Preferential trade agreements under uncertainty

Abstract

In Preferential Trade Agreements (PTAs), not all firms utilise preferential tariffs, suggesting the presence of fixed costs of using tariff preference. I develop a model where firms can trade in a PTA under the standard Most Favoured Nation regime or, after paying an additional fixed cost, under the Preferential regime. I show that if tariff preferences become uncertain, more firms export under the MFN regime, but with the option to switch to the Preferential regime in the future. The model extends the Handley and Limao one and nests its empirical equation under the restriction of a single trade regime. I apply the model to an excellent natural experiment: the Brexit referendum and UK trade with PTA partners. I find that ignoring the partial uptake of trade agreements understates the impact of uncertainty on trade and can lead to biased empirical results. Brexit uncertainty had a modest negative effect on UK imports from PTA countries. Continuity Agreements signed to replace the existing EU agreements only partly reduced the uncertainty introduced by the referendum.

Keywords: Trade agreements; Policy uncertainty; Brexit

JEL classification: F13; F14; F68

In Preferential Trade Agreements, firms can trade under a preferential regime, which gives them access to lower customs tariffs compared to the Most Favoured Nation (MFN) tariffs, the default for members of the World Trade Organization. However, although tariff preferences are available for trade between PTA members, not all firms use them. The partial uptake of tariff preferences suggests the presence of some additional costs related to the utilisation of PTAs. Such costs can be related to bureaucratic procedures or compliance with Rules of Origin, for instance, and they create regime heterogeneity.

In this paper, I ask what happens when tariff preferences granted under a trade agreement become uncertain, with the possibility for trade to revert to the MFN regime. While in general trade agreements are believed to reduce uncertainty between trading partners, in recent times trade policy uncertainty increased within trade agreements, with episodes such as Brexit and the NAFTA renegotiation.

To study uncertainty in trade agreements, I use a theoretical model that accounts for the I apply the model to the UK’s trade with its trade agreement partners following the 2016 Brexit referendum. As a member of the EU, the UK had trade agreements with about 70 countries beyond the EU, and by leaving the EU the UK was also leaving all these other trade agreements. This generated substantial uncertainty about the future trade regime between the UK and its trade agreement partners, with a concrete probability of trade reverting to the MFN regime. To counteract this uncertainty, the UK government started to re-negotiate those agreements and, to date, managed to roll over most of them by signing Continuity Agreements.

Empirical results show that the uncertainty generated by the 2016 referendum negatively affected UK’s imports from its PTA partners. Continuity Agreements mitigated the negative effect of uncertainty but did not offset it completely – they reduced uncertainty by a third on average. The analysis shows that accounting for the partial uptake of trade agreements is a key factor for identifying uncertainty effects in the data.

Finally, I use the parameters estimated from data on UK imports to construct a measure that tracks the evolution of uncertainty over time, and I quantify its effects on UK imports from trade agreement partners. I find that the major source of uncertainty has been the 2016 referendum. Uncertainty then fell in 2017-18 only to increase again and get close to referendum levels bythe end of 2019, in concomitance with a series of political events in the UK. A partial equilibrium quantification exercise shows that by the end of 2019, uncertainty reduced UK’s imports from PTA partners by -1.1%. Continuity Agreements helped to mitigate these effects but did not completely offset uncertainty, bringing the reduction in UK imports by the end of 2019 from -1.1% to -0.75%.

Read Working Paper 007