Recommendation 1: In the near future, maintain as much as possible the existing structure of applied tariffs.
Recommendation 2: Reducing firms’ imported input costs in order to increase their competitiveness is potentially a sensible strategy, but is not without its downsides.
Recommendation 3: Decisions regarding the structure of applied tariffs should be made bearing in mind the possible regional consequences of any changes.
Recommendation 4: The analysis of the impact of tariff changes on prices and households should be undertaken at a detailed level.
Recommendation 5: Before eliminating tariffs on goods not produced within the UK, the impact on developing countries should be evaluated.
Recommendation 6: We strongly support the idea of encouraging trade in environmental goods, but suggest that any such policy should not be based on the existing lists. Instead the government should produce a list based on scientific evidence.
Recommendation 7: The Government should develop a broader approach to environment-friendly or climate-friendly trade policy. It should consider both tariffs and non-tariff barriers to imports and exports, and should be closely integrated with the Government’s domestic environmental policies.
Recommendation 8: The Government should consult widely, not just on tariffs, but also on non-tariff measures and trade-related regulations, and the coherence of these policies with domestic policy objectives.
The UK left the European Union on January 31, 2020. While the UK remains bound by the EU’s trade policy during the transition period, preparations have now begun to establish the UK’s independent trade policy that will apply once the transition period ends.
One part of such preparations is establishing the UK’s independent tariff schedule that will apply to goods imported into the UK. Already in the summer of 2018 the UK submitted its schedule of bound Most Favoured Nation (MFN) tariffs to the WTO, replicating the tariff rates of the existing EU bound tariff schedule. These bound tariff rates represent the maximum tariffs that the UK could levy on imports from a third country (a country with which the UK has no trade agreement).
There is, however, nothing to prevent the UK from applying lower tariffs than those in its bound schedule if it so wishes. Indeed, in February 2020 the Department for International Trade (DIT) launched a public consultation concerning the UK’s applied Most Favoured Nation tariffs, which ran until the 5th of March 2020. In their consultation, the DIT invited views on a number of proposals of how to simplify and tailor the UK’s applied MFN tariff. In this briefing paper we outline the proposals under consideration, discuss their potential implications, and provide our recommendations on issues that we believe are important for the UK Government to consider when formulating the UK’s trade policy going forward.
The Taxation (Cross-border Trade) Act 2018 sets out the principles for the Government to take in to account in setting tariffs. These include bearing in mind the interests of consumers and producers in the UK, the desirability of maintaining and promoting the external trade of the UK and the productivity of the UK, and the extent to which the goods concerned are subject to competition. The Government should also aim to balance strategic trade objectives such as future free trade agreements with the commitment to developing countries to reduce poverty through trade.
The Government’s consultation document considers several changes to the UK’s applied MFN tariff. In this section of the paper we outline these proposals, and provide some summary calculations as to their impact on the structure of UK tariffs.
The Government’s proposals:
Before exploring what the impact would be of simplifying the tariff structure, it is useful to outline the current structure of the UK’s MFN tariff as a member of the EU. We take this as the starting point of our calculations below. As Table 1 shows, currently, around 26% of tariff lines are zero, and a further 11.5% are below 2.5%. Of the UK’s imports from ‘MFN countries’ (i.e. countries without any preferential agreement with the UK), 50% is in products where the current MFN tariffs are zero.
|Table 1: Current structure of EU MFN tariffs (2018)
|Number of 8-dig
|% of all
|% of imports from
|0.01% – 2.49%
|2.5% – 19.99%
|20% – 49.99%
|N.B. Based on 8-digit tariff data from UNCTAD TRAINS (including AVEs) and trade data from HMRC Overseas Trade Statistics. Both tariff and trade data for 2018. Imports are UK’s imports from countries without any preferential agreements with the UK (i.e. excludes EU, FTA countries, EBA/GSP countries). Any product where tariff data is missing has been excluded. AVEs are the ad valorem equivalent (percentage equivalent) of tariffs that are not currently expressed in percentage terms.
