Air departure tax in Scotland: an economic assessment

2. Current Aviation Market

In order to contextualise the impact of introducing ADT, it is important to set out the current passenger demand for air services to / from Scotland and the current supply side in terms of e.g. routes, frequency, aircraft etc.

Current Taxation levels / Background to APD

APD came into effect in the UK in November 1994, as a tax paid by airlines based on the carriage of chargeable passengers. The tax has evolved since its introduction (where it was initially applied at £5 per passenger to UK and European Economic Area destinations, £10 per passenger elsewhere) through a number of incarnations to a distance and carrying class based system.

APD is paid on air passenger trips beginning in UK airports, although some UK airports are exempt, including those in the Highlands and Islands. As such, APD is paid on one leg of flights to destinations outside the UK and both legs of most domestic flights.

The current APD rates for financial year 2017-18 are shown in the Table below.

Table 2.1: Current Air Passenger Duty Rates and Bands ^[8]

Destination Bands and distance from London (miles)	Reduced rate: (for travel in the lowest class of travel available on the aircraft)	Standard rate: (for travel in any other class of travel)	Higher rate: (for travel in aircraft of 20 tonnes or more equipped to carry fewer than 19 passengers)
Band A (0 to 2,000 miles)	£13	£26	£78
Band B (over 2,000 miles)	£75	£150	£450

If there is only one class of travel available, the ‘Reduced’ rate applies (unless the seat pitch exceeds 40 inches). This means that for a typical UK internal return flight, the UK APD paid would be £26 per adult (2 * £13 in each direction).

For reference, the recent history of APD rates that have applied in the UK (and will apply in the remainder of the UK from April 2018) is laid out here. ^[9]

Demand Side Baseline

In order to understand the net additional impact of the introduction of ADT, a 2016 demand side baseline was developed setting out travel through Scottish Airports at present.

The baseline was subsequently developed into a set of forecasts for each Scottish airport up to and including the year 2030 ^[10] . This provides the counterfactual position (i.e. what would happen if APD was retained in its current form), against which the impact of the introduction of ADT at a range of lower rates will be measured.

Scottish Airports – Passenger Numbers

By way of context, the recent trend in terminal passenger numbers at Scotland’s airports is shown in the figure below.

Figure 2.1: Terminal Passengers at Scotland’s Airports (‘000, 2005-16)

In 2016, total terminal passenger numbers were therefore nearly 27 million. Prior to the last recession, passenger numbers peaked in 2007 before bottoming out in 2010 following the global economic crisis of 2008. Since then there has been steady growth, particularly since 2012 when growth has averaged around 5% per annum. Aberdeen Airport has more recently however been affected by the downturn in oil and gas related activity, and has seen a 21% drop since 2014. The decline of Prestwick is also notable with a 72% reduction in terminal passengers since 2005. At a national level, these reductions at Aberdeen and Prestwick offset some of the growth seen at Edinburgh and Glasgow. The background to the introduction of ADT is therefore a position of five years of sustained growth in travel through Scotland’s airports driven primarily by Edinburgh and Glasgow who together account for around 80% of passenger numbers. Figures for quarters 1 and 2 in 2017 confirm this trend of strong growth. Note that these figures relate to ‘terminal passengers’ – and an arriving, departing or transferring passenger is counted as a terminal passenger. APD / ADT is however charged on departing passengers only.

Current Air Passenger Duty

At present, APD is payable on air travel journeys which begin in the UK, with the rate depending on the ultimate destination and class of travel, as set out above. Journeys which begin elsewhere and interline through the UK do not typically incur APD charges, and since 1 March 2016, journeys made by those aged under 16 are also exempt. The vast majority of APD payable flights from Scotland depart from Aberdeen, Edinburgh, Glasgow and Prestwick Airports ( APD is not payable on flights originating in the Highlands and Islands (including Inverness), whilst passenger numbers through Dundee Airport are very marginal (approximately 38,000 in 2016).

