The future of rail in New Zealand

The future of rail in New Zealand Dave HeatleyResearch FellowNZ Institute for the Study of Competition and Regulation ISCR February Seminar 19 February 2009

---

Outline  A quick history Poor economic performance for decades • Important because economic performance is a measure of net benefit to society  Can the problem be improved by changing governance or ownership? • Experience to date suggest otherwise  Do other social benefits (externalities) justify having rail anyway? • Externalities are smal , benefits unclear or are better dealt with by other mechanisms  Is there a future for rail? 2

---

History: how has rail evolved?  First railway opened 1863 Lines built by provincial governments but came under central government control from 1876  Initial y fragments of rail connected ports to the hinterland  Joining the fragments into one network took a century • North Island Main Trunk completed 1908 • South Island Main Trunk completed 1945 • Rail ferries joined islands in 1962 3

---

Rail network growth and decline 1908: current network Rail network 1952-92: contraction 1993: rail size first reached during public ownership privatised 6000 ) 5000 m (k e 4000 il lin a 3000 rnep 2000 O 1000 0 1870 1880 1890 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 Year 1936: limits on road freight 1983-86: limits on road imposed – al owing overshoot freight removed  The rail network has evolved as it adapted to competition Adaptation has slowed since 1991, but external conditions continue to change 4

---

Rail’s evolution driven by competition  Rail’s initial purpose was to connect ports to the agricultural hinterland – “opening up the country”  Branch lines up every major valley were eventually out- competed by roads and subsequently closed  As the country industrialised, rail connected ports to mines and industrial sites  Rail still has this role for transporting large volumes of non- urgent bulk goods  Moving people between urban centres was the primary motivation for connecting into one big network  Rail took over from sea travel, but was eventually supplanted by road and air  Moving general freight was a complementary use of the network  But as the road network grew, it offered better connectivity and even better economies of scope  Rail today needs road for door-to-door delivery (“intermodal”)  Intermodal is competitive over long distances only 5

---

Economic performance: characteristics  Rail is a very long-lived asset • Most rail today runs along formations built in the 1800s • Maintenance and renewal are expensive, and get more costly as the network ages  The majority of assets are fixed, and most of those are sunk  Individual lines have different profit characteristics • Need volume to be profitable • Network effects means that they cannot be evaluated in isolation  Evaluations of the economics of rail are very sensitive to: • Rates of maintenance and renewal • Deferring maintenance can make things look good in the short term • Valuation of assets • E.g. Is the fixed rail infrastructure worth $81m (market value) or $6bn (ONTRACK valuation)? • Choice of required rate of return • Choice of reference point(s) 6

---

Financial performance 1875-1982 Rail profit/loss as % of revenue: 1875-1982 50% 0% -50% 1875 1885 1895 1905 1915 1925 1935 1945 1955 1965 1975 7

---

Economic performance 1983-1997 R a il – C o m p o n e n ts o f E c o n o m ic S u rp lu s 200.0 100.0 0.0 (100.0) (200.0) 1997 $m (300.0) (400.0) (500.0) (600.0) 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 O perating profit O pportunity c os t on c apitalNet ec onom ic s urplus  “the history of rail in New Zealand has been one of enormous taxpayer and social cost entailed in retaining rail as a state-owned enterprise” (ISCR, 1999) 8

---

Running down the assets: 1991-2008 Sleeper replacement 1991-2008 160 140 120 )s00 100 (0 rs 80 epe 60 leS 40 20 0 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 Sleepers replaced (000s) Long-term replacement level Public Private ownership Public Track renewal 1991-2008 120 )m 100 (kde 80 wen 60 rekc 40 raT 20 0 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 Track renewal (km) Long-term replacement level 9

---

TranzRail, Mainfreight and NZX 800 700 600 500 400 ex (1996=100) dIn 300 200 100 0 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 NZX Al gross index TRH share price (adjusted with dividends) MFT share price (adjusted with dividends) 10 1

