Meanwhile, the protocol erects significant barriers to the vastly greater flow of trade between Northern Ireland and Great Britain. These barriers threaten to have a severe negative impact on living standards in Northern Ireland, leading to higher prices and large-scale inefficient trade diversion. Long-term losses of Northern Irish GDP of 3-4% – and possibly much higher – are plausible. These consequences warrant intervention by the UK government to radically dilute the operation of the protocol.
The Northern Ireland Protocol (NIP) has been in operation for just six weeks but is already causing serious problems, with growing calls in Northern Ireland (NI) and other parts of the UK for it to be scrapped. Yet not very long ago, many politicians were suggesting the NIP offered NI ‘the best of both worlds’ – free access to both UK and EU markets for its exports. What has gone wrong?
The main rationale for the protocol was to keep the border between NI and the Republic of Ireland (ROI) open. But to allow this after the UK left the EU’s economic sphere, the EU insisted that there would need to be customs and regulatory checks on goods moving from GB to NI to prevent potentially non-compliant goods from crossing the Irish border and ‘distorting’ the EU single market.
A quick look at cross-border trade flows in Ireland shows that the notion that these could seriously impact the operation of the EU single market is fanciful. According to UK government estimates, NI sales to the Republic total around £2.7 billion per year. Irish trade statistics show an even lower figure, of around £1.5 billion. These flows are tiny. They amount to about 0.02% of EU GDP, and 0.2% of EU imports from non-EU countries (Chart 1).
Indeed, NI exports to the Republic only amount to 2% of total ROI imports (half the share of forty years ago). These flows are incapable of destabilising even the small Irish market, let alone the continent-wide economy of the EU. The Irish Revenue made this point themselves back in 2017, stating that ‘something that will distort the EU market will not be sourced from Northern Ireland into the South’. They further argued that the volume of cross-border trade was sufficiently low that there would also be no justification for setting up customs posts to check it.
In contrast to miniscule Irish cross-border flows, NI’s trade flows with Great Britain are vastly greater and amount for the majority of the province’s external sales and purchases. Sales of goods by NI to GB account for about 55% of total NI external sales of goods and are four times larger than NI sales to ROI. On the import side, the dominance of GB as a trading partner is even greater. Around 65% of external purchases by NI come from GB, with the value of purchases from GB about six times higher than from ROI (Chart 2).
It is true that a substantial number of NI firms trade with ROI and thus have an interest in trade with ROI remaining unrestricted. But even in this group of firms, GB is a much more important trade partner – among firms exporting to ROI, sales to GB are almost double those to ROI and purchases from GB are four times higher (Chart 3). Similarly, claims sometimes made that NI firms selling to GB are dependent on inputs from ROI are wide of the mark; NI businesses selling to GB source five times more purchases from GB than they do from ROI (see here).
Given the dominance of the GB market for NI exporters, it is very positive that they retain free access to it under the protocol. However, the protocol does erect barriers to NI purchases from GB, which is a big problem for a small economy that is closely integrated with the large UK economy – these new barriers are bad news for both NI exporters and importers from GB.
There are problems for NI exporters because they import a lot of crucial inputs from GB – about 30% of the inputs used by NI businesses come from GB, twice as much as from ROI and the EU combined. These inputs are now potentially subject to customs and regulatory checks and possible delays. A recent survey of NI manufacturers showed around a quarter of them had seen a significant negative impact from new trade barriers which was likely to persist.
The structure of the NI economy also makes the imposition of new checks on purchases from GB particularly onerous. NI firms tend to be small, and so lack the capacity to deal with compliance with paperwork costs. NISRA data shows high dependence on GB-sourced inputs even for the smallest firms with 0-9 employees. The NI economy is also much more dependent on agri-food than the UK as a whole (agri-food is around 20% of NI external sales), and as noted earlier the additional costs for NI agri-food producers sourcing ingredients and raw materials from GB are likely to be high.
Problems for NI retailers, wholesalers, and consumers are even more serious. The retail and wholesale sectors in NI buy seven times more goods from GB than from NI and eight times more than from the rest of the EU (Chart 4). Analysis by the NI Department for the Economy shows enormous dependence across retail sectors on goods bought from or sourced via GB, including, crucially, foodstuffs.
Raising barriers to imports from GB is therefore potentially costly. But how costly? The main problem area is agri-food products which are now subject to the EU’s sanitary and phytosanitary (SPS) control system when moving from GB to NI. This is already causing significant problems, with some classes of goods (e.g. chilled meats, many plants) entirely excluded while others are subject to delays and additional costs.
The situation is set to get a lot worse in because various ‘grace periods’ have been put in place which reduce the extent of checks on agri-food. When these expire, food products from GB will face the full panoply of EU SPS checks and the additional costs associated with this could be very substantial. A variety of international studies (see for example Cadot & Gourdon, Disdier et al.) suggests such SPS and other regulatory controls on agri-food products impose costs equivalent to 20-30% of the value of the goods being traded. US studies suggest that in some cases, the costs are far higher and effectively prohibit trade flows.
