Saturday, 31 March 2018

A Brexit Turkey (part 3) - Customs Union is not a Brexit solution

Some bad ideas seem reluctant to die. The Institute of Directors (IoD) have published a "partial" Customs Union proposal (based on Turkey's Customs Union agreement with the EU), primarily to eliminate preferential Rules of Origin (RoO) in UK-EU trade.

In my previous post, I discussed how  preferential RoO are far from insurmountable. How significant are the costs associated with preferential RoO ? Open Europe assumed leaving the EU Customs Union would add 4% to cost of trade transactions (RoO costs & customs admin) leading to a 1% loss of GDP. Other studies suggest RoO adding less than 2% to transaction costs. One detailed study suggests RoO is a one-time cost that does not increase with increased volume of transactions.

A Customs Union seems a very blunt instrument to use, especially as trade with 3rd countries would still be subject to RoO. Are there other ways to address this issue ?

Rule of Origin Solutions

Blockchain technology has been touted as a way to simplify the administration & auditing of complex supply chains, hence minimising RoO administration costs. However, as promising as this technology may appear, we are probably a decade away from fulfilling this potential.

recent paper examined preferential RoO for the Food & Drink Federation, a sector which makes particular use of global supply chains:
- Five case studies are examined, identifying challenges provided by use of non-local content in qualifying for tariff-free trade via preferential RoO.
- Both Pan-European Mediterranean (PEM) RoO and CETA RoO are compared for each case.
- The first case study is UK wholemeal bread, which uses flour milled in the UK from grains sourced from UK, US and Canada: PEM RoO require grains to be sourced entirely locally;  CETA RoO limits non-local grains to 20% by weight of final product. Otherwise the EU tariff of 9% would be payable on exports to the EU.

Some context is required here: 
- Bread is primarily produced and consumed locally: 85% of wheat used by UK flour millers is home-grown, with most of the imports from Canada; UK domestic sales amount to £3.5bn per annum ("Bakery & Snacks"). 
- By contrast, UK bread exports to the EU amounted to just £87.5m for the first 9 months of 2016, i.e. ~£117m per annum ("BakeryInfo"). 
- The UK runs a massive trade deficit with the EU across the whole food and drink sector (Food & Drink Federation statistics).

The Food & Drink Federation paper makes a number of suggestions for addressing these RoO issues, which could also apply to others sectors:
- Allow 10% non-originating product content without losing originating status (already provided in PEM RoO, under article 5.2 (a) );
- Full bi-lateral cumulation in EU-UK trade so that EU content counts as local in UK exports to EU and vice versa - almost certain to be part of an UK-EU FTA;
- Full diagonal cumulation with EU FTAs, e.g. so that UK, EU and Canada can count each other's content as local in trade between themselves;
- Exempt Least Developed Country (LDC) origin content in UK-EU trade so that LDC content counts as local
- Exempt products / content where EU & UK have same MFN tariffs. i.e. count products as local where there is no risk of trade diversion arising from different MFN tariffs - a form of halfway house to the UK governments customs partnership concept);
- Allow origin to be determined on final value OR weight criteria (protects products where local processing adds high vale to a premium product);
- Simplify RoO documentation.
Another approach would be to abandon preferential RoO in favour of simpler non-preferential / Most Favoured Nation (MFN) RoO (as suggested by Hosuk Lee-Makiyama). Non-preferntial RoO typically determines origin on the simple basis of where last substantive processing was performed, rather than where inputs are sourced :
"... the world trading system could do without preferential rules of origin. Preferences can be granted on the basis of most favoured nation (MFN) rules of origin anyway. Empirical literature suggests that, if the purpose for enacting preferential rules of origin was to facilitate commerce or promote inward investment, then their implementation has in practice defeated the purpose. Beneficiaries of preferences often prefer to trade using MFN rules of origin, rather than going though cumbersome procedures to show that they can ‘benefit’ from preferential rules. Thus, in the end, preferential rules of origin are neither necessary for preferences to be granted, nor have they facilitated trade or investment. Our policy recommendation for the negotiators of the Harmonized Working Programme (HWP), which aims to establish common rules of origin for all WTO members, is to also decide to outlaw preferential rules of origin. "   Abstract from "The case for dropping preferential rules of origin" by Edwin Vermulst and P. Mavroidis 
Adopting amended or simplified RoO for UK-EU trade will of course be entirely dependent on EU agreement, which may not be forthcoming. Is there perhaps a more simple and radical way to avoid preferential RoO ?
- IoD's contention is that (i) preferential RoO is a significant trade barrier (ii) tariffs on manufactured products and processed foods are in any case low - so low that they offer no leverage in securing trade agreements with third countries.
- This actually chimes with arguments for "Unilateral Free Trade" as proposed by Economists for Britain, who argue that we should ignore preferential FTAs and instead unilaterally lower tariffs - favouring UK consumer interests (including  manufacturers sourcing inputs) over producer interests. Preferential RoO is removed as a consideration for all imports and exports. Attention can then be focused on removing regulatory / technical barriers to trade.
- Unilateral Free Trade offers a way to eliminate preferential RoO, without requiring EU agreement. If as IoD and others argue, tariffs are sufficiently low to not be an issue, why chase preferential trade deals - even with the EU ?