---

Comparison with other SOEs  As at 1 July 2008, government equity in KiwiRail & ONTRACK valued at $12bn – same as Meridian, Genesis, NZ Post, Mighty River Power, Transpower, Solid Energy and al other SOEs combined (CCMAU, 2008)  In 2007/08, the other SOEs had an after-tax profit of $637m on revenues of $9.4bn (after-tax ROE 5.3%)  Rail (+ferries) had negligible profit on revenues of $741m. Would need to triple rail revenue without increasing costs to get similar ROE  But rail ‘needs’ $1bn programme to rebuild rundown network  The $12bn valuation is indefensible $6bn land value is opportunity cost (but only available if network is shut down) 11 1

---

Other evaluations  Booz Al en Hamilton (2005) calculated economic deficit of rail freight was $424m in 2000/01.  “in the foreseeable future it is unrealistic to expect that [ONTRACK] wil earn a commercial rate of return on its approximately $12 bil ion of land and rail assets (CCMAU, 2008)  “…our rail system cannot survive without substantial government subsidies into the future” (Clark, 2008) 12

---

Performance: freight carried Growth in surface freight 1993-2007 25 Rail Road s) 20 nillio 15 s (b kmenn 10t toeN 5 0 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 13 1

---

Performance: summary  Rail has performed poorly since at least the 1920s  The rational response is a gradual rundown of long-lived infrastructure – and this has been happening  Rail is unfixable in its present form without “heroic assumptions” about its future economics  The network should be closed unless: • A change in ownership or governance can fix this problem • It has significant positive externalities that exceed the costs involved and these cannot be achieved at lower cost by other means; or • A viable subset of the network can be found 14

---

Changing governance and ownership  Ongoing experiment to find a model that works Orr (1981) observed a cycle which repeated on four occasions (1880s-1950s): • A perception of poor financial performance lead to a review finding political interference as a primary cause and recommending corporatisation as a remedy • Corporatisation implemented with the goal of making rail a viable business. It general y succeeded in improving the financial position • Within five years each corporatisation had been reversed, inevitably due to demand for increased political control • The financial performance gains were typical y short- lived 15

---

Governance changes 1875-1982 Profit/loss for rail in New Zealand: 1875-1982 (showing periods of corporatisation) 50% A B C D 0% -50% 1875 1885 1895 1905 1915 1925 1935 1945 1955 1965 1975 Profit/loss as % of revenue 16 1

---

Corporatisation #5 and privatisation  The fifth corporatisation was NZRC in 1982. Waves of line closures and restructuring over next 10 years  Privatisation 1993 (NZ Rail) and public float (Tranz Rail) in 1996  Tranz Rail initial y bul ish about investment but that slowed after 1997  Financial performance dived from 2000 Over 2002-2004 government bought the below-rail infrastructure (ONTRACK) – creating structural separation 17

---

Rail buyback completed in 2008  Ontrack and Tol unable to agree on a rail access charge  Government wanted increased investment – but Tol could not justify it based on expected returns  Government preferred to subsidise a SOE rather than a “foreign-owned” private company – but why? Government should be indifferent if its desired outcomes are achieved  Government wanted to prevent rail-line closures (“rail essential for regional development”) – but at what cost? 18 1

---

Governance and ownership: summary  Ownership models tried • Local ownership (1860-1876) • Independent private railways (various to 1908) • Government department (5 times) • Government corporation (5 times) • Privately held (1993-1996 & 2007-2008) • Public company (1996-2007) • State-owned enterprise (2008+)  Vertical integration • Integrated with ferries (1962-) • Integrated with trucking (1980s to 2008) • Structural separation (2004-2008)  Typical y economics improved around time of changes but reverted within a few years  Underlying economic problems persist 19

---

Externalities  Recent justifications for rail are based on reducing externalities: especial y accidents, congestion and greenhouse emissions  But the optimal level of externalities is not zero – at some point it becomes more expensive to lower them than the welfare created by their further abatement  The aim is to lower externalities towards this point at the least cost.  As rail investments are very expensive, they need to be evaluated according to this least cost principle 20

