Coordinating Monetary and Fiscal Policies in an Open Economy

Delegating monetary policy to an independent central bank will not incur coordination problems with fiscal policy. The fiscal authority faces no incentive to try and expand the economy, as its efforts will be quickly countered by a contractionary monetary policy to maintain the inflation target. We argue that this proposition is not true in an open economy. When the external sector is added to the analysis powerful incentives for a fiscal expansion may surface, with important implications for the trade balance. The exchange rate adds an important new dimension to the coordination story. Acknowledgements; The authors wish to thank The Leverhulme Trust for financial support for this research, under the heading “EMU and European Macroeconomic Policy in a Global Context” (F/08 519/A) 1 Management School, Imperial College London. (1) Introduction; The Background. One of the first acts of the Labour government elected in 1997 was to grant the Bank of England operational independence to set monetary policy in pursuit of a target inflation rate over a two year horizon. There are powerful arguments for granting full or –as in the case of the Bank of England instrument independence to a central bank, and these arguments have underpinned movements to reform many central banks around the world. Both theoretical and empirical questions have figured heavily in the literature that has grown up around this issue. The theoretical case for delegating monetary policy is based on an intellectually appealing argument first formalised by Kydland and Prescott (1977) and refined subsequently by Barro and Gordon (1983) concerning the time inconsistency of monetary policy. This pointed out that discretionary policy making leads to an inflation-bias because policy makers can spring an inflation surprise to secure higher short-run output and employment. Rational agents noting the apparent time inconsistency problem involved in low inflation announcements, use the policy maker’s (optimal) response when forming their expectations, hence produce the inflation bias. The subsequent, voluminous, literature has posed many solutions to this problem of inflation bias endemic in discretionary policy making. Although it is not our intention to review all these, some contributions are particularly relevant to this paper. One of these, and the one most directly of interest to us now, was that of Rogoff (1985) who proposed a solution based on delegating monetary policy to a conservative central banker. He argued that when the private sector became aware of the central banker’s incentives for low inflation they were compelled to reduce their inflation expectations. Walsh (1995) in turn recommended a contract with the central bank specifying a penalty for inflation as a way to eliminate inflation bias. Many of the justifications advanced by the Bank of England in support of their independence draw heavily on this implicit contract version of the elimination of inflation bias due to Walsh. The most important extension to these models of gains to central bank independence have also focussed on the assessment of gains from delegation, but – cruciallythese extensions allow for the presence of more than one policy-making authority. Where there are two authorities, fiscal and monetary, central bank independence and conservatism raise new issues, as noted in the compelling analysis by Dixit and Lambertini (2001). In general, where fiscal and monetary decisions are no longer co-ordinated by a separate authority (the Treasury), then the policy making raises strategic interactions between the central bank and the treasury, and with central bank independence, this policy “game” is non-cooperative. So these studies have drawn attention to the potential distortions that can result in this case of two authorities, and specifically the studies appear to undermine the conclusion from the earlier work that delegation of monetary policy to a conservative central bank is sufficient to ensure efficient macro outcomes. Non-cooperative outcomes are likely to be suboptimal if the two authorities have differing objectives. (See Nordhaus(1994), Blake and Weale (1998), and Hall, Henry and Nixon(2001)). Moreover, as Dixit and Lambertini (op.cit), and Hughes-Hallett and Viegi (2001) show, suboptimal outcomes arise where instrument independence is granted to the central bank, but fiscal policy is left unconstrained. Two other things are noteworthy in these examples however. They are usually long run, and they are for the closed economy. It is with these two limitations that the present article is concerned. In parallel with this theoretical literature, empirical studies have aimed to evaluate the relationship between central banking independence and macroeconomic performance. Cross-section evidence tends to suggest that greater independence leads to a better average, but lower variability of inflation performance.(Alesina and Summers (1993), and Grilli et al (1991)). Meanwhile, there seems little evidence of a stable relationship between independence measures and other real macro economic variables such as average GDP growth. In one of the few empirical assessments of the effects of monetary policy delegation and co-ordination, Hall, Henry and Nixon (2001), conclude that the lack of co ordination between monetary and fiscal which is a feature of the UK policy framework after the granting of instrument independence to the BoE, has resulted in over appreciation and medium term loss of traded goods output.