27 December 2018

Best of 2018: How to fix the Irish border problem


This week CapX is publishing its favourite articles of the year. This piece was originally published on March 19.

The problem

The UK leaving the EU will mean the reintroduction of a customs border between Northern Ireland and the Republic of Ireland, and between the Republic of Ireland and Great Britain. This entails a number of changes which could cause difficulties and delays for businesses and individuals on both sides of the Irish sea and land border.

Potential solutions to these problems exist, but are mischaracterised by people who seem determined to deny that they could work. The discussion, more than any in Brexit is plagued with inaccurate assumptions, emotional responses and misreporting.

The assumptions, real and false

The current discussion can sometime seem to be based on the assumption that there is at present no border on the island of Ireland. This is simply wrong. The current border marks the boundary between two different jurisdictions. Many things fundamental to daily life change when someone crosses the border – education provisions, health services, minimum wages and aspects of labour laws, excise duty rates, and personal taxation, for example. The task after Brexit will be to stop this border becoming significantly harder. But it will continue to exist. The existence of no border at all would be at odds with the Belfast Agreement, which is concerned with respecting the integrity of the territory of Northern Ireland as part of the United Kingdom, as well as the rights and interest of the nationalist and unionist communities. For there to be no change at all at the Irish border would require the entire UK to be in the Customs Union and Single Market, which would eliminate any possibility of an independent trade policy.

It is true that at present there are very few controls at the border. Such controls are normally needed at international borders to manage flows of people, for security, and to conduct trade-related functions. The trade functions are payment of import duties, including rules of origin compliance, confirmation that goods meet the standards of the market (manufacturing standards or animal and plant health requirements, known as sanitary and phytosanitary “SPS” measures), payment of VAT and excise duties, anti-smuggling and anti-counterfeiting operations, and any other international treaty functions (for example, the policing of wildlife trade rules).

In the case of the border between the UK and the Republic of Ireland, some of these functions still apply because of differences between the two jurisdictions outlined above. The UK and Republic of Ireland also have different immigration policies, but thanks to the historic and enduring Common Travel Area (CTA), because neither are part of the Schengen arrangement, there is no need for immigration controls at the Irish border. This is sustainable as long as the UK does not require visas for EU27 visitors, and that the Republic of Ireland does not join the Schengen area. The issue is therefore primarily about the trade border.

Then there is the question of what a customs border actually means in practice. Although customs checks are part of the requirements at the EU external border now, in practice most of the necessary formalities are conducted electronically, with checks taking place pre-arrival, followed up in a small minority of cases at warehouses, or even in the market and payments made on account.

At the moment HMRC physically checks only 4 per cent of consignments arriving in the EU at the UK external border; the Irish authorities only check 1 per cent. Consequently, even if a border operation along the lines of other third country frontiers were reintroduced, one could expect to see at least 95 per cent of goods pass the border without checks, albeit with physical infrastructure in place for those that are stopped.

The only goods that are required to go through specific entry points are animals and animal products, which must pass through ports/crossings with border inspection posts (BIPS) where veterinary checks can be done by accredited handlers and vets. It is important to point out that meat and animal trade within member states is not completely frictionless now, due to the sensitivity of health and safety concerns, and can only be done between accredited operators, accompanied by regulatory certifications.

This process of mutual recognition of accreditations and inspection regimes within member states removes the need for safety checks at the borders, and it is in the interests of both the EU and the UK to maintain this, which could be agreed in a free trade agreement (FTA). The EU has agreed mutual recognition of sanitary measures for meat and animals with Canada and New Zealand, and consequently reduced the number of border inspections of meat products from those countries.

Given the existing highly regulated trade between the UK and other EU member states, especially the Republic of Ireland, and the priority that both sides attach to an Irish land border without physical infrastructure, it should be possible to continue the current ways of enforcing and inspecting sanitary regimes for meat and animal trade inside the border.

If, in the future, the UK considers changes to its SPS regime in this area, it would be a matter for the government of the day to establish whether costs arising from trade barriers with the EU that might result outweigh the benefits to consumers and traders that changes could bring. It would also be possible for the matter to be devolved to the government of Northern Ireland (reflecting what is already effectively an all-island regime) and it would then be for the devolved powers to determine whether border checks for meat and animals between the Northern Ireland and the mainland would be acceptable to offset the possibility of stricter controls at the land border.

While some might say this sounds like a border in the Irish Sea, it is worth remembering that at present the rules in Northern Ireland are not the same as in the rest of the UK. There is a Single Electricity market for the island of Ireland. Carbon taxes imposed by UK carbon tax legislation do not apply in Northern Ireland. Corporate tax rates are allowed to differ under devolved powers. Maintaining the all-Ireland regime for meat and meat products would not be unique.

