Conclusion
Macroeconomics was born following a scientific revolution in the sense of Kuhn [1962]. It initially evolved on the basis of the ideas of Keynes [1936], which gained acceptance as a result of the Great Depression, and which were refined and adapted by his many subsequent followers.
In the 1960s, the Keynesian theory of aggregate fluctuations faced the monetarist counterrevolution, spearheaded by the rise of inflation and the success of the ideas of Friedman [1956, 1968] and Friedman and Schwartz [1963], in explaining both the Great Depression of the 1930s and the rise in inflation of the late 1960s in monetary terms. For some years, macroeconomics settled into a quieter state, in the form of the neoclassical synthesis, which was accepted as “normal science” by both Keynesians and monetarits. The empirical breakdown of the Phillips curve and the stagflation of the 1970s highlighted the remaining inadequacies of the microeconomic foundations of macroeconomics and resulted in the rational expectations revolution and the eventual emergence of DSGE models. These models were initially new classical competitive models (Lucas [1972], Kydland and Prescott [1982]). Soon, new Keynesian DSGE models, based on short-term price and wage stickiness emerged (Mankiw and Romer [1991]). Divisions around these developments seemed to have entered a quiet phase until the financial crisis of 2008–2009. The new neoclassical synthesis was on its way of becoming the new normal. The financial crisis and the Great Recession of 2008–2009 have disrupted this process and have led to a quest for models that also incorporate financial frictions and their interactions with the real economy.Growth theory has been following its own path and displaying impressive progress, with fewer divisions and less acrimony. It has focused on the determinants of the accumulation of physical and human capital and endogenous technical progress.
In recent years, it has been stressing the role of institutions in fostering these processes.In this book, we have reviewed the key dynamic general equilibrium models in both growth theory and aggregate fluctuations, including the benchmark dynamic general equilibrium models of growth and the new neoclassical synthesis DSGE models of aggregate fluctuations.
It remains to be seen how dynamic macroeconomic modeling will evolve. A number of avenues for future research appear to exist, including (1) the fuller analysis of the role of institutions for long-run growth, (2) the fuller incorporation of labor market and financial frictions in DSGE models of aggregate fluctuations, and (3) more satisfactory models of the role of political institutions and politics in the determination of monetary and fiscal policy.
One hopes that progress in these directions, and others suggested in this book, will contribute to the emergence of a theoretically more adequate and empirically more relevant dynamic macroeconomics.
1. The first-known written instance of this quote attributes it to an unidentified Danish parliamentarian of the 1930s. The quote has been repeated in various forms by, among others, the physicist Neils Bohr, the movie mogul Sam Goldwyn, and the baseball catcher Yogi Berra, to each of whom it is sometimes wrongly attributed.
2. For a more detailed discussion of the past evolution of macroeconomics, see chapter 1.
3. In the context of a widely publicized lecture at the London School of Economics in June 2009, Paul Krugman argued that much of the past 30 years of macroeconomics was “spectacularly useless at best, and positively harmful at worst.” The arguments of Krugman and other critics were taken up by the Economist magazine on July 16, 2009, in an article called “What Went Wrong with Economics,” to which there was a response by Robert Lucas, published in the Economist on August 6, 2009. Krugman’s criticisms reemerged with an article called “How Did Economists Get It So Wrong?” in the New York Times Magazine on September 2, 2009.
In this article, Krugman [2009b] argued that “the economics profession went astray because economists, as a group, mistook beauty, clad in impressive-looking mathematics, for truth.” A more refined version of these views was presented in Krugman [2009a], a popular book arguing for a return to traditional Keynesian macroeconomics, published at around the same time. See also Skidelsky [2009] for similar arguments from one of the biographers of Keynes. John Cochrane provided a detailed response to Krugman, in an article called “How Did Paul Krugman Get It So Wrong?” published on his blog on September 16, 2009, and eventually in Economic Affairs in 2011. This incident is a typical example of an old-style acrimonious exchange that generated more heat than light.4. However, note the strong disagreements with this view of, among others, Solow [2008], Krugman [2009b], and Stiglitz [2014, 2018].
5. As a result of the greater complexity of DSGE models, there is a serious dichotomy in the teaching of macroeconomics. Most undergraduate courses and textbooks still by and large rely on the IS-LM, AS-AD models of the previous generation, appropriately extended, but with a marginal role for the intertemporal approach. See, among others, Blanchard [2016b]; Mankiw [2016], and Jones [2017]. In contrast, graduate courses and textbooks rely almost exclusively on the intertemporal approach and DSGE models. See, among others, Blanchard and Fischer [1989]; Romer [2018], and Sargent and Ljungqvist [2018]. DSGE models have some way to go before they become fully accessible to undergraduate students, although Barro [1997b] and Chung [2015] are notable examples suggesting that the task can be accomplished.
6. See Shimer [2005, 2009, 2010], Blanchard [2009], Alogoskoufis [2016, 2018], and Christiano et al. [2018] on some of these points. Also note that one can in principle construct a hybrid new neoclassical synthesis model with staggered pricing, labor market distortions, and investment dynamics.
Some (mainly empirical) papers have undertaken this task. See, for example, Smets and Wouters [2003, 2007] and Christiano et al. [2016]. Yet this results in even greater complexity and significant losses in the degree of theoretical consistency and transparency compared to the original benchmark models.7. See Bernanke et al. [1999], Blanchard [2009, 2018], Stiglitz [2014], Ghironi [2018], Wright [2018], Vines and Wills [2018], Christiano et al. [2018], and Gertler and Gilchrist [2018] for discussions of this point. Linde et al. [2016], an example of recent attempts to incorporate financial market imperfections in estimated DSGE models, highlights the inherent difficulties.
8. See Blanchard [2016a, p. 1] who argues that “current DSGE models are seriously flawed, but they are eminently improvable and central to the future of macroeconomics.” Christiano et al. [2018, p. 135] argue that “the enterprise of dynamic stochastic general equilibrium modeling is an organic process that involves the constant interaction of data and theory. Pre-crisis DSGE models had shortcomings that were highlighted by the financial crisis and its aftermath. Substantial progress has occurred since then. …We don’t know where that process will lead. But we do know that DSGE models will remain central to how macroeconomists think about aggregate phenomena and policy. There is simply no credible alternative to policy analysis in a world of competing economic forces operating on different parts of the economy.”
9. In one sense, the atheoretical approach of Sims [1980] is the modern equivalent of the measurement without theory approach of Burns and Mitchell [1946], while the calibration approach of Prescott [1986], combined with Bayesian econometrics, has evolved into the modern equivalent of the structural econometrics of Klein [1950].
10. See Favero [2001], Canova [2007], DeJong and Chetan [2011], and Herbst and Schorfheide [2015] for detailed accounts of modern empirical macroeconomics.
11. See Alogoskoufis and Smith [1991] and Alogoskoufis [1992] for the empirical significance of the Lucas critique in the case of estimated Phillips curves.