WELFARE AND MACROECONOMIC INTERDEPENDENCE [*].

We develop a baseline model of monetary and fiscal transmission in interdependent economies. The welfare effects of expansionary policies are related to monopolistic supply in production and monopoly power of a country in trade. An unanticipated exchange rate depreciation can be beggar-thyself rather than beggar-thy-neighbor, as gains in domestic output are offset by deteriorating terms of trade. Smaller and more open economies are more prone to suffer from inflationary shocks. Larger economies benefit from moderate demand-led expansions, but may he worse off if policy-makers attempt to close the output gap. Fiscal shocks are generally beggar-thy-neighbor in the long run; in the short run they raise domestic demand at given terms of trade, thus reducing the welfare benefits from monetary expansions. Analytical tractability makes our model uniquely suitable as a starting point to approach the recent "new open-economy macroeconomic" literature.

I. INTRODUCTION

How are policy shocks transmitted in open economies? How do they affect national residents' welfare? This paper is part of the ongoing research program aimed at building a modem theoretical paradigm for international policy evaluation--based on micro-founded models of imperfect competition with nominal rigidities--to address such questions. In the new literature, as in the traditional Mundell-Fleming-Dornbusch model, nominal shocks increase domestic output and employment and depreciate the real exchange rate. In contrast to the traditional approach, policy analysis based on the new models can rely on an index of social welfare that is logically consistent with the behavioral analysis of the international transmission mechanism.

The first generation of contributions to the new literature [1] has emphasized the welfare and policy implications of monopoly distortions in production, extending to an open-economy setting the key conclusions of influential closed-economy models such as Blanchard and Kiyotaki [1987] and Ball and Romer [1990]. [2] In such models wages and prices are suboptimally high, while output and consumption are suboptimally low. Given nominal rigidities, a small demand shock that raises output toward its efficient levels also improves national welfare. Consistently, in the benchmark open-economy model by Obstfeld and Rogoff [1995], a small unanticipated monetary expansion raises consumption and welfare everywhere in the world economy, no matter whether the shock originates at home or abroad. Only if one considers other internal inefficiencies in addition to monopolistic supply (such as nonlump-sum income taxes) do money shocks have asymmetric effects on domestic and foreign residents.

In open economies, however, there exists an economic distortion that is directly associated with openness, namely, a country's power to affect its terms of trade by influencing the supply of its products. The literature still lacks a comprehensive exploration of the interplay between internal and external sources of economic distortion (that is, monopolistic supply in production and monopoly power of a country in trade) that lies at the very core of the "new open-economy macroeconomics." This paper attempts to fill this gap by providing a baseline choice-theoretic analysis of monetary and fiscal transmission in interdependent economies.

When we account for both internal and external distortions, the analogies of policy analysis in open- and closed-economy models break down in several crucial dimensions. Consider, for instance, the effects of unanticipated money shocks. In both closed and open economies with nominal rigidities, monetary surprises raise output toward its efficient level. In open economies, however, they also reduce domestic consumers' purchasing power in the global markets. Because of the latter effect, expansionary policies can be beggar-thyself Smaller and more open economies are more likely to suffer from domestic nominal shocks that worsen their terms of trade. Larger economies, that could benefit from moderate output and employment gains, are actually worse off if policy-makers engineer large monetary expansions in an attempt to close the gap between actual and potential output.

Our analysis has rather unorthodox policy implications. In contrast with the popular model of competitive devaluations, exchange rate shocks are not beggar-thy-neighbor: due to deteriorating terms of trade, the benefits from a domestic monetary expansion accrue primarily to foreigners. Also in contrast with the conventional wisdom, domestic fiscal expansions hurt the trading partners: fiscal linkages are in general beggar-thy-neighbor in the long run. In the short run, fiscal shocks modify the trade-off between employment and domestic inflation by raising domestic demand at unchanged terms of trade, thus reducing the welfare benefits from a monetary expansion.