Another common error, or wilful misrepresentation, is to overstate the volume of trade across the border. This is an enormously significant point.

Northern Ireland’s economy is still largely integrated with the UK. According to Northern Ireland government figures, 87 per cent of its turnover represents sales within Northern Ireland or to Great Britain. Just 5 per cent is sales to the Republic of Ireland, 3 per cent to the rest of the EU, and 6 per cent to the rest of the world. So Northern Ireland’s primary economic interest is in maintaining the UK single market. Similarly, Irish exports to Northern Ireland are just 1.6 per cent of total Irish exports. Nearly 10 times as much goes to mainland Great Britain.

In short, for both Northern Ireland and the Republic of Ireland, maintaining barrier free trade to Great Britain significantly outweighs the importance of free flows within the island of Ireland (although clearly minimising the costs and frictions at the land border is vitally important as well). For both economies intra-Irish trade is marginal, even after 40 years of EU membership.

Another false assumption in the Irish border debate is that there is complete homogeneity within the existing single market, whether for goods, taxation, and many other areas. That isn’t the case. For example, there are many products where UK safety standards exceed those of the EU members (one example is furniture safety).

Yet border checks within the EU are not necessary to prevent unsafe furniture entering the UK market. Compliance is maintained because responsible UK importers do not wish to commit a crime by placing sub-standard furniture on the market, and because there is effective enforcement by trading standards officers behind the border.

The conclusion to be drawn is that Brexit will not see Northern Ireland and the Republic of Ireland moving from a perfectly harmonised trading system, in which no checks are needed, to one where the two areas are entirely divergent and border controls are needed.

There is a spectrum. At one end is the current low visibility border (but a border nonetheless), and at the other is a very hard border such as between neighbouring countries that have no trade agreements and low trust. All parties must ensure that the solutions proposed move from the current low visibility border further along the spectrum as little as possible. There is no inherent reason why physical controls at the border are needed to manage this new situation. That has been pointed out by, among others, Jon Thompson, Chief Executive of HMRC in his evidence to the Select Committee on Exiting the EU.

A proposal

The first critical piece of the solution is for the EU and UK agree not to impose tariffs on goods traded between them. Once again, the Republic of Ireland has clear interests here.

If tariffs apply on food and agricultural produce, for example, the UK may be forced to unilaterally eliminate all tariffs on these products from all countries (as would be required by WTO rules that do not allow tariffs to be applied selectively between countries, other than pursuant to a full free trade agreement) to prevent damaging food price inflation. Irish farmers would then face direct, unprotected competition from farms in Argentina, Brazil and many other places. This could lead to a decimation of the Irish beef industry.

Logically, the Irish government should now be lobbying for an FTA with zero tariffs and maximum trade facilitation in Brussels, not in London where this is already the desired outcome. European Council draft guidelines issued this month, and the Irish government’s own position paper in May 2017 envisage a free trade agreement. We therefore assume a zero-tariff FTA is agreed between the UK and the EU for all goods, which removes the biggest single financial impact of leaving the customs union – as import duties on originating goods will not be payable.

The question then becomes how to manage the formalities associated with trading from outside of the customs union and single market.

A bilateral border arrangement can be agreed as part of an FTA, or even in the absence of a full FTA, under an exemption for frontier traffic under the General Agreement on Tariffs and Trade (GATT). The following sets out the foundation for a bilateral UK-EU highly facilitated solution for the border on the island of Ireland, much, though not all of which can also be applied to the other UK-EU borders, including the Irish sea border between the Republic and Great Britain.

This is not about avoiding a border (there is a border now) or waiving the rules that protect the integrity of the single market. It is not purporting to eliminate frictions entirely – there will be new administrative processes for businesses that currently do not trade outside the EU, and associated costs. The aim is to minimise the costs and frictions both for commercial reasons and, perhaps more importantly, to avoid the need for physical infrastructure at the border to respect the commitments made in the Phase 1 Joint Report and to avoid increasing tensions and disrupting daily life in border communities.

This can be done by bringing best practice and available technology solutions to bear in an environment of trust and regulatory compatibility.

It will be hugely beneficial to have a bilateral border process in operation, as is the case in the Norway/Sweden border, and as described in the “Smart Border 2.0” report for the European Parliament by customs specialist Lars Karlsson. But if this cannot be agreed, and the EU requires the Republic of Ireland to deploy a standard third country approach to its trade with the UK, including Northern Ireland, there are many options available to the UK in dealing with imports that will mitigate the impact.

The UK could reasonably take the view that goods on the EU market are safe on the UK market, thus eliminating regulatory compliance checks, and implementing the customs facilitation outlined below for imports.