---

Rail safety and congestion  “Rail is safer” – or is it? • Truck-related road deaths: 5.4/year • Rail-related deaths: 5.3/year (Average fatalities per billion NTK, 1998-2007) • Rail injury rates appear lower but the figures are not directly comparable  “Shifting freight to rail use reduces congestion” • Ignores congestion issues at transfer points • Ignores congestion issues at level crossings • Ignores congestion within the rail network • Ignores off-peak scheduling of the larger trucks that actually compete with rail  Congestion costs $1bn/year (BAH, 2005) • 70% in Auckland, 20% Wellington & CCH • Little reduction likely from increasing rail freight share. Rail carries non-urgent freight so off-peak trucking is a viable substitute • Commuter rail can help. Subsidies already internalise these costs 21 2

---

Rail and greenhouse emissions  Conventional wisdom: “rail is 4x better than trucking” (e.g. MOT, 2008b) • Considers averages, not the marginal values applicable to mode shifts • Typical y quoted figures are from Europe – may not be applicable to NZ • When used to justify projects, constant technology into the future is often assumed • Based on operating emissions only – ignores ful life cycle • Same sources suggest that sea freight is 2x better than rail • Rail needs road for door-to-door delivery 22 2

---

Rail CO : marginal comparison 2  Averages are dominated by non-contestable freight. For rail, this includes the point-to-point transfer of non-urgent bulk goods that it is preferential y suited for  Contestable portion is smal part of overal freight task (for Australia this estimated to be 4-15%)  Affleck (2002) studied actual emissions for 7 specific Australian long-distance corridors • Intermodal (rail+road) distances generally longer • Intermodal produced 41-69% (average 55%) of the road- only emissions  Gabal a & Cranley (2008) studied shipping consumer food products to Melbourne • Intermodal distances were 30% longer • Intermodal produced 61% of road-only emissions  These figures are less that 2x – nowhere near 4x general y quoted in NZ 23 2

---

Modal shift forces inefficient behaviour Freight task (NTK) 2006/07 (NFDS 2008) Air 0.3% Coastal Rail Road shipping 15%* 70% 15% Coastal Rail Road shipping 25% ~45% 30%** Freight task (NTK) 2040 (NZTS 2008) Air ? * NZTS 2008 uses 18%. **Target specified as a percentage of inter-regional freight. The proportion of total freight may be lower . •Implementation of the NZ Transport Strategy (MOT, 2008a) requires shipping to steal all of rail’s modal share – so rail must steal that plus more from road 24 2

---

Rail and indirect energy costs  Main inputs into rail infrastructure are steel, concrete and diesel • All are significant sources of CO emissions 2  For the Netherlands, Bos (1998) estimated that 45% of total rail energy use is in providing the rail infrastructure. (This proportion would be much higher on lightly-used tracks in NZ.)  O’Toole (2008) found urban rail energy payback periods were often longer than the project lifetime – up to 172 years for a light rail project in Portland, Oregon  O’Toole also found that rail projects tends to cannibalise the most efficient bus routes – lowering their average energy efficiency  Modal comparisons are misleading without ful lifecycle accounting of environmental costs 25 2

---

Greenhouse: summary  Operational emission benefits wil be negated by infrastructure emissions unless infrastructure wel -chosen and heavily used  Forced modal shift wil move rail into tasks where it is not natural y efficient, potential y increasing overal emissions  Greenhouse externalities best dealt with via carbon tax or equivalent emission- trading system which should drive an optimal modal al ocation 26

---

Externalities: summary  Results are mixed Main externalities are already internalised, or best internalised via other mechanisms  Remaining externalities relatively smal relative to the economic problem  Transport infrastructure should be decided on primary benefits not externalities (Glaeser & Gottlieb, 2008) 27

---

Is there a future for rail?  Commuter rail • Wel ington and Auckland rail should be evaluated independently – but require ful environmental and economic analysis  Tourist rail • Only viable on lines paid for by other rail users  Rail freight • Extensive pruning in the 1950s-1980s was not enough to achieve economic viability • Recommended approach is to find a viable subset, then apply marginal analysis to see if any additions worthwhile 28 2

