Economic impacts for Scottish and UK seafood industries post-Brexit: report

Appendix G: Data Sources and Derivation

G.1 Production data

The partial-equilibrium modelling requires data on production of each category modelled. Much of the model preparation needed to address the difficulty of bringing together landings data (by species) with trade data (by commodity code), and estimating production of each species based on both landings and processing (which may use landed and imported material). Due to the global nature of the trade modelling, data were required for all countries and therefore United Nations ( UN) Food and Agriculture Organization ( FAO) datasets were used.

The approach was as follows, for each country (or country group) in the model:

• Take production of processed fish from the UN FAO commodities and trade database (in tonnes);
• Convert to live weight equivalent using European Market Observatory for Fisheries and Aquaculture Products ( EUMOFA) conversion factors for the relevant trade codes;
• Take landings (or aquaculture production) from FAO production database (in tonnes);
• If landings is greater than the live weight equivalent of processed fish, the difference is allocated to the fresh/chilled category, and converted to final weight using EUMOFA conversion factors;
• Average price per tonne of each category is calculated from the UN Comtrade database (using world trade, based on export volumes and values) and applied to the final weight of each category to obtain the value of production; and
• Where production appeared to be less than exports, an adjustment to the data was made. One option would simply be to assume zero domestic consumption of domestic production, but that would appear to be an extreme assumption. Where this situation arose, we used our judgement as to the most appropriate way of adjusting the data to make it internally consistent. In some cases, we applied the average ratio of exports to production from other the countries. In one case for the EU, and in two cases for the Rest of the World (RoW) we made the adjustment on a different basis: for these we have actual data on intra- EU trade, and intra-RoW trade. That trade is part of the domestic consumption of domestic production, and therefore where the problem arose for the EU or the RoW we used the intra-regional trade flows to make the adjustment. For other cases, we estimated domestic consumption based on FAO fish consumption statistics per capita combined with population size, and an estimate of the proportion of fish consumption that might be attributable to that species.

Data on production of processed fish are only available from the FAO commodities database. The latest year available is 2013. We have therefore estimated 2015 production based on a scaling factor using 2013 and 2015 landings data as follows:

• Create scaling factor based on 2015 landings volume / 2013 landings volume;
• Apply scaling factor to 2013 live weight equivalent of production by category;
• Divide by EUMOFA conversion factor to obtain final product volume;
• Apply value per tonne per category (average worldwide export price for each category from UN Comtrade database); and
• Sum value of all categories for each country (group) to obtain total production value.

Where no landings were registered for a country, a scaling factor of 1 was applied (i.e. 2015 production was based on the same volumes as 2013 production). For salmon, the difference between live weight of production of processed fish and aquaculture production or capture landings was attributed to Atlantic or Pacific salmon according to the characteristics of each country's (or group of countries) industries.

The values of production of each species for each country (group) used in the model is shown in Table G.1.

G.1.1 Production under zonal attachment scenarios

Scenarios 1, 2 and 3 assume that UK (and EU) production adjusts based on the zonal attachment ( ZA) ^[28] of the species in UK waters. The new level of production for each species was calculated based on the additional UK landings anticipated from a change of quota distribution to the ZA principle, by stock.

The ZA estimate for each stock was taken from University of Aberdeen & SFF (2017), applied to the 2015 Total Allowable Catch ( TAC), and the change in landings was calculated as the difference between the ZA quota compared to quota uptake for 2015. Zonal attachment for Nephrops was estimated based on ICES stock advice for the individual Functional Units ( FU) and the proportion of each FU in UK waters. Where quota uptake was below 90% (Nephrops in North Sea and West of Scotland), it was assumed that there was no increase in landings despite a potential increase in quota under ZA. In most cases, a change in quota for the UK was assumed to result in the inverse change to the EU quota, with the exception of haddock in the North Sea, where the change in EU quota was limited to the absolute level of increase in UK quota.

The change in the value of production was calculated based on the change in landings and the average value of production per tonne of live weight landings, for each country. The average value of production takes into account the split between fresh and processed categories and the different values per tonne of those categories.