However, assuming that the political will committed by the EU in its Article 50 negotiating mandate, and by the Irish government in its position paper is acted upon, a cooperative bilateral solution can be implemented. It could work like this:

An Irish border protocol in the FTA would provide for two categories of trader in Northern Ireland and the Republic of Ireland: large traders based on a defined value of consignments imported, number of employees, turnover or similar; and small or local traders (essentially those trading below the large trader thresholds or whose trade was entirely confined to the border zone, for example dairy farmers and their suppliers).

Every company or individual wishing to trade on a commercial basis across the border would need to notify the authorities and agree which category they fell into.

It has been estimated that there are only 53 businesses employing more than 250 people established in Northern Ireland who export goods to the Republic. Some 92 per cent of cross-border businesses employ less than 50 people. If these smallest businesses could be largely removed from the border processes, support could be focussed on the larger business, especially those who do not trade outside of the EU at present, to enable them to manage customs declarations and rules of origin and meet the standards to qualify for trusted trader programmes.

These traders would have to meet capability and security requirements, which would be supervised by HMRC or the IRC under arrangements similar to the existing Authorised Economic Operators (AEO) scheme. In exporting to the other side of the land border, they would complete all necessary paperwork in their home jurisdiction, submit it to authorities on both sides through a joint platform certifying their goods were safe to be released, rules of origin requirements had been met (or if not, the amount of duty that would be due), and so on, and stating the destination of the goods (for example, the importer’s premises). The consignment would then be sealed to cross the border. The initial declaration would be sufficient to release goods onto the market with formal completion of the process. Any necessary payments of VAT and import or excise duties would be made afterwards.

This process will be familiar to businesses who currently import from outside of the customs union. The difference proposed for the Irish border here is that all imports from the other side will be covered either by the small trader exemption (under which no per consignment declarations will be required, but authorities will monitor and enforce in order to ensure that thresholds are not breached) or the large trader facilitation, so there will be no need to distinguish between carriers at the border itself.

Rather than requiring a full declaration and payment (if applicable) for every consignment imported, declarations and payments for large traders would be made on an aggregated basis over an accounting period, for example monthly. This self-assessment process is possible already under the Union Customs Code and we have already recommended that the UK implement it for all suitable traders. It is similar to how other taxes and duties are collected.

Small or local traders would be required to submit simpler returns, perhaps quarterly, showing the extent of their trade, its destination, standards conformity, and make any necessary payments (if they have imported goods that do not meet rules of origin).

It would be possible to monitor aspects of this process remotely with electronic or bar code tagging, number-plate recognition, secure smartphone apps, GPS tagging, and so on. The technology for all aspects of this already exists in other contexts (police in Northern Ireland already use ANPR for motoring offences; GPS tagging is used to get concessionary car insurance rates). However, this is not in itself necessary to make the system work. The low volume of cross border trade and the constraints of geography mean away from the border checks (routine at business premises and mobile spot checks) are viable for overseeing compliance.

The system could be controlled by ensuring matching of declarations for import and export; routine checks behind the border either at designated trade facilitation posts (as envisaged by the Irish Revenue Commissioners) or at importer and exporter premises; spot checks by mobile compliance units; and trading standards, food standards and other relevant agency enforcement.

If over time these methods are found to be ineffective, and customs revenues are being lost, technological means described above could be deployed on main roads. Behind the border checks could be used for small, local crossings.

This would clearly require close collaboration between the Irish and UK customs services, ideally a seamless electronic system operated jointly by the UK and the Republic a so-called “green channel”.

It would also ideally require the simplest possible data submission arrangements for companies, i.e. the “single window” which has been partially introduced in the UK but not yet in Ireland, which enables all data required by the multiple agencies operating at a border to be submitted through one portal. And it would require the EU to accept that its customs code (UCC) would be enforced differently with respect to small traders and to accept the minimal risk involved. A joint EU/UK committee could manage the issue (as is the case with the EU/Switzerland border) and the EU would have a right to inspect in Northern Ireland (and vice versa if controls were enforced both ways).

In principle this system would work in the same way for agriculture as for manufactured goods – the formalities would be analogous. They would work most smoothly if the EU and UK had mutually recognised SPS standards as, pursuant to WTO rules, they clearly should, since the objectives achieved by their SPS regulation (and at present even the regulations themselves) are the same.

In the absence of mutual recognition (of both regulations and conformity assessment), Northern Irish exporters of meat products and animals would need to be approved in line with EU requirements for imports and the goods would need to pass through (and potentially be inspected at) BIPS, as described above. The UK shouldn’t reciprocate, as it should be taking an approach of openness, both to minimise frictions for importing businesses and to establish a strong record of compliance with the WTO agreements. However, the risk of reciprocation of full third country requirements on meat and dairy products from the Republic of Ireland to the UK, even in transit, is viewed with serious concern by Irish farmers.

