Available online at www.sciencedirect.

com

Journal of Economic Theory 147 (2012) 545550 www.elsevier.com/locate/jet

Introduction to economic growth

Abstract This introduces the symposium on economic growth.

Economic growth continues to be one of the most relevant and exciting sub-areas of economics. Its relevance stems from the questions it focuses on. The problem of economic development remains a major one for humanity at large and for economics as a science. At the time Adam Smith laid many of the foundations of modern economics, there were likely small differences between the richest and the poorest nations in the world (e.g., Maddison [18], Acemoglu, Johnson and Robinson [3]). Since then, the gaps between the rich and poor have increased to a level that would have been incomprehensible to most 18th and 19th century economists. At the root of this great disparity is the differential growth experience around the world. Some, like many in western Europe and western European offshoots around the world, have grown rapidly during the 19th and early 20th centuries, while many others have stagnated. This differential growth led to a huge gap in income per capita and living standards that continues to this day. Naturally, economic growth also has the power to rapidly close such gaps as illustrated by the experiences of countries of Japan, South Korea, Singapore, and more recently China. Thus, the consequences of a few percent change in the growth rate of a nation can have huge consequences for the well-being and living standards of its citizens in one or two generations.
E-mail address: [email protected]. 0022-0531/$ see front matter 2012 Published by Elsevier Inc. doi:10.1016/j.jet.2012.01.023

546

D. Acemoglu / Journal of Economic Theory 147 (2012) 545550

Economic growth, and more broadly economic development, is also an exciting area to study for several reasons. First, despite the great relevance and importance of the questions at hand, there are still so many unknowns and so many major challenges that the area has attracted great intellectual activity and is likely to continue to do so in the future. Second, economic development, as Kuznets emphasized, is highly multi-faceted (Kuznets [16]). It is not just about growth of aggregate output, but also about the fundamental transformation of an economy, ranging from its sectoral structure, to its demographic and geographic makeup, and perhaps more importantly, to its entire social and institutional fabric. These processes naturally require a much more holistic approach to economic growth and development than in many other areas of economics. Thus the political, social and demographic elements in the process of growth are paramount. This leads to a rich array of questions and a variety of new approaches to fundamental questions of economic growth. Third, the theory of economic growth also enjoys a special place within economics because it combines micro and macro in an exciting way. Economic growth, by its nature, is about aggregate and macro outcomes. Yet, as opposed to many other sub-areas of macroeconomics, the study of economic growth has remained rmly based on microeconomic foundations (e.g., Cass [7], Koopmans [15]), and over time it has continued to build on recent developments in microeconomics, particularly in game theory and information economics (see Acemoglu [1] for a review). Finally, economic growth is also a largely empirical eld as more researchers have turned to investigating new empirical questions within the area of growth and development. Economists have recently brought much energy, and rich data, to the investigation of the microeconomic relationships underpinning the process of economic growth. This symposium, which honors and celebrates David Casss contributions to this area, focuses on theoretical aspects of the study of economic growth and showcases some of the recent and exciting research under this umbrella (the vast and rapidly expanding area of empirical development economics falls beyond the scope of the current symposium). As bets the richness of the eld of economic growth, the papers in this symposium cover a large set of topics with an even broader range of approaches. The papers include: applied studies developing new approaches to established questions within the eld of economic growth (such as technology diffusion or structural change), papers tackling new questions at the frontiers of economic growth (such as the role of organizations, entrepreneurship and the sources of asset price bubbles in the context of neo-classical growth models); and work answering open questions in the theory of aggregate growth. A major topic within the area of economic growth is the study of technology diffusion. It is well recognized that technology differences across nations, industries and rms are the main sources of productivity differences and there has been much advance in models of endogenous innovation and technology. Nevertheless, the forces shaping the diffusion of technology are still poorly understood. Three papers within the symposium investigate various different aspects of the process of technology diffusion. Investment in vintage capital by Boyan Jovanovic and Yuri Yatsenko [14] revisits models of vintage capital, originally introduced by Johansen [13], Arrow [4], and Solow [21]. They propose a rich model of investment in different vintages of heterogeneous capital goods, which are combined to produce a unique nal good with a constant elasticity of substitution production function. They provide an elegant characterization of equilibrium, determining how the investment is allocated across different vintages. An important result of this paper is that, in contrast to many other models of vintage capital, not all investment goes to the latest vintage. The reason is imperfect substitution between different vintages. The structure of equilibrium has the avor of staggered adoption of new technologies. In particular, the economy adopts and starts producing with new technologies gradually rather than immediately