Partial Customs Union & Irish border 

IoD claim that a partial Customs Union like Turkey's would help alleviate issues with the Irish land border. But Turkish experience does not support this argument, with long lorry queues at the Turkey-Bulgaria border. A number of issues arise:
- Basic agricultural goods are excluded from the Turkish Customs Union and require full customs clearance is still required for these goods. Meat / dairy constitutes a significant portion of UK / RoI trade.
- An A.TR movement form must be presented at the Turkey-EU border, declaring all customs formalities have already been completed.
- Turkey is outside the EU's VAT union, so import VAT is also payable at the border. Does the IOD intend for UK to remain in the EU's VAT union as well ? (meaning UK could not cancel the "tampon tax" or eliminate VAT on household energy bills for example).

    The IoD's proposed "partial" Customs Union would not remove the need for a customs border.

    The IoD's argument for "partial" Customs Union rests entirely on RoO. RoO does not provide an insurmountable barrier. RoO administrative costs are subject to debate, appear low and may be negligible beyond an initial cost.

    A "partial" Customs Union would only avoid RoO for trade with the EU - any preferential trade with third countries (including other European countries such as EFTA  members) would still require preferential RoO.

    A "partial" Customs Union would not allow an independent UK trade policy, and will keep the UK tied into an EU-centric trade policy. This is counter to UK's strategic interests. The UK runs a huge trade deficit with the EU, whereas UK trade with the Rest of the World is broadly in balance. UK exports to the Rest of the World have grown and overtaken UK exports to the EU, mirroring the EU's diminishing share of global GDP.

    There are other ways to mitigate the impact of RoO. The worst case is that some supply chains may be repatriated to the UK - given the huge trade deficit the UK has with the EU, this may be no bad thing.

    Ironically,  Unilateral Free Trade offers a way to eliminate preferential RoO, without requiring EU agreement. If as IoD and others argue, tariffs are sufficiently low to not to be an issue, why chase preferential trade deals - even with the EU ? Perhaps a hybrid solution could be adopted - eliminate tariffs where UK has no domestic industry to protect (e.g. tropical fruits !) and also for inputs to manufacturing (e.g. car parts would be zero-rated), but retain higher-rated EU tariffs for key sectors / finished products to use as leverage in trade agreements (e.g. retain 10% tariff on finished cars, ~40% tariffs for meat etc.). The UK could then seek to negotiate FTA's to eliminate these high rated tariffs using non-preferential / MFN RoO.

    In short, we should reject  the IoD's proposal to keep the UK locked into a 19th century Zollverein Customs Union, designed primarily to promote European political integration. As Brexit solutions go, the IoD's proposal is a real turkey.


    Posts in this series:


    1. Paul - the IoD are being misleading as we have discussed before. Their proposal would do very little to solve the Irish border issue; it is single market membership that is the main reason for the open goods border.

      On rules of origin - the literature on this is thin and inconclusive but more recent papers have tended to suggest they are rather less of a barrier than previously thought and preference utilisation higher. One reason for this may be that FTAs have other elements(on customs, regulatory coop) that make them useful.

