Air Transport Research Society Key Results of the ATRS Global Airport Benchmarking Tae H. Oum University of British Columbia, Canada And The Global Airport Benchmarking Task Force of Air Transport Research Society (ATRS) www.atrsworld.org Asia Pacific: P. Forsyth, Yeong-Heok Lee, Yuichiro Yoshida, Japhet Law Europe: Jaap de Wit, Nicole Adler, Hans-Martin Niemeier, Eric Pels North America: David Gillen, Tae Oum, Bijan Vasigh, Chunyan Yu
Outline Objective of the Benchmarking Study Airports Included Methodology Key Results on Efficiency and Costs Airport User Charge Comparisons Effects of Business Strategies and Ownership forms Conclusions Air Transport Research Society
Objective of the Study To provide a comprehensive, unbiased evaluation of airport performance, including: Productivity and Efficiency Unit Cost Competitiveness Aviation User Charges Levels (Financial Performance) Air Transport Research Society
Objective of the Study Con t. Identify effects of Business environment within which an airport operates Ownership and management forms Extent of focus on non-aviation (commercial) activities Extent of outsourcing Ownership forms On productivity, unit cost, other performance measures; and Airport User Charges Air Transport Research Society
Air Transport Research Society Many Previous Studies have one or more of the following shortcomings Limited to analysis of airports in a certain country or continent. Used less than comprehensive measures; Ignored Multiple Outputs/Services especially ignoring Non-Aeronautical Services (Commercial; Development, etc) Ignored so-called soft-cost inputs including outsourcing services Did not control for the factors beyond managerial control
ATRS Global Benchmarking Work To measure and compare: productivity and efficiency, cost competitiveness, and financial results, of the world s major airports To provide the most comprehensive performance benchmarking results on various dimensions of airport activities To develop a method for computing the overall efficiency index after removing the effect of factors beyond managerial control To examine the sources of productivity and unit cost differences between airports Air Transport Research Society
Airports Included in the 2006 Report* Canada-U.S. Europe 63 airports 38 airports and 10 airport groups Asian Airport 24 airports and 4 airport groups Australia and NZ 9 airports ------------------------------------------------------------- Total 134 airports and 14 airport groups * Mostly use 2004 data (except 2005 data used for airport user charges) Air Transport Research Society
Air Transport Research Society Data Sources: 2001-04 Airport s Annual Reports, Financial Statements, and direct data requests; US FAA, DOT statistics; Association of European Airlines (AEA) Statistics ICAO Digest of Statistics: annual and monthly traffic data annual financial data -- not for all airports ACI; IATA annual traffic statistics Capacity information general information surveys (Asia Pacific and Europe) occasional and not complete IMF and World Bank various price indices including GDP deflators for service sectors and PPP
Airports Characteristics Air Transport Research Society
90 80 70 60 50 40 30 20 10 0 Figure 3.4.1c: Passenger Traffic (2002-04) - North America Millions ATL ORD DFW LAX DEN LAS PHX JFK MSP IAH DTW SFO EWR MCO MIA SEA YYZ PHL BOS LGA CLT IAD CVG FLL BWI MD HNL SLC TPA SAN YVR DCA OAK STL PIT PDX CLE SJC MEM YUL MCI MSY SNA YYC SMF RDU BNA IND AUS SAT ONT MKE PBI ABQ JAX RNO YEG YO SDF YHZ ALB YW RIC Mea 2004 2003 2002 Air Transport Research Society (ATRS)
80 70 60 50 40 30 20 10 0 Figure 3.4.1b: Passenger Traffic (2001-04) - Europe M illio n s LHR CDG FRA AMS MAD LG W FCO MUC BCN ORY MAN STN CPH MXP ZRH DUB ARN IS T BRU DUS OSL VIE ATH TXL HEL LIS HAM PRG BHX GVA CGN EDI WAW MLA CIA LJU TLL BTS M ean 2004 2003 2002 2001 Air Transport Research Society (ATRS)
45 40 35 30 25 20 15 10 5 0 3.4.1c: Passenger Traffic (2001-04) - Asia Pacific Millions HKG BKK PEK NRT SIN SYD ICN DXB KUL PVG CAN TPE MEL MNL BOM KIX SHA SEL BNE SZX DEL AKL PER MAA CHC ADL WLG HKT MFM CNS PEN CNX HDY CEI Mean 2004 2003 2002 2001 Air Transport Research Society (ATRS)
Figure S-5: Non-Aeronautical Revenue Share (2004) 90% 80% 70% 60% North America Europe Asia Pacific 50% 40% 30% 20% 10% 0% Air Transport Research Society (ATRS)
Airport Productivity Analysis Inputs Outputs Labor Other non-capital (soft cost) inputs Runways Terminals Gates Aircraft movements Passengers (Cargo) Other revenues including concessions Air Transport Research Society (ATRS)
Methodology for productivity MEASUREMENT Partial Factor Productivity (PFP) Measures: Labor Productivity Capital Input Productivity Productivity of Soft Cost Inputs Need to go beyond PFPs PFP does not tell the whole story Variable Factor Productivity (VFP); Total Factor Productivity (TFP: capital input accounting problem) Cost Competitiveness Air Transport Research Society
Air Transport Research Society Methodology for productivity MEASUREMENT cont d Multiple Outputs: Aircraft movements, passengers, Cargo Tonnes, and non-aeronautical services output The first step for computing any productivity (partial, multi-factor, or total factor productivities) is to aggregate these multiple outputs into a single output index Similarly, airports use multiple inputs which need to be aggregated into a single input index.