D. Acemoglu / Journal of Economic Theory 147 (2012) 545550

547

with all new investment owing into the latest technology. Moreover, Jovanovic and Yetsenko show that the pattern of adoption resembles the well documented S-shape. The paper Competing engines of growth: innovation and standardization by Daron Acemoglu, Gino Gancia and Fabrizio Zilibotti [2] emphasizes another source of slow technology diffusion: the interplay between skill-intensive innovation and the process of standardization. They argue that the process of standardizing new technologies, which enables them to be used by more abundant and cheaper lower skilled workers, is a major part of the growth process in practice. But endogenous standardization is both an engine and a barrier to growth. As standardization takes place rapidly, existing technologies are utilized better, increasing productivity and income per capita. However, the anticipation of standardization discourages innovation because innovators will have shorter life spans during which to protably use their new technologies. As a result, equilibrium growth is an inverse U-shaped function of the standardization rate and thus the degree of competition. This particular pattern implies that the growth and welfare maximizing rates of standardization are intermediate and thus need to be supported by optimal, but not full, protection of intellectual property rights. The paper also shows how the interplay between innovation and standardization leads to a new type of multiplicity of equilibria, with some equilibria having the avor of development traps with low innovation because of high rates of standardization faced by innovators, though paradoxically this also leads to lower range of new technologies used by less skilled workers as the economy does not produce many new technologies. The third paper in the symposium addressing issues of technology diffusion is by Erzo Luttmer, Technology diffusion and growth [17]. Luttmer considers a neoclassical model with rms subject to idiosyncratic productivity shocks and subject to xed costs to stay in business. The interplay of productivity shocks and xed costs implies that rms that receive a sequence of negative shocks are no longer protable and choose to exit. They are replaced by new rms. The equilibrium process of entry and exit leads to a stationary rm size distribution under a variety of regularity conditions. Luttmer also shows that if entrants are of relatively low productivity (i.e., their productivity is close to those of rms at the margin of exit), then the rm size distribution approximates a Pareto distribution with an exponent of 1, which is a good approximation to the rm size distribution in the US data. A new area of research within the theory of economic growth focuses on the role of organizations in economic growth. Two papers within this symposium address various facets of this problem. Organizing growth by Luis Garicano and Esteban Rossi-Hansberg [9] provides a new model featuring organizations within the process of economic growth. They construct a tractable framework in which productivity growth results from the interplay between accumulation of knowledge and information in communication technology. Agents accumulate knowledge both to use available technologies and to invent new technologies. The rst use of knowledge also necessitates organizations, which is an original aspect of the model considered by Garicano and Rossi-Hansberg. Organizations play the role of coordinating economic activity and facilitating the use of existing technology. Using this framework, the authors conduct a variety of comparative static exercises. They show that information technology, by increasing both innovation and the effectiveness with which this innovation is used, always increases growth. However, improvements in communication technology, arguably one of the more common forms of technological advances in the recent past, may have the converse effect and reduce growth, because they increase the return to using available technologies through organizational improvements. This work not only emphasizes the role of organizations, but also complements ideas related to standardization both as an engine and a barrier to economic growth.