(leading to concerns about the so-called “imbalance” in the economy). The purpose of the present article is to analyse further the effects of a monetary and fiscal policy choices in the open economy, using a dynamic model to show that there is an incentive for the fiscal authority (“the treasury”) to push the economy above the equilibrium level of output, even where it knows the independent central bank will tighten monetary policy and successfully offset the inflation effects of the fiscal expansion. In this wider, open economy, setting, we show that incentives exist for the treasury to engage in an expansion which in the medium term yields higher output, in spite of the attendant monetary contraction. What is more important, this incentive exists even in the case where the central bank always succeeds in offsetting the inflation effects of the stimulus, and where the treasury always knows this to be the case. We contend that the presence of these incentives is the basic premise from which game theoretic analyses by the authors just cited proceed. Open economy considerations are thus crucial to the assessment of the optimality or otherwise of central bank independence with conservatism. This already begs some important questions. ‘How much can the treasury expand?’, and ‘How long will the real expansion last?’ are the most serious. In the rest of this paper, we set out a simple dynamic model which addresses these questions. It uses the presence of open economy effects on domestic inflation to show that in the short run the treasury can get output above the natural rate, with spending increase fully financed by tax increases, even where the central bank succeed in keeping inflation on target. There is a cost in doing this of course, which limits both the amount of the fiscal stimulus and the duration of the expansion. In our model this cost is due to the consequent real exchange rate appreciation as monetary policy is tightened, which if large enough produces a trade deficit and hence lowers the stock of overseas assets held by domestic residents. This overseas asset stock disequilibrium cannot continue in the long run, so the fiscal stimulus must be reversed. But, if the initial stock is substantial, there may not be a need for this policy reversal for a considerable period of real time, hence our description of this a medium term non-neutrality. [A counter argument is that we are assuming non-Ricardian equivalence. Thus, although the treasury funds expenditure by taxes, consumers do not revise down their spending due to expected tax increases in the future. If they did then the output increase would disappear. We discount this, though we might want to revert to some more formal defence-e.g. the OLG model we had in a earlier version of the paper] The existence of this medium term non-neutrality is central to the existence of rival objectives held by the central bank and the treasury, which is a necessary condition for suboptimal macroeconomic outcomes. Before moving to our model, we briefly note other, related studies. Related work to ours includes Giovannini (1988) who presents a model of the joint determination of the exchange rate and the current account in response to changes in government spending. The movement in the exchange rate is largely determined by the relative size of crowding out of private consumption compared to the fiscal stimulus. He finds that if private sector wealth falls sufficiently then an increase in government spending will generate a real exchange rate depreciation. But, if on the other hand, the fiscal stimulus increases the demand for domestic output, movements in the exchange rate move to crowd out foreign demand through a real appreciation. Leith and Wren-Lewis (2000) also consider the dynamic interactions of monetary and fiscal behaviour. Their results suggest that where fiscal policy does not ensure [long run] solvency, monetary policy needs to be relatively passive if the model is to be stable. Likewise, if monetary policy is fairly active in seeking to raise interest rates in response to excess inflation then fiscal policy too must be active and self-stabilising to ensure model stability. Therefore the interaction between the two arms of economic policy means that there are only two stable policy regimes; where both are passive, or where both are active. Finally, in policy simulations based on a small econometric model of the UK economy, Wren-Lewis, Darby, Ireland and Ricchi (1996) investigate the long run effects of a balanced budget increase in government spending on the domestic demand for domestic goods and hence the real exchange rate. Their model implies a rise in the demand for domestic products results, as the import propensity of government is only half that of the private sector. Output initially rises given this impact multiplier, and remains above base over the medium term due to the combined effects of nominal inertia and a gradual crowding out of private consumption. Note that in