The policy assignment principle with wage indexation *.

Chang, Wen-ya ; Lai, Ching-chong

1. Introduction

Originating from the seminar contributions by Mundell (1962, 1964) and Swan (1963), much of the discussion on stabilization policies achieving simultaneously internal and external balances under fixed exchange rates has centered around the so-called assignment problem. (1,2) Recently, Frenkel (1986, pp. 615-6) claims that an increased degree of capital-market integration may simplify, rather than complicate, the solution to the "policy assignment problem." In accordance with Frenkel's argument, Ramirez (1988) shows that in the general case of imperfect capital mobility, the signs of relevant policy multipliers are ambiguous and the assignment problem cannot be resolved. When capital mobility is perfect, the ambiguity vanishes and it is possible to choose an appropriate mix between fiscal and exchange rate policies to achieve given levels of desired output and balance of payments. However, Lai, Chang, and Chu (1990) demonstrate that, with perfect capital mobility and fixed exchange rates, the proper coordinat ion between fiscal and exchange rate policies suggested by Ramirez (1988) is not feasible. They propose that the government should give up the use of its exchange rate as a policy instrument and instead introduce monetary policy associated with fiscal policy to achieve desirable internal and external targets. (3) A common feature of the existing contributions on the assignment problem is that they ignore the role of the supply side, in particular the role of wage flexibility in the labor market.

In practice, Sachs (1980) and Bean (1988) provide the fact that real wage rigidity (full wage indexation) is a prevalent feature of many western countries. (4,5) As claimed in Pitchford (1990, p. 157), "In Australia the practice of partially indexing wages has become an art form." In addition, Miller and VanHoose (1998, pp. 339--40) provide further evidence that in the aggregate, nominal wages have been only partially indexed to the inflation rate in the U.S. In the literature, Sachs (1980), Ahtiala (1989), Devereux and Purvis (1990), Pitchford (1990), Chang and Lai (1994), and Lawler (1996) explore the role of wage indexation rules on the robustness of policy effectiveness in the Mundell (1963) model. Missing from these studies is an analysis of alternative wage indexation schemes in the assignment problem.

In view of the practice in the real world and theoretical interest, this paper tries to highlight the role of wage indexation schemes in the issue of policy assignment. The relevant problem we attempt to deal with is: Under fixed exchange rates with perfect capital mobility, what is the appropriate coordination between fiscal and exchange rate policies? By introducing a completely specified supply function characterized by alternative wage indexation rules into Mundell's (1963) framework, we show that the policy mix of fiscal and exchange rate policies can successfully work if the government undertakes a wage indexation scheme. Specifically, if there is zero wage indexation (rigid money wage), a mixture of fiscal and exchange rate policies is not feasible to achieve simultaneously internal and external balances. Alternatively, when there is partial wage indexation or full wage indexation (real wage rigidity), an appropriate mix between fiscal and exchange rate policies can attain given levels of desired outp ut and official foreign reserves.

The remainder of the paper is organized as follows. The theoretical structure of the model embodying alternative wage indexation rules is outlined in section 2. Section 3 re-examines the well-known policy assignment problem. Section 4 explores the linkage between policy assignment and wage indexation rules under a managed floating regime. Finally, the main findings of our analysis are presented in section 5.

2. The Model

The theoretical structure can be viewed as a variant of Pitchford (1990). The open economy under investigation operates under fixed exchange rates, and is assumed to be small in the sense that it cannot affect the foreign price level and interest rate. Domestic production is limited to a single final commodity, which is partly consumed domestically and partly exported. Domestic consumers have access to both domestic goods and imported goods. These goods are regarded by domestic residents as imperfect substitutes. Moreover, assume that there is a single traded bond, with the domestic bond market perfectly integrated with that in the rest of the world. In an environment of perfect capital mobility and fixed exchange rates, the authorities cannot sterilize any balance-of-payments surplus or deficit. (6) Accordingly, the economy can be characterized by the following macroeconomic relationships:

y = C(y) + I(r) + G + B(y, q), (1)

