S INCE the development of formal macroeconomic models for open economies, most of the relevant literature has been characterized by a dichotomy in the specification of the economy's supply of real output: either a simple Keynesian case of perfectly elastic supply, or an inelastic output supply specified.' Only recently attempts have been made to study the intermediate, and plausible, case in which changes in demand for a country's output lead both to a price and output response. The key element in such attempts the explicit incorporation of factor markets and of a Phillips-curve type of analysis into the macroeconomic model of the open economy.2 The importance of explicitly studying output supply in the open economy primarily lies in that many well-known results on the effectiveness of different policies, as well as the economy's dynamic adjustment to internal and external shocks, critically depend on the specification adopted. For example, Salop (1974) has shown that the incorporation of a classical labor supply function, one that specifies a positive relationship between labor supply and real wages, into an otherwise standard macro model of the open economy leads to the prediction that a successful devaluation causes a fall in output and employment. As noted by Salop, this result contradicts the conclusion of more conventional models (see, for example, Meade (1951) and Tsiang (1961)) that devaluation increases output and employment. Similarly, Casas (1975) has recently studied the case of an economy operating under floating exchange rates and perfect capital mobility. He finds that to the extent that wages are fully indexed to the price level, changes in the domestic money supply will not affect the level of output, but fiscal expansion will raise real output. Such a result is a complete reversal of the well-known theorem (see Mundell (1968) and Fleming (1962)) that only monetary changes affect employment under flexible exchange rates and perfect capital mobility (Casas (1975, p. 697); parentheses are mine).3 These considerations suggest that the aggregate output-supply function has a key role in the macroeconomics of the open economy. Yet no empirical estimates of this function, comparable to estimated import and export functions, for example, are available.4 The main purpose of this study to derive and estimate simple specifications of output supply in the open economy. In particular, I investigate below the responsiveness of output supply to actual and unanticipated changes in domestic and import prices for the seven main industrialized countries (Canada, France, Germany, Italy, Japan, the United Kingdom, and the United States), based on annual data from 1955 to 1975. Rather than attempting to obtain the best possible results for each country, the main emphasis of this study has been on getting comparable results for the countries considered.5 The plan of this paper as follows. Section II Received for publication September 29, 1978. Revision accepted for publication December 17, 1979. * Tel-Aviv University and Boston University. I would like to thank R. Barro, M. Blejer, K. Clements, A. Stockman, seminar participants at the University of Rochester, and two anonymous referees of this REVIEW for helpful comments on an earlier draft. Any errors are mine. ' For studies embodying the simple Keynesian case of perfectly elastic output supply; see, for example, Meade (1951), Tsiang (1961), Mundell (1968), and Fleming (1962). On the other hand, basic versions of the monetary approach to the balance of payments have usually assumed the inelasticoutput-supply specification; see, for example, Frenkel and Johnson (1976, ch. 6 and part II). 2 See, for example, Branson (1975, 1976), Dornbusch and Krugman (1976), Kenen (1978), Kingston and Turnovsky (1978), Laidler (1976), Leiderman (1978), and Salop (1974). 3 For analysis along these lines using a model that includes non-traded goods see Makin (1978). 4 For surveys of econometric work in international trade see Leamer and Stern (1970), and Magee (1975). Output-supply elasticities, in the context of open-economy models, have recently been estimated by Clements (1978) for the United States, and by Leiderman (1979) for Italy. On the estimation of output reduced-forms for several of the countries included in our sample, see Brunner and Meltzer (1978), and Stockman (1978). 5 Lucas (1973) and Taylor (1979) have empirically examined output determination by applying a given model across different countries. These studies, however, have abstracted from explicit open-economy considerations.
Read full abstract