The assumed changes to production values as a result of ZA distribution of quotas are shown in Table G.2. Further details of the calculations are provided in Appendix H.

Table G.1. Value of production in 2015 for each country (group) for each species (group) to be modelled ($000)

Cod		Crab		Haddock		Hake		Herring
Country	Production	Country	Production	Country	Production	Country	Production	Country	Production
China	1,152,845	China	8,289,815	Canada	36,892	Argentina	400,099	China	82,067
EU27	548,165	EU27	216,854	China	213,780	EU27	339,042	EU27	794,838
Faroe Islands	100,244	Indonesia	428,351	EU27	71,453	RoW	590,994	Nigeria	10
Iceland	588,636	RoW	944,184	Faroe Islands	17,843	South Africa	209,780	Norway	238,653
Norway	1,056,351	UK	205,936	Iceland	85,715	USA	374,014	RoW	476,774
RoW	754,980	Viet Nam	457,671	Norway	190,828	UK	33,206	UK	77,156
Russia	964,062	Thailand	223,333	RoW	2,755
UK	79,871	China, Hong Kong SAR	6,707	Russia	192,761
				UAE	12,692
				UK	69,336
				USA	17,601
Mackerel		Nephrops		Saithe		Salmon		Scallop
Country	Production	Country	Production	Country	Production	Country	Production	Country	Production
China	536,421	China	8,064	China	43,241	Canada	713,659	Argentina	373,116
EU27	543,825	EU27	198,821	EU27	57,678	China	534,787	Canada	482,125
Nigeria	10	India	905	Faroe Islands	52,191	EU27	2,364,840	EU27	288,714
Norway	500,987	RoW	28,351	Iceland	88,745	RoW	8,029,374	Japan	3,283,815
RoW	942,054	UK	204,683	Norway	258,432	UK	926,927	Peru	154,686
Russia	147,576	Vietnam	1,156	RoW	765,006	USA	1,948,464	RoW	19,639,925
Ukraine	665			UK	23,202			UK	189,558
UK	301,665							USA	888,155

N.B. These figures include the adjustments for situations where production appears to be less than exports.

Table G.2. Assumptions of changes to production values as a result of UK– EU quota allocation based on zonal attachment

Species	Current UK Landings (t)	Current UK Production ($000)	Change to UK Landings (t)	Change to EU Landings (t)	Price UK ($/t)	Price EU ($/t)	Change to UK Production ($000)	Change to EU Production ($000)
Cod	15,638	78,599	2,538	-2,538	2,731	2,057	6,929	-5,219
Crab		205,936	-	-			0	0
Haddock	32,321	69,315	7,039	-3,598	2,084	2,066	14,667	-7,432
Hake	10,875	33,175	21,699	-21,699	2,552	1,879	55,365	-40,769
Herring	93,595	68,075	205,634	-205,634	726	1,130	149,327	-232,320
Mackerel	247,979	273,409	132,335	-132,335	1,103	1,287	145,901	-170,348
Nephrops	25,874	204,683	1,244	- 1,244	7,899	9,337	9,831	-11,620
Saithe	12,312	23,202	24,237	-24,237	1,781	1,806	43,178	-43,776
Salmon		926,927	-	-			0	0

G.2 Trade data

The data on trade in the model are derived from the UN Comtrade database. In order to ensure compatibility with the latest available production data, the data for 2015 are used. For each species the bilateral trade flow is required, for example the level of UK imports from the EU, and exports to the EU. These bilateral flows are needed for every pair of countries that are included for any given species. Note that in the trade data there is information which is reported by the UK, as to the level of imports from the EU; and there is also information which is reported by the EU, as to the level of exports to the UK. These are the mirror flows. In principle the mirror flows should be the same, in practice this is never the case. Partly this is because the flows are reported on a different basis – this is the difference between cost, insurance and freight (c.i.f.) and free on board (f.o.b.) reporting requirements. However, in part this is also simply because of differences in the way data is collected and reported by different countries (in particular where there may be re-exports or re-imports), and differences may arise because of errors made in collecting the data. Given these inherent difficulties in the data for each bilateral flow the average of the two mirror flows was used. For the trade of the EU in each case the external trade of the EU minus the UK was used, hence excluding intra- EU trade flows, as this forms part of the EU-minus- UK domestic production. With fish there is the added complication that landings made e.g. by UK-flagged vessels into a foreign port should be treated as UK exports, but there is evidence to suggest that such landings are not always well documented. Landings of herring and mackerel to non- UK ports which are not recorded in the export data have been adjusted for.

