dynainics ollly, from now on we shall assume that x stands for
ti me, and use t instead of x. We are now ready to understand the second
part of the definition of dynamical econornics.
In fact, y'(t) = y(t) can be considered as a relation that involves the value
of y at any point of time and the value it has at an arbitrarily close point,
1.3. References 3

determined by y'. The `different points of time' clause is necessary to exclude

the case, already rnentioned abo y e, of quantities dated at the same point of
ti me. Time must enter in an 'essential' way: for example, if it enters only as
a unit of measurement (i.e. because we are dealing with quantities that are
flows per unit of time), the system is not dynamical.
The simplest types of functional equations, widely used in dynamical eco-
nomics, are linear, constant-coefficient difference and differential equations
(the meaning of these words will be clarified in the following treatment) and
the relativo sirnultaneous systems. These functional equations will be dealt
with in Part I and Part II respectively. In Part III other types will be treated,
together with rriore advanced material concerning the qualit.ative theory of
LI
differential and difference equations.

1.3 References
Throughout this book, the end-of-chapter references will be indicated only
by name(s), date, title. Complete information as to publisher, place of pub-
lication, etc., is contained in the Bibliography at the end of the volume.

Batimol, W.J., 1970, Economic Dynamics, Chap. 1 and Appendix to Chap.

8.
Day, R., 1994, Complex Economic Dynamics, Vol. I, Chaps. 1, 2.
Frisch, R., 1936, On the Notion of Equilibrium and Disequilibriurn.
Hicks, J.R., 1939, 1946, Value and Capital, Chap. IX, Sect. 1.
Hicks, J.R., 1965, Capital and Growth, Chap. 1.
Machlup, F., 1959, Statics and Dynamics: Kaleidoscopic Words.
Samuelson, P.A., 1947, Foundations of Economic Analysis, Chap. XI, pp.
311-17.
Samuelson, P.A., 1949, Dynamic Process Analysis, Sects. I, II.
Volterra, V., 1959 (1929), Theory of Functionals and of Integral and Integro-
Differential Equations, Chap. 1, pp. 1-7.