The proposal to eliminate tariffs that are 2.5% or less, and round down tariffs to the nearest standardised band, would involve changing the tariff on over 70% of the 8-digit product codes, which is more than 6,500 tariff lines. This reveals that in terms of the number of products that will be affected, the changes would be substantial. However, the overall impact on the UK’s MFN tariff will be extremely small; indeed only 1.8% of all tariff lines would see a change bigger than 5 percentage points. The weighted average applied tariff would see only a modest reduction, from 2.5% to 1.8%, and the simple unweighted average from 7.6% to 6.6% if we consider all goods (see Table 2 below). Following this change, UK imports of just under £27 billion from ‘MFN countries’ which currently face tariffs of up to 2.5% would be tariff free. If we exclude non ad-valorem tariffs (i.e. those not expressed as simple percentages) the unweighted average tariff on imports from MFN partners would fall from 4.8% to 3.9%.
Roughly 12% of all UK MFN tariff lines are so-called non-ad-valorem tariffs. These tariffs are not measured in percentage terms but rather levy a fixed charge per unit of a good (such as the EU’s tariff on fresh bananas of 114.00 EUR per 1000 kg). Some goods also face a two-part tariff, combining both an ad valorem tariff and a specific tariff (such as the tariff on goat meat (02045031), which faces a two-part tariff of 12.80% + 222.70 EUR / 100 kg). The percentage burden of a specific tariff therefore depends on the price the importer pays for the good, which may vary across time and across countries. Various methods have been suggested for estimating the ad-valorem equivalent (AVE) (i.e. percentage rates) of such specific tariffs; here we use AVEs estimated by UNCTAD.
Since the AVEs are a function of the unit price, which can vary considerably from year to year, the percentage equivalent of non-ad-valorem varies likewise. For example, the EU’s MFN tariff on item 2205901000 – Vermouth and other wine of fresh grapes – is 9 EUR per 100 litres; it has an estimated AVE tariff of 8.4% in 2017, but 46.1% in 2018. Similarly, the tariff on certain types of buttermilk, curdled milk and cream (0403903900) is 1.62 EUR/kg/lactic matter + 22 EUR/100 kg, which was converted into an AVE tariff of 152.2% in 2017, but 136.2% in 2018.
Thus, while converting the current non-ad-valorem tariffs to percentage terms may be helpful for importers and exporters, the time-variability of AVEs will make it a difficult task to decide at which percentage rates to set them if the aim is to maintain a rate of protection equivalent to that which these industries currently enjoy.
The second item of consideration in the consultation document is whether tariffs should be eliminated on intermediate goods. The underlying logic is that cutting tariffs on intermediate goods supports UK producers, as it lowers the cost of imported inputs, which in return may make producers of final goods more competitive.
The impact on the UK’s tariffs partly depends on which definition of intermediates is used for the analysis, and the consultation document suggests three options:
Table 2 summarises our estimates of the changes from simplifying the UK’s tariff schedule as per the consultation proposal (i.e. rounding down tariffs according to their relevant tariff band), and from setting all intermediate goods to zero. All values are based on tariff and trade data for 2018, although little changes when using 2017 data. The tariff simplification proposal applied on its own reduces simple (weighted) average tariffs to 6.6% (3.9%). In comparison, setting all intermediates tariffs to zero without the tariff simplifications would have a greater effect both on the % of imports affected (79.4% v 71.4%), and on the average tariff which goes down to 4.5% (2.8%). Overall, when combining the simplification and the elimination of tariffs on intermediates, around 84% of all the UK’s imports from ‘MFN countries’ would be zero, and average tariffs fall to 4.0% (2.4%). This shows that there is considerable overlap between the goods on the tariff simplification list, and when we interpret intermediates using the wider definition. It also show, that there is a larger impact on average tariffs from the change in intermediate tariffs than from the simplification element of the proposal.