548

D. Acemoglu / Journal of Economic Theory 147 (2012) 545550

Occupational choice and development by Jan Eeckhout and Boyan Jovanovic [8] turns to another role of organizations in the context of the global division of labor. They consider an assignment model of occupational choice and investigate how globalization, corresponding to the ability of workers that are geographically separated to match together, affects productivity and welfare. The same types of information and communication technologies studied by Garicano and Rossi-Hansberg [9] can thus be thought of as increasing the geographic span of control of managers, enabling matching between different types of workers across wider geography. Eeckhout and Jovanovic show that welfare gains in such a framework depend on the skill heterogeneity of worlds labor work force, which is being better exploited by the increasing geographic span of control. Middle income countries experience the smallest change in welfare because they have the smallest change in the factor price ratio from the resulting re-organization of the world economy. Francisco Buera and Joseph Kaboski in Scale and origins of structural change [5] turn to another major question in the theory of economic growth, the origins of structural change. In addition to documenting a rich set of facts related to sectoral reallocations, scale of productive units, and the shifting of production from home to market, they provide a new model of structural change based on the interaction between scale economies and mass consumption. In their model, different sets of goods, satisfying different wants, can be produced at home or in the market, and using a traditional or modern technology. A key factor affecting the mode of production is the scale of activity. Market technologies make better use of scale economies. This creates both a connection between growth and a shift in activities to the market technology, and a link between structural change and mass consumption. Their model thus provides a simple explanation for structural changes based on endogenously changing production relationships, and can account for the rich set of facts that Buera and Kabolski themselves document. In Random walk to innovation: why productivity follows a power low, Christian Ghiglino [11] turns to another central question in the theory of economic growth: the source of new and more productive ideas. Ghiglino provides a simple search model in which innovators, who have limited information about how new ideas will result from recombinations of existing ideas, search for better ways of producing output. Combining better ideas leads to better ways of using existing resources. Ghiglino shows that, under reasonable assumptions on the search process, this searchrecombination process leads to new ideas that have a non-degenerate distribution with a thick tail, meaning that the distribution contains a signicant fraction of ideas with high productivity. In fact, it can be approximated by a power law. This approach has the promise of explaining why the combination of existing ideas can lead to a large supply of new high productivity ideas, and thus be an integral part of the process ensuring sustained growth in the economy. The possibility of bubbles of various forms in neoclassical growth models has long been recognized. Bubbles are generally viewed as a consequence of dynamic inefciency, resulting from a capital stock above the golden rule level, and also often as a potential source of additional inefciency and the allocation of resources. In Bubbles and capital ows, Jaume Ventura [22] challenges many of the long held views about bubbles. First, he shows how bubbles (even stochastic bubbles) can arise not because of dynamic inefciency in the entire world equilibrium, but because of a version of dynamic inefciency in some part of it. Second, he shows how such bubbles may play a useful role by acting as a substitute for international capital ows when international trade in assets is limited. Put differently, bubbles in low productivity countries relax the constraints faced by these countries in accumulating capital. Ventura shows how this view of asset price bubbles leads to a rich set of implications which are potentially in line with several recent trends.

D. Acemoglu / Journal of Economic Theory 147 (2012) 545550

549

Entrepreneurships is often viewed as a key input into economic growth. Exactly what entrepreneurs do and how entrepreneurial ideas and attitudes develop are poorly understood. In Evolution and the growth process: natural selection of entrepreneurial traits, Oded Galor and Stelios Michalopoulos [10] provide an original evolutionary model of the emergence and selection of entrepreneurial traits. Entrepreneurial spirit is modeled as risk tolerance, and evolves non-monotonically in the course of the development process. In the early stages of development, there is positive selection in favor of entrepreneurial spirit because risk tolerant individuals are more productive. They in turn contribute to the knowledge stock and the growth process. However, Galor and Michalopoulos also show that later in the stages of development, risk aversion gains an evolutionary advantage, endogenously slowing down the process of growth. Their model, thus, predicts that entrepreneurial spirit should be more common in countries at middle stages of development. Another central set of issues concerns how policy can be best used to encourage innovation. In Prizes and patents: using market signals to provide incentives for innovation, V.V. Chari, Mikhail Golosov and Aleh Tsyvinski [6] construct a general and exible framework for using a mechanism design approach to thinking about when prizes are better suited to encourage innovation than patents. Patents encourage innovations by providing ex post monopoly power to innovators. Prizes, on the other hand, reward innovation according to a pre-specied menu. Prizes can be improved upon by using market signals in determining the compensation for innovation. Chari, Golosov and Tsyvinski show that mechanisms relying on prizes allocated on the basis of market signals can be manipulated. When such manipulation is not possible, prizes based on market signals emerge as the best mechanism. However, patents emerge as part of the optimal design on incentives when manipulation is possible. In addition to new questions and new approaches to existing questions within the broad area of the theory of economic growth, there are important technical questions that are still open. Three of the papers in this symposium make signicant advances in resolving open questions in the literature. Martin Kaae Jensen in Global stability and the turnpike in optimal unbounded growth models [12] provides a conclusive answer to a question that has long been open: whether in unbounded models on economic growth, optimal growth paths will have the turnpike property and approach a balanced growth path (provided that there is sufciently limited discounting). For general unbounded technologies (nesting non-smooth technologies, joint production and several sectors), Jensen provides a complete characterization of the solution and shows that under mild assumptions it will indeed approach optimal growth path. Tapan Mitra and Debraj Ray in On the PhelpsKoopmans theorem [19] turn to another major result of growth theory, the PhelpsKoopmans theorem, which characterizes the inefciency of growth on the basis of whether the capital stock path is above and bounded away from the golden rule stock of capital. A well-known result shows that every stationary path with capital stock in excess of the golden rule is inefcient, as reducing the capital can increase the consumption level in all dates. A natural conjecture would be that a path is inefcient if it converges to a capital stock in excess of such a minimal golden rule. However, Mitra and Ray show that in the presence of non-convexities in the production technology, this is not true. This surprising result suggests that a strong form of the PhelpsKoopmans theorem is false. Mitra and Ray then formulate and establish a variant of the PhelpsKoopmans theorem for the case of non-convex technologies. They uncover a new set of restrictions on the technology set for the theorem to be valid. These restrictions are automatically satised for convex technologies. Finally, Tapan Mitra and Santanu Roy in Sustained positive consumption in a model of stochastic growth: the role of risk aversion [20] turn to another important problem in the theory