In order to understand what may happen to passenger demand and revenue in the future, it is essential to firstly review the functioning of the APD regime in the present day. As such, we sought to estimate the number of APD payable trips and APD tax revenue collected on flights from Aberdeen, Edinburgh, Glasgow and Prestwick airports in 2016, as follows:

• Data was obtained from the most recent CAA Passenger Survey which covered Scottish airports, which included surveys at Aberdeen, Edinburgh, Glasgow and Inverness (the 2013 survey).
• Based on the destination country and class of travel, we identified which APD band would apply to each respondent, excluding interlining trips on which APD is not charged.
• Given that different market segments are likely to respond differently to a change in price, a set of 60 passenger categories was developed, reflecting:
  • Traveller place of residence – UK or non- UK;
  • Travel purpose – business or leisure;
  • Destination Type, i.e. destination of journey leaving Scotland – Domestic (Scotland), Domestic (Other UK), Western Europe, OECD (outside Europe), Newly Industrialised Countries or Less Developed Countries ^[11] .
  • Airline type – Network / National / Regional, Chartered or Low Cost Carrier.
• On this basis, we calculated the number of APD payable trips falling into each category, taxation band and airport. The analysis has assumed that travel in Premium Economy, Business Class and First Class incur APD charges at the 2017-18 standard rate.
• We then calculated the proportion of passenger trips which had been captured by the surveys, based on 2016 annual airport passenger volumes also obtained from the CAA.
• The number of passengers falling into each category at each airport was then correspondingly factored up to 2016 annual levels. Ideally, we would have applied factors specific to the passenger category, but this information was not available.
• The CAA Passenger Survey 2013 did not include responses from travellers through Prestwick Airport - we therefore generated our own estimates based on published statistics. We assumed that all travellers made return journeys, all were bound for destinations in Western Europe, all were travelling economy and none were interlining (Ryanair is the only operator at Prestwick currently). Total terminal passengers in 2016 were obtained from the CAA website and the proportions of UK / Foreign residents and Business / Leisure passengers was derived from CAA Passenger Survey data.
• In May 2015, the UK Government removed APD charges on flights undertaken by children aged less than 12, and in March 2016, they also removed APD charges on flights undertaken by those aged 12-15. Based on the proportions of travellers falling into each age band reported by the CAA Passenger Survey, and assuming an even spread of trips across the year, we correspondingly factored down the number of passengers assumed to be liable for APD payment.
• We then multiplied the number of trips in each tax band by the APD rates in 2016, to estimate total APD receipts that year.

Based on the CAA Passenger survey sample, it is estimated that in 2016, APD was payable on 12.2 million air passenger journeys, raising approximately £251m.

Table 2.2: 2016 APD Summary, 2016

	Band A	Band B	Total
APD Payable Trips	10,750,824	1,421,411	12,172,235
APD Receipts	£141,206,297	£109,522,285	£250,728,581

Source: PBA Analysis

Band A (broadly short-haul) taxation raises 56% of total APD receipts, and the remaining 44% is raised by Band B (broadly long-haul) taxation.

Subsequent to the main elements of this analysis being undertaken a new Government Expenditure and Revenue Scotland ( GERS) estimate of £264m for 2016-2017 was published ^[12] . Consultation with Scottish Government has identified that the methodology applied in estimating the PBA and GERS figures is largely the same – however an additional assumption has been made in this analysis regarding the tax collected on trips from Scotland to destinations using two separate flights (and therefore liable for two separate ADT / APD payments), and this accounts for the majority of the difference between the two figures. Note though that all calculations of APD accruing in Scotland at present are estimates and the true figure will not be known until the tax is devolved and collected in Scotland from 2018-19 onwards.

Development of Demand Side Counterfactual

On 1 April 2018, it is expected that APD will be ‘switched off’ in Scotland by HMRC and will no longer be payable on flights departing from Scotland, with the new devolved ADT regime coming into effect. In order to isolate the effect of the introduction of ADT, we have estimated how passenger volumes and tax receipts would change if the current APD regime had remained in place in Scotland. In the subsequent analysis, this counterfactual position will be compared against a range of scenarios where ADT is introduced, with a view to identifying the net impact of the policy (i.e. the impact less what would have happened in any case).

A number of ‘Do Nothing’ cases ^[13] were developed, which reflect the range of possible demand scenarios likely to emerge if the current APD regime is retained. These ‘Do Nothing’ cases represent the counterfactual, against which the intervention ‘Do Something’ scenarios (i.e. the introduction of ADT under a number of different scenarios) have been tested. The potential Do Nothing cases are:

• DfT Low, Central and High constrained passenger growth forecasts, obtained from UK Aviation Forecasts 2013;
• DfT central unconstrained passenger growth forecasts, again from the UK Aviation Forecasts 2013; and
• Forecasts based on DfT Central unconstrained forecasts, which also reflect the emergence of a Low Cost Long Haul ( LCLH) market in Scotland;

These forecasts are developed by DfT at the individual airport level. The initial intention was to directly apply the DfT’s forecast future passenger volumes at each of Scotland’s main airports. However, some of the airports in question, including Edinburgh, have already surpassed the constrained and unconstrained ‘Central’ DfT forecasts for 2020. In order to work around this issue, the underlying year-on-year growth rates were developed and applied to 2016 estimates of APD Payable Movements. APD receipts for the period 2017-2027 were then estimated by multiplying the number of movements by 2016 APD rates. The chart below compares the various growth forecasts (for terminal passengers) derived from the DfT’s UK Aviation Forecasts 2013, in the context of the recent trends noted above.

