Much Ado about Nothing? – An analysis of economic impacts and ecologic effects of the EU-emission trading scheme in the aviation industry

Transportation Research Part A 45 (2011) 1066–1076

Contents lists available at ScienceDirect

Transportation Research Part A journal homepage: www.elsevier.com/locate/tra

Much Ado about Nothing? – An analysis of economic impacts and ecologic effects of the EU-emission trading scheme in the aviation industry Jan Vespermann a,*, Andreas Wald b,1 a b

Aviation Management Institute, European Business School, Rheingaustr. 1, 65375 Oestrich-Winkel, Germany European Business School Paris, 37/39 Boulevard Murat, 75016 Paris, France

a r t i c l e

i n f o

Keywords: Emissions trading Air transport and the environment Future scenarios for the air transport market and industry Quantitative research in air transport: modeling and applications

a b s t r a c t From 2012 on, all CO2 emissions from flights departing from or arriving at airports within the European Union have to be offset. We analyze the economic and ecological impacts that are caused by an inclusion of the aviation industry into the proposed emissions trading scheme (ETS). Building on the now fixed system design we employ a simulation model to estimate the impacts of the scheme. Our results indicate that financial impacts are highly dependant on external settings, such as allowance prices and demand growth. We show that the financial burden on the aviation industry will be rather modest in the first years after the introduction of the system and therefore induce only low competition distortions. Likewise, emission reductions within air transportation will be comparably low. While aviation will induce a decline of emissions in other sectors, significant absolute reductions within air transportation can only be reached by a more restrictive system design. Ó 2010 Elsevier Ltd. All rights reserved.

1. Introduction Aviation is one of the fastest growing industries of the global economy. Over the past 20 years, the industry grew by an average annual rate of around 5%. Aviation is an important contributor to the world’s gross domestic product, generating substantial employment across all nations. Despite the current recession and the crises in air transportation, the industries’ growth is estimated to remain a global phenomenon with an average projected annual growth rate of about 4.2–5.1% (IATA, 2009; Boeing, 2009; ACI, 2008; Airbus, 2007). While this development has a positive economic influence, the growth rates of international aviation do not seem to be ecologically sustainable.2 Most studies agree that the negative environmental impacts of aviation are largely uncompensated (Goetz and Graham, 2004; Chapman, 2007). This inadequate reflection of the ‘real’ costs of air transportation by current prices would represent a market failure and lead to sub-optimal activity levels and low investments in more efficient operational procedures and technologies (Wit et al., 2005). Negative external effects of aviation thereby exist both at the local and global level. At the local level, particularly in the airport proximity and under flight paths, noise pollution is an environmental problem (Schipper, 2004). At the regional and global level, aircraft emissions from the burning of kerosene are important (Janic, 1999). Overall, air transportation is believed to be responsible for about 2–3.5% of all anthropogenic greenhouse gas emissions (Oxford Economics, 2008).

* Corresponding author. Tel.: +49 6723 8888316; fax: +49 6723 8888301. E-mail addresses: [email protected] (J. Vespermann), [email protected] (A. Wald). 1 Tel.: +33 0155 187775. 2 For an introduction to the concept of sustainability and its application to the transportation/aviation sector see Gudmundsson and Höjer (1996), Banister (2008), Amekudzi et al. (2009) and Hoyos (2009). 0965-8564/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.tra.2010.03.005

J. Vespermann, A. Wald / Transportation Research Part A 45 (2011) 1066–1076

1067

Even though aviation has been able to achieve substantial efficiency improvements, these gains have regularly been offset by even higher numbers of traffic growth. If this development continues, air transportation will further increase its share of emitted greenhouse gases and thereby its contribution to climate change. Some studies estimate that emissions from aviation may increase more than threefold until the year 2050 (Olsthoorn, 2001), and may even threaten international aims for the reduction of anthropogenic emissions as agreed on in the Kyoto Protocol (Macintosh and Wallace, 2008). To address these problems, the European Union (EU) has elaborated plans to subject air traffic within its member states to an emissions trading scheme (ETS). The system would cover one third of all CO2 emissions caused by international aviation and impact the industry from January 2012 on. From that time, all emissions of flights departing from or arriving at airports within the EU need to be offset. These additional costs will incur both ecological effects as well as economical burdens on the air transport industry. First studies have analyzed – at an aggregated level – the effects of such a scheme under fixed parameter constellations or for individual airlines. Carlsson and Hammar (2002) explore the possibilities of using incentive-based environmental regulations of CO2 emissions from international civil aviation. Wit et al. (2005) develop concepts for amending the existing EU emissions trading scheme to address the climate change impact of aviation. The study by Morrell (2007) focuses on the method of allocation of emissions permits in the EU context. Scheelhaase and Grimme (2007) analyze the potential cost impact of an ETS system. Their research covers four selected carriers on the European market. Scheelhaase et al. (2009) present a case study on two selected carriers from the US and Europe, analyzing potential impacts of the EU emissions trading scheme on competition structures between European and non-European network airlines. While all of these studies provide important insights into the ETS system and potential impacts of such a scheme, they rely on assumptions and parameters that have recently been overtaken by the increasing concretion of the system design. Also, rather restrictive scenario considerations on the impacts of the ETS system lead to a lack of validity and explanatory power. Many important cause-and-effect chains and feedback loops, such as higher overall costs of flying resulting into less demand for air transportation, remain unconsidered. Furthermore, many questions on the ETS impact on the overall competitive environment of the industry remain open. From an ecological perspective, there is – to our knowledge – no analysis on the environmental effects of the ETS system. It is not yet clear to which extend an inclusion of air transportation into the EU–ETS system will induce emission reductions. This research aims at quantifying both economical and ecological impacts of an inclusion of the air transportation industry into an ETS system. We therefore sub-divide our general research question into two sets of questions. The first set addresses economical impacts. By quantifying these costs, we draw conclusions concerning potential distortions of traffic flows and competition structures. The second set of questions looks into the ecological effects of the ETS system on air transportation. As there is no detailed analysis of these effects, our objective is to quantify overall emission reductions, closing a gap in the literature. The paper is set out as follows. We begin our analysis by introducing the EU–ETS, its political background, system design and main parameters (Section 2). Section 3 introduces the simulation model. We present the model parameters and interdependencies and lay out the method to compute the financial and ecological impacts of the ETS system. We thereafter analyze and interpret the results of the simulation model and discuss potential model limitations (Section 4). Section 5 provides a conclusion.