(D + R)/p = L(y, r), (2)

r = [r.sup.*] (3)

y = S(p, e, [p.sup.*]), (4)

where y = domestic output, C = consumption expenditure, I = investment expenditure, r = domestic interest rate, G = government expenditure, B = balance of trade, D = domestic credit, R = foreign exchange reserves, p = domestic price level, L = real money demand, [r.sup.*] = foreign interest rate, e = exchange rate defined to be the price of foreign exchange in terms of domestic currency, [p.sup.*] = foreign currency prices of imports, q = [ep.sup.*]/p = terms of trade, and S = aggregate supply function. Using subscripts with the relevant variables to indicate partial derivatives, as is customary, we impose the following restrictions on the behavior function: 1 > [C.sub.y] > 0, [I.sub.r] < 0, [B.sub.y] < 0, [B.sub.q] > 0, (7) [L.sub.y] > 0, [L.sub.r] < 0.

Equations (1) and (2) are the IS equation and the LM relation, respectively. The interest rate parity is given in equation (3). Under fixed exchange rates with perfect capital mobility, this implies that the domestic interest rate is equal to the foreign interest rate. (8) Equation (4) is the aggregate supply function. Because the aggregate supply function embodying alternative forms of the wage indexation scheme will play a crucial role in assessing the assignment problem, we now turn to derive this function in detail.

Define labor employed as N, and the aggregate short-run production function as:

y = y(N), (5)

where [y.sub.N] > 0 and [y.sub.NN] < 0. Demand for labor can be derived from equation (5) by setting the marginal revenue product of labor equal to the nominal wages w:

p[y.sub.N] = w.

The demand for labor [N.sup.d] is therefore a decreasing function of real wages in terms of the domestic price:

[N.sup.d] = f(w/p), (6)

with f' = df/d(w/p) = l/[y.sub.NN] < 0.

On the other hand, in line with Salop (1974) and Pitchford (1990), we assume that the supply of labor [N.sup.s] is an increasing function of real wages for which the relevant price deflator is the general price level g. Hence,

[N.sup.s] = h(w/g), (7)

with h' = dh/d(w/g) > 0, where g = [alpha]e[p.sup.*] + (1 - [alpha])p and [alpha] = the fraction of expenditure spent on imports.

Following Sachs (1980) and Pitchford (1990), the minimum money wages w are assumed to be indexed to the general price level:

dw/w = k(dg/g); 0 [less than or equal to] k [less than or equal to] 1. (8)

In the case of k = 0 there are sticky wages, whereas in the case of k = 1 the wages are fully indexed. Specifically, when wages are automatically indexed to the general price level, the nominal wages at once jump with an adjustment of g. In contrast, when wages are not automatically indexed to the general price level, the nominal wages remain intact with a jump in g.

It is assumed here that the economy has not yet reached full employment such that the wages are set at the level of minimum money wages and the employment is determined by labor demand, i.e.,

w = w, and N = f(w/p). (9)

From equations (5), (8), and (9), we can derive an aggregate supply function:

y = S(p, e, [p.sup.*]), (4)

with

[S.sub.p] = -f'[y.sub.N]w[(1 - k) + k[alpha]] > 0, (4a)

[S.sub.e] = [S.sub.[p.sup.*]] = f'[y.sub.N]wk[alpha] [less than or equal to] 0. (4b)

Without loss of generality, it is assumed that initially p = e = [p.sup.*] = 1 (and hence g = 1) throughout this paper.

