P. Aghion and A. Banerjee. Volatility and Growth: The Clarendon Lectures in Economics.

P. Aghion and A. Banerjee. Volatility and Growth: The Clarendon Lectures in Economics. Oxford University Press, 2005, Oxford, UK. ISBN 0-19-924861-3, Price $35.00.

In February of 1999, Philippe Aghion from Harvard University and Abhijit Banerjee from MIT delivered the Clarendon lectures in Economics at Oxford University. In those lectures, Aghion and Banerjee explored the foundations of connections between aggregate volatility and growth, in real GDP. It took the authors six more years to formalize the content of those lectures, update them, add in a bunch of new material on open economies, and thematically unify it all as a six-chapter book titled Volatility and Growth.

Almost all the material in the book comes out of papers previously published by the authors, sometimes jointly with other researchers. As the authors put it, some of the book's contents took a while to mature and meet the publication test. This time to build makes the book superadditive, i.e., makes the book bigger than the sum of its parts. Arguably, the reader is getting the more mature and thematically unified view of the literature than is ever available from the journal version of the individual papers; case in point, the formal models outlined in the chapters are presented in a less rigorous but far more readable fashion than their journal counterparts. This makes the book more accessible to graduate students who are thinking of starting work in this area. Such students can get the necessary rigor from the journal papers but can rarely get the crucial big picture from them. A book like this fills that hole. It opens up many questions and suggests some frameworks in which to pose those questions.

So, what is the book all about? In one sentence, it is a detailed exploration of the causal connections between the trend in real GDP and volatility surrounding that trend, or more simply, the connection between growth and volatility (business cycles). One is of course allowed to ask, why is that an interesting question. It turns out that the cross-country empirical evidence on this issue seems to converge to the stylized fact that volatility hurts growth; it is this correlation that makes the pursuit of a causal connection between growth and volatility a worthwhile endeavor.

The topic itself is a very old one. Way back, Marx wondered if the mad orgy of accumulation in a capitalist economy periodically sowed the seeds of its own destruction. More recently, real business cycle analysts using the neoclassical, market clearing, rational expectations paradigm have argued that exogenous shocks to total factor productivity (the Solow residual) cause volatility in output and some of that volatility may lead to growth. Aghion and Banerjee are not satisfied with the neoclassical answer because (a) that framework ignores the reality that markets do not always work, and (b) that the magnitude of the aforementioned productivity shocks needed to produce realistic business cycles are not those we observe.

The authors believe that the right way to proceed would be to introduce some market failures into the neoclassical paradigm. Previous work on the financial accelerator by Ben Bernanke, Mark Gertler, Nobuhiro Kiyotaki, and John Moore have assured the authors that introducing credit market frictions, to prevent the Modigliani-Miller irrelevance theorem from working, may be prudent. The idea is that the source of finance for business investment, whether it is internal or external, may be important in the aggregate. Standard agency cost concerns ensure that external finance carries a premium over internal finance. Productivity shocks may affect borrowers financial health and change the magnitude of this premium. Borrowers who rely substantially on external finance will face a much higher premium and hence exhibit very divergent investment behavior from those with low external stakes. In this setting, even a small negative productivity shock can generate persistence and amplification of the initial shock thereby creating large output variability.

In Chapter 1, the authors start by introducing productivity shocks into the neoclassical growth model. In that environment, the average growth rate in the economy turns out to depend solely on the savings rate. Higher volatility has two opposing effects on the savings rate: on the one hand, more volatility increases precautionary savings raising the savings rate and growth, while on the other hand, more volatility reduces the average return on capital causing consumption to rise and growth to fall. For realistic parametric configurations, the model suggests that heightened volatility raises growth, contrary to the received empirical wisdom. The authors go on to introduce productivity shocks into a Schumpeterian environment. Here too, the model predicts counter to reality, for now a slump reduces the opportunity cost of making long-term investments (such as R&D) thereby promoting growth. It is now readily apparent how credit market frictions help to match the predictions of the Schumpeterian model with reality. Aghion and Banerjee unite the financial accelerator story with the Schumpeterian story, the subject matter of their Chapter 2, yielding the following: in a slump, credit constrained firms may not be able to attract enough external funds to make the aforementioned productivity-enhancing long-term investments. This is a strong and clear prediction, one the authors are able to empirically verify, at least for financially underdeveloped economies.

Thus far, all volatility came from outside the models in the form of exogenous productivity shocks, a remnant from the real business cycle literature. The authors go on to abandon this remaining neoclassical relic and seek to generate endogenous persistent volatility arising from credit market frictions and their interaction with a growing economy. For this to work, the authors have to rely heavily on strong price effects, and here is why. In their setup, faster growth is accompanied by a rise in interest rates that raises the debt to be repaid by entrepreneurs. From the financial accelerator story, it is evident that this increase in debt raises agency costs and further reduces the wealth of the entrepreneurs and hence their future investment ability. The latter hurts growth and a growth cycle has emerged in the model. This work is similar to a lot of recent research by Kiminori Matsuyama and some earlier work by Costas Azariadis and Bruce Smith, although the authors seem unaware of this other research.

The remaining chapters in the book extend these ideas to open economy settings. In a small open economy setting, the interest rate will be determined in world markets and will not be influenced by domestic growth. Here the exchange rate will be asked to perform the role played earlier by the domestic interest rate.

Overall, the book presents an excellent albeit selective treatment of the growth-volatility nexus, one heavily influenced by the research program of the authors. Both Aghion and Banerjee are masters of the art of writing down simple and parsimonious models (some may argue too "reduced form") that get to the main point very quickly. Graduate students and researchers wishing to venture into this area of research will get to see the masters at work and learn a lot about their craft.

Skeptics may argue that the final product did not yield many truly new insights. After all, the financial accelerator story was well known at the time and its introduction into the Schumpeterian environment, while instructive, raised more questions than it answered. For example, if long-term investments are allowed in the model, why are the credit contracts restricted to a single period in length? Skeptics may also point out that the authors exercise of generating endogenous persistent volatility, while of sufficient theoretical interest, may not be that empirically relevant since the required price effect would never obtain in reality. But none of these comments can detract from the main point the authors have made forcefully and in a highly reader-friendly manner: credit market imperfections are key to understanding the growth-volatility nexus.

Joydeep Bhattacharya

Iowa State University