|Table 2 : Impact on average tariffs and trade from DIT’s tariff proposal, applying changes to all UK’s imports from ‘MFN countries’
|% of imports with tariff = 0
|Total tariff revenue
|Current EU MFN tariff
|Zero on all intermediates
|Simplifying tariffs + zero on all intermediates
|N.B. Tariff data sourced from UN Comtrade at 10-digit (TARIC) level. 10-digit tariff lines have been aggregated into 8-digit codes using a simple average. Trade data sourced from HMRC’s Overseas Trade Statistics at the CN 8-digit level. Both tariffs and trade data reported here are for 2018. Data categorised into BEC categories using Eurostat concordance tables from CN to BEC. ‘AVE’ refers to Ad-Valorem Equivalent tariffs, and indicates where we have included non-ad-valorem tariffs and where these have been excluded.
The final element of the DIT’s proposal is whether to set tariffs to zero where there is no, or very limited, UK production. Although this is not stated in the consultation document presumably the justification for this is that, while it would lower prices for UK consumers and those UK producers who use these goods as intermediates, it would have no impact on UK production.
Based on the analysis above, we provide eight recommendations with regard to the proposed changes in the UK’s tariff structure.
Recommendation 1: In the near future, maintain as much as possible of the existing structure of applied tariffs. This require less adjustment by UK firms, will have less of an impact on Northern Ireland, and allows more negotiating flexibility in FTA negotiations.
First, for producers all tariffs are, in some sense, a nuisance. Hence eliminating those that are already low could be seen as a straightforward gain, because that nuisance is being eliminated and the level of protection is already very low. However, for a producer the nuisance associated with a tariff is less to do with the direct impact on price than to do with the bureaucracy involved. Given the large number of products for which tariffs would change, and given the small net effect of the change, one should question whether deviating from the existing tariff structure is worthwhile. Alternatively put, what are the costs and benefits of such a change? The proposed benefits of the simplifications under consideration are first to remove ‘nuisance’ tariffs (of 2.5% or less), and secondly that tariff banding will ‘support UK importers and remove complexity’. The proposed benefits of reducing tariffs on intermediates are to lower input costs and thus support UK manufacturing. However, there may also be also costs associated with these changes.
For example, approximately 50% of UK goods exports are destined for the EU. The more the UK MFN tariffs differ from EU MFN tariffs, the more important will be ‘rules of origin’ in any future free trade agreement (FTA) with the EU, and the greater the administrative nuisance cost. If the UK introduces widespread changes to its tariff structure, it is inevitably going to increase those nuisance costs – not just for those goods where tariffs are less than 2.5% but for almost all goods. Further, rules of origin are frequently a contentious issue in FTA negotiations. The greater the extent to which the UK diverges from the EU’s existing tariff structure, the more difficult those negotiations are likely to be.
Second, lowering the UK’s applied tariffs will have implications for trade between Great Britain (GB) and Northern Ireland (NI). Under the Protocol on Ireland/Northern Ireland in the Withdrawal Agreement EU tariffs would need to be levied on all goods being exported from GB to NI which are ‘at risk’ of then entering the Republic of Ireland. In the event of a free trade agreement with the EU, this could apply to all goods where the UK’s applied tariff is lower than the EU’s applied tariff. However, it might not need to apply to those goods where the applied tariffs are the same. There is currently considerable uncertainty as to how the border between GB and NI will operate, and which goods will be deemed by the Joint Committee to be ‘at risk’. If the UK Government wishes to minimise the impact on Northern Ireland, there are very strong grounds for at least initially maintaining as much as possible the existing tariff structure.
Third, while it is easy to assume that low tariffs have little effect and that it is therefore relatively costless to remove them, industries and/or in firms with low profit margins could suffer significant consequences from removing even a low tariff. This is not an argument for necessarily retaining those tariffs, but it is an argument that low tariffs could matter.
Fourth, and relatedly the Government is planning to negotiate a series of FTAs over the next three years. If all tariffs which are currently less than 2.5% are removed, this would remove tariffs on over 1000 8-digit product lines (see Table 1). In the hard world of commercial diplomacy these are 1000 product lines that the UK could have used as a non-negligible bargaining chip. If these tariffs are retained then they could help to improve market access for UK exporters arising from future negotiated free trade agreements.