      Final point - a lot of this customs union push is from the car industry. With EU MFN tariffs at 10% there is an issue here. But what happens if these get slashed as a result of pressure from Trump (eg to the US level of 2.5%). Possibly a bit of game changer.

    2. At the turn of the century, the car industry were strong backers of joining the euro, on the basis that UK car industry would be decimated due to extra burden of currency transactions compared with european competitors. Since then, UK car industry has thrived and UK content in UK cars has increased from low 30's% to mid 40's%.

      It's not clear why the car industry would believe a Turkey solution is essential compared with an FTA with bi-lateral cumulation and continued customs co-operation / AEO scheme (as per Japanese letter to HMG). As you say, RoO costs are lower than has been historically supposed, car companies have advantage of scale to minimise RoO costs and do have small (but growing) sales via preferential FTAs (~3% to Turkey, ~1.7% to South Korea, ~1.2% to Israel - see

      A Turkey style Customs Union with the EU does not help with third country FTAs and meeting local content threshold. I suspect what car companies really want is to remain in "the" EU's customs union and continue to benefit from EU's FTAs and count EU content as local (aka EU membership). The obvious answer is that the UK needs to boost the current program of increasing UK car content in UK cars, and here UK government could/should help with investment in UK supply chains.

      Alternatively the UFT approach at least provides car companies lowest input prices on world market. If US did succeed in pressuring Eu to lower its imports tariffs of 10% on finished cars, that would make UFT option very viable for UK exports to EU.

    3. Paul - yes I am puzzled by that too. Based on my calculations it's hard to believe that any of the major producers would have a problem with the content thresholds under an FTA. At least for exports to the EU. Conceivably there could be a problem with exports beyond the EU but I wonder what share of these is even covered by preferential FTAs where RoO would apply.

      Not a lot I think - looking at the detailed data, of UK car exports of $42bn in 2016, $16bn were to China, Japan, Australia and the US. That is already the bulk of the non-EU total. RoO would not apply in any of these cases.

    4. I've looked at lots of reports over last few years trying to understand car industry position on Brexit, including this recent one one:

      The ACEA's main concern RE customs union appears to be speed through customs processing and impact on JIT supply chains. But they also recognise that via customs facilitation (including self-assessment, electronic information exchange between customs authorities, expansion of AEO scheme etc) "the existing invisible border between the EU and the UK can be maintained". This chimes with Japan's letter to HMG, which mentioned RoO only in the context of concerns over delays at customs (i.e. checking of paperwork at customs border including RoO paperwork) and called for continuation of the AEO scheme post-Brexit to continue fast processing through customs.

      The only mention of RoO in the recent ACEA report is in context of local content threshold in third country trade via FTAs (which would not be resolved via a Turkey-style Customs Union with the EU).

      The ACEA report also majors on regulatory issues, notably whole car vehicle approval, an issue which will of course cut both ways (EU car sales to UK) and is of course unrelated to the customs union question. I hope (probably forlornly) that ACEA & the UK car industry realises regulatory issues are based primarily on UNECE WP.29 - where automatic component type approval already covers much of the EU car approval scheme & UNECE are working towards an international whole car approval system (at behest of the EU). The EU's own interests RE their own car industry and their regulatory initiatives at UNECE should be basis for agreement.

    5. Interesting Paul, and I agree this really shouldn't be a problem. You might also note that some of the recent literature about NAFTA renegotiations has a strangely familiar ring to it, notably comments along the line that NAFTA has allowed the creation of an integrated North American auto industry where 'components sometimes cross borders many times'...(see link below)

      Yet of course NAFTA is only an FTA, not a customs union and not a single market.

      1. Unfortunately, I can't see that article behind the FT paywall. I'll have to wait for my next free pass. Talking of USA, it is of course interesting to note that German car manufacturers already make use of USA in some of their supply chains.

    6. Paul indeed- and you'll note UK carmakers also do this. Sorry re. the FT; try this instead

      ‘ the U.S., Canada and Mexico are economically very integrated with goods and services often crossing their borders many times in order to produce a final product.’

    7. Paul you should read, and tweet, this