Chart against a background of shipping containers in a port
VoxEU Column International trade Monetary Policy

Central bankers and trade ministers: How monetary policy can prevent trade wars

Trade wars can have large macroeconomic consequences, but there is limited focus on the interaction between optimal trade and macroeconomic policies. This column studies the optimal design of trade policies under different monetary policy rules. It finds that in a trade war, a monetary policy rule which targets Consumer Price Index inflation leads to lower average tariff rates and higher welfare than producer price inflation targeting or a policy of fixed nominal exchange rates. Furthermore, an optimally designed and delegated monetary policy rule that internalises the formation of trade policy can eliminate the trade war and even offset some of the monopoly distortions in production.

The world has recently experienced a reversal of the decades-long process of trade liberalisation within the framework of a multilateral rules-based trading system (e.g. Fajgelbaum et al. 2020). There is increasing fear of a global trade war, 1 where most countries follow unilateral non-cooperative tariff policies. 2 This would likely have large macroeconomic consequences for the world economy. One key question is how high would we expect tariffs to climb in a global trade war? While there is a long-standing trade theory literature on the determination of optimal non-cooperative tariffs for small and large economies, 3 there has been much less focus on the interaction between optimal trade policy and macroeconomic policy.

On the one hand, the literature on optimal macroeconomic policy in open economies has for the most part abstracted from trade policy. In the New Keynesian open-economy literature, many papers have explored the nature of monetary policy spillovers and the benefits of international monetary policy cooperation. The recent increase in trade frictions among major countries led to papers investigating the impacts of exogenous tariffs or export subsidies within these types of models (e.g. Barattieri et al. 2021, Bergin and Corsetti 2023). In contrast, there has been much less exploration of optimal trade policy within the standard New Keynesian frameworks, where prices and wages are sticky.

On the other hand, most of the literature on trade policy and trade agreements has implicitly abstracted from short-run considerations when exploring the determinants of tariffs and other trade barriers, in either cooperative or non-cooperative environments. The ‘new world’ of discretionary trade policy calls for an integrated framework that considers the optimal choice of tariffs within the broader context of the short-run economic policy agenda.

In a first paper along these lines (Auray et al. 2022), we have extended the simple two-country model of Galì and Monacelli (2005) to incorporate endogenous tariff formation within a dynamic non-cooperative game. Using the assumption of Bagwell and Staiger (2003), where tariffs are determined as an equilibrium in interacting dynamic incentive constraints for each country, we find that tariffs should be acyclical if prices and wages are flexible, so that trade restrictions are independent of macroeconomic shocks. However, with wage and price rigidities, tariffs vary with the business cycle since governments attempt to use tariffs to correct for the absence of nominal wage or price adjustment. However, we find that tariffs may be either procyclical or countercyclical, depending on the nature of shocks. Empirical results supporting these patterns of conditional cyclicality of trade protection are presented in a follow-up paper (Auray et al. 2024a).

An extended analysis in Auray et al. (2024b) compares the outcome for optimal tariffs and trade wars given different assumptions about the degree of monetary policy commitment. In that study, we show that an inflation targeting monetary framework is associated with significantly lower tariffs and a less intense trade war than a situation characterised by discretionary monetary policy. In a quantitative application, we find that average global tariffs would be 30% lower with an inflation targeting monetary rule than under discretion.

In a more recent study (Auray et al. 2024c), we extend the analysis of optimal non-cooperative trade policies to explore the consequences of alternative monetary policy rules and their optimal design. The key assumption behind the analysis is that trade policy is made in a second-best setting, where the tariff-setter takes into account pre-existing monopolistic distortions and nominal rigidities. When monopolistic markups are positive, a tariff improves terms of trade but is likely to exacerbate the distortion due to aggregate output being below the socially optimal level. With sticky prices, a tariff may also have a secondary negative effect due to its deflationary impact. We show that a monetary rule that targets the Consumer Price Index (CPI) or stabilises the nominal exchange rate exacerbates the domestic output costs of tariffs and leads to lower average tariff rates in a trade war, and thus to higher welfare.

Clarida et al. (2002) and Engel (2011) use New Keynesian open-economy models with currency producer pricing and free trade to show that an optimal monetary policy should target the inflation rate of domestic goods (or the producer price index). In Auray et al. (2024c), we show that this result no longer applies in the context of a trade war. Our first result is that CPI inflation strictly welfare-dominates PPI targeting, because it leads to lower trade-war tariffs. While CPI targeting stabilises the response of the overall CPI to any shock, it leads to a greater deflationary response of the PPI. This leads to a greater cost of a tariff in terms of the deviation of aggregate output from its socially optimal level. In the context of endogenous trade policy, this will reduce the desired trade-war tariff. In a fully symmetric global economy, welfare is higher under CPI targeting because average tariffs are lower in all countries.

Figure 1 illustrates the implications of three different simple forms of monetary policy for optimal tariffs and welfare in a symmetric trade war for various degrees of price rigidity. With fully flexible prices, the monetary stance is irrelevant for the outcome of a trade war. But as prices become stickier, CPI targeting leads to lower tariffs and higher welfare than either PPI targeting, or, more surprisingly, a policy of fixed nominal exchange rate.