2. The EU–ETS system: background and system design 2.1. Political background of emissions trading schemes The EU–ETS has been developed to facilitate the implementation of the Kyoto Protocol. The Kyoto Protocol is a legally binding international agreement under the United Nations Framework Convention on Climate Change (UNFCCC), aimed at combating global warming through the stabilization of greenhouse gas emissions. The protocol was adopted in 1997 and entered into force in 2005. It requires the 37 so called ‘Annex-I countries’ (industrialized countries) to a reduction of emissions – including the greenhouse gas carbon dioxide. As the Annex-I countries commit themselves to reduce their emissions below a certain baseline (a reduction of 5.2% from the 1990 level by the year 2012), the protocol can be considered a cap system with national-level commitments. The Kyoto Protocol authorizes three cooperative implementation mechanisms that involve tradable permits – the Clean Development Mechanism (CDM), Joint Implementation (JI), and Emission Trading Schemes. While the Clean Development Mechanism allows the creation of carbon credits (so called Certified Emission Reductions – CER) by developing emission reduction projects in non-Annex-I countries, Joint Implementation projects allow project-specific credits (Emission Reduction Units – ERU) to be transferred within Annex-I countries. Emissions trading schemes represent a third policy approach aimed at reaching the goals of the Kyoto Protocol with market-based mechanisms. Market-based instruments in particular have the potential to encourage efficiency improvements at emitters and provide economic incentives for technological innovation. The basic line of argument is that behavioral sources of pollution problems could be traced to an ill-defined set of property rights (Tietenberg, 1980). However, by making property rights (here: the right to emit) explicit and transferable, the market would gravitate towards an ultimate cost-effective allocation of the permits (Montgomery, 1972). This can be ascribed to the fact that as long as marginal abatement costs

1068

J. Vespermann, A. Wald / Transportation Research Part A 45 (2011) 1066–1076

differ, incentives for trade exist. Polluters that feature high marginal abatement costs would be incentivized to buy permits from those polluters with low marginal abatement costs. The price for emission permits would then be established by the interaction of demand and supply for allowances in the market. Since allowances are instruments with a transparent price, investors and emitters can trade and buy them on markets not only for ‘consumption’ purposes but also for speculation. The EU–ETS is the currently largest cap-and-trade emission program. The present emissions trading scheme entered into force in October 2003 (Directive 2003/87/EC) and has been operational since January 2005. It covers about 50% of carbon dioxide emissions within the European Union and includes about 12,000 energy-intensive installations from five sectors (iron/steel, refineries/coke ovens, cement, glass/ceramic, and pulp/paper). Emission permits can be purchased on an exchange, from a broker, or directly from another party, with individual companies as the ultimate buyers of allowances. Both CERs and ERUs are valid for meeting ETS obligations. 2.2. Integration of aviation into the EU–ETS While the Kyoto Protocol excluded emissions from international aviation services from national emissions targets, Article 2.2 of the protocol stipulated the industrialized countries (‘Annex-I countries’) to stabilize greenhouse gas emissions brought about by kerosene separately through the International Civil Aviation Organization (ICAO). Even though the ICAO in principle favored market-based regulatory approaches such as emissions trading schemes and has opted to pursue the development of such schemes, the EU went ahead and decided to integrate aircraft operators flying into or out of any EU airport into its existing emissions trading system from 2012 on (Directive 2008/101/EC and 2009/339/EC). However, contrary to the existing sectors which are covered under National Allocation Plans (NAPs),3 emission permits for the aviation industry will be distributed directly through the European Commission. Under the new directive, a limit on the amount of CO2 emissions from aviation activities is set. The cap on the total number of allocated allowances creates the scarcity needed for a trading market to emerge. With the help of market mechanisms, these emission rights are expected to flow to wherever specific avoidance costs are higher than the market price for emission rights (‘cap-and-trade’). As we formulate our simulation model based on the actual design of the proposed ETS system, we introduce the main specifications of the system:  Allocation method (2008/101/EC, §1(4) amending 2003/87/EC, §3c-e): Emission allowances are allocated based on a combined benchmarking/auctioning approach, with the airline operator as the trading entity. For the first trading period in 2012, the quantity of allowances to be allocated among airlines will be equivalent to 97% of the mean average of the historic annual emissions from air transportation in the years 2004–2006. Eighty-five percent of these allowances are granted for free, the remaining 15% of allowances are to be auctioned. From 2013 on, the total quantity of allowances will reduced to 95% of historic aviation emissions. The initial allocation benchmark is calculated based on historic CO2 emissions per revenue tonne kilometer (RTK) (see Section 3 for its calculation).  System design (2008/101/EC, §1(3) amending 2003/87/EC, §3b): The EU–ETS scheme is an open cap-and-trade system, as aircraft operators are allowed to purchase emissions allowances from other sectors (i.e. stationary sources). Market participants that keep their emissions below the level of their allocated allowances are able to sell their excess allowances. Contrariwise, companies that emit in excess of their allowances have to buy additional permits on the market.  Interplay with other ‘Flexible mechanisms’ of the Kyoto Protocol (2008/101/EC, §1(8) amending 2003/87/EC, §11a): In the first trading period in 2012, up to 15% of credits by the Clean Development Mechanism (through the purchase of Certified Emission Reductions – CER) and Joint Implementation (through the transfer of Emission Reduction Units – ERU) are allowed. From 2013 on, credits from these sources are restricted to a maximum of 1.5%.  Threshold/exemptions (2008/101/EC, §1(22) amending 2003/87/EC, Annex I): Aircraft with a certified maximum take-off weight (MTOW) of less than 5.7 tonnes; military, fire fighting, training, search and rescue, humanitarian, customs, and police flights are not included in the ETS system. Additionally, flights on routes to so called ‘outermost regions’ and airlines that fall under the ‘de minimis’4 clause are excluded from the system. 3. Modeling the impacts of the EU–ETS system To assess the financial and ecologic impacts of the ETS system we employ a simulation model. The outlay of the model reflects the now fixed EU emission scheme design according to Directives 2008/101/EC and 2009/339/EC (Fig. 1). The simulation model consists of input variables and dependant variables. Since feedback processes are considered, two initial input variables (fuel consumption, airline passengers) become endogenous. As we try to assess the economic and ecologic impacts of the emissions trading scheme, the target values for the analysis are (1) emission costs and (2) emission reductions. Our level of investigation is the flight level, i.e. the most detailed level possible. We include all scheduled flights 3 From 2013 on, the design of the existing EU–ETS will be altered (Directive 2009/29/EC). The emissions cap will be set at the level of the European Union (Phase III of the scheme, 2012–2020). 4 Annual CO2 emissions less than 10,000 tonnes or less than 243 flights per 4-month-period for three consecutive months.