Figure 2.2: Forecast Do Nothing Terminal Passengers (2016-2027) ^[14]

It can therefore be seen that the different growth scenarios create a range of circa 27 million to 33 million terminal passengers by 2027, a significant range of 6 million passengers. DfT growth forecasts were produced prior to the emergence of a viable LCLH market in the UK, and so the decision was taken to use the LCLH adjusted DfT Central Growth as our core growth scenario.

For simplicity, average fares were assumed to be as per 2016 levels throughout the study period.

The output of this process is a highly disaggregated spreadsheet-based model which:

• represents 2016 terminal passengers, ADT payable passengers, ADT revenue and average fare paid by market segment;
• provides a range of underlying growth projections to 2030; and
• is capable of testing a range of ADT scenarios as tax rate amounts and bands have not yet been set and decisions will be partly informed by this report.

Supply Side Baseline

This section sets out the supply side baseline of the Scottish aviation market, covering:

• Scheduled Passenger Operations;
• Scheduled Destinations;
• Based Aircraft; and
• Benchmarking.

As previously noted, APD is only applicable from Aberdeen ( ABZ), Dundee ( DND), Edinburgh ( EDI), Glasgow ( GLA), and Prestwick ( PIK) Airports, although it must be acknowledged that APD collection on flights from Dundee is negligible (estimated at <0.1% of Scottish APD receipts in 2016).

This analysis therefore provides context for the consideration of supply side responses and could also form the ‘baseline’ for the subsequent monitoring and evaluation of ADT.

Scheduled Passenger Operations

This section provides an outline of the range of scheduled services from Scotland. Scotland’s primary airports have a portfolio of airlines serving them both year-round and with additional seasonal services, as is illustrated in the Table below:

Table 2.3: Scheduled Passenger Operators with Services to / from Scotland

	ABZ	DND	EDI	GLA	INV	PIK
Year Round Services	18	1	27	25	5	1
Seasonal Services [15]	7	0	18	14	4	1

The operators offer a range of destinations, some of which are seasonal. The tables below set out the scheduled and seasonal destinations by airport split by market segment:

Table 2.4: Scheduled Destinations from Scottish Airports (Year-Round)

	ABZ	DND	EDI	GLA	INV	PIK
Scottish Internal	4	0	4	8	4	0
UK Domestic)	15	1	15	18	8	0
Short-haul Europe (excluding UK domestic)	16	0	70	34	1	8
Long-haul (Asia, Africa and Americas)	0	0	4	3	0	0
Scheduled Totals	35	1	93	63	13	8

In terms of year-round connections, Edinburgh serves the widest variety of destinations and is particularly dominant in the year round ‘Short-Haul Europe’ market, given the strong business connections to key European hubs, such as Paris and Amsterdam. Aberdeen has a roughly even split between ‘ UK Domestic’ and ‘Short Haul Europe’. Inverness is dominated by ‘ UK Domestic’ traffic, with key connections to London and other UK hubs.

Year-round long-haul flights are only available from Glasgow and Edinburgh, connecting with hubs in the Middle East (e.g. Abu Dhabi, Dubai, Doha) and North America (e.g. New York Newark). Overall, however, the long-haul market from Scotland is currently very limited compared to Manchester, London Gatwick etc.

Table 2.5: Scheduled Destinations from Scottish Airports (Seasonal)

	ABZ	DND	EDI	GLA	INV	PIK
Scottish Internal	0	0	0	0	0	0
UK Domestic (excluding short-haul)	3	0	1	2	1	0
Short-haul Europe	10	0	35	38	3	8
Long-haul (Asia, Africa and Americas)	0	0	5	11	0	0
Scheduled Totals	13	0	41	51	4	8

There is a significant seasonal route network operating from Scotland, predominantly to warm-weather destinations in e.g. Spain, Greece, Turkey etc. This is dominated by Glasgow and Edinburgh, although Aberdeen, Inverness and Prestwick have seasonal connections to continental Europe.

Long-haul connectivity increases significantly during the summer months from Edinburgh and Glasgow, although again this is predominantly supporting the long-haul tourism market (e.g. Florida, Mexico, Canada etc.).

