Dynamic Macroeconomics

1.4 Conclusion
This introductory chapter provides a brief overview of the evolution and the
current state of macroeconomics as well as some key facts about long-run
economic growth and fluctuations in output, unemployment, and inflation.
It thus sets the stage for an examination and assessment of the main
macroeconomic models of long-run growth and aggregate fluctuations that
are presented in the remainder of this book.

---

1
. The term “macroeconomics,” coined by Frisch [1933], only prevailed beginning in the 1950s. For
example, in the first edition of the most successful introductory economics textbook of the postwar
period, Samuelson [1948], the terms “macroeconomics” and “microeconomics” were not used at all.
Samuelson titled the (unabashedly Keynesian) macroeconomic section of his textbook “Determination of
National Income and Its Fluctuations,” and the (unabashedly neoclassical) microeconomic section “The
Composition and Pricing of National Income.” However, by the third edition, in 1955, Samuelson uses
the terms “macroeconomics” and “microeconomics” repeatedly to distinguish between the two main
subdisciplines of economics.
2
. See Samuelson [1978] for an analysis of the canonical classical model of economic growth that
synthesizes the views of Smith [1776], Malthus [1798], Ricardo [1817], and Mill [1848].
3
. These include Irving Fisher of the Yale School (see I. Fisher [1896, 1911, 1930]), the Cambridge
School (Pigou [1917], Keynes [1923], Marshall [1923]), the Stockholm School (Wicksell [1898]), and the
Austrian School (von Mises [1912]). Monetary theories of aggregate fluctuations included, among
others, those of Pigou [1913], Hawtrey [1919], Marshall [1923], and Keynes [1923]. These economists,
like Hume and Thornton before them, stressed the short-run real effects of money when there is partial
adjustment of nominal wages and prices. I. Fisher [1896], Wicksell [1898], and von Mises [1912],
among others, also stress monetary factors in aggregate fluctuations through short-run deviations of the
market interest rate from the Wicksellian “natural” (equilibrium) rate of interest, which causes
fluctuations in the real interest rate. Humphrey [1991] discusses how many otherwise “classical”
monetary economists had suggested reasons for the short-run nonneutrality of money.
4
. Pre-Keynesian trade cycle or business cycle theories are surveyed and analyzed extensively by
Haberler [1937].
5
. It is again useful to quote Frisch [1933, p. 3], who also wrote that “as far as I know no determinate
macro-dynamic analysis is yet to be found in the literature.” This statement was about the state of the
pre-Keynesian business-cycle models, which were mainly heuristic and nonrigorous. The IS-LM
framework was given its name by Hansen [1949] and was popularized by Hansen [1953]. The IS curve
is the equilibrium relation between investment I and savings S, and the LM curve is the equilibrium
relation between the demand for money L and the supply of money M. See chapter 15 for more details.
6
. For the concept of scientific revolutions, see Kuhn [1962]. Macroeconomics itself was established
through such a revolution, the Keynesian revolution (Klein [1947]). The Keynesian orthodoxy was later
challenged by the monetarist counterrevolution (Johnson [1971]), the rational expectations revolution
(Begg [1982]), to end up with the revolution implied by the adoption of dynamic stochastic general
equilibrium models. For recent scholarly analyses of the key theoretical developments in
macroeconomics since Keynes, see Woodford [1999], Blanchard [2000], and DeVroey [2016].
7
. We analyze the Solow model in chapter 3, the representative household model in chapter 4, and the
overlapping generations models in chapter 5.
8
. It has to be noted that Klein [1950] expends a lot of effort on providing microeconomic foundations
for his econometric model of the United States. For the most part, the 45 pages of chapter II are
devoted to economic theory, with profit-maximizing firms and utility-maximizing households. However,
when shifting from economic theory to his statistical models, Klein uses the theory relatively loosely, as
a rationalization of “simple” aggregate behavioral equations.
9
. We analyze the various traditional Keynesian models and the Samuelson multiplier-accelerator model
in chapter 15.
10
. The term “neoclassical synthesis” was first used in the 1955 edition of Samuelson [1948]. Modigliani
[1944] and Patinkin [1956] develop this synthesis, which was nonetheless derived from the General