G.3 Tariffs

The data on tariffs derives from the UN TRAINS database, which provides information on the 6-digit tariffs levied by each country on each importer. In the absence of a free trade agreement between countries the tariffs will be the MFN applied tariffs; where there is a free trade agreement then the tariffs will be the preferential tariffs. Where the data was available the tariff data is based on the 2015 tariffs. These tariffs were not always available, as countries do not always report their tariffs every year, especially where there have been no changes from the preceding year(s). Where the data for 2015 was missing data from an earlier year was therefore used.

Figure G.1 below gives the average EU MFN tariff for each species. This is an average across the underlying 6-digit HS codes which make up each of the species. In constructing such averages the user needs to decide whether to use simple average tariffs, or weighted average tariffs, where the weights are given by the share of imports of each of the relevant tariff lines. The difficulty in using weighted average tariffs is that the level of the import flow is endogenous to the level of the tariff. Suppose there were a very high tariff which drove imports to zero; that tariff would receive a zero weight (because imports are zero), although clearly the tariff is highly restrictive. For this reason we have used simple average tariffs.

Figure G.1 shows that EU MFN average tariffs on the fish species are in most cases quite high, except for salmon where it is just under 5%. For fresh/chilled, frozen and fillets of salmon, the MFN tariff is only 2%, to facilitate supply for the EU processing industry, but this rises to 13% for smoked salmon.

Figure G.1. EU MFN average tariffs by species

G.4 Non-tariff measures

On leaving the EU the UK faces a broad range of possible trade relationships with the EU27. As far as NTM are considered:

• In scenarios where the UK leaves the single market and does not agree mutual recognition of standards and testing and certification, traders will face SPS and TBT rules and be required to show that shipments meet EU standards. Estimates of ad valorem equivalents ( AVE) are difficult to estimate (Fugazza, 2013) nor are product-level AVEs easy to find at the 6-digit product and country level. A recent United Nations Commission on Trade and Development ( UNCTAD) paper on Fish NTMs restricted itself to coverage ratios and similar qualitative measures (Fugazza, 2017). Ghodsi et al. (2016) is an exception and has estimated simple average AVE for a group of 40 importers by main measures (notably SPS, TBT and Contingent Protection) for total trade). EU members are shown separately rather than as an average. The country range of AVE for SPS is –2.9% to 14.7% (a negative meaning that complying with the SPS standard leads to an increase in trade not a reduction), although this is not specific to fish. Estimates specific to fish, for all countries worldwide, are –3.7% to –1.3%. For TBT the equivalent range is 0.6% to 16% for EU countries (all trade), and 0.7% to 2.1% for fish trade worldwide. The differences across Member States likely arises from trade composition differences and perhaps from different customs procedures across the Member States.
• In scenarios where the UK agrees a Free Trade Agreement with the EU27 (e.g. Scenario 2), UK traders will be required to produce Certificates of Origin (CoO) crossing into EU territory. As the Marine Scotland (2017) study shows the cost of certificates per se is low ^[29] but for processed products there may be costs in tracing origin of purchased inputs and possible costs of delay in crossing borders as customs check origin. The stylised fact on the tariff equivalent costs of Rules of Origin are in the range of 4–8% (Cadot & Gourdon, 2015).
• If the UK leaves the single market and/or the Common Commercial Policy it will face the full rigours of EU Contingent Protection (anti-dumping measures). Turning to Ghodsi et al. (2016) once more, their simple average AVE for contingent protection varies across EU Member States from 5.4% to 65.7%.
• When the UK leaves the EU, catch certificates will be required for fish exports to the EU. UK catch certificates are currently required for UK catches that are exported to third countries for processing and subsequently re-exported back to the EU. Catch certificates are also required for exports to some other countries (including Iceland, Norway, Thailand and Ukraine) ^[30] . There is currently no charge for issuing a UK catch certificate, with the costs to the exporter being only the administrative time in obtaining the documentation. For importers, there are charges for checking the validity of catch certificates from third countries ^[31] , which range from £21–45 per country (with mixed consignments potentially requiring catch certificates from multiple countries).

