Aviaion Policy and Performance in China: A Comprehensive Evaluaion Anming Zhang Sauder School of Business Universiy of Briish Columbia January 2007
Ouline 1. Aviaion Policy Developmens 2. Indusrial Performance 3. Evaluaion of Airpor Privaizaion on Efficiency 4. Corporae Governance & Compeiion 5. Conclusions
1. Aviaion Policy Developmens - Early aviaion reform separaed civil aviaion from he air force - Subsanial reforms began in lae 1980s: Six runk airlines: o separae CAAC from direc airline operaion Enry of non-caac carriers - In 1990s, Big Three carriers were given furher discreionary power (& allowed o lis in sock marke) - Consolidaion o Big Three
- Airline operaion was separaed from airpor operaion - Hisorically, CAAC, represening Cenral Governmen, owned & operaed airpors - Airpor funding comes from fiscal spending - Rapid raffic growh in 1980s-90s pressured on airpor infrasrucure - To overcome funding shorage, CAAC se up: Airpor infrasrucure fee (1992) Fund of infrasrucure consrucion for civil aviaion (1993)
- To encourage local iniiaive & funding, localize airpor conrol & managemen - Tes cases: Xiamen in 1988 Shanghai Hongqiao in 1993 - Localizaion program acceleraed in early 2000s & was compleed by 2003 excep Beijing and Tibe
- From 2006, Cenral Governmen sopped subsidizing airpors Funding will be mainly from local governmens I was also hoped airpors would improve efficiency and arac funds from privae secor - Since 2002, foreign invesors allowed o inves in airpors e.g. HKIA invesed in Hangzhou airpor (35%) & Zhuhai airpor Frapor in Ningpo airpor (25%); ec. - Foreign invesmen in airlines; JVs in cargo business
- Airpor privaizaion via IPO (Iniial public lising): In 1996, 25% shares of Xiamen airpor floaed in Shanghai Sock Exchange So far, 6 airpor companies incl. 7 airpors are lised - Airline IPOs - Asses managemen ransferred o SASAC under he Sae Council
- CAAC serves as he indusry s regulaor, aiming o mainain a fair marke environmen: Compeiion Proec consumers ineress: Safey; Air raffic conrol Inernaional affair & cooperaion - Thus, aviaion policy liberalizaion paves he way for more marke-oriened airline & airpor managemen
2. Indusrial Performance China ranked 2 nd in he world in 2005, in erms of boh pax-km (33 rd in 1980) & on-km (35 h in 1980) The Air proporion of domesic iner-ciy pax-km of all modes Highway, Rail, Waer, Air has increased from 1.7% in 1980 o 9.0% in 2002 Aviaion has been he fases growing ranspor mode 1980-2005, annual growh rae: passenger: 16.8% air cargo: 16.6%
Aviaion growh Passenger (Pax) Pax-km Air Cargo Cargo Tonkm Ave. ann. growh 1981-1985 16.8% 24.2% 17.0% 24.2% 1986-1990 17.3% 14.6% 13.6% 14.5% 1991-1995 25.3% 24.2% 22.3% 22.2% 1996-2000 5.6% 7.3% 14.2% 17.7% 2001-2005 17.5% n/a 13.8% 15.9% 1980-2005 16.8% n/a 16.6% 18.2%
Toal facor produciviy of airlines averaged abou 3% ann. growh higher han 1-2% for oher SOEs Ann. growh of labour produciviy: 11.4% for 1978-2000 Good performance is a resul of he policy liberalizaion E.g. Enry of non-caac carriers; laely, privae carriers Ensuing compeiion; hence, improved produciviies Relaxed fare conrol: up o 45% below he official base rae; no limi for ouris roues Simulaing demand and flee expansion
- Airpor decenralizaion: local iniiaive & invesmen in airpor infrasrucure - This objecive appears o be achieved E.g. In 1999, 40% of invesmen in civil-aviaion fixed asses from local governmens Shanghai governmen ook iniiaive o build Pu-dong airpor & conribued 80% of funds Similarly, new Guangzhou airpor
Safey performance CAAC conrols financial finess & safey qualificaion of new carriers; ATC Invesigaive group by he Sae Council afer a major acciden A major concern: when gov sared o wihdraw from airline & airpor operaions, aviaion safey would be compromised? Has he policy liberalizaion increased acciden risk?