1069

J. Vespermann, A. Wald / Transportation Research Part A 45 (2011) 1066–1076

Price elasticity of demand ( a) Average ticket prices

Demand growth (dgamy)

(atpay) Airline passengers

Market growth (gm)

Efficiency gains (eg)

Emissions factor (ef)

T Transport t activity ti it (ta t ay)

Emission reduction

(paxay)

Fuel consumption (faiy)

CO2 emissions at airline (eay) Required allowances

Increase in passenger

(awray)

prices

B Benchmark h k (b) Input variable Dependant variable

F Free allowances ll (awffay)

Allowance price (awp)

Bought allowances Emission costs (cay)

Interdependency Allowance costs per passenger (awcay)

Feedback process

Fig. 1. Simulation model.

Table 1 Indices. Indices a i m s y

Comment Airline Flight event Market Simulation Year

a = 1, . . . , A i = 1, . . . , I m = 1, . . . , M s = 1, . . . , S y = 1, . . . , Y

All airlines covered under the ETS system (313) as specified in Commission Regulation No. 748/2009 All projected airline specific flight events relevant under the ETS system Individual air traffic markets, classified according to Boeing (2009) Model simulations Year after ETS introduction/horizon of simulated projection (2012–2020)

Table 2 Input variables. Input variable

Distribution/data source

awp atpay ef eg mgm

Allowance price (EUR) Average ticket prices (EUR) Emissions factor Efficiency gains (% p.a.) Market growth (% p.a.)

taay

Transport activity (RTK) Price elasticity of demand

ga

awp  N (25, 6.25) Official airline data 3.15 eg  N (1.0, 0.25) gm  N (lm, r2m ) Official airline data; ICAO data ga  N (0.8, 0.2)

Table 3 Dependant variables. Dependant variable awcay awfay awray b cay

Allowance costs per passenger (EUR) Free allowances Required allowances Benchmark (kgCO2/RTK) Emission costs (EUR)

dgamy eay faiy

Demand growth (% p.a.) CO2 emissions at airline (t) Fuel consumption on flight level (t). Based on typical pay-load and distancesa

paxay

Airline passengers (mil. p.a.) Sources: Official airline data; IATA

a Typical payloads are based on average seat load factors (75–82% for full-service network carriers, 80–90% for low-cost carriers) and freight load (60–65%) (Source: ICAO data). Since great circle distances only provide an approximation to actual traveled distances (Source: OAG data), inefficiency factors are used (short-haul: 10%; long-haul: 5%). The System for assessing Aviation’s Global Emissions (SAGE) project provides average fuel consumption data for all current aircraft types.