In our view the benefits of the proposed simplifications do not outweigh the costs. This is not an argument for saying that the UK should forever simply apply the EU’s entire MFN tariff structure, but rather to say that simplification per se is not a sufficient rationale in a period of intense uncertainty and adjustment. Given the changes and uncertainties arising from Brexit and the on-going negotiations, it would be better not to make wide-ranging changes to the existing tariff structure. This does not preclude the UK from subsequently lowering its applied tariffs should it wish to do so in a more considered fashion.
Recommendation 2: Reducing firms’ imported input costs in order to increase their competitiveness either domestically or in export markets is potentially a sensible strategy. However, this requires more detailed consideration of the effective rate of protection in given sectors and industries, and a blanket approach e.g. based on the BEC classification may well not yield the desired results. Increasing effective protection is also likely to reduce the incentives for firms to improve productivity.
The proposal to remove tariffs on intermediate inputs raises two broad issues. The first concerns the impact this proposal would have on average on UK tariffs, and the second concerns the extent to which this policy would be effective in supporting UK firms.
As detailed earlier, with regard to the impact on average UK tariffs the additional effect of proposed changes would be relatively modest. The reason is partly because of the degree of overlap (circa 3500 products) between the so-called ‘nuisance’ tariffs and intermediates products, and partly because the remaining MFN tariffs on intermediates tend to be low.
While the impact on average tariffs is modest, there is a bigger impact on the value of imports affected. Our analysis indicates that with the existing tariffs just over 50% of the value of UK imports face a zero tariff. Based on existing trade flows the tariff simplification proposal would increase this to just over 71%, and removing tariffs on intermediates would further increase it to around 85% (Table 2). These latter figures are almost certainly an underestimate as they are based on existing trade flows. As tariffs are reduced, one would expect some reorientation of trade towards the lower tariff lines, which would increase the share in the value of imports coming in duty-free. This analysis suggests that removing the tariffs on intermediates could therefore have some impact on UK producers. However, it is important to recognise that whether this is positive or negative will depend on the ‘effective rate of protection’ (ERP).
The ERP takes into account the overall impact of a given tariff structure on a firm, including tariffs on both the final good and the intermediates the firm purchases. The effective rate of protection captures the effect of the tariffs on the value added of the industry. The lower the tariff on intermediates is relative to the tariff on the final good, the bigger is the degree of effective protection being granted to the domestic producer selling in the domestic market. Similarly, the lower the tariffs on intermediates, the more competitive domestic producers could be in export markets. All this suggests that lowering tariffs on intermediates could be beneficial to domestic producers, but it also indicates that the net effect will depend on any changes in the tariffs on final goods.
It is also important to remember that increasing effective protection allows firms to pay themselves more money and hence might reduce the incentives for them to innovate and seek efficiencies in their actual production methods as opposed to just benefitting from changes in the relative prices of inputs and outputs.
The box explains effective protection in more detail and gives some results. Briefly, the tariff simplification plus the elimination of tariffs on intermediates would reduce the average rate of effective protection from 3.9% to 3.6%, reducing the rate in two-thirds of the sectors we can measure and increasing it in the other third.
Box: the effective rate of protection:
As an example, consider a shoemaker that can sell a pair of shoes (the final product) at £100 in the international market. To produce a pair of shoes the shoemaker uses £50 worth of imported leather. In the absence of tariffs, the value added to the shoemaker is £100-£50 = £50. Now suppose that the tariff on shoes in the domestic market is 20%, but there is no tariff on leather. Then the price of shoes in the domestic market will be £120. The domestic producer matches the foreign price and sells at £120, increasing value added by £20. While the nominal rate of protection on shoes is 20%, the effective protection rate (i.e. the impact on value added as opposed to on the final price of the good) is 40% (20/50). If there were tariffs on the intermediate – the imported leather – then this would raise the shoemaker’s costs, and therefore reduce the effective protection rate. Suppose the tariff on both the shoes and also the leather is 20%, this would reduce the effective protection afforded to the shoemaker to 20%.