550

D. Acemoglu / Journal of Economic Theory 147 (2012) 545550

of growth: whether in the presence of aggregate stochastic shocks, long run consumption can be guaranteed to stay away from zero. This is a minimal condition on stochastic growth models, but surprisingly, even Cobb-Douglas production technology, which ensures that the standard Inada conditions hold, does not guarantee that the economy will be able to stay bounded away from zero consumption. Mitra and Roy provide a complete characterization of the conditions under which stochastic growth models, even with the Inada conditions ensuring arbitrarily large productivity near zero capital stock, can lead to zero consumption with probability 1 and when they will ensure that aggregate consumption remains bounded away from zero. We hope that the range of diverse, ambitious, original and exciting papers in this symposium is the best way for us to honor David Casss memory and contributions to the theory of economic growth. References
[1] D. Acemoglu, Introduction to Modern Economic Growth, Princeton University Press, Princeton New Jersey, 2009. [2] D. Acemoglu, G. Gancia, F. Zilibotti, Competing engines of growth: innovation and standardization, J. Econ. Theory 147 (2) (2012) 567598, this issue. [3] D. Acemoglu, S. Johnson, J. Robinson, Reversal of fortune: geography and institutions in the making of the modern world income distribution, Quart. J. Econ. 117 (4) (2002) 12311294. [4] K.J. Arrow, The economic implications of learning by doing, Rev. Econ. Stud. 29 (2) (1962) 155173. [5] F. Buera, J. Kaboski, Scale and the origins of structural change, Working paper, Fed. Reserve Bank Chicago, 2008. [6] V.V. Chari, M. Golosov, A. Tsyvinski, Prizes and patents: using market signals to provide incentives for innovations, J. Econ. Theory 147 (2) (2012) 778798, this issue. [7] D. Cass, Optimum growth in an aggregative model of capital accumulation, Rev. Econ. Stud. 32 (3) (1965) 233240. [8] J. Eeckhout, B. Jovanovic, Occupational choice and development, J. Econ. Theory 147 (2) (2012) 654680, this issue. [9] L. Garicano, E. Rossi-Hansberg, Organizing growth, J. Econ. Theory 147 (2) (2012) 620653, this issue. [10] O. Galor, S. Michalopoulos, Evolution and the growth process: Natural selection of entrepreneurial traits, J. Econ. Theory 147 (2) (2012) 756777, this issue. [11] C. Ghiglino, Random walk to innovation: why productivity follows a power law, J. Econ. Theory 147 (2) (2012) 710734, this issue. [12] M.K. Jensen, Global stability and the turnpike in optimal unbounded growth models, J. Econ. Theory 147 (2) (2012) 799829, this issue. [13] L. Johansen, Substitution vs. xed production coefcients in the theory of economic growth: a synthesis, Econometrica 27 (2) (1959) 157176. [14] B. Jovanovic, Y. Yatsenko, Investment in vintage capital, J. Econ. Theory 147 (2) (2012) 548566, this issue. [15] T.C. Koopmans, On the Concept of Optimal Economic Growth, Cowles Foundation Paper, 238, 1965. [16] S. Kuznets, Modern Economic Growth, Yale University Press, 1967. [17] E. Luttmer, Technology diffusion and growth, J. Econ. Theory 147 (2) (2012) 599619, this issue. [18] A. Maddison, The World Economy: A Millennial Perspective, Development Centre, Paris, 2001. [19] T. Mitra, S. Ray, On the PhelpsKoopmans theorem, J. Econ. Theory 147 (2) (2012) 830846, this issue. [20] T. Mitra, S. Roy, Sustained positive consumption in a model of stochastic growth: the role of risk aversion, J. Econ. Theory 147 (2) (2012) 847877, this issue. [21] R.M. Solow, Substitution and xed proportions in the theory of capital, Rev. Econ. Stud. 29 (3) (1962) 207218. [22] J. Ventura, Bubbles and capital ows, J. Econ. Theory 147 (2) (2012) 735755, this issue.