3. The Assignment Problem

Under the fixed exchange regime, equations (1) - (4) can be simultaneously solved to determine y, r, R, and p in terms of policy instruments G and e. The solutions of y and R are given by:

y = H(G, e), (10)

R = K(G, e), (11)

where [H.sub.1] = [S.sub.p]/[DELTA] > 0, [H.sub.2] = [B.sub.q]([S.sub.p] + [S.sub.e])/[DELTA] [greater than or equal to] 0, (9) [K.sub.1] = (D + R + [S.sub.p][L.sub.y])/[DELTA] > 0, [K.sub.2] = {(D + R)[[B.sub.q] - [S.sub.e](1 - [C.sub.y] - [B.sub.y])] + [L.sub.y][B.sub.q]([S.sub.p] + [S.sub.e])}/[DELTA] > 0, and [DELTA] = [S.sub.p](1 - [C.sub.y] - [B.sub.y]) + [B.sub.q] > 0.

It is clear from equations (10) and (11) that the effect of fiscal expansion on domestic output and foreign reserves is expansionary regardless of the degree of wage indexation. This outcome is conformable to Mundell's (1963) assertion. On the other hand, a devaluation of the domestic currency will lead to an accumulation of foreign reserves, but it may contribute a zero or positive impact on domestic output depending on the degree of wage indexation. Specifically, if there is full wage indexation (k = 1), a devaluation has zero effect on output; but a devaluation will lead to an expansion in output when there is zero or partial indexation (k = 0 or 0 < k < 1). These results are not identified in the existing literature.

We now turn to the assignment problem, which is the main focus of this paper. We assume that policymakers wish to achieve the two targets of a desirable output y and a desirable level of foreign reserves R, (10) by using two policy instruments (fiscal policy G and exchange-rate policy e).

Following Levin (1972) and Turnovsky (1977), a general adjustment rule of policy instruments is

G = [a.sub.11][H(G, e) - y] + [a.sub.12][K(G, e) - R], (12)

e = [a.sub.21][H(G, e) - y] + [a.sub.22][K(G, e) - R], (13)

where an overdot on the relevant variables denotes the rate of change with respect to time, and [a.sub.ij] are the speeds of adjustment of the policy tools. The necessary and sufficient condition for the system to be stable is that

[a.sub.11][H.sub.1] + [a.sub.12][K.sub.1] + [a.sub.21][H.sub.2] + [a.sub.22][K.sub.2] < 0, (14)

([a.sub.11][a.sub.22] - [a.sub.12][a.sub.21])([H.sub.1][K.sub.2] - [H.sub.2][K.sub.1]) > 0. (15)

After substituting the comparative-statics results reported in equations (10) and (11) into (15), we have

[H.sub.1][K.sub.2] - [H.sub.2][K.sub.1] = -(D + R)[S.sub.e]/[DELTA] [greater than or equal to] 0. (16)

Equation (16) with equation (4b) reveals that the degree of wage indexation is a key factor to determine whether pairing of government spending and exchange rate policies with targets is feasible. Specifically, if nominal wages are rigid (k = 0), one obtains [H.sub.1][K.sub.2] - [H.sub.2][K.sub.1] = 0 or equivalently

[FORMULA NOT REPRODUCIBLE IN ASCII] (17)

implying that they y = y and R = Rloci in a G--e space will either be parallel to or coincide with each other. As a result, the stability condition expressed in equation (15) is definitely violated, and it is impossible to assign the appropriate policy mix for achieving desirable domestic output and foreign reserves.

If there is partial or full wage indexation (0 < k [less than or equal to] 1), it can be easily inferred from equation (16) that

[H.sub.1][K.sub.2] - [H.sub.2][K.sub.1] > 0.

Equation (15) thus can be reduced to

[a.sub.11][a.sub.22] - [a.sub.12][a.sub21] > 0. (18)

Given the stability conditions reported in equations (14) and (18), the possible stable assignments of fiscal and exchange-rate policies turn out to be:

(i) [a.sub.11] < 0, [a.sub.21] > 0, [a.sub.12] = [a.sub.21] = 0; that is, the fiscal authority should decrease its spending whenever output exceeds its desired level, and the foreign exchange authority should appreciate the home currency whenever foreign reserves exceed their target level.