We have undertaken detailed analysis using the latest UK input-output tables, and detailed tariff and trade data at the 8-digit level of aggregation. These are then aggregated to the SIC07 categories used in UK input-output tables, using either simple averages or weighted averages within each industry. This enables us to calculate effective rates of protection (ERP) for 46 SIC07 industries. The results are summarised in Table 3. The first two columns of results give the simple and average rates of effective protection for the current MFN tariff schedule, the DIT tariff simplification proposals, the DIT intermediate tariff proposals, and finally the combined effect of simplification and intermediate tariff removal.
Table 3: Effective Rate of Protection for domestic producers
|Number of sectors where ERP goes up
|Simplification + intermediates
Our analysis suggests that the impact of the proposed tariff simplifications/banding would result in a modest reduction in the domestic ERP from 3.9% to 3.2% for ERP computed using simple unweighted tariffs, and from 3.7% to 3.0% for weighted tariffs. Reducing the tariffs only on intermediates results in a small increases in the ERP to 4.4% and 4.1% respectively. Adding the tariff simplification proposals to the reductions in tariffs on intermediate goods, reduces the ERP to 3.6 (simple) and 3.3% (weighted). Indeed the current MFN tariffs offer a higher degree of effective protection than the combined DIT proposals. However, if we add the BEC intermediate tariff liberalisation (row 4) to the tariff simplification proposal (row 2) then we see a modest increase in the ERP.
The last two columns of Table 3 indicate for how many of the 46 sectors the ERP rises, where we are comparing the current MFN structure with each of the two variants of the DIT’s proposals. Tariff simplification would reduce effective protection in six industries using simple average tariffs, and eight industries when we use weighted average tariffs. With the removal of tariffs solely on BEC intermediates all sectors see an increase in effective protection. However, when we combine the two proposals there are 14 (15) sectors for whom the ERP rises, leaving over 30 industries with a decrease in effective protection despite the reductions in intermediate tariffs.
This might seem paradoxical. However, the effective rate of protection depends on the difference between the final tariff and the intermediate tariff and both are changing here, and many goods, while notionally classified as intermediates are also the final outputs of many firms. We recognise that the actual change in the effective protection rate will be industry and firm specific and the data are not sufficiently disaggregated to capture this.
Nevertheless, the analysis sends an important message: Reducing tariffs on intermediates (under the BEC classification) could result in a modest increase in the domestic ERP. However, when combined with the simplification proposals, for some industries the ERP may decrease, and so fail to provide ‘support’ to UK industries. On the other hand, we find that reducing tariffs on intermediates would raise the competitiveness of UK producers in export markets (albeit by a very small amount). Exporters and domestic suppliers may therefore have conflicting interests.
Recommendation 3: Analysis of the impact of the proposed tariff changes on prices and households should be undertaken at a detailed level in order to assess the possible benefits to UK consumers.
One of the Governments’ principles underlying the consultation is the objective ‘to bear in mind the interests of consumers and producers’. While this is laudable, it is also vague. There are various plausible consumer interests – price, quality, variety, public health – and a given tariff change may be good for consumers in one dimension but not in another, and there may be different interests in the short run versus the longer run. One direct impact of changing the tariff structure could be on consumer prices. To assess this we have undertaken some detailed analysis which leads to several conclusions.
First, we have calculated an estimated change in consumer prices for 81 ISIC 4-digit sectors (73 of which are manufacturing and 8 agricultural sectors). Averaging the change in UK domestic prices across sectors, suggests that applying the tariff simplification / banding proposal and eliminating tariffs on products with zero UK production would reduce consumer prices on final goods by 0.24%. Once we weight sectors to reflect their importance in people’s consumption, the change in the average price of goods consumption is -0.15%. And if we then allow for the fact that UK consumers only spend around 40% of their income on goods as opposed to services we find an average impact of -0.06% on the cost of living. So, while there may be some products with particularly high ad-valorem equivalents which see a much larger reduction in prices we do not anticipate that the DIT proposals would have much effect on average.