Figure 1. Faal acciden rae and number of flighs (1950-2005) Faal Fligh Acciden Rae per 10,000 Flighs 1.2 1 0.8 0.6 0.4 0.2 0 8,000 1.11 7,000 6,000 5,000 4,000 3,000 2,000 0.209 0.192 0.098 1,000 0.021 0.006 0 1950-59 1960-69 1970-79 1980-89 1990-99 2000-05 Commercial Flighs (in 1,000)
Figure 2. Inciden rae (1996-2005) 1.8 1.6 1.5772 1.6007 Inciden per 10,000 Flighs 1.4 1.2 1 0.8 0.6 0.4 1.4444 1.1184 0.8151 0.894 0.9434 0.8726 0.802 0.7777 0.2 0 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005
Acciden rae has dropped since 1980s Inciden rae dropped significanly during 1998-2000 a period when mos airlines were lised in sock markes and remained sable aferwards Thus, safey improvemen & liberalizaion may acually go hand by hand Commercially-oriened firms (airlines, airpors) hemselves have srong incenives o mainain safey
Air raffic conrol (ATC) Three-level managemen: CAAC ATC Bureau, 6 regional bureaus, and he ATC ceners Beer coordinaion beween miliary and civil aviaion auhoriies is imporan 30% of air space available for civil aviaion Sub-opimizaion in airspace uilizaion: airpor congesion & delays become a major problem Daily limis on ake-offs/landings in Beijing & Shanghai Alhough underinvesmen in ATC faciliies, foreign invesmen is sill resriced
3. Evaluaion of Airpor Privaizaion on Efficiency - Privaizaion via IPO: improve efficiency? - Prior expecaion Lised airlines/airpors would be required o fulfil higher corporae governance sandard, and be subjec o capial marke discipline Hence, improve managemen and efficiency - A major problem of Chinese airpors: Low produciviy
Table 1. ATRS airpor produciviies, 2004 (Vancouver = 1.0) Beijing 0.487 Shanghai 0.413 Guangzhou 0.776 Shenzhen 0.555 Asia Pacific Mean 0.682 Europe Mean 0.684 Norh America Mean 0.731 Hong Kong 0.931 Residual Variable Facor Produciviy
- Performance oucome? - No an easy ask o evaluae: Lack of daa & exising sudies - Fung, e al. (2006) compued produciviy for 25 major airpors in China over 1995-2004 They asked: Convergence of produciviy among airpors from differen regions? - Here, I asked wheher public lising improves produciviy?
Table 2. Sample airpors & lising saus Airpor Beijing Changsha Chengdu Chongqing Dalian Guangzhou Hailar Harbin Hefei Hohho Jinan Kashi Kunming Lanzhou Nanning Qingdao Sanya Shanghai Shenyang Shenzhen Taiyuan Tianjin Urumqi Xiamen Xian Lising Year/ Sock Exchange 2000/ Hong Kong 2003/ Shanghai 1998/ Shanghai 1998/ Shenzhen 1996/ Shanghai
Table 3. Lised airpors Airpor Beijing Guangzhou Shanghai Shenzhen Xiamen Lising Year/ Sock Exchange 2000/ Hong Kong 2003/ Shanghai 1998/ Shanghai 1998/ Shenzhen 1996/ Shanghai Sae Share in 2003 65.0% 60.0% 63.0% 64.0% 75.0% * Noes: * Sae share in 1997.
Table 4. Efficiency scores from DEA analysis Lised Airpors Non-Lised Airpors Year Coun Mean Sandard Sandard Coun Mean deviaion deviaion 1995 0 N.A. N.A. 25 0.4811 0.2892 1996 1 1.0000 N.A. 24 0.4590 0.2891 1997 1 0.4184 N.A. 24 0.5628 0.3423 1998 3 0.7021 0.3003 22 0.5194 0.3213 1999 3 0.6727 0.3469 22 0.4740 0.3177 2000 4 0.6770 0.2929 21 0.4991 0.2965 2001 4 0.6976 0.2899 21 0.4487 0.3105 2002 4 0.6974 0.2391 21 0.4284 0.3020 2003 5 0.7586 0.2272 20 0.3950 0.2578 2004 5 0.7716 0.2521 20 0.4135 0.2638 Pooled 30 0.7161 0.2471 220 0.4703 0.2985
Regression analysis Furher run he following regression: e = a + a L + bh + cy O 0 1 j i i i e O = Efficiency score of airpor O in year L = Dummy variable for lising (1 if airpor is lised; 0 oherwise) j Y = Year dummy H i = Hub dummy wih i=1,2 represening he inernaional and regional hubs, respecively.