(ii) [a.sub.12] < 0, [a.sub.21] > 0, [a.sub.11] = [a.sub.22] = 0; this situation indicates that a contractionary fiscal policy should be adopted, as foreign reserves are above their target, and a currency devaluation should be adopted, as output is above its target.

(iii) [a.sub.11] < 0, [a.sub.21] > 0, [a.sub.11] = [a.sub.22] = 0; this situation implies that an expansion in G should be undertaken if R exceeds R and a currency appreciation should be carried out if y exceeds y.

Before ending our discussion, we should explain why the degree of wage indexation is a crucial factor to determine the feasibility of policy mix. This point can be made clearly through combining the money market equilibrium condition with the aggregate supply function. First, we transform equation (4) as

P = S(y, e, [p.sup.*]), [S.sub.y] = 1/[S.sub.p] > 0, [S.sub.e] = [S.sub.[p.sup.*]] = -[S.sub.e]/[S.sub.p] [greater than or equal to] 0. (19)

Substituting equation (19) into equation (2) with r = [r.sup.*], we then have

(D + R)/S[y, e, [p.sup.*]) = L(y, [r.sup.*]). (20)

When there is zero wage indexation ([S.sup.e] = [S.sup.[p.sup.*]] = 0), then equation (20) reduces to

(D + R)/S(y) = L(y, [r.sup.*]). (20a)

Obviously, if D is assumed exogenous, then the relationship between y and R in equation (20a) is mutually dependent on each other. How can one then possibly use the mixture of fiscal and exchange rate policies to achieve given targets of y and R, when these two desired targets y and k are not consistent with the resulting money market equilibrium condition? However, if there is some degree of wage indexation, then equation (20) will prevail. Under such a situation, we have an additional degree of freedom, that is, adjusting the exchange rate policy satisfies both targets and the resulting money market equilibrium condition.

4. A Managed Floating Regime

In the previous section our discussion focused on the framework of a pure fixed exchange rate regime. Recently, leading industrialized countries such as the G7 have adopted the system of a managed floating regime. They now periodically intervene in the foreign exchange markets to ensure stability of exchange rates [Daniels and VanHoose (1999, pp. 87-9)]. An interesting question naturally arises: Are the implications in the previous section robust enough to stand on their own if the analysis shifts to the system of a managed floating regime?"

Under a regime of managed floating rates, the model will be composed of equations (1)-(4) and the following behavior of foreign exchange intervention. In order to stabilize its nation's currency, the foreign exchange authority adopts a rule of leaning against the wind. According to the intervention policy of leaning against the wind in the spot market, the foreign exchange authority sells (buys) foreign reserves whenever the domestic currency tends to depreciate (appreciate) sharply. This implies

R - [R.sub.0] = -[xi](e - [e.sub.*]), [xi] > 0; (21)

where [R.sub.0] = the initial foreign exchange reserves, [xi] = degree of intervention, and [e.sub.*] = the pre-announced publicly-known target exchange rate that the foreign exchange authority attempts to defend, exogenously determined. One point should be mentioned here. If [xi] [right arrow] [infinity], then equation (21) reduces to e = [e.sub.*]; that is, the authority is determined to intervene in the foreign exchange market to maintain the officially-announced exchange rate. Equations (1)-(4) and (21) with [xi] [right arrow] [infinity] constitute a pure fixed exchange rate model in the previous section. In reality, [xi] is within 0 and [infinity] and corresponds to a managed float.

