Introduction

Economists, going back to Karl Marx, have been interested in the connection between the dynamism of capitalist economies and their apparent instability, marked by the recurrence of the business cycle.

The economists of the real business cycle school (see Kydland and Prescott 1982; Long and Plosser 1987, for the classic formula­tions of this view), very far from Marx in almost every other way, share with him a belief that these fluctuations are an integral part of the growth process. In their world, markets always do the best that is possible, given the resource constraints that the economy faces, and therefore fluctuations happen because they have to. The reason why they do, in this view, is that productivity grows in fits and starts, mostly because big new ideas drive productivity and big new ideas are rare. Periods of rapid growth are often followed by a lull, and occasionally, a negative productivity shock (such as an oil shock), so that the entire process looks like a series of fluctuations around an upward trend. Trying to get rid of these fluctuations may come at the cost of killing growth.

Traditional Keynesians hold that there is nothing necessary about many of the ups and downs that we observe: Appropriately chosen policies would get rid of most of them. In their world, how­ever, it is assumed that these fluctuations have nothing to do with growth, which has its own autonomous dynamic.

This monograph grew out of our dissatisfaction with all of these positions. With the Keynesians, for not taking the question ser­iously: After all, the key insight of modern growth theory is that growth happens through the decisions of individuals and it is hard to imagine that these decisions are entirely insulated from forces that cause the business cycle.^[I] With the real business cycle school, in part because our reading of the micro evidence radically under­mines their presumption that markets always work, and partly because the data suggests that there are not enough identifiable productivity shocks of the right magnitude to explain by them­selves the many ups and downs of capitalist economies.

We set out to build a model of the aggregate economy that takes the interactions between volatility and growth seriously, while being open to the possibility of market failures. Specifically, we begin by building a model of macro fluctuations and growth along conventional lines, where growth is driven by R&D and productivity is subject to shocks. What we add is the possibility of imperfections in the credit markets. There is mounting evidence that credit markets, especially in developing countries, but even in the developed world, do not come anywhere close to the neo­classical ideal of a single market rate at which anyone can borrow or lend as much as they want.

We introduce credit market imperfections into the model very simply: We assume that there is a limit to how much credit anyone can get, which is a multiple of their wealth. We allow this multiple to vary across economies, and use it as a natural measure of finan­cial development. Ajustification for assuming that individuals face a constraint of this kind, including a simple model that generates exactly this kind of credit supply function, is given in Chapter 0.

Chapter 1 starts the main body of the monograph. We investigate what happens when we introduce productivity shocks respect­ively into the AK model and the Schumpeterian model of endo­genous growth. We argue that the main message from both of them is that if volatility has an effect on growth, it is probably positive: it encourages precautionary savings in the AK model, and slumps reduce the opportunity cost of long-term R&D investments in the Schumpterian model. Moreover, when volatility hurts growth in the AK model, it is because it leads to lower total investment.

The problem is that in the cross-country data the correlation between volatility and growth is clearly negative (Ramey and Ramey 1995).

Chapter 2 introduces credit constraints into the Schumpeterian model developed in Chapter 1. Firms that face credit constraints run the risk that they may not necessarily be able to raise money for profitable investments. In an environment where productivity shocks create uncertainty in a firm's liquidity position, this can dis­courage firms from undertaking long-term investments that might generate a call on the firm's liquidity. If long-term investment (like R&D) are the investments that generate productivity growth, a combination of shocks and credit constraints can undermine growth.

The theoretical part of Chapter 2 demonstrates this formally, and argues that these effects are large enough to explain a sub­stantial reduction in the growth rate in the most financially underdeveloped countries. The empirical part shows that the cor­relation between growth and volatility is indeed more negative in the less financially developed countries. Moreover, productivity shocks have a bigger impact on growth in these countries.

So far the presumption has been that the sources of volatility are given from outside, though their effect may be accentuated by credit constraints. The next two chapters set out to investigate whether there can be volatility in this world without any shocks. Chapter 3 shows that this kind of intrinsic volatility can arise in a closed economy if credit constraints are tight but not too tight: The economy does not settle down to a steady state, but fluctu­ates forever around a trend growth rate, even when there are no shocks. Periods of fast growth tend to push up interest rates, which squeezes profits. As profits shrink, investment falls, because firms are credit constrained and their investment is constrained by how much money they have at hand. This slows growth, which allows profits to be reconstituted, and so on.

This can only happen at certain levels of financial development. In the most developed economies, credit constraints are too slack to matter. These economies grow the fastest and do not fluctuate. In the least developed ones, they are so tight that growth is always slow and the cycle never starts. These economies do not fluctuate either, but grow slower than the economies at an intermediate level of financial development, which do fluctuate.

The idea that it is the economies at an intermediate level of financial development that are most likely to be endogenously volatile carries over to open economies as well. This is the sub­ject of Chapter 4, which should be read as an application of the ideas in Chapter 3 to a specific context where it is easy to match up the theory with the data: The empirical discussion in Chapter 4, largely based on Gourinchas-Valdes-Landeretche (2001), referred to as GVL in the Chapter, suggests that the mechanism underlying our model of endogenous fluctuations matches up well with GVL empirical findings concerning lending booms.

At the heart of these endogenous fluctuations is the idea that shocks, positive or negative, to the borrowing capacity of firms have spillovers on the rest of the economy. An extreme version of what such spillovers can do is a self-fulfilling currency crisis. These happen when people expect a currency crisis and therefore expect the interest cost of foreign currency loans to go up. This leads them to cut back on production, precipitating the crisis. This possibility is the subject of Chapter 5, which also looks at the role of monetary policy in dealing with financial crises in open economies.