There is another possible consumer impact and that is the impact on variety. We have not undertaken empirical analysis of this but it is worth noting that, first, variety is much more likely to be dependent on regulations than on tariff levels. Secondly, reductions in tariffs could increase variety by allowing for more imports of foreign varieties, or less variety if increased import competition drives reduces the variety offered by domestic producers.
Recommendation 4: Before eliminating tariffs on goods produced outside the UK the impact on developing countries should be evaluated.
Recommendation 5: Decisions regarding the structure of applied tariffs should be made bearing in mind the possible regional consequences of any proposed changes.
It is important to consider the impact of eliminating tariffs on goods not produced in the UK on developing countries. Many of the agricultural products which the UK does not produce are imported from developing countries (e.g. bananas, tea, or coffee). Most developing countries currently benefit from preferential access to the UK market through the Generalised System of Preferences (GSP) or Everything But Arms (EBA) schemes. The UK Government has indicated that it wishes in principle to maintain such schemes, though of course they may differ from that of the EU. The tariff preferences are in place to help poorer countries compete against richer ones. Eliminating MFN tariffs on such products would erode the preferences that developing countries currently benefit from, which could be detrimental to these countries, and to key sectors within those countries. This possibly unintended effect on developing countries further illustrates that trade (tariff) policy should be aligned with other policy objectives
With regard to both consumers and producers the Government needs to consider the coherence of tariff policy with its strategic structural objectives regarding the future direction of the economy. For example, not all producers will benefit from any given tariff change, be this with regard to nominal or effective rates of protection, and therefore the Government needs to bear in mind that’s policies may impact positively on some but not others, and thus needs to have clear criteria for prioritising the decisions taken. These impacts may be both on different firm types within sectors, as well as across sectors.
Trade policy, and within that tariff policy, should be seen as part of domestic policy, and should not be formulated independently of domestic policy. Thus the broad strategic objectives of domestic policy should be determined before designing trade policy. For example, the Government has indicated a strong wish to regionally ‘rebalance’ the UK economy. Yet there is no sense of this in the proposed tariff changes. If the Government is indeed serious about regional rebalancing, then some assessment should be undertaken of the regional consequences of the proposed tariff changes, and these should then be discussed with the key stakeholders, the business community and devolved administrations before decisions are made.
Recommendation 6: We strongly support the idea of encouraging trade in environmental goods, but suggest that any such policy should not be based on the existing lists. Instead the government should produce a list based on scientific evidence. Such a policy is an opportunity for the UK government to show genuine innovation and leadership in formulating a more climate-friendly trade policy.
As discussed earlier, any change to the existing tariff structure should only be undertaken if there is a clear reason for that change. One possible reason could be to better align the tariff structure with the UK’s sustainability objectives. This leads to a consideration of whether the UK should reduce import tariffs on ‘environmentally-friendly’ goods. There are several existing lists of environmental goods drawn up variously by the WTO, APEC, the OECD and UNCTAD. Many of the goods on these lists are already included in the UK Government’s proposal on tariff simplification or with regard to tariffs on intermediates.
The average tariff for goods on these existing lists is between 2%-3%. Hence even if these were the only goods which the Government chose to reduce tariffs on, it is unlikely that this would have a big net impact on UK prices, or on UK producers. Further, the average share of UK imports and exports covered by the products in these lists is typically small.
However, the existing lists appear to have been drawn up with more respect for countries’ production and export priorities than environmental considerations. Many of the existing lists were heavily influenced by mercantilist concerns, and as a result the list predominantly features goods where average tariffs were already low in most developed countries.
Recommendation 7: We recommend that the Government develops a broader approach to an environment-friendly or climate-friendly trade policy which should consider both tariffs and non-tariff barriers to imports and exports, and be integrated with the Government’s domestic environmental policy.