Following the same solution procedure stated in the previous section, equations (1)-(4) and (21) can be simultaneously solved to determine y, r, R, p, and e. The solutions of y and R are stated as follows:

y = H(G, [e.sub.*], D), (22)

R = K(G, [e.sub.*], D), (23)

where [H.sub.1] = [[xi][S.sub.p] - (D + R)[S.sub.e]]/[ DELTA] > 0, [H.sub.2] = [xi][B.sub.q]([S.sub.p] + [S.sub.e])/[DELTA] [greater than or equal to] 0, [H.sub.3] = [B.sub.q]([S.sub.p] + [S.sub.e])/[DELTA] [greater than or equal to] 0, [K.sub.1] = [xi](D + R + [S.sub.p][L.sub.y])/[DELTA] > 0, [K.sub.2] = [xi]{(D + R)[[B.sub.q] - [S.sub.e](1 - [C.sub.y] - [B.sub.y])] + [L.sub.y][B.sub.q]([S.sub.p] + [S.sub.e])}/[DELTA] > 0, [K.sub.3] = -[xi][[B.sub.q] + [S.sub.p](1 - [C.sub.y] - [B.sub.y])]/[DELTA] < 0, and [DELTA] = [xi][[S.sub.p](1 - [C.sub.y] - [B.sub.y]) + [B.sub.q]] + [L.sub.y][B.sub.q]([S.sub.p] + [S.sub.e]) + (D + R)[[B.sub.q] - [S.sub.e](1 - [C.sub.y] - [B.sub.y])] > 0.

From equations (22) and (23), the effects of fiscal expansion and currency devaluation on output and foreign reserves are qualitatively similar to equations (10) and (11) and are identical to the outcomes of equations (10) and (11) if [xi] [right arrow] [infinity]. However, domestic credit expansion will lead to a decrease in foreign reserves, but may contribute a zero or positive effect on domestic output depending on the degree of wage indexation. Specifically, if there is full wage indexation (k = 1), then monetary expansion has zero effect on output; but monetary expansion leads to an increase in output when there is zero or partial indexation (k = 0 or 0 < k < 1).

Two points should be noted here. First, if [xi] [right arrow] [infinity], then domestic credit expansion impotently affects output and leads to an equal decrease in foreign reserves ([H.sub.3] = 0 and [K.sub.3] = -1). This is the Mundell (1963) proposition under fixed exchange rates. Second, if there is real wage rigidity (k = 1), then the effect of domestic credit expansion on output under a managed float is the same as that under a pure fixed regime.

We now turn to the assignment problem. As is done in the fixed exchange regime, assume that policymakers wish to achieve the two targets of a desirable output y and a desirable level of foreign reserves R, by adjusting two policy instruments (fiscal policy G and exchange-rate policy [e.sub.*]). The general policy rule for achieving y and R is thus

G = [a.sub.11][H(G, [e.sub.*], D) - y] + [a.sub.12][K(G, [e.sub.*], D) - R], (24)

[e.sub.*] = [a.sub.21][H(G, [e.sub.*], D) - y] + [a.sub.22][K(G, [e.sub.*], D) - R]. (25)

The stability condition requires that

[a.sub.11][H.sub.1] + [a.sub.12][K.sub.1] + [a.sub.21][H.sub.2] + [a.sub.22][K.sub.2] < 0, (26)

([a.sub.11][a.sub.22] - [a.sub.12][a.sub.21])([H.sub.1][K.sub.2] - [H.sub.2][K.sub.1]) > 0. (27)

Substituting the comparative-statics results reported in equations (22) and (23) into (27), we obtain

[H.sub.1][K.sub.2] - [H.sub.2][K.sub.1] = -[xi](D + R)[S.sub.e]/[DELTA] [greater than or equal to] 0. (28)

With [S.sub.e] = f'[y.sub.N]wk[alpha] in equation (4b), equation (28) also shows that the key factor determining a feasible assignment rule of G and [e.sub.*] under the system of managed floating rates is the degree of wage indexation. A comparison of equation (28) with (16) clearly indicates that the policy assignment principle under a managed floating regime is equivalent to that under a pure fixed regime. The implications in the previous section are thus robust enough to sustain changes in the system of managed floating rates.

