The last few months have seen a growing awareness of the challenge facing both the EU and the UK in adapting their customs processing systems for the reimposition of a hard border between the two sides once they are no longer linked and merged in the EU’s free movement zone for goods. With around 10% of the EU27’s goods exports going to the UK market and around 40-50% of the UK’s currently routed into or through the EU27, this is clearly a big potential shift in the conditions attached to movement of goods and a potential source of new frictional costs for traders.
In whatever ways the EU does not function as a unitary state, in its internal goods trade it most certainly does, and the recreation of an external frontier for goods trade between the EU and the UK means the reimposition of clearance costs and potential delays on a frontier at which the latency currently created by these things is pretty much zero. This is especially the case across the land frontier between Northern Ireland and the Irish Republic, which does not even have the built-in delays (however minimal) of a channel crossing or an airfreight journey for policymakers to exploit in minimising the impact of trade processing on goods flow.
All this makes the ‘trade facilitation’ problem for the EU and the UK quite different from the one around which much of the practice of negotiated trade facilitation has developed over the last 30 years. While trade facilitation has developed materially as a core discipline of the World Trade Organisation (WTO) and World Customs Organisation (WCO) repertoire over this period, the focus of this work has been on reducing opacities, creating rights of recourse and eliminating the scope to extract rents from customs processing. This model is encapsulated in the WTO Trade Facilitation Agreement (TFA) agreed in Bali in 2013 and in most trade facilitation chapters in modern FTAs. The former assumes that customs systems are an inevitable source of frictional costs, but can be made incrementally better. The latter tend to try and lock in a baseline of already existing good practice, but little more.
The EU-UK problem is that customs processing is currently non-existent and needs to be reimposed in the least disruptive way. This is a question of managing the logistics of border crossings in a completely new way; one that will demand new facilities and new processes. But it is also a question of preserving some of the efficiency gains that have been built into EU-UK cross-border supply and distribution chains over the last 20 years as firms have squeezed the potential of the single market to develop inventory and distribution models built on low or zero border latency.
There are broadly three sets of tools that the EU and the UK will be able to use to mitigate the impact of this change. The first, are commitments to trade facilitation in an EU-UK trade agreement or a standalone customs cooperation agreement. The second, is bilateral cooperation between the UK authorities and the authorities of the key systems that will manage the overwhelming bulk of UK sea and road-borne trade with the single market, especially France, the Netherlands and Ireland. The third, is unilateral reform of their respective systems to increase their processing speed.
The problem of negotiating efficiency
Negotiated commitments in an FTA or customs agreement on trade facilitation are likely to be the weakest of these tools. This is a reflection of the general problem of binding trading partners to efficiency targets. The TFA, like the limited references to trade facilitation in the General Agreement on Tariffs and Trade itself, implicitly encourages process efficiency and rapid customs clearance, but does not commit its signatories to provide it. Only in a handful of bilateral trade agreements, have commitments been made to rapid processing. The strongest of these are in the dormant Trans-Pacific Partnership at US urging, and these are heavily caveated. While the EU’s past FTAs have sought commitments in a range of areas that contribute to faster clearance times, it has not sought rapid processing commitments, unlike the US.
This reflects a number of things. Trade facilitation commitments are generally systemic rather than preferential, so states are tied to their baseline rather than best case performance in making binding commitments. The EU is a single customs space overseen by 28 separates customs systems of varying levels of efficiency and reliability. The EU makes binding commitments with this in mind. The same constraints have limited its capacity to offer binding commitments to deliver a ‘single window’ online system for traders, and to deliver the commitments it has made in this area. Even unitary systems can be reluctant to make commitments that can be breached by underperforming elements in their customs services, and frame their offers with wiggle room accordingly.
There is nevertheless useful work to be done by bilateral agreement at the EU-UK level in this area. It should, for example, be used to cement mutual recognition of Authorised Economic Operator systems that confer expedited treatment on traders and their supply chains, for example, to remove obstacles to clearance time compression on the ground by allowing green lane treatment on both sides of the frontier, or ensuring enhanced data sharing arrangements between customs authorities so that they can better assess the risk level of cargoes and, in turn, whether physical checks are warranted. But commitments that simply codify existing automation standards, transparency or single point of clearance protocols that already exist in practice will be of limited value, even if negotiators argue that they provide a useful baseline. More importantly, even ambitious commitments to efficiency will not get anywhere near the kind of optimised protocols that EU and UK traders need on the ground. The ground-breaking (but caveated) targets for clearance times in texts like TPP – 2 hours for express processing and 48 hours for general processing - look expeditious when set against WCO or even some OECD averages. They would still represent substantial delays for a system that is essentially free of any policy-related delays at all.