Conrol for hub saus - Definiion of a hub airpor: Eiher in a large ciy or sraegically locaed so ha i has large connecing raffic, or boh - Inernaional hubs: Shanghai, Beijing, Guangzhou - Regional hubs in he sample: Shengyang, Xian, Chengdu Shenzhen (No.4), Kunming (No. 5), Uramuqi
Figure 3. Six regions & hubs 烏魯木齊
Table 5. Regression analysis of produciviy level Models 1 2 (excl. Xiamen) 3 (yr dummies) 4 (yr & oher dummies) Inercep 0.3980** 0.3960** 0.3907** 0.7585** (19.9097) (19.7620) (7.4898) (4.7791) Lising 0.0932* 0.1330** 0.1042* 0.2454** Hub Saus (1.7318) (1.9952) (1.8669) (4.2085) Inernaional Hub 0.4818** 0.4653** 0.4777** 0.4052** (8.9546) (8.1186) (8.7529) (7.5486) Regional Hub 0.1640** 0.1660** 0.1650** 0.1635** (4.0118) (4.0749) (3.9933) (4.0963) Adjused R 2 0.3080 0.3230 0.2933 0.4054 No. of Observaions 250 240 250 250
Produciviy growh The Malmquis index, 2 1 1 1 1 1 1 1 1 1 1 1 1 1 ), ( ), ( ), ( ), ( ), ( ), ( ),,, ( = + + + + + + + + + + + + O O O O O O O y x D y x D y x D y x D y x D y x D y x y x M O D = Oupu disance funcion of airpor O, which is he same as he efficiency score
Table 6. Mean produciviy changes (1995-2004) & componens Malmquis Produciviy Technical Efficiency Technological Change Lised Airpors 1.2082 0.8499 1.4215 Non-lised Airpors 1.3395 1.1045 1.2128
Regression analysis of produciviy growh = + + + O 0 1 j i i i M a a L b H cy
Table 7. Regression analysis of produciviy growh 2 Models 1 (yr dummies) Inercep 1.0829** 1.139** (46.8068) (21.2063) Lising -0.0307-0.0488 Hub Saus (-0.4979) (-0.8191) Inernaional Hub -0.0165-0.0109 (-0.2713) (-0.1887) Regional Hub -0.0174-0.0191 (-0.3679) (-0.4278)
Figure 4. Malmquis produciviy change before & afer IPO Produciviy Change 1.2 1 0.8 0.6 XMN SHA SZX PEK CAN 0.4 0.2-3 -2-1 0 1 2 3 Year before and afer lising
4. Corporae Governance & Compeiion 4.1 Dominan sae ownership of airpors - Sae share: dominan & non-radable - Lile incenive for SOE managers o exer effor - and for Governmen, as he owner, o monior he managemen - On he oher hand, dominan sae ownership likely creaes inefficiency
4.2 Corporae governance problems - Info disclosure: accuracy & compleeness of info Criical o effecive (exernal) marke discipline Ligh punishmen on missaemen IPO approval & supervisory handled by same Deparmen - Inernal conrol sysem E.g. Lack of independence of he Supervisory Board - Insiuional invesor For he 4 lised airpors: Larges insiuional invesor only holds abou 5% (Shenzhen airpor)
4.3 Inernaional aviaion policy & compeiion - Used o be very conservaive, due largely o: Weak domesic carriers Large imbalance beween Chinese and foreign ravelers - As a resul, limied inernaional raffic righs - Limied marke opening & compeiion
4.4 Compeiion & HKIA s success - In erms of produciviy, HKIA (Hong Kong In l Airpor) performs well relaive o: Airpors for Asia, Europe & N. America Mainland s airpors - Compeiive airpor charges - World s No. 1 in inernaional air cargo & No. 5 in inernaional passengers
Imporan reason: Compeiive pressures - In 2005, 41 million passengers go hrough HKIA (6 imes HK populaion), of which 1/3 are hub raffic - Of he 3.44 million air cargo, 86% are gaeway or hub raffic - Compeiion for gaeway raffic wih PRD (Pearl River Dela) airpors: E.g. Over 70% air cargo from/o PRD - Compeiion for hub raffic Asian compeiors: Seoul, Tokyo, Osaka & Singapore Laely, from Beijing, Shanghai & Bangkok Threa of Mainland-Taiwan Direc ranspor links
4.5 Recen policy relaxaions Over pas decade: Rising domesic income Chinese carriers joining global airline alliances (e.g. Air China in Sar Alliances) Less resricions on ciizens ravelling abroad Visa waive policy Liberal bilaeral ASA wih U.S., Korea, Japan, ec. Cargo liberalizaion firs Improve corporae governance: E.g. arac foreign insiuional invesors Spli share reform: sae shares in lised airpors have been reduced o jus over 50% (from 60% plus)
5. Conclusions - Significan effors have been exended o he reform of adminisraive and regulaory frameworks, wih Cenral Governmen gradually wihdrawing from direc airline & airpor operaions and managemen - The policy liberalizaion has conribued o dramaic growh in air raffic and produciviy, and has improved marke compeiion and air safey - Airpor localizaion has been successful in encouraging local invesmens in airpor infrasrucure
- Mixed resuls regarding he efficiency impac of parial privaizaion via IPO: Lised airpors are producively more efficien han nonlised airpors Lile evidence ha lised airpors produciviy performance has improved significanly afer IPOs In effec, lised airpors produciviy grows slower han nonlised airpors produciviy
- Given majoriy sae ownership will remain in foreseeable fuure, srenghening corporae governance and inroducing compeiion should be vial for efficiency improvemen - Implicaions of Norheas Asia aviaion liberalizaion and inegraion - Finally, Cenral Governmen needs o srenghen ATC managemen and funding
Thank You