We finally discuss the usual assignment principle in the existing literature where policymakers wish to achieve the two targets of y and R, by adjusting two policy instruments (fiscal policy G and domestic credit policy D). In this case, the general policy rule for achieving y and R is thus composed of equation (24) and

D = [a.sub.21][H(G, [e.sub.*], D) - y] + [a.sub.22][K(G, [e.sub.*], D) - R]. (25a)

It follows from equations (24) and (25a) that the stability condition requires

[a.sub.11][H.sub.1] + [a.sub.12][K.sub.1] + [a.sub.21][H.sub.3] + [a.sub.22][K.sub.3] < 0, (26a)

([a.sub.11][a.sub.22] - [a.sub.12][a.sub.21])([H.sub.1][K.sub.3] - [H.sub.3][K.sub.1]) > 0. (27a)

Substituting the comparative-statics results reported in equations (22) and (23) into (27a) gives

[H.sub.1][K.sub.3] - [H.sub.3][K.sub.1] = -[xi][S.sub.p]/[DELTA] < 0. (28a)

As a consequence, equation (27a) reduces to

[a.sub.11][a.sub.22] - [a.sub.12][a.sub.21] < 0. (29)

Given the stability condition reported in equations (26a) and (29), stable assignments of fiscal and domestic credit policies turn out to be:

(i) [a.sub.11] < 0, [a.sub.22] > 0, [a.sub.12] = [a.sub.21] = 0; that is, the fiscal authority should decrease its spending whenever output exceeds its desired level, and the monetary authority should increase domestic credit whenever foreign reserves exceed their target level.

(ii) [a.sub.12] < 0, [a.sub.21], < 0, [a.sub.11] = [a.sub.22] = 0; this situation indicates that a contractionary fiscal policy should be adopted, as foreign reserves are above their target, and a contractionary monetary policy should be adopted, as output is above its target.

If [xi] [right arrow] [infinity], then equation (28a) degenerates to

[H.sub.1][K.sub.3] - [H.sub.3][K.sub.1] = -[S.sub.p]/[DELTA] < 0, (30)

where [H.sub.1] = [S.sub.p]/[DELTA] = [H.sub.1] > 0, [H.sub.3] = 0, [K.sub.1] = (D + R + [S.sub.p][L.sub.y]/[DELTA] = [K.sub.1] > 0, and [K.sub.3] = -1. A comparison of equation (30) with (28a) clearly reveals that the fixed exchange rate regime and a managed floating regime will lead to identical policy assignment principles. In addition, even if we consider an explicit specification of the aggregate supply function, the above results regarding stable assignments of fiscal and monetary policies under alternative exchange rates are conformable to Case 1 of Lai, Chang, and Chu (1990, p. 819).

5. Concluding Remarks

It is well known in the literature of the policy assignment problem that, under fixed exchange rates with perfect capital mobility, fiscal and exchange rate policies cannot be relied upon to achieve desirable targets of output and official reserves. This paper reexamines this assignment problem by explicitly introducing alternative wage indexation schemes proposed by Sachs (1980) and Pitchford (1990). It is found that a mixture of fiscal and exchange rate policies can indeed be employed to stabilize desirable internal and external targets when there is some degree of wage indexation or full indexation. However, if there is zero wage indexation (fixed money wages), a mixture of fiscal and exchange rate policies is not feasible. Furthermore, we also show that the results under fixed exchange rates are robust when the analysis shifts to the system of a managed floating regime. Consequently, our conclusion suggests that in a very open economy with international mobility of capital, a wage indexation scheme may b e a potential vehicle to successfully coordinate macroeconomic policies with desirable targets.

* We are grateful for the constructive comments and suggestions from an anonymous referee. Any errors and/or omissions remain our responsibility.

Notes

(1.) In their popular textbook, Caves, Frankel, and Jones (1996, pp. 542-6) clearly illustrate the original contribution of Mundell's (1962) analysis.

