پویایی بیکاری با تحرک سرمایه بین المللی

We study the response of domestic unemployment rates to shocks in total factor productivity for economies with high capital mobility and low labour mobility. We show that high capital mobility amplifies the impact on the domestic unemployment rate of domestic fluctuations in total factor productivity, shortens the lag of the response to shocks and raises the variability of unemployment. But average unemployment is unaffected. Capital flows increase the riskiness of labour income and reduce the riskiness of capital income but do not reduce mean welfare.

One of the most striking recent changes in the world economy is the speed with which the capital markets of industrial countries have become integrated. Table 1 gives sample means for a measure of the penetration of foreign capital in OECD economies, the inflow of foreign direct investment as a fraction of total domestic investment. This measure is likely to understate the extent to which foreign capital penetrates domestic labour markets, as it includes only foreign capital that remains under foreign control. Yet, the striking feature of the data is the big increase in international capital mobility after the mid 1980s, which is about four times as large as the increase in trade flows. The increase is world-wide, although it is on average bigger in European countries, following the single market process that started in 1986 and culminated in 1992. The rise is much bigger in small economies than in the G7.1 Concurrently with this increase in international capital flows there has been an increase in both the average rate of unemployment and the standard deviation of its cyclical component (henceforth, volatility). Table 2 gives data for unemployment in OECD countries. On average unemployment increased by 60% between 1970–1985 and 1986–1998 and volatility increased by about 50%. Moreover, as in the case of FDI flows, the increase in both the average rate and volatility is larger in smaller economies. Is there any connection between the rise in international capital mobility and the higher and more volatile unemployment? Usually, a big foreign investment such as the relocation of an industrial plant from one country to another is greeted by the receiving country as good news and by the losing country as bad news. Is this reaction justified? How does international capital mobility affect the job creation and job destruction flows of the domestic economy? How does it affect the equilibrium unemployment rate and its volatility? These are the questions addressed in this paper. The paper focuses on theory, with some quantitative examples to illustrate the strength of the identified links.2 The paper shows first how a time series for unemployment and the capital stock can be jointly derived from an equilibrium model. In this respect the paper borrows from the model of Bean and Pissarides (1993), although the wage setting mechanism is new. The paper's second and main objective is to introduce international capital mobility and show that more international capital mobility affects the dynamics of unemployment, although not, generally, the mean level of unemployment. We show that as more international capital mobility takes place, unemployment responds faster and with more amplitude to shocks to total factor productivity, so over long periods of time both unemployment and workers’ incomes are more volatile than in an economy without international capital mobility. The intuition behind the results is simple. We explain it here for a one-off negative shock that lasts for ever, although extension to other shocks is straight-forward. In the absence of international capital mobility the capital stock of a country is historically given (a state variable). When its economy is hit by a negative TFP shock, it reduces its demand for labour, but the capital stock initially remains at the historical high level. Its demand for labour does not fall very much initially because the high level of the capital stock cushions it. But if the negative shock persists the capital stock falls gradually, as firms find it unprofitable to replace depreciating equipment. So eventually, and if the shock persists long enough, the capital stock falls to a new low level and the demand for labour gradually falls with it. Suppose now there is perfect international capital mobility and the negative shock that hits the country is not world-wide. The gradual adjustment of the capital stock in the case of the closed economy is now replaced by faster adjustment because now capital leaves the country in the pursuit of higher rates of return elsewhere. The cushion to labour demand provided by the historically high capital stock disappears and so the demand for labour falls faster than in the closed economy. But in the long run the fall in labour demand is the same in the two economies because the new level of the capital stock reached quickly in the case of perfect capital mobility is the same as the level reached gradually in the case on no mobility. The implications of our analysis are that when there is international capital mobility employment and unemployment are more volatile but on average they are the same as with no capital mobility. Consequently, international capital mobility makes workers’ incomes and jobs less secure and capitalists’ incomes more secure. We do not pursue the analysis of welfare-improving policy but an obvious response is that there should be more protection of workers’ incomes against shocks. We show, however, that for any given level of tax-financed unemployment compensation international capital mobility improves the welfare of the representative consumer. The intuitive reason is that in an economy with a lot of international capital mobility the response of employment and output to productivity shocks is less dependent on the historical capital stock. Section 2 is a formal statement of a life-cycle model of savings with labour market matching. Unemployment in this model is an equilibrium outcome of costly job creation. Section 3 studies the dynamics of employment and unemployment when there is no international capital mobility and Sections 4 and 5 study dynamics and welfare in small economies with perfect international capital mobility. Section 6 studies the large economy when TFP shocks are imperfectly correlated across countries. Section 7 provides numerical examples of how capital mobility affects the dynamic response of the unemployment rate to fluctuations in TFP. We conclude with a discussion of some testable implications of the model.

We have shown that international capital mobility can substantially amplify fluctuations in unemployment and output. Calibrations show that the variance of unemployment with perfect international capital mobility can be up to three times as large as the variance of unemployment without capital mobility, implying that cyclical peaks and troughs in unemployment overshoot those in economies without capital mobility by up to 1.7 percentage points of unemployment. Small OECD economies have experienced a rise in the variance of their unemployment rates of this order of magnitude sometime in the mid 1980s, which coincided with a fast rise in international capital mobility. Such effects imply that small economies trading in a world with large international capital flows need to devise ways to insure the income of workers whose wealth is poorly diversified across countries, because international capital flows tend to shift income risk from capital to labour. But it would be a mistake to achieve this security by restricting capital movements because on average capital mobility does not reduce welfare. Future work in this area needs to address this policy question within reasonably estimated or calibrated models. Estimation should test whether the dynamic properties of unemployment are different when there is international capital mobility from those without (see Vallanti, 2005, for some results). In particular, is adjustment to domestic shocks faster when the economy is small? Are foreign shocks transmitted to the domestic economy through capital flows when the economy is large? In the case of large economies, trade flows need also to be taken into account in empirical tests, because foreign shocks may influence trade prices and thereby be transmitted to the domestic economy. A higher foreign shock would then increase the domestic demand for labour, working against the effect of higher capital mobility. This trade channel should be less important in small economies as they are more likely to be faced with a perfectly elastic demand for their exports.