Further, it is important to note that encouraging trade in environmental goods is but one way of linking trade and the environment, or trade and climate change. It is highly likely that future governments will soon need to consider their policies with regard to trade in environmentally unfriendly goods, and for example the role of border carbon adjustments.
Recommendation 8: To consult widely not just on tariff policy but also on policy with regard to non-tariff measures, and the coherence of such policies with domestic policy objectives.
We welcome the opportunity to consider and respond to the Government’s consultation on the UK Global Tariff. However, trade policy should be seen as part of overall economic policy and needs to be coherent with it. In turn tariff policy is but one component of UK trade policy, including the UK’s overall trade objectives and its commitments to developing countries.
Tariffs are an important part of any government’s trade policy, and will of course be part of any negotiations over free trade agreements. However, in today’s world and particularly for the UK, the key barriers to trade rarely lie in the tariffs that are levied. For goods trade, non-tariff barriers and notably regulations and standards are in many cases more significant. The UK is primarily a service economy, and with regard to services trade the primary barriers centre around diverging regulations. We urge DIT, therefore, to consult widely on these more significant trade policy issues and to avoid focussing overly on the issue of tariffs both in its public consultations and in its own work.
 Some changes were necessary to the Tariff Rate Quotas, which had to be divided between the UK and the EU27. The proposals for these have been contested by a number of WTO members and remain under negotiation.
 Information about the consultation on UK’s Global Tariff can be found here: https://www.gov.uk/government/consultations/the-uk-global-tariff
 The detailed calculations we undertook in the preparation of this work can be found as a technical appendix to this briefing paper on-line here: http://blogs.sussex.ac.uk/uktpo/files/2020/03/Technical-Appendix_BP39.pdf
 This analysis uses UNCTAD tariff data for 2018, downloaded from WITS at the TARIC (10-dig) level and aggregated into 8-digit codes using unweighted averages. The is at the 8-digit level, as this is the most detailed level for which UK trade data is available. In reality, tariffs would be rounded from the 10-digit (most detailed) level. However, it is unlikely that this would make any significant difference to the results presented here.
 Note that these averages are calculated at the CN 8-digit level, calculating the averages at a different aggregation level (e.g. HS 6-digit) would yield somewhat different results.
 Information about how UNCTAD calculates their AVEs can be found in Box 2.1 (p.65) here: https://www.wto.org/english/res_e/publications_e/wto_unctad12_e.pdf
 We use concordance tables from CN 8-digit to BEC from Eurostat, accessed here: https://ec.europa.eu/eurostat/ramon/other_documents/index.cfm?TargetUrl=DSP_OTHER_DOC_DTL#cn
 Note that we have not made any attempt at predicting the percentage levels at which agricultural tariffs would be set if the DIT goes ahead with this part of the proposal. Instead we have run our analysis both including ad-valorem-equivalents (in which case the AVEs have been rounded down the same way as all other tariffs), and excluding non-ad-valorem tariffs.
 In this table we define intermediates in the widest possible way by combining the suspension list, the inward processing list and those products listed as intermediates using the BEC categories.
 See for example: https://blogs.sussex.ac.uk/uktpo/2020/01/14/determining-goods-at-risk/ ; https://blogs.sussex.ac.uk/uktpo/2019/12/09/eu-tariffs-could-cover-75-of-imports-of-goods-into-northern-ireland/ and https://blogs.sussex.ac.uk/uktpo/2019/10/24/better-than-the-status-quo-for-northern-ireland-not-quite-so-simple/
 As an estimate of products with zero production, we identify 223 6-digit products where the UK had no exports over 2016-2018, or where the UK had zero production according to data on agricultural production sourced from FAOSTAT. Further details are available in the technical appendix to this Briefing Paper.
 Earlier work by the UKTPO which looked at increases in trade costs between the UK and the EU showed that changes in tariffs impact differentially across regions. See: Gasiorek, M, Serwicka, I & Smith, Alasdair (2019) ‘Which manufacturing industries and sectors are most vulnerable to Brexit?’, World Economy, 42 (1). pp. 21-56.