1070

J. Vespermann, A. Wald / Transportation Research Part A 45 (2011) 1066–1076

of those airlines that will be covered under the ETS system. As some of the input parameters are modeled based on distributions, we adopt a simulation model using risk (see Tables 1 and 2) analysis software.5 We give an overview on the employed indices and variables (see Table 3). 3.1. Input variables and used data Since the input (i.e. independent) variables are fundamental for all further calculations, we explain these variables as well as the origin of the underlying data in more detail: Allowances prices (awp): Most studies estimate an average of EUR 20-30 for the mid-term development of allowances prices (Wit et al., 2005; Scheelhaase and Grimme, 2007). Based on current prices and those for carbon dioxide futures, we assume a similar development, with an expected value of l = EUR 25. However, to allow for changes of allowance prices, we model these prices based on a normal distribution. Average ticket prices (atpay): Average ticket prices are taken from official airline data. If this data was not directly derivable from official airline information, we estimated average ticket prices based on the type of business model (e.g. average of ticket prices of all other airlines within the group of low-cost or full-service carriers). Ticket prices are assumed to increase at 2% per year. Emissions factor (ef): Set at a value of 3.15 as specified in EU Directive 2009/339/EC. Efficiency gains (eg): Aircraft fuel efficiency gains are often estimated to amount to about 1% annually (IATA, 2004; Wit et al., 2005). A recent study by Macintosh and Wallace (2008) calculated an annual rate of improvement of 1.2% since 2000. We assume efficiency gains to be normally distributed and fluctuate around 1% per year. Market growth (mgm): Growth numbers are modeled based on market specific Revenue Passenger Kilometers (RPK) traffic growth forecasts (lm) according to Boeing (2009). Other estimations from aircraft manufacturers (Airbus, 2007) or those of industry trade groups (IATA, 2009; ACI, 2008) predict similar numbers concerning the long-term growth of aviation. We differentiate 23 different air traffic markets, such as traffic from Europe to the Middle East or to South America. All flights on a specific market are assumed to grow according to the growth numbers set for that market. Since most traffic forecasts published at the end of 2008 or at the beginning of 2009 do not yet incorporate the current global recession, we adjust the projections, incorporating a decline in traffic in 2009. In 2011, the level of 2008 is reached again. Again, we allow for fluctuations of demand, assuming demand growth for each market to be normally distributed around the expected value. Transport activity (taay): The transport activity of each airline – used for the calculation of the benchmark – is based on official airline data. If this data was not available we used data provided by ICAO. Price elasticity of demand (ga): Prior research generally assumes that policy-induced cost increases to airlines are passed onto consumers (Wit et al., 2005; WBCSD, 2004). It is argued that due to the competitiveness of the air transport industry, there is no scope for airlines to absorb the cost increase and reduce already small profit margins. To quantify the impact of an increased cost base on demand, we base our assumptions on previous findings for the airline industry. Lu (2008) provides an overview of price elasticities measured in various research papers. She concludes that elasticities depend on parameters such as travel distance and booking class, with business passengers generally being less sensitive to price changes. Brons et al. (2002) provide a (meta-)analysis of price elasticities of demand for passenger air transport and the ‘expected’ level of price sensitivity in a specific setting. Based on these studies we employ a uniform price elasticity for our analysis with an expected value of 0.8 (see Table 2). Overall, we believe our model to be more detailed and more accurate than previous studies. The model is based on the now fixed EU–ETS system design and builds on the most detailed level of analysis possible, i.e. the flight event. Thus, for the initial year after the introduction of the ETS system (2012), we model the fuel consumption and subsequent emissions for over 8 million flights covered under the system. Additional explanatory power is gained, since variable distributions are incorporated in the simulation model. Consequently, the model does not rely on restrictive assumptions and fixed parameter settings. By interpreting risk profiles, we are able to analyze fluctuations of input variables and subsequent impacts on the values of analysis. 3.2. Dependant variables Dependant variables are directly or indirectly calculated from input variables. Their computation is explained in detail as we proceed through the steps of our simulation model. Fuel consumption (faiy) and airline passengers (paxay) become dependant variables due to the consideration of feedback processes. 3.3. Calculation of the benchmark The calculation of the benchmark is the basis for all further analysis. Its calculation is based on historic aviation emissions of the years 2004–2006. Ninety-seven percent of these emissions will be distributed, of which 85% are granted for free and 5

@Risk (Palisade Corporation).

J. Vespermann, A. Wald / Transportation Research Part A 45 (2011) 1066–1076

1071

15% are to be auctioned. Since emissions at an airline are directly proportional to the fuel consumption of individual flight events (faiy), total emissions for any given airline and year (eay) can be calculated at I X

faiy  ef ¼ eay ;

with ef ¼ 3:15:

ð1Þ

i¼1

The total number of free allowances (awfay) for the first trading period equals A X

awfay ¼ 0:85  0:97 

a¼1;y¼2012

A X X 1 2006 eay : 3 y¼2004 a¼1

ð2Þ

For all subsequent trading periods, 95% of historic emissions are taken into consideration (of which 85% are granted for free). The allocation methodology for the distribution of allowances to aircraft operators is based on the transportation activity6 of airlines (taay) in 2010 and the total quantity of free allowances for the air transport sector calculated above. Thus, the benchmark (b) is quantified as

b¼

,

A X

awfay

a¼1;y¼2012

A X

taay :

ð3Þ

a¼1;y¼2010

Consequently, for the first trading period, an airline (a) is assigned the following number of allowances free of charge:

awfa;y¼2012 ¼ b  taa;y¼2010 :

ð4Þ

For all subsequent trading periods the amount of free allowances is based on the reduced factor of 95%. 3.4. Emission costs The required allowances per airline (awray) match the total emissions of that airline, specified in Eq. (1). The emission costs (cay) can therefore be calculated as the product of bought allowances and allowance price (awp)

cay ¼ ðawray  awfay Þ  awp: Total costs for aviation and a specific year are the sum of the individual airline burdens

ð5Þ PA

a¼1 cay .

