Costs & Quality Is higher punctuality more costly?

Transcription

1 Costs & Quality Is higher punctuality more costly? Brussels, 4 nd November Andrew Smith & Manuel Ojeda-Cabral ITS, University of Leeds

2 Rail Context & Aims Proportion of trains arriving at destination on time Reliability: a key element of a transport network Big issue in the UK railway Plenty of research looking at demand s perspective But very little from the supply s point of view Lack of theory and evidence on the quality-cost link (e.g. Civity, 2012) Aim: explore and understand the relationship between costs and reliability of train operating companies.

3 Marginal cost and valuation of quality

4 Definitions OECD (2010) defines reliability as the ability of the transport system to provide the expected level of service quality upon which users have organized their activity (Kouwenhoven and Warffemius, 2015). Reliability of travel times: how faithful the actual service is, compared to the promised timetables. An element of Quality but importantly, a promised quality TOCs must keep them to some extent. Several measures: punctuality, delay minutes, standard deviation, Public Performance Measure (PPM)

5 Quantifying (un)reliability 100% Reliability Scenario (or Reliability Scenario): Delay = Lateness Lateness = the difference between the actual and publicly timetabled arrivals at destinations Buffer time = additional travel time (in excess of the minimum necessary) added as contingency < 100% Reliability Scenario (or Unreliability Scenario):

6 Articulating the theory (II) How do (un)reliability and costs relate to each other? Direct effects Different types of costs! 1 2 BUT ALSO Indirect effects 3 BUT THEN

7 Articulating the theory (III) Different type of costs: Reactive costs = f(lateness) Lateness = f(preventive costs) Reactive Preventive Staff extra hours Compensations, penalties. Reactive maintenance + Foregone revenue Staff & Rolling stock. Capacity Buffer time related Preventive maintenance + Foregone revenue Decision: Preventive Costs Outcome: Reliability = X Consequence: Reactive costs

8 Articulating the theory (IV) Implications It is possible to improve Quality AND reduce costs. If Reactive Costs are too high If the PC-RC balance is optimized (note: solely from the TOC costs perspective), Reliability improvements must come at an extra cost (extra PC).

9 Articulating the theory (V) 300 Costs vs. Lateness Illustration numerical example Preventive Lateness Reactive Costs 150 Reactive Total Costs Preventive ,9 0,8 0,7 0,6 0,5 0,4 Lateness 0,3 0,2 0,1 0

10 Empirically

11 Data 18 Train Operating Companies in the UK 4 years of data ( ) plans to extend the sample. Can we learn something from this short period of time for the same 18 firms? Several measures of Reliability available: Total delay minutes per year, split by who was responsible (TOC, IM or another TOC) % delayed trains (for several delay thresholds) PPM: Public Performance Measure - % of scheduled trains arriving at their terminating station 'on time (within 5 or 10 minutes) Partial analysis of the data first

12 Reliability data (I) 10,00% Distribution of trains by arrival time for one TOC 9,00% 8,00% 7,00% 6,00% 5,00% 4,00% Govia Thameslink Railway - % trains 5 to 10 minutes late Govia Thameslink Railway - % trains 10 to 29 min. late Govia Thameslink Railway - % trains 30 min. to 2 hours late Govia Thameslink Railway - % trains cancelled 3,00% 2,00% 1,00% 0,00%

13 Empirically (II) What can we observe with the available data? 1) Across companies for a given year % Punctual Trains vs. Total Costs 1,2 % of trains arriving within 5 minutes 1 0,8 0,6 0,4 0, Total Costs of the TOC

14 Empirically (II) What can we observe with the available data? 1) Across companies for a given year - TOCs nature

15 Empirically (III) What can we observe with the available data? 1) Across companies for a given year

16 Empirically (IV) What can we observe with the available data? 2) Within company across several years Unit costs vs. Reliability - Within TOC (Govia Thameslink example) 16 [CELLRANGE] 14 [CELLRANGE] [CELLRANGE] 12 [CELLRANGE] Unit costs PPM (Quality measure)

17 Conclusions Where have we got to? What are the next steps? 1) Finalize the generic theoretical economic framework 2) Develop the framework for rail context accounting for the existing incentives. 3) Operationalize econometric framework based on the above. 4) What policy and regulatory implications could follow?

18 Thanks!