(2.) The literature of the policy assignment problem includes Ott and Ott (1968), Levin (1972), Nyberg and Viotti (1976)(1979), Turnovsky (1977), Kenen (1985), Boughton (1989), Jha (1994, Ch. 6), and Miller and VanHoose (1998, Ch. 12), among others.

(3.) Most studies in the literature of the assignment principle focus on how fiscal and monetary policies are utilized successfully to attain both internal and external balances. Few have devoted effort to discussing the pairing of fiscal and exchange rate policies to targets. To our knowledge, Swan (1963) is a pioneering contribution [see Kenen (1985, pp. 649-54) and Caves, Frankel, and Jones (1996, pp. 397-403)].

(4.) For an empirical description of wage indexation, see Emerson (1983).

(5.) Agenor (1996) provides an excellent survey and a lot of evidence on the role of the labor market in the transmission process of stabilization policies for developing countries.

(6.) As under perfect capital mobility, it would be impossible for the monetary authorities to sterilize indefinitely; otherwise, the resultant loss of foreign reserves would be infinite. For a detailed explanation, see, for example, Swoboda (1972) and Lai, Chang, and Chu (1990).

(7.) Condition [B.sub.q] > 0 indicates that the Marshall-Lerner condition is satisfied.

(8.) The return of foreign bonds holdings should be [r.sup.*] + ([e.sup.E] - e)/e, where [e.sup.E] is an expected future exchange rate and ([e.sup.E] - e)/e denotes the expected rate of changes in the exchange rate. Mundell (1963) assumes that static expectations prevail, i.e., ([e.sup.E] - e)/e = 0. Hence, the return of foreign bonds holdings is [r.sup.*].

(9.) From (4a) and (4b), we have [S.sub.p] + [S.sub.e] = -f'[y.sub.N]w(1 - k)[greater than or equal to] 0, as k [less than or equal to] 1.

(10.) Since the balance of payments is always in equilibrium in a world of perfect capital mobility, it no longer appears as an endogenous variable and cannot be treated as a target variable. The level of official foreign reserves thus takes its place.

(11.) The experiment in this section was suggested by an anonymous referee, to whom we are grateful.

References

Agenor, Pierre-Richard, (1996). "The Labor Market and Economic Adjustment," IMF Staff Papers 43, pp. 261-335.

Ahtiala, Pekka, (1989). "A Note on Fiscal Policy under Flexible Exchange Rates," European Economic Review 33, pp. 1481-86.

Bean, Charles R., (1988). "Real Wage Rigidity and the Effect of an Oil Discovery," Oxford Economic Papers 40, pp. 451-62.

Boughton, James (1989). "Policy Assignment Strategies with Somewhat Flexible Exchange Rates," in: M. Miller, B. Eichengreen, and R. Portes, eds., Blueprints for Exchange-Rate Management, London: Academic Press.

Caves, Richard E., Jeffrey A. Frankel, and Ronald W. Jones, (1996). World Trade and Payments: An Introduction, 7th Ed., New York, NY: Harper Collins.

Chang, Wen-ya and Ching-chong Lai, (1994). "Anticipated Shocks and the Acceleration Hypothesis: The Implication of Wage Indexation," Southern Economic Journal 61, pp. 453-64.

Daniels, Joseph P. and David D. VanHoose, (1999). International Monetary and Financial Economics, Cincinnati, OH: South-Western College Publishing.

Devereux, Michael B. and Douglas D. Purvis, (1990). "Fiscal Policy and the Real Exchange Rate," European Economic Review 34, pp. 1201-11.

Emerson, Michael, (1983) "A Review of Current European Indexation Experience," in: R. Dombusch and M. H. Simonsen, eds., Inflation, Debt, and Indexation, pp. 160-79, Cambridge, MA: MIT Press.

Frenkel, Jacob A., (1986). "International Interdependence and the Constraints on Macroeconomic Policies," Weltwirtschaftliches Archiv 122, pp. 615-45.