Table G.3. Estimates of NTM in fisheries

NTM Type	Tariff Equivalent Estimate	Source
SPS, health/hygiene certification	General trade ( EU countries): -2.9% to 14.7% Fish (worldwide): –3.7% to –1.3% £5–30/tonne, depending on shipment size (tariff equivalent will depend on shipment size and value per tonne).	Ghodsi et al. (2016) Marine Scotland (2017)
Other TBT	General trade ( EU countries): 0.6% to 16% Fish (worldwide): 0.7% to 2.1%	Ghodsi et al. (2016)
Rules of Origin	4 to 8%	Cadot & Gourdon (2017)
Anti-dumping measures (trade defence)	General trade ( EU countries): 5% to 65% Fish (worldwide): 1.7% to 1.9%	Ghodsi et al. (2016)
Catch certificates	Currently no charge for issue Check of validity for importers: £21–45 per country	See footnotes [30,31]
Freshness /border delays	2 nights delay = 5% price reduction 3 nights delay = 10% price reduction 4 nights delay = 20% price reduction	Marine Scotland (2017)

The wide range of these estimated AVEs, taken together with the lack of fish-specific estimates for the EU, suggests caution in modelling changes in NTMs and in interpretation of the results. The modelling therefore uses estimates based on these AVEs to explore possible impact of different levels of NTMs. A common level of NTMs in the base equilibrium is assumed across all species, equivalent to a 15% ad valorem tariff equivalent ( AVE) for non- EU countries. This AVE is applied for all species across all bilateral trade flows for trade with non- EU countries. The exception to this is that in order to capture the EU’s Single Market we assume that the NTMs within the EU are equal to a 5% AVE equivalent. While the Single Market has achieved a very substantial amount of non-tariff measure reductions, nevertheless these have not been reduced to zero.

In modelling the scenarios, the AVE is assumed to be 10% for scenarios with a moderate level of NTMs (non- EU countries in Scenario 1, the EU in Scenario 2), and 15% for scenarios with a higher level of NTMs (non- EU countries in Scenario 2, EU and non- EU countries in Scenarios 3 and 4).

G.5 References

Cadot, O. and Gourdon, J. (2015). NTM, Preferential Trade Agreements and Prices: New Evidence, Working Papers 2015-01, CEPII Research Center.

Fugazza, M. (2013). The Economics behind Non Tariff Measures: Theoretical Insights and Empirical Evidence, Study Series No 57, Policy Issues in International Trade and Commodities, UNCTAD Geneva.

Fugazza, M., (2017). Fish Trade and Policy: A Primer on Non Tariff Measures, UNCTAD Research Paper No 7.

Ghodsi. M., Gruebler, J. and Stehrer, R. (2016). Estimating Importer-Specific Ad Valorem Equivalents of Non Tariff Measures, Working Paper 129, September. The Vienna Institute for International, Economic Studies.

Marine Scotland (2017). Estimating the potential costs of non-tariff barriers to fisheries following Brexit. Marine Analytical Unit, September 2017.

University of Aberdeen & SFF (2017). The Spatial Distribution of Commercial Fish Stocks of Interest to Scotland in UK Waters. A report prepared by the University of Aberdeen for the Scottish Fishermen’s Federation. January 2017.