The US trade deficits will have to be reduced materially to prevent a crisis down the road. This column argues that fiscal consolidation, in association with a weaker dollar, appears to be the right way for the US to improve its trade balance. In addition, the US should pursue an outcome in which there is an expansion in aggregate demand in China – and in some other surplus countries – at the same time as fiscal consolidation in the US, rather than attempting to sustain the demand for home-produced goods by means of a fiscal policy which remains expansionary, or indeed by means of tariffs.
One does not have to be a Trumpist to believe that the large and persistent trade deficits of the US need correction. In 1980, the US was a net creditor to the rest of the world by an amount equal to 20% of GDP. In 2025, it is a massive debtor. US external debt is now 90% of GDP and rising unsustainably (Bayoumi and Gagnon 2025). Trade deficits will have to be reduced materially to prevent a crisis down the road. The question arises: How best could Trump have achieved this objective when he took office? This is a non-trivial question in counterfactual history.
We think that for any such policy, the starting point must be fiscal consolidation. Such fiscal consolidation could be combined with either tariffs or with a currency depreciation. The latter would be better. But in the case of the US such a currency change could only be achieved in the presence of some kind of ‘Mar-a-Lago currency accord’.
Some observers have been calling for fiscal consolidation and others have been arguing for a ‘Mar-a-Lago accord’. The contribution of this column is that we identify a need for both approaches. Thus, our column is in the same spirit as the articles by Obstfeld (2025) and Clarida (2025). We develop our argument by going back to the ideas in James Meade’s book on the balance of payments (Meade 1951), as set out by Trevor Swan, and as developed by Mundell, Fleming, and Dornbusch.
Trump has imposed tariffs on US imports to improve the US trade balance. But tariffs without concomitant macroeconomic retrenchment would not do this. If the exchange rate is floating, as in the US, tariffs tend to reduce exports. This is because the reduction in demand for imports would tend to strengthen the exchange rate directly; and exchange rate appreciation is also likely since the rise in interest rates caused by higher domestic economic activity would lead to higher capital inflows. Moreover, the central bank is likely to reinforce the rise in interest rates in response to the tariff-induced rise in the price level. In a fully employed economy such as the US, the success of tariffs as a policy to switch demand towards domestic goods is thus conditional on there being sufficient fiscal contraction to make room for the extra demand for domestic goods. (Note that the US dollar has in fact depreciated. Needless to say, this is because of the chaotic manner in which Trump’s tariffs were introduced, which created huge uncertainty, reduced the incentive to invest, raised the risks of a US recession, and created doubts about US creditworthiness.)
Currency depreciation is an alternative way to switch demand towards domestic goods and improve the trade balance. But that too requires fiscal consolidation. If the exchange rate is floating, the only way to bring about a currency depreciation without causing inflation is if there is also – at the same time – a fiscal contraction. That is because in a fully employed economy, the central bank will only be prepared to reduce interest rates if there is, at the same time, some kind of fiscal tightening. Given that the currency is floating, such a depreciation of the currency will divert demand away from imports and stimulate demand for exports, replacing the demand for domestic goods which has been reduced by the fiscal consolidation.
There are other, less conventional ways to weaken the dollar – for example, by taxing capital inflows (Pettis 2025, Pettis and Hogan 2024, Tett 2025) – but these would be inadvisable. Interfering with the market for government securities would run the risk of destroying confidence in the dollar as a safe haven. That would certainly damage dollar dominance in the longer term and, even worse, might lead to an immediate financial crisis.
So, both tariff increases and currency depreciation require fiscal consolidation if they are to be used as a policy that switches demand towards domestic goods and so improves the trade balance. But tariffs are likely to be worse than currency depreciation as a form of expenditure-switching policy. There are at least four reasons for this (Dornbusch 1987, Bordo and Levy 2025, Baldwin 2025a, 2025b). First, a tariff acts on imports alone, while a currency depreciation acts both to reduce imports and to increase exports. Second, tariffs bring with them efficiency costs in that US consumers pay more to obtain goods at home than from cheaper sources abroad (and the higher the tariff, the larger the cost). Third, a tariff on imports is likely to lead to a reduction in the profitability of exporting, to the extent that protected import-competing goods are inputs into the production of exports, reducing the degree to which the expenditure-switching policy will be successful. Finally, the US economy may become less productive over time as a result of import-competing firms being sheltered from foreign competition. Of course, a tariff will cause foreigners to supply their exports at a cheaper price, especially for a large country such as the US, thereby improving the US terms of trade. This ‘optimum tariff’ argument has long been recognised as valid from a purely national perspective. (The height of the optimum tariff depends on the degree of market power possessed by the tariff-imposing country.) Nevertheless, all tariffs, including optimum tariffs, hurt trading partners and therefore invite retaliatory action by foreign policymakers. A trade war is virtually certain to bring large losses all round, including to the tariff-imposing country (in this case, the US).
It appears, therefore, that fiscal consolidation in association with a weaker dollar, is the right way for the US to improve its trade balance.
Despite what we have just said, there has been a tendency to assume (e.g. Pettis and Hogan 2024, Pettis 2025) that the US is powerless to resist a flood of excess savings from China and other surplus countries. Martin Wolf (2025) has also argued along similar lines:
” The analysis suggests that the benefit to the US of its persistent net capital inflows is the ability to have a larger fiscal deficit and so grow its public debt. This does not look like a good bargain. But if the government cut its deficit, while the external inflow continued, the outcome could be to drive the private sector into deficit, either via a slump in its income or a surge in its spending. The former means a recession. The latter means asset price bubbles … …”.