It is important to note that the above is a snapshot of the aviation industry in Scotland in late 2017. Aviation is however a dynamic industry with highly mobile assets and this baseline will change and evolve over even short periods of time.

Whilst the overall number of scheduled flights is an indicator of air market vitality, the type of airports served is also key. Scheduled flights to airports that are attractive business destinations in their own right (e.g. Zurich), and / or key hubs (e.g. Amsterdam Schipol) are economically more valuable than smaller airports (e.g. Kaunas). The figure below shows the number of scheduled links between Scottish airports and Europe’s busiest airports.

Figure 2.3: Number of Scheduled Links with Europe’s Top 100 Airports ^[16,17]

Based on the above measure, Edinburgh is very well connected with the main European hubs with almost two thirds of links in this category, whilst just under 50% of flights from Glasgow and just over 20% from Aberdeen connect with Europe’s busiest airports, the top 10 being: London Heathrow, Paris Charles de Gaulle, Amsterdam, Frankfurt, Istanbul, Madrid, Barcelona, London Gatwick, Munich and Rome.

Scheduled Destinations

Building on the above analysis, the tables below show weekly frequencies from the Scottish airports in summer and winter 2017 to key hubs / international business destinations. ^[18]

Less than 10 rotations per week indicates less than 2 per day which is usually considered a minimum base for strong business passenger use (i.e. the ability to make a day return trip). More than 15 rotations per week usually indicates at least 3 rotations per day and suggests strong business and interlining use.

Table 2.6: Summer 2017 Weekly Frequencies to Key Hubs from Scottish Airports

	ABZ	DND	EDI	GLA	INV	PIK
Amsterdam	28		56	34	14
Brussels			13	2
Paris – Charles De Gaulle	14		32	14
Copenhagen	5		15
Dublin	11		67	51	9
Dusseldorf			3	6
Frankfurt [19]	14		17
Madrid			15
Milan [20]			17	3
Munich			13	1
Oslo	5		9
Rome [21, 22]			7
Stockholm			12
Reykjavik			6	7
Lisbon			3	2
Total	77	0	285	120	23	0
London Airports
Heathrow	68		101	67	7
Gatwick	6		51	43	18
Stansted		11	49	39		11
Luton	5		22	15
City	12		67	30
Total	91	11	290	194	25	11

Table 2.7: Winter 2017 Weekly Frequencies to Key Hubs from Scottish Airports

	ABZ	DND	EDI	GLA	INV	PIK
Amsterdam	26		42	31	7
Brussels [23]			16	2
Paris – Charles De Gaulle	21		43	7
Copenhagen	6		12
Dublin	10		50	46	3
Dusseldorf			5	4
Frankfurt	14		12
Madrid			9
Milan			14	2
Munich			9	4
Oslo	10		4
Rome			5
Stockholm			5
Reykjavik			9	7
Lisbon			3	2
Total	87	0	238	105	10	0
London Airports
Heathrow	80		109	70	7
Gatwick	11		49	40	15
Stansted		11	48	36		11
Luton	4		24	16	7
City	11		74	38
Total	106	11	304	200	29	11

From these tables, it is clear that Edinburgh and Glasgow (in that order) provide higher frequencies to the key business hubs of London, Amsterdam, Paris and Dublin, although Aberdeen has strong connections with Frankfurt. Scotland also has highly developed connections with a second tier of key hubs such as Milan, Copenhagen, Oslo and Stockholm. Aberdeen’s close links with Scandinavia are also apparent.

Seasonality is a prominent feature in the Scottish aviation market – there is an outbound ‘school holidays’ exodus and a summer inbound tourist surge, whilst there is also a ‘flight to the sun’ effect in the winter.

Scotland Based Aircraft

The planned introduction of ADT is in part intended to stimulate the Scottish economy on a variety of levels – one of these is to encourage additional ‘based aircraft’ in Scotland. Based aircraft are considered to be of economic and social value because they create local direct and indirect employment (this is explored further in Chapter 6), and also offer early morning departures to key destinations.

As part of the industry consultation, a group of key airlines operating in Scotland was requested to provide a record of based aircraft in Scotland. The information received by return was relatively patchy. In order to counter this, we took a sample of aircraft departing Scotland’s main airports before 07:30, which was taken to be a proxy for the number of aircraft that were overnighting in Scotland.