3.5. Emission reductions Costs attributed to the ETS system are generally expected to be at least partly passed through to the passengers (Wit et al., 2005; WBCSD, 2004). Resulting higher ticket prices will lead to less demand for air transportation and consequently to lower growth rates. While this will result in decreasing revenues for airlines, the lowered transportation activity will also involve reduced emissions. We integrate these processes into the simulation model by introducing feedback loops. The model calculation then becomes a stepwise process, each time allowing for increasing ticket prices in 1 year to result into lower demand and thus fewer emissions in the following years. Building on the overall number of passengers of an airline (paxay) and taking into account the total emissions costs for an airline calculated before, the average allowance costs per passenger (awcay) can be estimated at

awcay ¼ cay =paxay :

ð6Þ

As emissions and costs passed through to the passengers differ among airlines, future demand growth rates also differ (dgamy). Based on market growth rates (mgm) and average ticket prices (atpay), the price elasticity of demand (ga) can be used to estimate the demand growth in the next period

dg am;yþ1 ¼ mg m þ ga ðawcay =atpay Þ:

ð7Þ

We assume that reduced growth rates proportionally impact both passenger numbers and fuel consumption in subsequent years. The ecological effect of the ETS system – and the line of argument put forward by the European Union – was introduced above: less demand for air transportation due to increased costs will involve reduced fuel consumption and with it lowered emissions. Emissions for the next trading periods can therefore be calculated by a projection of historic fuel consumption, adapted by the demand impact of the ETS system. Considering efficiency gains (eg), emissions for each subsequent period can thus be calculated as A X a¼1

ea;yþ1 ¼

A X I X ðfaiy  ðdg am;yþ1  egÞÞef :

ð8Þ

a¼1 i¼1

6 Transportation activity = (Great circle distance between departure and arrival airport + 95 km) * (Payload). Payload means the total mass of freight, mail, baggage and passengers excluding crew members.

1072

J. Vespermann, A. Wald / Transportation Research Part A 45 (2011) 1066–1076

Furthermore, the demand side effects of the ETS system have to be supplemented by those emission reductions that – while ‘induced’ by the aviation industry – are observable in other sectors. These reductions are determined by the ‘outflow’ of emission permit from other industries to air transportation. The number of transferred allowances thereby equals those permits that are required by aviation less those allowances that are allocated for free and those allowances that are auctioned within air transportation (15% based on historic emissions): A X a¼1

awr ay 

A X

awfay  0:15  0:97 

a¼1

006 X A 1 2X eay : 3 y¼2004 a¼1

ð9Þ

For each year, the reduction effect of the EU–ETS in aviation is equal to the difference between aggregated emissions under the ETS system (8) and those under an unrestricted growth scenario (‘‘business as usual”). The total ecologic effect of an integration of air transportation into the EU–ETS is the sum of effects within aviation (8) and within other sectors (9).

4. Results and discussion 4.1. Economic impacts and effects on competition structures Based on our model, we calculate the benchmark at b = 0.68 kgCO2/RTK.7 We compute the average industry burdens and CO2 emissions until 2020: The financial industry burden is expected to rise from EUR 2.25 billion after the introduction of the ETS system to about EUR 3.67 billion in 2020. For the years 2012–2020, the mean financial impact of the ETS system on the aviation industry is EUR 2.98 billion. The figure illustrates that the aviation industry will be a strong net buyer of emission certificates from other sectors. Since free allowances are fixed at historic emissions, the share of allowances allocated free of charge (black bars) is constantly decreasing to about 52.8% in 2020. The slope of the graph featuring the emission costs reduces after the first year of the system introduction, since the quantity of allowances to be allocated decreases from 97% in 2012 to 95% in 2013 based on historic emissions (of which 15% are to be auctioned). While this analysis gives an overview on average expected burdens, it does not picture the uncertainty of future events. The following risk profile therefore gives an overview on the expected average annual industry costs until 2020 (S = 1000): Caused by the input variables distributions, expected financial burdens fluctuate around the mean of EUR 2.98 billion calculated before. Figure 3 indicates that 90% of all results range between EUR 1.88 and 4.07 billion. An analysis of the model robustness shows that fluctuating allowance prices and the growth of demand in particular have an effect on the financial impacts. Changes of these parameters such as sharply decreasing allowance prices would thus lead to strongly changing financial effects of the ETS system. Consequently, extreme parameter constellations like significantly increased allowance prices result into dramatically changed financial burdens for airlines. Changes in realized efficiency gains and fluctuating demand elasticities only have a minor impact. It has to be added that other input parameters – such as overall market growth – are in turn influenced by overall economic developments such as prices for energy. External shocks therefore often have an indirect influence on the demand for allowance prices, and with it ultimately on the economic and ecologic impact of the EU–ETS system on aviation. For 2009, total operating costs of international aviation are estimated to account for about EUR 360 billion (IATA, 2008). Assuming that of these costs, about 50%8 arise at airlines that are covered under the proposed ETS system, these airlines would show a cost base increased by about 1.25%. More specifically for the year 2012 and for the major European carriers, we estimate an additional cost burden of 0.92% for Air France/KLM, of 0.82% for Lufthansa German Airlines and of 2.1% for British Airways based on these carriers’ current cost structure. Additional costs as calculated in the simulation model may thus not have a substantial negative industry impact in the early years of the trading scheme. Nevertheless, as the number of free allowances is fixed at historic emissions, continuing growth in aviation may lead to more significant financial burdens, and putting additional pressure on an industry that regularly suffers under severe competition and widespread losses (Oum and Yu, 1998; Givoni and Rietveld, 2009). Also, an increased auctioning degree of the emissions scheme or a geographical system expansion (e.g. North America, Oceania) may lead to additional costs for airlines. On the revenue side, passenger growth is slowed down by an estimated average annual rate of 0.8% until 2020. While for 2013 demand is estimated to only be reduced by 0.7% compared to the unrestricted base scenario, this demand reduction is expected to rise to about 6% until 2020. Assuming a transferability of these numbers for revenue impacts, those may change likewise. Besides general economic consequences, revenue and cost effects of the ETS system are expected to impact competition structures within the industry. Since the ETS system is applied equally to EU and non-EU carriers some authors estimate that the introduction of such a system would not affect the operating efficiency of EU carriers significantly (Wit et al., 2005). This however, does not take into account three aspects: competition with other modes of transportation (Carlsson and Hammar, 7 b = (167.9 million tonnes CO2)/(246,921 million revenue tonnes kilometres). Based on a set of representative flight missions, Scheelhaase et al. (2009) calculate a benchmark of 0.61 kgCO2/RTK. The official benchmark will be published by the European Commission on September 30th, 2011. 8 The EU market covers about 1/3 of international traffic. However, the ETS system includes both EU airlines and international airlines offering flights to the EU, including most major North American and Asian carriers.