Efficiency Measurement Method: Our Choice Index number approach: Productivity = Output Index / Input Index VFP = Output Index / Variable Input Index VFP (Variable Factor Productivity) Since multiple outputs need to be aggregated into an index, there has to be weights for aggregating multiple outputs; in applied studies we normally use revenue shares as weights for aggregation Since multiple inputs are used, they need to be aggregated to form an input index. Usually cost shares of these inputs are used as weights for aggregation. Air Transport Research Society
Potential Reasons for the Measured Productivity (gross VFP) Differentials (A) Factors Beyond Managerial Control: Airport size (Scale of aggregate output) Average aircraft size using the airport Share of international traffic Share of air cargo services Extent of capacity shortage - congestion delay Connecting/transfer ratio We compute residual (net) productivity measures after removing effects of the Factors Beyond Managerial Control (A) (B) Factors within Managerial Control: Emphasis on commercial activities (non-aeronautical) Quality of Service (incl. passenger satisfaction) The Extent of outsourcing activities Managerial and technical efficiency (which we are trying to measure)
Air Transport Research Society Results on Effects of Airport Characteristics on gross VFP (Airport size: some mixed results on the effects of airport output scale: larger airports are expected to have higher gross VFP, however, the effects may level off once airports reach certain size); %International: airports with higher proportion of international passengers are likely to have lower gross VFP. %Cargo: airports with larger proportion of cargo traffic are expected to have higher VFP. Capacity Constraints: congested airports are likely to have higher gross VFP. Therefore, these factors which are beyond airport s managers control have been removed from our VFP measure before computing the residual VFP.
Residual (Net) Variable Factor Productivity: Overall operating efficiency measure Asia-Pacific Figure S-4c Residual Variable Factor Productivity - 2004 Asia Pacific HKG=1.0 1.2 1 0.8 0.6 0.4 0.2 0 ICN HKG SIN SEL BNE KIX SYD CAN MEL ADL MFM AKL WLG PER CHC SZX NRT PEK CNS Mean AOT MAHB SAA AAI Mean(group) HKG=1.0 Air Transport Research Society (ATRS)
1.2 1.0 0.8 0.6 0.4 0.2 0.0 Residual (Net) Variable Factor Productivity: Overall operating efficiency measure - Europe Figure 4.5.2b Residual Variable Factor Productivity - 2004 Europe CPH=1.0 CPH = 1.0 CPH ATH GVA OSL BRU ZRH LGW LJU AMS LHR DUS EDI VIE MAN HAM BHX MUC WAW PRG CGN MLA Mean ANA CAA - Finland ADR BAA ADP FRA DAA CAA-Sweden PPL Berlin Mean (group) Air Transport Research Society (ATRS)
1.8 1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0 Residual (Net) VFP: Overall operating efficiency measure North America Figure 4.5.2a Residual Variablre Factor Productivity - 2004 North America YVR=1.0 YVR=1.0 ATL TPA RDU FLL RNO LAS MCO YVR MKE MSP BNA YOW PBI CLT YUL HNL SMF SDF DFW OAK AUS BOS MCI PDX YWG YYC IND SAT PHX SNA DEN ABQ RIC LGA DCA MDW SJC SLC SEA IAH ALB SFO YHZ MSY CVG SAN JAX ORD STL MEM CLE IAD EWR DTW YEG BWI LAX JFK PIT PHL ONT YYZ MIA Mean Air Transport Research Society (ATRS)