It should be noted that some of these aircraft may not be based in Scotland and may only be overnighting, with their crew staying in hotels and their engineering support being accommodated elsewhere. However, this does give some indication of the main operations that are of value to Scotland and these include Ryanair, easyJet, Flybe, Loganair, Eastern, Jet2 and Thomas Cook. The key European feeder flights are largely served by aircraft that are overnighting.

Table 2.8: Overnighting Aircraft by Airline (Sample taken on Friday 14 ^th July 2017)

	ABZ	DND	EDI	GLA	INV	PIK
FlyBe	4		8	4
easyJet			8	4	1
Ryanair			7	2		3
Loganair	6	1	2	6	6
Eastern Airways	6			1
BMI Regional	2				1
Thomas Cook				5
Brussels Airlines			1
KLM	1		1	1	1
Jet2	5		5	5
TUI Airlines UK	1		1	1
Hop!			1
Total	25	1	34	29	9	3

It is notable from the above table that based aircraft tend to be concentrated in the low cost carrier, charter and regional airlines sector. Edinburgh, Glasgow and, to a lesser extent, Aberdeen are more dominant in this respect, although this is to be expected given the higher overall flight frequency.

Benchmarking Against Comparable Airports

In developing the analysis, a degree of benchmarking was undertaken with ‘secondary’ comparator airports on the European periphery, with a view to identifying the level of Scottish connectivity compared to other ‘peripheral’ European airports. The results are shown in the table below:

Table 2.9: Benchmarking with EU Peripheral Secondary Airports (2015)

City / Airport	Passengers (Million)	Long Haul Routes	Short Haul & Domestic Routes	National Population (Million)	International Tourists (000s), 2013 [24]
Edinburgh & Glasgow (Combined)	19.8	6 & 10 Seasonal	113	5.3	2,700
Lisbon	20.1	25 & 5 Seasonal	137	10.5	8,324
Copenhagen	26.6	36 & 7 Seasonal	139	5.6	8,068
Helsinki	16.4	20 & 8 Seasonal	96	5.4	4,226
Dublin	25.0	18 & 16 Seasonal	132	4.6	7,550
Keflavik	4.9	16 & more seasonal	48	0.3	807

It is important to note that the Scottish airports do not represent an entirely ‘like-for-like’ comparison with the other cited airports for the following reasons:

• The benchmarked airports are national capitals and thus represent the primary airports in those countries. The long-haul connectivity of the Scottish airports is impacted heavily by the dominance of London Heathrow, and to a lesser extent London Gatwick and Manchester.
• Lisbon and, to a lesser extent, Dublin serve large national diaspora further afield – e.g. Brazil and a number of Atlantic islands in the case of the former and the east coast of the USA in the case of the latter.
• Scotland does not have a national carrier, unlike these other countries.
• There will be different taxation and charging regimes with respect to air travel and airports in these countries.

Whilst not direct comparators, it can be seen from the above table that the main two Scottish airports significantly lag other European peripheral airports in terms of overall connections, but particularly in terms of long-haul connections. This in effect means that Scotland is less well connected to key current and growing markets. Whilst APD is only one of a large number of factors determining the scope and scale of the route network, the reduction of taxation under the ADT proposals is likely to act as a stimulant to route development.

Development of Supply Side Counterfactual

The supply side counterfactual would ideally describe how the destinations served, frequency of connections and airline seat capacity will change between 2016 and 2027. However, this information was not available to the study team beyond published plans, and in many cases is commercially confidential information. We are aware that several low-cost carriers with operations in Scotland have placed substantial orders for new aircraft, and so it is reasonable to assume that there will be some route growth in the absence of changes to APD. We have however assumed that trend supply side expansion and corresponding changes in passenger demand are implicitly accounted for within growth factors derived from the UK Aviation Forecasts 2013.

Supply Side Summary

The supply side baselining has demonstrated that Scotland is relatively well-connected to key short-haul destinations in Europe, with direct flights to the likes of Amsterdam, Paris and Frankfurt amongst others. Scottish passengers can also access an extensive network of global connections through the hub airports to which Scotland is connected, including London Heathrow, Amsterdam Schipol, Dubai and Paris Charles de Gaulle.

However, the direct long-haul connections from Scotland are generally limited to hubs in the Middle East and the east coast of the USA. This compares unfavourably to other countries on the European periphery, such as Iceland, Ireland, Denmark and Finland which have a much more extensive route network ( Table 2.9).

The airline industry in Scotland is also dominated by low cost, regional and charter carriers, with only a few of the higher value network carriers (e.g. Emirates and Qatar) operating a significant route network from Scotland (and none of whom have based aircraft in Scotland).