Jha, Raghbendra, (1994). Macroeconomics for Developing Countries, London: Routledge.

Kenen, Peter B., (1985). "Macroeconomic Theory and Policy: How the Closed Economy Was Opened," in: R. W. Jones and P. B. Kenen, eds., Handbook of International Economics, Vol. II, pp. 625-77, Amsterdam: North-Holland.

Lai, Ching-chong, Wen-ya Chang, and Yun-peng Chu, (1990). "The Composition of Government Spending and the Assignment of Instruments to Targets in a Small Open Economy: Comment," Southern Economic Journal 56, pp. 815-21.

Lawler, Phillip, (1996). "Real Wage Rigidity and Anticipated Fiscal Policy," Journal of Economic Studies 23, pp. 3-17.

Levin, Jay H., (1972). "International Capital Mobility and the Assignment Problem," Oxford Economic Papers 24, pp. 54-67.

Miller, Roger L. and David D. VanHoose, (1998). Macroeconomics: Theories, Policies, and International Applications, Cincinnati, OH: South-Western College Publishing.

Mundell, Robert A., (1962). "The Appropriate Use of Monetary and Fiscal Policy for Internal and External Stability," IMF Staff Papers 9, pp. 70-9.

Mundell, Robert A., (1963). "Capital Mobility and Stabilization Policy under Fixed and Flexible Exchange Rates," Canadian Journal of Economics and Political Science 29, pp. 475-85.

Mundell, Robert A., (1964). "A Reply: Capital Mobility and Size," Canadian Journal of Economics and Political Science 30, pp. 421-31.

Nyberg, Lars and Staffan Viotti, (1976). "The Balance of Payments and Optimal Economic Policy: The Assignment Problem Revisited," Journal of International Economics 6, pp. 337-46.

Nyberg, Lars and Staffan Viotti, (1979). "Limited Instrument Flexibility and the Assignment Principle," Oxford Economic Papers 31, pp. 177-86.

Ott, David J. and Attiat F. Ott, (1968). "Monetary and Fiscal Policy: Goals and the Choice of Instruments," Quarterly Journal of Economics 82, pp. 313-25.

Pitchford, John D., (1990). "Macroeconomic Policy and the Current Account in an Open Economy with Partial Wage Indexation," Economic Record 66, pp. 157-70.

Ramirez, Miguel D., (1988). "The Composition of Government Spending and the Assignment of Instruments to Targets in a Small Open Economy," Southern Economic Journal 55, pp. 432-42.

Sachs, Jeffrey D., (1980). "Wages, Flexible Exchange Rates, and Macroeconomic Policy," Quarterly Journal of Economics 94, pp. 731-47.

Salop, Joanne, (1974). "Devaluation and the Balance of Trade under Flexible Wages," in: G. Horwich and P. A. Samuelson, eds., Trade, Stability, and Macroeconomics, pp. 129-51, New York, NY: Academic Press.

Swan, T. W., (1963). "Longer-Run Problems of the Balance of Payments," in: H. W. Arndt and W M. Corden, eds., The Australian Economy: A Volume of Readings, pp. 344-95, Melbourne: Cheshire Press.

Swoboda, Alexander K., (1972). "Equilibrium, Quasi-Equilibrium, and Macroeconomic Policy under Fixed Exchange Rates," Quarterly Journal of Economics 86, pp. 162-71.

Turnovsky, Stephen J., (1977). Macroeconomic Analysis and Stabilization Policy, London: Cambridge University Press.

Wen-ya Chang ** and Ching-chong Lai ***

** Professor, Department and Graduate Institute of Economics, Fu-Jen Catholic University, and Institute of Economics, National Sun Yat-Sen University, Taiwan.

*** Research Fellow, Sun Yat-Sen Institute for Social Science and Philosophy and Institute of Economics, Academia Sinica, and Professor, Department of Economics, National Taiwan University, Taiwan.