بهبود رشد بالا، موجودی و اعتدال بزرگ

We present evidence about the disappearance of the high-growth recoveries from recessions with intense job creation typically observed until the eighties. This result matches the belief that recessions now have an L-shape as opposed to the old-time recessions that always had a V-shape. We also show how this change in business cycle dynamics can explain part of the Great Moderation. We postulate that these two phenomena may be due to changes in inventory management brought about by improvements in information and communications technologies.

There is a traditional debate in economics on how recessions finish, i.e., whether they are “V-shaped” or “L-shaped”. The former type of recession refers to the case in which the economy springs back rapidly from its slump, whereas the latter refers to the case in which the economy faces a period of flat or at best slowly improving performance. Facing either V-shaped or L-shaped recessions has both economic and econometric implications. The economic implications of facing each of these types of recoveries are evident. V-shaped recessions are viewed as evidence in favour of Friedman's (1993) plucking model, in which output cannot exceed a ceiling level but is occasionally plucked downward by recessions which have only temporary effects. On the contrary, recessions which are followed by flat recoveries (L-shaped) are viewed as having permanent effects on the level of production. The econometric implications of facing these two alternative recoveries have to do with the traditional discussion regarding whether US output exhibits either a deterministic or a stochastic trend. On the one hand, Kim and Nelson (1999b) document that US recessions were usually followed by periods of very high growth which have been called the “third phases of business cycles”. Having rebounds after recessions which spring the economy back to pre-recession levels can be viewed as the economic interpretation of the papers that find evidence that GDP is trend stationary and that the effects of recessions are mainly transitory. In this context, Cheung and Chinn (1999) conclude that, with a long span of data, one can obtain evidence of trend stationarity. On the other hand, if one cannot observe the rapid recoveries in output, the negative effects of recessions can be viewed as more permanent. Supporting this view, Campbell and Mankiw (1987) show that there is a considerable permanent effect of a surprise change in output. In this paper we show that, after the eighties, business cycle recoveries have turned out to be L-shaped, so the periods following recessions are now characterized not by high growth but by lower growth than in the course of the expansion. The change in this business cycle feature roughly coincides with the jobless recoveries from the recessions since the nineties, as documented by Groshen and Potter (2003). To provide a view of the form of US recoveries, Fig. 1 shows the growth rates of GDP together with the recessions as documented by the NBER. In the graph, we can observe the decline in the relevance of the high-growth recovery phase of the cycle in the last three recessions. While the end of the seven recessions prior to the mid-eighties were characterized by above-average growth rates, the recessions after that date were followed by quarterly growth rates below the average. According to our discussion above, an important consequence of the disappearance of this high-growth recovery phase is that recessions now have the potential to have long run effects on the economy.1 Full-size image (37 K) Fig. 1. GDP growth rates 1953.1–2010.2. Note: Shaded areas correspond to recessions as documented by the NBER. The vertical discontinuous line refers to 1984.1. The horizontal discontinuous line represents the average growth rate of the whole sample. Figure options The sluggish pace of recovery in output during the recoveries since the nineties contributes to the sluggishness of job creation observed after the latest recessions. In these recoveries, Groshen and Potter (2003) find more evidence of permanent job losses than of temporary layoffs and reallocation of jobs from one industry to another. Schreft et al. (2005) show that one common feature of the recent jobless recoveries was the greater use of just-in-time employment practices, the employment of temporary and part-time workers and the use of over time to achieve a more flexible workforce. Noticeably, the loss of the high-growth recovery phase of business cycles and the evidence of jobless recoveries roughly coincide with the period of the Great Moderation previously documented by McConnell and Perez Quiros (2000) and Kim and Nelson (1999a), which has been dated in the first quarter of 1984. In this paper we present evidence to show that these two features of US business cycle dynamics may be related. According to our measures, part of the high volatility of output growth before 1984 can be explained by the existence of the high-growth recovery phase. By means of a counterfactual exercise, we show that when this phase is removed from business cycle dynamics, the statistical evidence for a structural change in the volatility of output decreases dramatically. In addition, we postulate that both the volatility reduction and the loss of the high-growth recovery phase can in part share the same economic sources, which are related to changing business practices. Kahn et al. (2002) analyze the role of inventory management as the source of reduction in output growth volatility. In addition, Davis and Kahn (2008) and Kahn (2008) directly relate the Great Moderation to changes in the role played by inventory accumulation from avoiding stockouts (see Kahn, 1987) to smoothing production. But this change in inventory management could also explain the loss of the high-growth recoveries since the eighties. According to these authors, firms maintained inventories to avoid stockouts until the eighties. In periods of low demand, inventories are low because the probability of stockout is low. As the economy exits the recessionary period, firms increase production not only to satisfy growing demand but also to replenish inventories above the level they had during the recession which would lead to recoveries with rapid growth. Using this view, Sichel (1994) stated that the high-growth recovery phase of business cycles until the eighties could be linked to inventory accumulation. However, the rapid improvements in information technology in the eighties have led firms to rationalize the use of inventories, which are now used to smooth production. In periods of low demand, firms maintain their production levels and accumulate inventories to respond to future periods of high demand. As the economy exits the recessionary periods, increasing demand would be serviced out of inventories, which explains why rapid-growth recoveries from the recent recessions have not been observed since this change in inventory management came into place. Although we do not analyze in the paper the marginal effect of the different sources of the Great Moderation put forward in the literature, we consider that the evidence shown in McConnell and Perez Quiros (1998), Kahn et al. (2002), Davis and Kahn (2008) and Kahn (2008) is enough to seriously consider the hypothesis of better inventory management as a potential explanation of the reduction in volatility.2 Instead, this paper goes further on the implications of this hypothesis by putting together the effect of changes in inventory management not only on the reduction in volatility but also on the loss of the high-growth phases of business cycles and, therefore, on the evidence that recoveries are now L-shaped and with sluggish job creation. In this sense, linking the reduction in volatility observed after 1984 with the changes in the pattern of the recoveries provides some fresh insights with which to examine other theories about the causes of the Great Moderation. Basically, these other theories fall under three groups. The first group associates the Great Moderation with “good luck”, understood as a reduction in the size of shocks hitting the US economy since the mid-eighties. This is the conclusion, among others, of Ahmed et al. (2004). A second group of papers contends that the Great Moderation is merely a consequence of “better economic policies”. For example, Clarida et al. (2000) argue that it is the more aggressive response of the Federal Reserve to inflation that lies behind economic stability. Finally, a third group incorporates better financial intermediation as in Dynan et al. (2006). None of these theories have predictions about how the alleged cause of the Great Moderation changes the shape of the recovery. According to Clark (2009), these theories are also difficult to reconcile with the fact that, in spite of the severity of the 2008–2009 recession, the Great Moderation is not over. The rest of the paper is organized as follows. Section 2 provides support for the disappearance of the high growth recovery with data on GDP. Section 3 presents a counterfactual exercise to gauge the importance of the high recovery phase in explaining the volatility of GDP growth. Section 4 relates these two phenomena to inventory management and Section 5 concludes.