their model there is the possibility of long run non-neutrality, due to the vintage production model they use, as the appreciation of the exchange rate reduces the cost of imported machinery which subsequently increases in the equilibrium level of output. Turning to normative analysis with game-theoretic aspects, Nordhaus (1994) is an important example suggesting the possibility that independent central banks and fiscal authorities become locked in an interaction over policy that leads to high deficits and tight money. If the monetary and fiscal authorities move simultaneously, then the resulting Nash equilibrium may be far removed from the contract curve and the bliss points of the two policy makers individually. He shows that when the two authorities play in a non-cooperative manner it is easy to move to a high interest rate – budget deficit equilibrium. Monetary policy is tight because fiscal policy is loose, and fiscal policy is loose because monetary policy is tight. This outcome results from the competing objectives of the two parties. Undoubtedly the nature of the game changes when the monetary authority no longer acts in a discretionary manner but follows clear and explicit rules as to how and when it conducts monetary policy. In this respect the policy games move away from a Nash towards a Stackelberg form of game, where the fiscal authority moves first by choosing the best position on the reaction function of the monetary authority. Whilst this can certainly offer an improvement on the Nash outcome, it may not be an efficient outcome, if it leads to a position off the contract curve. Even if a monetary authority follows a clear and simple monetary rule, lack of coordination can still be a problem. Blake and Weale (1998) undertake a similar analysis of the co-ordination problem in a dynamic model where the monetary authority seeks to control inflation and the fiscal authority has a budgetary target. Once again the Nash solution to this policy game may yield a position that is not on the contract curve. In a dynamic game, both players may, however, recognise the advantage of cooperation, and given there are sufficient gains to be made and there is a low enough discount rate, cooperation may evolve as a solution. They point out that if each authority has to learn about the other’s behaviour, through the use of a learning rule for example, then co-ordination may take a long time to achieve. Beetsma and Bovenburg (1997), provide a general analysis of the central bank independence and public debt policy, and show that where the two arms of macro policy are co-ordinated, then a debt target is necessary. Finally, in a set of papers, Dixit and Labertini (2001a, 2001b) provide a penetrating analysis of monetary and fiscal interactions, both in the single country case where the two authorities have differing output and inflation objectives, and in a monetary union. Their conclusions are far reaching, and have a direct bearing on the present work. In the single country case, allowing for two policy making authorities they show that monetary commitment can be undermined by fiscal discretion, and to overcome this either by commitment by both authorities or by coordination between them. In the monetary union, generally similar results hold; in particular, making the central bank (the ECB) conservative is still likely to lead to an adverse outcome, unless there strategic fiscal behaviour can be eliminated. In the next section, we analyse the effects of a fiscal stimulus, undertaken to raise output above the natural rate, using a small analytical model of an open economy. We postulate that the monetary authority responds by raising interest rates, by a sufficient amount to ensure that the target inflation rate continues to be met. We show that nonetheless, output remains above the natural rate over the medium term. Hence, an incentive exists for the fiscal authority to undertake such a stimulus, a feature which is the basis of the game analysis in Nordhaus (op.cit.) and Hall, Henry, and Nixon (op.cit). The fiscal authority has an incentive to make a “politically” motivated expansion (i.e. to attempt to expand the economy above the natural rate) in the medium term. Even where this is successfully countered by an independent central bank charged with targeting inflation, the fiscal authority can succeed in keeping output above the natural rate, possibly for a considerable period, due to the effects of the exchange rate on domestic inflation. This medium term non-neutrality will be accompanied by a further effect on the overall balance of the economy as the traded goods sector is squeezed by a higher real exchange rate. 2. The Dynamics of Monetary and Fiscal Policies in an Open Economy 2.1 The Open Economy Model The model we use is a simple open economy one, which has dynamics only in the external sector. There is no investment. The model is pared down to capture essentials, and so that we can provide some insight into the analytical properties of its behaviour. Extensions would not change these in any important way. The relevant equations are given next by equations (1) –(7). M X G C Y − + + = (1) ( ) ∗ − + − = q q r F C M d φ σ γ (2) ( ) ( ) ∗ ∗ − Λ + − = q q Y Y 0 (3) ( ) ∗ − = q q X X φ (4) ( ) ∗ − − = = q q C M M M φ (5)