1073

J. Vespermann, A. Wald / Transportation Research Part A 45 (2011) 1066–1076

CO2 Emissions [Mil. tonnes]

Emission costs [Billion EUR] 4.0

375

3.67 3.49 350

3.31

3.5

3.14 2.98

325

2.81 2.66 2.50

300

2.25 275

258.0

270.4

264.2

276.8

289.9

283.3

296.8

310.9

303.7

3.0

2.5

2.0

250 1.5 225 1.0

200 175

65.1* %

62.2 %

60.7 %

59.3 %

58.0 %

56.6 %

55.3 %

54.1 %

52.8 %

2012

2013

2014

2015

2016

2017

2018

2019

2020

150

0.5

0,0

*Share of allowances allocated free of charge. Fig. 2. Estimated financial impact 2012–2020.

2002), potential cross-subsidies of routes at non-EU carriers, or advantages of carriers that are able to completely redirect international traffic flows. Competition structures in general are affected as air transportation is becoming less competitive compared to other modes of transportation – such as car or high-speed rail – on short-haul markets. Since short-haul flights usually feature high specific fuel consumption, a cost allocation according to induced emissions would have to assign higher costs to these flights (about EUR 3-7 per passenger depending on trip length). The exact impact of an increase in ticket prices however, is difficult to estimate, since travel decisions are generally based on a bundle of transport characteristics such as punctuality, flexibility

Percentiles 1.88

4.07

5%

90%

Mean

5%

1

2.98

0.9

Standard deviation 0.655

0.8

Minimum

0.7

1.52

0.6

Maximum

0.5

4.46 0.4 0.3 0.2 0.1 0 1.5

2.0

2.5

3.0

3.5

4.0

Fig. 3. Risk profile for average estimated financial impact 2012–2020.

4.5

Emission costs [Billion EUR]

1074

J. Vespermann, A. Wald / Transportation Research Part A 45 (2011) 1066–1076

On overall EU-ETS level incl. aviation:

-2.7%

-3.4%

In aviation only: 375

-3.9%

-4.4%

-5.0%

-5.5%

-6.2%

CAGR 3.4%

Unristricted scenario

CAGR 2.4%

Ristricted scenario (ETS system)

-6.8%

-7.7%

350

-6.8%

-5.8% 325

CO2 Emissions [Mil. tonnes]

-7.5%

-4.8% -3.9% -2.9%

300

-1.9% -0.9% 275

258.0

275.7270.4

285.0 276.8

2014

2015

304.7

294.7

325.8

315.0 296.8

336.9

310.9 303.7

289.9 283.3

266.7264.2

250

225

200 2012

2013

2016

2017

2018

2019

2020

Fig. 4. Estimated CO2 emission reduction from 2012–2020.

and schedule (Teichert et al., 2008). Furthermore, in aviation, pricing strategies are often detached from costs (e.g. subsidization of short-haul flights by long-haul flights). Secondly, non-EU carriers like North American or Asian airlines might choose to cross-subsidize routes covered by the ETS system. These airlines might also respond to the system by deploying their most efficient aircrafts on routes falling under the scheme. While second-order effects could in principle have an impact on competition structures, we assume these effects to be limited in practice either by strategic considerations or operational constraints, supporting the results of prior studies (Wit et al., 2005). A third effect associated with the ETS system is the emergence of artificial stops outside EU territory or the complete redirection of traffic flows. Artificial stops may occur in nations not covered by the system but close to EU territory (e.g. Switzerland, Turkey), and may be particularly interesting for non-European long-haul carriers on high density routes (Albers et al., 2009). While these additional stopovers could lead to reduced emission costs for airlines, they would lead to longer traveling times and incur additional costs like increased airport charges. The complete redirection of traffic flows could lead to a strengthening of hubs within the mentioned territories (e.g. Zurich, Istanbul) and their respective carriers such as Swiss and Turkish Airlines or of more distant hubs (e.g. Dubai, Doha) with carriers such as Emirates and Qatar Airlines. While the former airlines already operate a hub-and-spoke network serving many cities within the EU, carriers particularly from the Middle East may increasingly serve international traffic markets (e.g. US–India), bypassing the European hubs and thereby completely avoiding emission costs (Vespermann et al., 2008). In this case redirected traffic flows will lead to increased environmental costs, foiling the objectives of the emissions trading scheme. Even though we assume that emission costs are not high enough to instigate major route reconfigurations among European airlines (see also Albers et al., 2009) our results support findings put forward in recent studies and expecting a competitive advantage for carriers outside the EU market, such as airlines from the Middle East (Scheelhaase et al., 2009; Scheelhaase and Grimme, 2007). 4.2. Ecologic impacts Building on the projected emissions under an ETS system (Fig. 2) and those under an unrestricted base scenario, we calculate the ecological effects of the emissions trading scheme9: Concerning the total EU–ETS impact, air transportation would evoke emission reductions that amount to 2.7% in 2012 and rise to about 7.5% in 2020 (Fig. 4). Aviation therefore has a substantial effect on emissions reduction in industry sectors that 9 The total industry impact as presented in Fig. 4 comprises CO2 reductions induced by air transportation in all sectors currently subject to the EU–ETS (i.e. stationary sources) and aviation. Calculations are based on the design and allocation quantities of the existing EU–ETS system, allowing the emission of 2.08 bn tonnes CO2 in 2012 and 1.97 bn tonnes CO2 in 2013. From 2014 on, this number further decreases by 1.74% annually (Directive 2009/29/EC amending Directive 2003/87/EC).

