The global financial crisis which began in east Asia in 1997 is not over, neither is the inquest into its implications for adjustment policy. In the wake of this crisis, we focus here on the role of capital controls, which formed a much publicised part of the crisis-coping strategy in one country (Malaysia) and, less openly, were also deployed by other crisis-afflicted countries. Evaluation so far has examined different target variables with different estimation methods, generally concentrating on efficiency and stability indicators and ignoring equity measures; it has also typically treated `control´ as a one-zero dummy variable, ignoring the `quality´ of intervention and in particular the extent to which efficiency gains are obtained in exchange for controls. Partly because of these limitations, the literature has reached no consensus on the impact of controls, nor therefore about where they fit within the set of post-crisis defence mechanisms. We propose an approach in which the government plays off short-term political security against long-term economic gain; the more insecure its political footing, the greater the weight it gives to political survival, which is likely to increase the probability of controls being imposed. The modelling of this approach generates a governmental `policy reaction function´ and an impact function for controls, which are estimated by simultaneous panel-data methods across a sample of thirty developing and transitional countries between 1980-2003, using, for the period since 1996, the `new´ IMF dataset which differentiates between controls by type. We find that controls appear to cause increases in income equality, and are significantly associated with political insecurity and relatively low levels of openness to trade. They do not, in our analysis, materially influence the level of whole-economy productivity or GDP across the sample of countries examined, although they do influence productivity in particular sectors, in particular manufacturing. But the dispersion around this central finding is wide: the tendency for controls to depress productivity by encouraging rent-seeking sometimes is, and sometimes is not, counteracted by purposive government policy actions to maintain competitiveness. Whether or not this happens – whether, as we put it, controls are `smart´, and the manner in which they are smartened - is vital, on both efficiency and equity grounds. We devise a formula for, and make the case for capital controls which are time-limited, and contain an inbuilt incentive to increased productivity, as a means of improving the sustainability and equity of the adjustment process whilst keeping to a minimum the cost in terms of productive efficiency.