In Wolf’s view, US fiscal policy needs to remain expansionary in order to prevent the unemployment that would ensue from the deflationary impact of excessive savings in China and elsewhere.
We disagree. Our answer is that an external inflow to the US would not continue at the previous level because of the fall in US interest rates which we have described above; it is this fall in foreign inflow which would enable the fall of the dollar. Our view is thus that an appropriate fiscal/monetary policy mix could deliver some improvement in the US trade deficit even if the US acted unilaterally in the way that we have described.
Nevertheless, since the US is a large country, it is important to analyse the effects of US action on the rest of the world, and to consider how foreign countries might respond. The best – globally cooperative – outcome would be one in which there is an expansion in aggregate demand in China – and in some other surplus countries – at the same time as there is fiscal consolidation in the US. In that case, extra demand from China and elsewhere would absorb the excess of exports over imports from those countries to the US – which is what has been enabling the US to run a trade deficit in the first place. And in that case the depreciation of the dollar would also ensure that, within the US, demand switched from the purchase of imports to the purchase of home-produced goods, thereby sustaining the demand for such goods even although there had been fiscal consolidation. No unemployment of resources would emerge, either in the US or anywhere else. The need for a cooperatively agreed outcome of this kind is what a ‘Mar-a-Lago accord’ should seek to achieve, which is why such an accord is needed. The idea of some kind of Mar-a-Lago accord thus has to be taken much more seriously than it has been, by, for example, Paul Krugman (Krugman 2025).
This is sort of outcome which the US should pursue, rather than attempting to sustain the demand for home-produced goods by means of a fiscal policy which remains expansionary, or indeed by means of tariffs. A flood of imports into the US is not inevitable. But it can only be avoided if there is global cooperation of the kind which we have described.
Of course, such cooperation may not be forthcoming. If the US were to act on its own, without a contemporaneous expansion of aggregate demand abroad, the outcome would, indeed, be a slump in demand and output in the US. This would then be spread abroad by a reduction in the US demand for imports. Such deflationary effects on foreign countries would be amplified by the depreciation of the dollar which the US policy was seeking to bring about.
In these circumstances there would, of course, be a danger of foreign retaliation. Currency wars are possible, just like tariff wars (Corden 1994: Appendix 13.1). The Chinese monetary authorities – and those in other countries – might resist the depreciation of the dollar, sending the effects of fiscal austerity straight back to the US, in the way that is feared by Wolf and Pettis. This is the outcome which it is crucial to avoid.
How might the world actually achieve the kind of orderly international adjustment which we have described, thereby avoiding a currency war and currency chaos? In particular, how might a weaker dollar be sustained in the face of the likely monetary policy reactions of trading partners?
Any dollar depreciation that is large enough to reduce the US trade deficit materially (say, 33%) would have significant foreign repercussions. The reactions of key countries, such as China, Japan, and the euro area, would be critical. Here it is relevant that ever since the amendment of the IMF Articles in 1987, the world has had a ‘non-system’ of exchange rates, in which each country can have any exchange rate regime of its choice (Corden 2004). Imagine, therefore, that China and Japan shadow the US dollar in response to a dollar depreciation. Adjustment would obviously be impeded. (Note that both these countries have resisted floating at various times in the past – China systematically and Japan sporadically). Moreover, since most of the adjustment would then be forced on to the euro area, Europe would also have good grounds to resist. That is a recipe for economic chaos.
A material reduction in the US trade deficit would thus require macroeconomic cooperation between the key countries. (Countries in the rest of the world would be likely to shadow one or other of the key currencies.) The good news is that the appropriate direction of policies in the medium run is clear enough, given the macroeconomic positions of the countries in the accord. In the US, there would have to be fiscal contraction and monetary expansion. In China, measures would be required to increase public and private consumption (not investment), combined with exchange-rate appreciation (Gourinchas et al. 2024). In Japan, measures to increase household and corporate consumption are called for, along with monetary tightening. In the euro area, there would need to be measures to increase public and private spending (on both consumption and investment), together with tighter monetary policy. The bad news is that the appropriate medium-run direction of policies may conflict with various short-term considerations. This is clearly true of fiscal consolidation in the US, but the point is more general than this. Richard Clarida argued recently that the Plaza accord of 1987 was only successful because it involved a cooperative policy adjustment of this kind (Clarida 2025). He is surely right.
Finally, how would the foregoing considerations affect dollar dominance? It is pertinent to note that exchange rate changes in the dollar so far (including dollar depreciations) appear not to have made much difference to dollar dominance. However, these exchange rate changes have been largely market-determined. What if, as above, the dollar depreciation is brought about by deliberate policy? We suggest that, at the present conjuncture, dollar dominance is much more at risk from the inexorable build-up of US public and external debt. If dollar depreciation and associated exchange rate appreciations were explained to the market as designed to help reduce global imbalances and thereby put the world economy, including the US, on a more stable macroeconomic footing, this is likely to be accepted by the market as a necessary move. Changes in currency dominance will nevertheless happen over time, but they will be gradual, not sudden (Eichengreen 2025).
Of course, this approach is a far cry from Trump’s actual policy, and from the policy advocated by his advisors. But it would have had a much greater chance of working out well for the US economy, and the world economy.
To read the column as it was published by the Centre for Economic Policy Research, click here.