Catalog

Proximate causes of slower growth

One way of dissecting growth is to identify how changes in the application of labor, human capital, physical capital, and technology have influenced its path. Changes in output growth can occur only if there has been a change in one or more of these components.

Employment

Figure 1.4.3 shows average growth of the labor force and employment for the periods 1990-1995 and 2000-2005. In all countries but Thailand, the growth of the labor force slowed between the pre- and postcrisis periods. In the Philippines and Thailand, the growth of employment accelerated between the two periods, barely changed in Indonesia, and slowed in Korea and Malaysia.

Clearly, this mixed picture cannot explain a general trend of slowing growth. Estimated labor shares in national income tend to be quite low in the crisis countries (some as low as 0.35). Taken at face value, a low labor share suggests that it would take quite large changes in employment or labor force growth to move growth rates by the observed magnitudes. Malaysia is the country where employment and labor force growth change most between the pre- and postcrisis periods: the rate of employment growth drops by 1.8 percentage points. This possibly could account for anything in the range of 0.7-0.9 percentage points of GDP growth a year, depending on the true labor share.

In the other countries, imputations would give much lower impacts, and in the Philippines and Thailand would suggest accelerating rather than slowing GDP growth. But unraveling causality is not straightforward. Labor force participation rates (and of course employment) are sensitive to economic conditions, which in turn will be influenced by growth and policies. In Malaysia, for example, where effects seem big, the gap between employment and labor force growth rates has, possibly, been influenced by policies on temporary workers and immigration.

Human capital

Measuring the quality of labor is fraught with difficulty. Estimates of average years of education or educational attainments do not capture critical factors linked to quality (see the chapter Education and structural change in four Asian countries, also in Part 1). Leaving this shortcoming to one side, attempts to estimate the contribution of human capital to growth using growth accounting methods have come up with estimates that are generally quite small (Young 1995, Collins and Bosworth 1996). If these estimates are considered reliable, it would require an implausibly abrupt slowing of human capital accumulation, or even reversals, to account for the shifts in aggregate growth that have been observed in the crisis countries. Human capital's impact on the deceleration in growth is unlikely to have been big.

Productivity growth

Technological progress is usually measured by growth in total factor productivity (TFP). TFP growth captures how much additional output can be generated for a given set of labor, human capital, and physical capital inputs. It is well known that estimates of TFP growth require a large number of assumptions and that they can be contaminated by errors in measurement of other inputs. For these reasons, TFP estimates should be considered with caution. Even for the same countries over the same time period, estimates of TFP growth often vary widely (see, e.g., Crafts 1999).

At an aggregate level, the effect of the crisis on TFP growth is likely to have been negative. There are at least two reasons for this. First, given significant fixed costs of capital investment and of hiring and firing of workers, it is likely that firms initially adjusted capacity utilization rates (see Capacity utilization, below) and workers' hours in response to lower demand. Second, as workers who lost their jobs moved into informal activity and back to agricultural work, this would have registered in declining aggregate productivity. Estimates of TFP growth by APO (2004) confirm that TFP growth collapsed during the crisis, but also suggest that TFP growth has since reverted to earlier trends (Figure 1.4.4). Taken at face value, these estimates imply that slower technical progress is an unlikely cause of the deceleration of growth.

Fixed capital

If changes in labor force growth, human capital accumulation, and TFP growth cannot easily account for the observed deceleration of output growth, it follows that capital accumulation has slowed. Slower capital accumulation requires either a lower ratio of fixed investment to output, or a decline in capital productivity, or both.

Figure 1.4.5 plots fixed investment rates. It is clear that investment rates declined steeply in the wake of the crisis. This experience fits with a much broader international pattern in which growth decelerations have been tied with declining investment ratios (Hausmann, Rodrik, and Velasco 2005; Rogers 2003; Attanasio, Picci, and Scorcu 2000). Having fallen, investment ratios have been broadly flat, showing little inclination to return to the levels seen in the precrisis period. In fact, declining capital productivity in Indonesia, Korea, Malaysia, and Thailand would require higher investment rates to deliver the same growth. Only in the Philippines might rising capital productivity have allowed investment rates to come down without pinching growth.

Summary

In accounting for lower growth in the postcrisis period, it is possible that demographic factors and changes in employment growth play a role, but a minor one and one that is differentiated by country. Changes in the rate of accumulation of human capital or in total factor productivity seem unlikely explanations. In all countries, investment rates have fallen and, except in the Philippines, impacts on growth are unlikely to have been compensated by higher capital productivity. In the next section, possible explanations for the fall in the fixed investment rate are canvassed.