If these costs were passed on to NI consumers this would represent a big hit to their disposable incomes. The Family Expenditure Survey shows 15-20% of consumer spending in NI is on food & drink, so that a 25% rise in costs for these items would imply a 4-5% drop in consumer living standards. This could be mitigated to an extent by NI retailers could turn to other sources of supply, but there would be significant additional costs from doing so – the ROI market is small and would have only limited capacity to replace GB suppliers in NI. Purchases from the rest of the EU would have to come the long route by sea to ROI and then across the border, also adding costs. These purchases would also involve a currency risk component. In this context, it’s worth noting that supermarket grocery costs in ROI have been estimated to be some 20% higher than in NI.
Another way to look at potential new border costs is to compare with estimates of the costs arising from new trade barriers between the EU and UK post-Brexit. Such a comparison suggests the potential costs to NI from the Irish Sea border to be much higher. UK goods trade (exports plus imports) with the EU is around 18% of UK GDP, while NI trade with GB is almost 50% of NI GDP – almost three times higher (Chart 5).
With NI sales and purchases from GB also more skewed towards agri-food than UK-EU trade, the weighted average rise in trade costs for NI from the Irish Sea border would be somewhat larger than for UK-EU trade. These trade costs would be concentrated more on the import side for NI but would be non-negligible for exporters too given dependence on GB inputs.
If we assume Brexit raises trade costs for non-food goods by 5% and for food goods by 20% (broadly in line with the international evidence on the incidence of non-tariff barriers) then this implies a cost for the UK of around 1% of GDP. But for NI if we assume a full incidence of such costs on imports from GB and 1/3 of the costs for NI exporters to GB (reflecting pricier GB inputs), the overall rise in trade costs for NI would be closer to 3% of NI GDP. That implies costs of around £1.5 billion for an economy of £49 billion, or about £820 per person per year in NI.
This methodology is conservative – in line with our general cautious approach to these border costs issues. Other approaches could produce much bigger numbers. A recent study of the costs of independence for Scotland – which has a similar share of trade with GB to that of NI – posited costs of 6-7% of Scottish GDP. Adding possible hits to productivity in the long run (possibly more plausible for a small economy like NI, with trade skewed towards goods, than for a large and diversified one like the UK where services exports are far more prominent) even larger numbers are possible.
From the above analysis it should be clear that the NIP is economically irrational for NI. We do not think there are many offsetting upsides. While some observers have been keen to talk up NI as an attractive destination for future FDI, we are doubtful about this. NI is not getting any extra market access from the protocol, and given its unfavourable geographical location and disruptions to supply chain links with GB, it is questionable how attractive it looks as a base either for EU companies selling to GB or GB companies selling to the EU.
What is to be done?
What can be done to remedy the situation? There are three main approaches. The first, which the EU is keen to push, is for the whole UK to align with EU agri-food regulations. This would certainly ease NI-GB trade flow issues. But it would also negate a large part of the potential trade upside for the UK as a whole from Brexit. It would render trade deals with the likes of Australia and New Zealand of limited value, as while tariffs might be cut by such deals, the remaining SPS barriers would still be prohibitive for many goods (as Canada has discovered post the CETA deal). SPS alignment with the EU would also almost certainly scupper the UK’s bid to join the CPTPP Pacific trade deal.
The second is for the UK to either negotiate major easements and derogations to the protocol with the EU. We would not hold out much hope on this score. A veterinary agreement between the UK and EU such as the EU has with New Zealand has been suggested as a way forward, and this would ease checks to some extent. But the UK pursued such an agreement in last year’s talks with the EU and the EU refused. It is not obvious why they would agree now. Similarly, while the ‘mutual enforcement’ approach recently suggested by the Centre for Brexit Policy is intriguing, there is little chance the EU would agree to it.
The third is for the UK to override key aspects of the NIP and impose a very light-touch regime of checks on goods flow to NI – along the lines the UK government itself previously outlined during 2020. The protocol arguably allows this, in particular because it specifies trade diversion as being a ground for activating Article 16 which allows parts of the protocol to be suspended. Trade diversion, and highly inefficient trade diversion at that, is definitely going to be the result if the NIP stays in place and as the various grace periods expire.
The obvious approach for the UK government would be to expand the ‘at risk’ goods concept which is currently related to customs issues only to cover regulatory issues too – and then exempt all GB to NI goods flows going to a recognised end-customer in NI from any checks. Careful monitoring of trade flows and in-market surveillance either side of the Irish border could be put in place to ease any concerns about non-compliant goods spilling into ROI.
Harry Western is the pen-name of a senior economist working in the private sector.
