Академический Документы
Профессиональный Документы
Культура Документы
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}
^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
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
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
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.
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.
PREFACE. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VII
1 Introduction 1
1.1 Definition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1 .2 Functional equations . . . . . . . . . . . . . . . . . . . . . . . 2
1 .3 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
7 .5 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
7 .5.1 Example . . . . . . . . . . . . . . . . . . . . . . . . . . 91
7.5.2 Other exercises . . . . . . . . . . . . . . . . . . . . . . 92
7.6 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
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
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
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
1.2 Fui
Let us now ti
f unctional eS
the scope of
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
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.