---

A viable subset?  Concentrate on rail’s strengths • Primary role: point-to-point transport of large quantities of bulk goods • Integrate with – rather than duplicate – coastal shipping routes and state highways  Rail’s potential environmental and economic advantages can be realised if only heavily-used lines are retained • 30% of the network (Auckland, Waikato, Bay or Plenty, Canterbury and West Coast) carries 70% of freight • These are also the regions with the majority of future predicted increases in demand (Richard Paling Consulting, 2008) • Closing 70% of the network would free up land (possibly worth billions) which could fund the upgrading of the core network to 21st century standards 29

---

Summary  Poor economic performance for almost a century  Changing governance or ownership is unlikely to help  Externalities are not sufficient to justify having the current rail system  A future for rail requires finding a configuration that is economical y viable • Smal er and more targeted than current network 30 3

---

Questions Final report and slides wil be available from www.iscr.org.nz 31 3

---

References 1 Affleck Consulting. (2002). Comparison of Greenhouse Gas Emissions by Australian Intermodal Rail and Road Transport. Brisbane: QR Network Access. Booz Al en Hamilton. (2005). Surface Transport Costs and Charges Study: Main Report. Ministry of Transport, New Zealand. Bos, A. J. M. (1998). Direction indirect : the indirect energy requirements and emissions from freight transport. Groningen. Retrieved from http://irs.ub.rug.nl/ppn/17069478X. Clark, H. (2008, July 1). Ful text: Helen Clark’s speech on the launch of KiwiRail. The New Zealand Herald. Retrieved from http://www.nzherald.co.nz/nz/news/article.cfm?c_id=1&objectid=10519328. Crown Company Monitoring Advisory Unit. (2008). Briefing to Incoming Ministers on entities monitored under Vote State-Owned Enterprises. Retrieved from http://beehive.govt.nz/sites/al /files/CCMAU_VSOE_BIM_0.pdf. Gabal a, S., & Cranley, L. (2008). Food Miles in Australia: A comparison of emissions from road and rail transport. Melbourne: CERES Community Environment Park. Retrieved from http://www.ceres.org.au/projects/FoodMiles_A_comparison_of_emission_from_road_and_rail_transport.pdf . Glaeser, E. L., & Gottlieb, J. D. (2008). The Economics of Place-Making Policies. Discussion Paper. Harvard Institute of Economic Research. Retrieved from http://ssrn.com/abstract=1299046. 32

---

References 2 ISCR. (1999). The Privatization of New Zealand Rail Part 2: Quantitative Cost Benefit Analysis. New Zealand Institute for the Study of Competition and Regulation. Retrieved October 2, 2008, from http://www.iscr.org.nz/f251,5018/5018_tranzrail_part_2_100799.pdf. Ministry of Transport. (2005). National Rail Strategy to 2015. Retrieved from http://transport.govt.nz/assets/NewPDFs/nrs.pdf. Ministry of Transport. (2008a). The New Zealand Transport Strategy 2008. Retrieved from http://www.transport.govt.nz/assets/Downloads/NZTS-final-PDF.pdf. Ministry of Transport. (2008b). Sea Change: transforming coastal shipping in New Zealand. Retrieved from http://www.transport.govt.nz/assets/Images/NewFolder-2/Sea-Change-2008.pdf . Richard Paling Consulting. (2008). National Freight Demands Study. Ministry Of Transport, New Zealand. Retrieved from http://www.transport.govt.nz/assets/Images/NewFolder-2/FREIGHT-STUDY-EXEC-SUMMARY1.pdf . Orr, S. (1981). New Zealand Railways Corporations. A Short History of Previous Experiences With New Zealand Railways as a Corporation.Economic Division, Ministry of Transport. O’Toole, R. (2008). Does Rail Transit Save Energy or Reduce Greenhouse Gas Emissions? SSRN eLibrary. Retrieved February 10, 2009, from http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1263426. 33

---