Figure 1 The effects of different monetary policy rules in a symmetric trade war

Figure 1 The effects of different monetary policy rules in a symmetric trade war

Notes: This figure shows the outcome for tariffs, consumption, employment, and welfare for different monetary policy rules across different degrees of price stickiness.

But in fact, one can do better than that, as none of the exact rules (PPI targeting, CPI targeting, or fixed nominal exchange rates) maximises aggregate welfare from the perspective of a cooperative authority that would internalise the motive behind national trade policies and the subsequent impact of the monetary rules on tariff choices. We think of this situation as one where the monetary rules are delegated to independent central banks, but the rules are designed ex-ante, taking into account the nature of trade policy and the implementation of monetary policy. In this case, we show that an optimal cooperatively designed rule will place a high weight on stabilising a function of the tariff-adjusted terms of trade – higher than CPI inflation targeting. This cooperative rule acts to fully offset the incentive to impose tariffs, and so in fact eliminates the trade war completely. It actually leads to small negative tariffs, which partly undo the pre-existing monopoly distortions in production. Hence, remarkably, in the presence of endogenous tariff setting, a particular monetary policy rule can not only prevent the trade war, but also partly alleviate the underlying production distortion in each economy, and in so doing actually dominates a free-trade outcome in welfare terms.

Finally, we show that the optimal cooperatively designed monetary policy is very close to a non-cooperative outcome where each country designs its monetary rule and delegates it to its own central bank. First, if a rule must be transparent in the sense that it targets an official price index, then we already found that CPI rather than PPI targeting will be unilaterally optimal for each country. However, if the rule allows for more flexibility, we show that in this case (as under a cooperative target choice), the optimal rule leads to an endogenous elimination of the trade war and a partial offsetting of the monopoly distortion and implies small welfare losses compared to the cooperative rule. Hence, we may conclude that a purely self-oriented optimal monetary design and delegation can achieve major welfare gains in an environment of endogenous non-cooperative trade policy and monopoly distortions in production.


Auray, S, M B Devereux and A Eyquem (2022), “Self-enforcing Trade Policy and Exchange Rate Adjustment”, Journal of International Economics 134 (C).

Auray, S, M B Devereux and A Eyquem (2024a), “The Demand for Trade Protection over the Business Cycle”, Journal of Money Credit and Banking 56(4): 865-898.

Auray, S, M B Devereux and A Eyquem (2024b), “Trade Wars, Nominal Rigidities and Monetary Policy”, Review of Economic Studies, forthcoming.

Auray, S, M B Devereux and A Eyquem (2024c), “Trade Wars and the Optimal Design of Monetary Rules”, NBER Working Paper 32451.

Bagwell, K and R W Staiger (2003), “Protection and the Business Cycle”, The B.E. Journal of Economic Analysis & Policy 3(1): 1–45.

Bergin, P and G Corsetti (2023), “The Macroeconomic Stabilization of Tariff Shocks: What is the Optimal Monetary Response?”, Journal of International Economics 143: 103758.

Barattieri, A, M Cacciatore and F Ghironi (2021), “Protectionism and the Business Cycle”, Journal of International Economics 129:103417.

Bowen, T R, J L Broz and B P Rosendorff (2023), “A theory of trade policy transitions”,, 31 October.

Caffarra, C and B Kilic (2024), “Re-joining trade with antitrust”,, 7 March.

Caliendo, L and F Parro (2021), “Trade Policy”, NBER Working Paper 29051.

Clarida, R, J Gali and M Gertler (2002), “A Simple Framework for International Monetary Policy Analysis”, Journal of Monetary Economics 49(5):879–904.

Engel, C (2011), “Currency Alignments and Optimal Monetary Policy”, American Economic Review 101(6): 2796–2822.

Fajgelbaum, P D, P K Goldberg, P J Kennedy and A K Khandelwal (2020), “The Return to Protectionism”, The Quarterly Journal of Economics 135(1): 1–55.

Galì, J and T Monacelli (2005), “Monetary Policy and Exchange Rate Volatility in a Small Open Economy”, The Review of Economic Studies 72(3): 707–734.

Novy, D and D Jacks (2019), “Trade wars may ‘bloc up’ world trade”,, 23 July.

Ossa, R (2015), “WTO success: No trade agreement but no trade war”,, 11 June.

Szczerbowicz, U, C Jardet, D Siena and A Berthou (2019), “The macroeconomic implications of a global trade war”,, 8 February.


  1. See Szczerbowicz et al. (2019) about the macroeconomic implications of a global trade war.
  2. See Ossa (2015) and Caffarra and Kilic (2014) for a discussion of the impact of WTO on trade war and Novy and Jacks (2019) and Bowen et al. (2023) for a discussion of the impact of trade war on the intensification of US and China trade blocs.
  3. See Caliendo and Parro (2021) for a recent survey.