J. Vespermann, A. Wald / Transportation Research Part A 45 (2011) 1066–1076

1075

are already covered under the existing EU–ETS system, provided that these industries are not ‘‘oversupplied” with emission permits. If focusing on emissions reductions that are limited to the aviation industry, we compute a CAGR for CO2 emissions that is about 1% lower under an ETS system compared to an unrestricted growth scenario (2012–2020 period). While in the first years after the introduction of the ETS scheme emissions reductions will be comparably low (<3%), the system will unfold its impact in the mid- and long-term. For 2020 we estimate an emission reduction of 7.7%, nearly equaling current emissions of the Lufthansa Group (26 million tonnes CO2). Our results show that the ETS system will bring environmental benefits. However, the results further indicate that the underlying intention of the political bodies, to significantly restrict growth in aviation, is not achieved. If political bodies consider the ETS scheme as a tool for substantially lowering the external effects of air transportation itself, this can only be achieved by either adapting the current scheme design (prohibit or limit the purchase of emission permits from other sectors; lower the cap; increase the auctioning degree), or by establishing additional policies covering so far unconsidered damages (such as the industries’ responsibility for other greenhouse gases and cirrus cloud enhancement). In this regard, the Intergovernmental Panel on Climate Change (IPCC) estimated the total climate change effect of aviation to be up to four times greater than the effect of its CO2 emissions alone. Janic (2003) assumes that considering all environmental costs of aviation, some routes may even become unprofitable. Alternatively, instead of attempting to constrain aviation it could be considered to raise the competitiveness of alternate transport modes with a better environmental performance, an approach suggested by Poudenx (2008) for urban passenger transportation. 4.3. Limitations The modeling of travel demand and resulting emissions is often subject to errors (Rodier and Johnston, 2002). Although our model is based on the most detailed data publicly available, real data on ticket prices, flight fuel consumption and demand elasticities is only available at airlines. Therefore, some of the data used for this analysis is an approximation. As a consequence of this, potential limitations with respect to the validity of the results lie within the data assumptions. While parts of the data will eventually be published, other data is unlikely to ever be publicly available. Estimated demand elasticities or information on flight fuel consumption represent highly confidential company data, allowing direct or indirect conclusions on customer loyalty or aircraft load factors. While some of the data uncertainty could be compensated for by the usage of simulation techniques and an analysis of variable distributions and risk profiles instead of fixed parameter constellations, an introduction of additional variables (e.g. average ticket prices distinguishing among different routes and booking classes; demand elasticities depending on travel distances and booking class) could come at the expense of an increased model complexity and parameter inaccuracy, leading to a lower predictive ability of the model (Flood and Carson, 1993). Further research could address these problems either by including data that will be available in the future (i.e. publication of the official benchmark in 2011), or by conducting studies building on internal airline data, such as case-based analyses in cooperation with individual industry players. 5. Conclusion This research aimed at quantifying both economical and ecological impacts of an inclusion of the air transportation industry into the EU–ETS system, taking the various parameters of the now fixed system design into account. We employed a simulation model to compute the effects of the ETS system. Until 2020, total financial burdens on the industry are expected to average about EUR 3 billion per year, but may significantly vary due to altered business conditions such as fluctuating allowance prices and demand growth. In the early years after the ETS system introduction, overall carbon permit costs are expected to account for about 1.25% of total industry costs. Competition distortions as a result of the introduction of the ETS system depend on the amount of these costs, though in general, they are estimated to be rather low. The ecologic effect of the ETS system is based on the assumption that higher costs in aviation will lead to a reduced demand for air transportation and less air traffic activity and thereby to lower emissions. Furthermore, since aviation is considered to be a strong net buyer of emission permits, the industry will induce emission reductions in other sectors. Compared to an unrestricted growth scenario, the annual growth rate for CO2 emissions within air transportation is expected to be about 1% lower under the ETS system. Due to the outflow of emission permits from stationary sources to aviation, more significant emission reductions can be achieved in these sectors. The ETS scheme: Much Ado about Nothing? Our research has shown that while the system will unfold both its economic and ecologic effects in the long run, the current system design will not evoke a substantial reduction of emissions from air transportation. If significant absolute reductions of environmental industry costs are intended, these reductions can only be reached by a more restrictive system design. References ACI – Airports Council International, 2008. Global Traffic Forecast Report 2008-2027, Geneva. Airbus, 2007. Global Market Forecast 2007-2026. Toulouse/Blagnac.