While complementing and expanding the findings of the "first-generation" new open-economy macroeconomics, these novel results do not come at the price of introducing additional technical difficulties: if anything, our setting can be viewed as a useful simplification of the models epitomized by Obstfeld and Rogoff [1996, Ch. 101. Our two-country, general-equilibrium model can be solved in closed form without imposing symmetry upon the economic structures of the two countries and without resorting to local log-linear approximations--a quantum leap over alternative specifications in terms of analytical tractability. While somewhat restrictive, a modeling strategy leading to a closed-form solution makes our analytical framework uniquely suitable to "inspect the mechanism" of international policy transmission and assess its welfare implications in both deterministic and stochastic settings. Because of this, our contribution provides a good starting point to approach the recent fast-growing micro-founded theory of macroeconomic interdependence.

The paper is structured as follows. Sections II introduces the model. Section III describes its structural form and solution in an application to the case of permanent monetary and fiscal shocks. Section IV studies the welfare effects of domestic monetary shocks, contrasting results in closed- and open-economy policy analysis. Sections V and VI delve into the analysis of monetary and fiscal interdependence. Section VII provides a synthesis of our results. Section VIII concludes.

II. THE MODEL

A. Preferences and Consumption Indexes

Our theoretical framework consists of a general equilibrium two-country model. Two key features of our framework, i.e., nominal rigidities and monopolistic competition, are standard in the new open-economy macroeconomic literature. The introduction of short-run nominal rigidities is motivated in terms of empirical plausibility and realism. Characterizing economywide distortions in terms of monopoly power is logically coherent with the assumption that output is demand-determined when prices are fixed. The main departure of our framework from the previous literature is a specification according to which the substitutability among nationally produced goods is higher than between domestic goods and foreign goods as a group. It is this very feature that allows us to examine the macroeconomic and welfare implications of terms of trade externalities.

The model includes two countries, Home and Foreign, each specialized in the production of a traded good. In each country there is a continuum of economic agents, with population size normalized to 1.

The lifetime utility of Home agent j [epsilon] [0, 1] is given by

(1) [U.sub.t](j)

= [E.sub.t] [[[sigma].sup.[infinity]].sub.[tau]=t] [[beta].sup.[tau]-t][[C.sub.[tau]][(j).sup.1-p/1 - [rho] + [chi] ln [M.sub.[tau]](j)/[P.sub.[tau]] + V([G.sub.[tau]]) - [kappa]/2 [l.sub.[tau]][(j).sup.2]]

[beta], [rho], [chi], [kappa] [greater than] 0.

Here [beta] is the discount rate, equal to 1/1(1 + [delta]), where [delta] [greater than] 0 is the rate of time preference, and 1/[rho] is the elasticity of intertemporal substitution. The consumption index for the Home agent, C, is defined as

(2) [C.sub.t](j) [equivalent] [([C.sub.H,t](j)).sup.[gamma]][([C.sub.F,t](j)).sup.1-[gamma]] 0 [less than] [gamma] [less than] 1,

where [C.sub.H](j) and [C.sub.F](j) are, respectively, consumption of the Home good and consumption of the Foreign good by individual j. Domestic real money holdings, M/P, provide liquidity services that enter the utility function. The function V is individual utility from public goods G, while l is the amount of labor supplied by the agent.

The utility of Foreign agent [j.sup.*]--foreign variables are indexed by asterisks--is similar to (1). Preferences over consumption goods are symmetric both within and across countries: the elasticity of substitution 1/[rho] and the rate of time preference [delta] of Foreign agents are identical to those of Home agents. The weight [gamma] in agent [j.sup.*]'s consumption index is the same as in (2):

(3) [[C.sup.*].sub.t]([j.sup.*]) [equivalent] [([[C.sup.*].sub.H,t]([j.sup.*])).sup.[gamma]][([[C.sup.*].sub.F,t]([ j.sup.*])).sup.1-[gamma]].

Yet, there are two notable differences between the two countries. First, the weight of the Home good in preferences can be different from the weight of the Foreign good ([gamma] [not equal to] 1/2). Second, …