1.8 1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0 Global Comparison Net VFP Figure S-4 Residulal Variable Factor Productivity - 2004 North America Europe Asia Pacific YVR=1.0 AOT MAH PEK CNS SAA AAI ATL TPA RDU FLL RNO LAS MCO YVR MKE MSP BNA YOW PBI CLT YUL HNL SMF SDF DFW OAK AUS BOS MCI PDX YW G YYC IND SAT PHX SNA DEN ABQ RIC LGA DCA MD SJC SLC SEA IAH ALB SFO YHZ MSY CVG SAN JAX ORD STL MEM CLE IAD EW R DTW YEG BW I LAX JFK PIT PHL ONT YYZ MIA CPH ATH GVA OSL BRU ZRH LGW LJU AMS LHR DUS EDI VIE MAN HAM BHX MUC WAW PRG CGN MLA ANA CAA( ADR BAA ADP FRA DAA CAA( PPL Berli ICN HKG SIN SEL BNE KIX SYD CAN MEL ADL MFM AKL WLG PER CHC SZX Air Transport Research Society (ATRS)
Alternative Methods show similar results; Comparison of Net VFP and Stochastic Variable Cost Frontier Results Figure 4B-1a Residual VFP versus Cost Frontier Efficiency Index 1 1.8 1.6 1.4 Residual VFP 1.2 1 0.8 0.6 0.4 0.2 0 0 0.2 0.4 0.6 0.8 1 1.2 Cost Frontier Efficiency Air Transport Research Society (ATRS)
Air Transport Research Society Cost Competitiveness consists of: Productive Efficiency Residual (Net) VFP Variable Input Prices: Labor price Soft cost input price (incl. outsourcing price)
Figure 5.4a Cost Competitiveness 2004 - North America YVR=0.0 SAT HNL SDF MKE BNA SMF PBI RIC YUL AUS YOW MCI OAK ABQ YYC IND PHX MSP YWG SLC ALB MSY DFW YHZ IAH DCA JAX SNA PDX BOS MEM YEG CLE BWI SJC CVG SEA LGA MDW SAN ORD LAX DEN STL PIT IAD DTW PHL EWR ONT SFO YYZ JFK MIA Mean FLL RDU LAS CLT RNO MCO YVR TPA ATL 0.8 0.6 0.4 0.2 0-0.2-0.4-0.6-0.8-1 Air Transport Research Society (ATRS)
Figure 5.4b Cost Competitiveness 2004 - Europe CPH=0.0 MLA LGW BRU HAM MAN OSL EDI LHR BHX VIE AMS ZRH MUC GVA CGN Mean CAA(FIN) ADR BAA FRAPORT DAA ADP Berlin CAA(SWE) Mean (group) DUS CPH ANA PPL TLL LJU ATH PRG 0.4 0.3 0.2 0.1 0-0.1-0.2-0.3-0.4-0.5-0.6 Air Transport Research Society (ATRS)
Figure 5.4c Cost Competitiveness (2004) - Asia Pacific HKG=0.0 CAN SZX NRT KIX BNE HDY ADL AKL WLG SYD MEL CHC CNS PER mean MFM SIN HKG PEK ICN SEL SAA AAI MAHB AOT mean (group) 0.8 0.6 0.4 0.2 0-0.2-0.4-0.6-0.8-1 -1.2 Air Transport Research Society (ATRS)
Air Transport Research Society Top Performers Top Operating Efficiency Performers based on Net VFP (Labor + Soft cost inputs only): Canada/US: Atlanta, Tampa, Raleigh-Durham Europe: Copenhagen, Athens, Geneva Asia-Pacific: Incheon (Korea), Hong Kong, Singapore Oceania: Brisbane, Sydney Top Performers Based on Unit Cost Competitiveness Index Canada/US: Atlanta, Tampa, Fort Lauderadale, Europe: Tallinn, Ljubljana, Athens Asia: AOT, Guangzhou, Shenzhen, Beijing, Oceania: Brisbane, Adelaide
Outline Objective of the Benchmarking Study Airports Included Methodology Key Results on Efficiency and Costs 2005-Airport User Charges Comparison Conclusions Air Transport Research Society
12000 10000 8000 6000 4000 2000 0 Landing Charges for Boeing 747 - North America 2005 US$ BNA ATL CLT SLC TPA RDU FLL MSY SAT PHX RIC LAS PBI MCO OAK SDF SJC MDW CVG RNO MSP MEM SNA MCI JAX ONT MKE HNL PHL IND IAD PIT SAN BWI MIA ABQ DCA SMF PDX ALB LAX ORD SEA AUS BOS YUL DTW IAH SFO YOW YWG DEN YEG YHZ YVR YYC STL DFW JFK EWR LGA YYZ Air Transport Research Society (ATRS)
Landing Charges for B747 - Europe 2005 $8,000 $7,000 $6,000 $5,000 $4,000 $3,000 $2,000 $1,000 $0 US$ LGW-off peak Man-Off Peak TXL LGW-Peak STN LHR-off peak MXP FCO CIA BRU DUS FRA LHR-peak MLA PRG ATH GVA CGN MUC LIS HAM ZRH IST Man-Peak VIE HEL ARN EDI MAD BCN OSL CDG ORY CPH MBX DUB WAW TLL BTS AMS LJU BHX Air Transport Research Society (ATRS)
$8,000 $7,000 $6,000 $5,000 $4,000 $3,000 $2,000 $1,000 $0 Landing Charges for B747 - Asia Pacific 2005 US$ KUL DXB BKK CNX HDY HKT CEI ADL CNS CGK CHC SIN MNL PER MAA DEL BOM MFM BNE TPE AKL HKG IC N PEK PVG SHA CAN SZX NRT KIX Air Transport Research Society (ATRS)