The EU has a formal Mutual Recognition Agreement with New Zealand that lists a number of sanitary measures in animal health and animal derived products where the parties have agreed recognition of equivalence, and associated reductions in border checks. CETA also includes a protocol in similar terms. For the UK and EU it would be reasonable to expect, at least while the respective regulations are still harmonised, to recognise not just the regulations but also the testing and enforcement regimes and thus avoid the need for controls at BIPs.

Special Economic Zones

Special Economic Zones (“SEZs”) could be also established to ease border concerns by allowing import, processing and export to be carried out with minimal formalities by businesses established in the zones. They would also be a focus for investment on both sides of the border with a preferential tax and regulatory environment. The UK should prioritise this, to generate economic activity in Northern Ireland, in the way that SEZs are used in the rest of the world. If the EU could be persuaded to help the Republic by agreeing a contiguous SEZ in the border region, perceptions of a border on the island of Ireland would be further reduced. At any rate, it would be up to the EU to explain why they chose not to cooperate in this manner.

The Belfast Agreement and the Joint Report

The joint report of progress during phase one of the Article 50 negotiations issued by the UK and EU negotiators in December 2017 did not effectively compel the UK to agree to Northern Ireland staying in the EU customs union and single market, as has been claimed by commentators in Britain and, judging from the draft withdrawal agreement it produced, by the European Commission.

It is notable that in any event, phase one was only intended to achieve “sufficient progress” towards dealing with the priority matters of citizens, finances and the Irish border, not to represent a final deal and the joint report itself envisages working towards agreed solutions. In its position paper the Irish government acknowledged that:

“The great bulk of these issues [faced by Irish businesses] will be addressed not in the initial Article 50 withdrawal agreement but in the subsequent EU-UK future relationship agreement or agreements, and also in the context of any transitional arrangement designed to bridge the gap between the exit agreement and the future relationship agreement or agreements.”

This pragmatic position was seemingly side-lined for political reasons, aimed at pushing the UK government to agree to the whole country, or at least Northern Ireland, staying in the customs union and single market.

That is not to argue that the UK should renege on the commitments in the joint report. Rather, it is clear that it is possible to achieve them, in particular a border with no physical infrastructure, without compelling Northern Ireland to remain in a regulatory zone with the EU. Doing so in fact goes against the letter and the spirit of the joint report and the Belfast Agreement. #

Even if the fall-back position of “full alignment” set out in paragraph 49 of the joint report, is required to apply in the absence of agreed solutions, it is wrong to assume that alignment is intended to mean complete adherence to the EU acquis. Terms like “regulatory alignment” must be interpreted according to standard principles of international trade and regulatory law and practice. It does not mean identical or complete acceptance of the acquis.

Generally, in a trade context, alignment means alignment or regulatory goals. Furthermore, the alignment envisaged here is stated to be required only where it supports the protection of the 1998 Agreement, North-South cooperation and the all-island economy. It has been suggested that the fields of cooperation under the Belfast Agreement are so wide ranging, that the alignment would be comprehensive.

In fact the establishment of forums for cooperation under the agreement was to accommodate differences, and coordinate policy implementation so the existence of divergence does not undermine the Belfast Agreement – quite the reverse. If all policies and regulations were harmonised it would not be necessary to operate a joint committee for coordination at all. And as noted above, except in relation to the electricity market and, arguably, agriculture, while there is significant integration within the immediate border areas, there is no real all-island economy as such.


No one denies the importance of the Irish border issue or the critical importance of ensuring that nothing disturbs the peace process. However, the kind of rhetoric that has been deployed is not constructive, and deals in emotion rather than facts.

Across a border of relatively limited economic significance (in international terms) such as the Irish border, it is possible to manage the necessary customs and related controls on a point to point basis rather than at the border itself.

The technology and processes exist for larger traders, and can be developed for smaller local traders. But it will need goodwill (on all sides) and a determination to make it work. A report commissioned by the European Parliament has suggested that the resulting solution could be a highly streamlined border that works better than any other border in the world.

It makes no sense for the UK to see its entire customs or goods arrangements constrained by the supposed needs of the Irish border, given that the problem of that border can be resolved.

Assuming an implementation period, the UK and the EU have between now and December 2020 to work on this. A lot can be achieved in that time, but only if the political will is there. There is no good reason why the Irish border question should dictate both the EU-UK FTA, and the future of the UK’s economic, legal and constitutional order for generations to come.

Shanker Singham is Director of the International Trade and Competition Unit at the IEA.