1076

J. Vespermann, A. Wald / Transportation Research Part A 45 (2011) 1066–1076

Albers, S., Bühne, J.A., Peters, H., 2009. Will the EU–ETS instigate airline network reconfigurations? Journal of Air Transport Management 15 (1), 1–6. Amekudzi, A.A., Jotin Khisty, C., Khayesi, M., 2009. Using the sustainability footprint model to assess development impacts of transportation systems. Transportation Research Part A (Policy and Practice) 43 (4), 339–348. Banister, D., 2008. The sustainable mobility paradigm. Transport Policy 15 (2), 73–80. Boeing, 2009. Current Market Outlook 2009-2028. Market Analysis, Boeing Commercial Airplanes, Seattle. Brons, M., Pels, E., Nijkamp, P., Rietveld, P., 2002. Price elasticities of demand for passenger air travel: a meta-analysis. Journal of Air Transport Management 8, 165–175. Carlsson, F., Hammar, H., 2002. Incentive-based regulation of CO2 emissions from international aviation. Journal of Air Transport Management 8 (6), 365– 372. Chapman, L., 2007. Transport and climate change: a review. Journal of Transport Geography 15 (5), 354–367. Flood, R.L., Carson, E.R., 1993. Dealing with Complexity: An Introduction to the Theory and Application of Systems Science. Springer, New York. Givoni, M., Rietveld, P., 2009. Airline’s choice of aircraft size – explanations and implications. Transportation Research Part A (Policy and Practice) 43 (5), 500–510. Goetz, A.R., Graham, B., 2004. Air transport globalization, liberalization and sustainability: post-2001 policy dynamics in the United States and Europe. Journal of Transport Geography 12 (4), 265–276. Gudmundsson, H., Höjer, M., 1996. Sustainable development principles and their implications for transport. Ecological Economics 19 (3), 269–282. Hoyos, D., 2009. Towards an operational concept of sustainable mobility. Sustainable Development and Planning 4 (2), 56–71. IATA – International Air Transport Association, 2004. Environment Review 2004, Montreal. IATA – International Air Transport Association, 2008. Financial Forecast 2009, Montreal. IATA – International Air Transport Association, 2009. Global Traffic Forecast, Montreal. Janic, M., 1999. Aviation and externalities: the accomplishments and problems. Transportation Research Part D (Transport and Environment) 4 (3), 159– 180. Janic, M., 2003. Modelling operational, economic and environmental performance of an air transport network. Transportation Research Part D (Transport and Environment) 8 (6), 415–432. Lu, C., 2008. The implications of environmental costs on air passenger demand for different airline business models. Journal of Air Transport Management 15 (4), 158–165. Macintosh, A., Wallace, L., 2008. International aviation emissions to 2025: can emissions be stabilised without restricting demand? Energy Policy 37, 264– 273. Montgomery, W., 1972. Markets in licenses and efficient pollution control programs. Journal of Economic Theory 5, 395–418. Morrell, P., 2007. An evaluation of possible EU air transport emissions trading scheme allocation methods. Energy Policy 35, 5562–5570. Olsthoorn, X., 2001. Carbon dioxide emissions from international aviation: 1950–2050. Journal of Air Transport Management 7 (2), 87–93. Oum, T.H., Yu, C., 1998. Cost competitiveness of major airlines: an international comparison. Transportation Research Part A (Policy and Practice) 32 (6), 407–422. Oxford Economics, 2008. Aviation – The Real World Wide Web. Report, Oxford. Poudenx, P., 2008. The effect of transportation policies on energy consumption and greenhouse gas emission from urban passenger transportation. Transportation Research Part A (Policy and Practice) 42 (6), 901–909. Rodier, C.J., Johnston, R.A., 2002. Uncertain socioeconomic projections used in travel demand and emissions models: could plausible errors result in air quality nonconformity? Transportation Research Part A (Policy and Practice) 36 (7), 613–631. Scheelhaase, J.D., Grimme, W.G., 2007. Emissions trading for international aviation – an estimation of the economic impact on selected European airlines. Journal of Air Transport Management 13 (5), 253–263. Scheelhaase, J.D., Grimme, W.G., Schaefer, M., 2009. The inclusion of aviation into the EU emission trading scheme – impacts on competition between European and non-European network airlines. Transportation Research Part D (Transport and Environment) 15 (1), 14–25. Schipper, Y., 2004. Environmental costs in European aviation. Transport Policy 11 (2), 141–154. Teichert, T., Shehu, E., von Wartburg, I., 2008. Customer segmentation revisited: the case of the airline industry. Transportation Research Part A (Policy and Practice) 42, 227–242. Tietenberg, T., 1980. Transferable discharge permits and the control of stationary source air pollution: a survey and synthesis. Land Economics 56, 391–416. Vespermann, J., Wald, A., Gleich, R., 2008. Aviation growth in the Middle East – impacts on incumbent players and potential strategic reactions. Journal of Transport Geography 16 (6), 388–394. WBCSD – World Business Council for Sustainable Development, 2004. Mobility 2030: Meeting the Challenges to Sustainability. Full Report 2004, Geneva. Wit, R.C.N., Boon, B.H., van Velzen, A., Cames, M., Deuber, O., Lee, D.S., 2005. Giving Wings to Emission Trading: Inclusion of Aviation Under the European Emission Trading System (ETS): Design and Impacts. Report for the European Commission, Delft.