Summary Landing/Takeoff Charges Global Results: Airports charging highest aircraft movement charges: Boeing 747: Toronto, Narita, Kansai, Birmingham Airbus 320: Toronto, Kansai, Narita, Birmingham North American Results (both B747 and A320): Lowest aircraft movement charges: Nashville, Atlanta, Charlotte, Tampa, Salt Lake City Highest charges: Toronto, LaGuardia, Newark, JFK European Results: Lowest charge for B747: Gatwick-offpeak; Manchester-offpeak; Berlin Tegal, Gatwick-Peak, Stansted; Heathrow-offpeak; Lowest charges for A320: Berlin Tegal, Milan Malpensa, Rome Fiumicino, Rome Ciampino, Gatwick-offpeak, Frankfurt, Malta Highest charges for B747: Brimingham; Amsterdam, Bratislava, Tallinn, Warsaw Highest charges for A320: Ljubljana, Oslo, Heathrow-Peak, Vienna Asia-Pacific Results: Lowest charges: Kuala Lumpur, Thailand airports; Dubai; Adelaide, Jakarta Soekarno-Hatta, Singapore Highest charges for B747: Kansai, Narita, major Chinese airports, Incheon; Highest charges for A320: Kansai, Narita, Chrischurch, Major Chinese airports, Hong Kong Air Transport Research Society
Air Transport Research Society Outline Objective of the Benchmarking Study Airports Included Methodology Key Results on Efficiency and Costs Airport User Charge Comparisons Effects of Business Strategies and Ownership Forms Conclusions
Results on Business Strategies Diversification of Revenue Source is good: Airports with larger share of nonaeronautical revenue achieves higher Net VFP (efficiency) Outsourcing: Airports who contract out their terminal operations to outside operator achieve higher efficiency Air Transport Research Society
Results on Business Strategies Diversification of Revenue Source is good: Airports with larger share of nonaeronautical revenue achieves higher Net VFP (efficiency) Outsourcing: Airports who contract out their terminal operations to outside operator achieve higher efficiency Air Transport Research Society
Air Transport Research Society Empirical Results on Ownership Forms Airports with private majority ownership achieved significantly higher profit margins than others, despite the fact that they generally levy lower aeronautical charges (because of their vigorous pursuit of commercial opportunities). Privatized airport with govt majority is less efficient than those owned by private majority or 100% gov t corporation; Airports with government majority and airports owned by multiple governments are the least efficient. Choice: either Majority Private Ownership or 100% Government Ownership
Empirical Results on Ownership Forms cont d About U.S. City Operated Airports: However, Airports with private majority are not necessarily more efficient than airports owned/operated by U.S. city governments or 100% public corporations (note: privatized airport also has monopoly power, not necessarily more efficient). Airports operated by U.S. and Canadian Airport Authorities are no more efficient than the airports operated by US city departments; Air Transport Research Society
Please Note The ATRS Global Airport Performance Benchmarking Report : 3 volumes, over 400 pages with valuable data and analysis Can be purchased by visiting www.atrsworld.org 2007 ATRS World Conference: 21-23 June, to be hosted by Univ. of California at Berkeley Air Transport Research Society
Thank You Air Transport Research Society