---

Theory itself. The neoclassical synthesis is presented in chapter 15.
11
. We analyze deterministic and stochastic versions of these theories of consumption in chapters 4 and
10.
12
. We analyze optimizing models of investment and the cost of capital in chapter 11.
13
. We analyze optimizing general equilibrium models of the demand for money and the determination of
the price level, the nominal interest rate, and inflation in chapters 7 and 12.
14
. We discuss developments involving the Phillips curve and the role of stabilization policy in Keynesian
models in chapter 15.
15
. We discuss the rational expectations hypothesis and compare it to the adaptive expectations
hypothesis in chapter 9.
16
. See King and Rebelo [1999] for an overview of such models. We discuss the stochastic growth
model in chapter 13 and a competitive equilibrium model without capital in chapter 14. Both are models
in the new classical tradition.
17
. Woodford [2003a] and Gali [2008] are based on such models. I introduce and analyze two
alternative new Keynesian models in chapters 16 and 17.
18
. See Arrow and Debreu [1954], Debreu [1959], and Arrow [1964] for the Arrow-Debreu model of
competitive equilibrium.
19
. The new growth theory sprang to life as a result of three important papers: by P. M. Romer [1986,
1990] and Lucas [1988]. It is introduced and discussed in chapter 8. Unified growth theory is surveyed
in Galor [2011], and the role of institutions is surveyed in Acemoglu et al. [2005].
20
. One cannot help noticing the proliferation of the adjective “new” to describe the theories that have
emerged since the 1980s: “new classical” and “new Keynesian” macroeconomics, “new neoclassical
synthesis,” “new growth theory,” and so on. There is little doubt that these theories are based on older
foundations. However, the DSGE models used are new and improved versions of the simpler and more
ad hoc older models. We take up this subject again in chapter 23, the final chapter of this book.
21
. Theterm “econometrics” is also due to Frisch, who was one of the founders of the Econometric
Society and the academic journal Econometrica.
22
. For example, in chapter 3, I devote a lot of space to the Solow [1956] model, even though its
consumption function is postulated and not derived from explicit microeconomic foundations. I also
devote chapter 15 to the traditional Keynesian models and the Phillips curve because of their
significance for the development of macroeconomics and to compare their structure and implications
with the new Keynesian models of chapters 16 and 17.
23
. See Keynes [1938] and, more recently, Rodrik [2015] for a similarly eclectic approach to the role of
models in economics in general and to the rules for setting up models and evaluating them.
24
. It is worth repeating that the rigorous and full empirical evaluation and testing of alternative theories
and models is outside the scope of the present text. This is a text on macroeconomic theory, not
empirical or applied macroeconomics. Texts on empirical macroeconomics and macroeconometrics exist
and could complement the present text for empirically inclined economists. Good examples of such texts
are Favero [2001], Canova [2007], DeJong and Chetan [2011], and Herbst and Schorfheide [2015].
25
. The key facts discussed in this section are the essential facts about long-run growth. Additional
facts, which require some prior theoretical background, will be discussed in chapters 3 and 8. For a
fuller and up-to-date survey of the key empirical facts about economic growth, see Jones [2016].

---

26
. GDP is defined as the market value of final goods and services produced in an economy over a
given period, say, a quarter or a year. Measuring how GDP changes over time or conducting cross-
country comparisons requires separating out differences in prices and differences in quantities of goods
and services. In contrast, GNI is defined as the value of income of domestic residents over a period of
time. This may differ from GDP to the extent that there is income from abroad. For the exact
methodology of measuring GDP and GNI for intertemporal and cross-country comparisons, see the
World Bank website (
www.worldbank.org
), Summers and Heston [1991], or Maddison [1982]. World
Bank data, the data of the Penn World Tables (which are based on the Summers and Heston
methodology), and the data of the Maddison Project (based on the Maddison methodology) are revised
at regular intervals. For the Maddison Project, see Bolt and van Zanden [2014].
27
. Between the eighth and twelfth centuries, China experienced a unprecedented period of economic
and cultural prosperity. It was in this period that gunpowder, printing, and the hydraulic hose reel were
invented. Coal was used in the production of steel, and a series of canals and waterways for water
supply and transportation of products was constructed. The magnetic compass was invented in that
period, and the imperial fleet of China had reached as far as the eastern coast of Africa by the early
fifteenth century. However, China gradually withdrew from the world economy, following a switch to a
policy of introversion, and around 1750, the standard of living in Europe had overtaken the Chinese
standard of living. One hundred years later, in the nineteenth century, China was unable to defend itself
during the Opium Wars against Great Britain, and in the twentieth century, it had become one of the
least-developed economies of the world.
28
. The number of years required for a magnitude like per capita output or income to double in value
depends on its growth rate. If per capita real income grows at g% per year, it takes approximately 70/g
years for it to double. Thus, if g is 1%, it takes 70 years for per capita real income to double; if g is 2%,
it takes 35 years; if g is 3%, it takes 23 years, and so on. This rule of 70 is derived from 100 × ln 2 =
69.3, which is approximately equal to 70, and from the approximation that g% is approximately equal to
ln(1 + g%). For growth rates higher than 5%, most economists prefer to use the rule of 72, because
dividing 72 by the growth rate is a more accurate approximation for growth rates between 5% and 10%.
29
. In what follows, log refers to natural logarithms.
30
. Crafts and O’Rourke [2014] present and try to account for the twentieth century growth experience
of some additional economies in their effort to explain the growth experience of the past century.
31
. For more formal econometric approaches to the stylized facts regarding business cycles in the
United States and the United Kingdom, see Hodrik and Prescott [1997] and Blackburn and Ravn
[1992], among others. For international evidence, see Backus and Kehoe [1992] and Basu and Taylor
[1999], among others.
32
. For an extensive discussion of the available empirical evidence with regard to the key shocks driving
aggregate fluctuations, see Ramey [2016].
33
. Unemployment rates are measured as the number of those who declare themselves as unemployed
(i.e, who are actively seeking a job), divided by the labor force, defined as the sum of the numbers of
the employed and the unemployed.
34
. Darby [1976] and C. Romer [1986] have provided lower revised estimates of the pre–World War II
unemployment rates in the United States, but the general picture does not change.