Compressing clearance times in practice
Bilateral cooperation between the UK authorities and the key customs systems that face on to the UK for the purposes of road-based trade with the single market will be a more practical route to enhancements in clearance times. Member state customs authorities in the EU already establish bilateral cooperation frameworks with counterparts in other states as a legal basis for data sharing and other forms of institutional cooperation. This is a model that the UK, France, Ireland and the Netherlands in particular will need to exploit to the full. The Swedish-Norwegian border provides some practical experience in the efficient use of, for example, co-location of officials to keep trade moving as fast as possible. While trade across this border benefits from the streamlined documentation protocols enabled by the EEA common regulatory framework in areas such as animal health, and EU-UK trade may not, it is still a good case study for compressing transit times.
Such solutions could be especially important for the UK-Ireland land border, where avoiding stoppages, or even delays for traffic and the paraphernalia of intrusive monitoring at the frontier is practically and politically imperative for both sides. In principle, such cooperation could extend to integrated online systems and even powers to instruct customs officials of the other side, provided that the EU states in question respected their wider obligations under the Union Customs Code and regulatory framework and duties under UK law.
Finally, unilateral improvements to the key UK and EU systems are both possible and will ultimately be the key to compressing clearance times. While operating within the obligation to process exit and entry of goods, and respecting the need to maintain secure borders, there is ample scope to optimise clearance times here. Authorised Economic Operator (AEO) systems linked to expedited treatment can be expanded in their coverage and firms assisted in meeting their requirements. Their benefits can be solidified in areas such as simplified documentation and pre-arrival clearance to avoid delays at the border. The use of self-assessment for AEOs is already foreseen in some contexts such as meeting origin rules and has considerable potential to reduce the documentation requirements and processing protocols attached to a border crossing. These tools can also be combined with Automatic Number Plate Recognition (ANPR) and iris scanning technology at the approach to the frontier itself and behind the border facilities on one or both sides for randomised physical inspections in areas like agriculture where they will remain necessary in some form.
To be sure, these will be easier hurdles for large traders to clear, but these companies also account for the bulk of trade, and certainly the bulk of highly optimised supply chain management. Targeted assistance will be required for SMEs. The biggest challenge here is not the technology, but the time and money to deploy it effectively.
Beside facilitative schemes and technology investments, there are also unilateral administrative decisions that the UK could take to optimise border efficiency. There is a need for an administrative reorganisation to bring the 20-plus regulatory actors involved in border processing (HMRC, border police, port health, etc.) under one umbrella, as a single government interlocutor at the border would render traders’ interaction with the customs system more efficient and predictable. Equally, a decision by HMRC to maintain the EU’s customs lexicon to avoid any confusion for traders after Brexit would help. This includes adopting the same definitions for goods as the UCC and preserving the name ‘AEO’ for any trusted traders scheme.
Trust and time
There are a number of conclusions we can draw here. The first is that binding commitments in an EU-UK bilateral agreement are unlikely to do much more than establish a simple baseline of efficiency at the UK-EU frontier. That baseline will be far below both the status quo and the kind of optimised clearance times that the two sides will want to establish in practice, even if they do not wish to codify them in a binding agreement. What a bilateral agreement can, and should, do is use regulatory alignment and mutual recognition of EU and UK standards in areas such as AEO systems and food safety or animal health.
The real heavy lifting will be done a level down in bilateral cooperation between customs authorities and, above all, in unilateral upgrade and reform of processing protocols and systems. This will often mean exploiting currently underused potential, such as widening the use of AEO systems to expand the use of expedited processing. This is one of the reasons why trust and time will matter in the exit process. Trust because customs cooperation is built on it and a well optimised system facing onto a poorly optimised one is a partial solution. Time because a period of transitional arrangements - coupled with the necessary investment - would allow systems on both sides to adapt and restructure not just for higher volumes of processing, but for optimised clearance times at those higher levels.
The almost-invisible customs border
Not easy…but this is how it might be done.
 In this context, we take the term hard border to denote a line over which the movement of goods defined as imports and exports is required to be processed and approved in some form, digital, physical or a combination of both. Such a border can be treated as inevitable in some form once the UK is no longer embedded in the free movement of goods zone created by the combination of the EU’s common external tariff, internal customs union and the standards harmonisation provided by the EU acquis. It is worth noting that a customs union in itself does not imply the absence of a hard border — the internal EU customs union has successfully eliminated all of its inner borders, but the EU-Turkey customs union has not for instance.
The views expressed in this note can be attributed to the named author(s) only.