 |
Project Information |
 |
Project Description
This project will study the port productivity for major container ports in the world. We attempt to answer two questions: First, what are the major determinants
affecting shippers' choices among ports; Second, given this knowledge, what are the best practices to improve port efficiency (productivity) for a port. We will
answer these two questions based on empirical studies. By collecting data from ports around the world, discrete choice models to address the first question will
be built, and measures to compare port performance under different policy scenarios will also be developed. The results from this project will be very useful for a
port to design appropriate policies. For example, we can answer the following ongoing heated debate in Hong Kong Ports: What if a 35% reduction of port charges
is implemented at Hong Kong terminals in response to the low pricing policy by the ports in southern China (e.g., Yantian Port in Shenzhen).
Project Objective
• To study the production efficiency of container port industry
• To understand the determinants of port production efficiency, especially how different deregulation practices affect the efficiency deregulation practices affect the efficiency
• To develop econometric framework to study ports' production/investment/performance
|
|
Project methodology |
|
Introduction
As logistics/supply-chain is evolving to be the artery of global economy from a typical business function, efficiency of ports has become an important factor affecting a nation's international competitiveness. Thus, monitoring and comparing one's ports with other ports in terms of overall efficiency has become an essential part of many countries' microeconomic reform programs. As indicated by UNCTAD (2006), the pressure to boost port competitiveness has triggered an ever-increasing number of port benchmarking studies, especially on global container ports.
In all efficiency studies, efficiency is measured by comparing observed and optimum costs, production, revenue, or whatever the organization is assumed to pursue, subject to the constraints on quantities and prices. The optimal quantity is termed frontier and the efficiency is then the distance between the observed quantity and the frontier. In empirical research, two methods are widely used to calculate or estimate the frontier functions and thereby measure efficiency: data envelope analysis (DEA) and stochastic frontier analysis. DEA is a deterministic method based on linear programming and was first introduced by Charnes, Cooper and Rhodes (1978), whereas stochastic frontier analysis is an econometric method accounting for random shocks and measurement errors. Stochastic frontier model was proposed by Aigner, Lovell, and Schmidt (1997) and Meeusen and van den Broeck (1977). As summarized, many studies use these two methods to study the port efficiency.
Today, port industry in the world has a very complex organizational structure. Each port has many terminals, which are operated by one or several operators. The operators are just like the firms to operate the terminals for their own objectives. For example, the Hong Kong Port has 10 terminals operated by six operators – Cosco-HIT Terminals (Hong Kong) Ltd, Hong Kong International Terminals Ltd, Kowloon Wharf Terminal & Warehouse Ltd, Asia Container Terminals Ltd, Modern Terminals Ltd, and DP World.
Administrative and ownership
structure of Hong Kong ports
(Source: Cullinane and Song, 2001.)
On one hand, since each operator in a port makes his own operation and investment decisions, the efficiency level should be different among different operators in a port. On the other hand, different operators in a port should have certain similarity in production efficiency because they share the same water way and other natural conditions. On a higher level, the operators from different ports within the same country should also have certain similarity in production efficiency because they are subject to the same government regulations and legal systems. More complicated than this, most terminal operators in the world today belong to several major port groups such as Hutchinson Port Holdings (HPH), Port of Singapore Authority Corp (PSA), P&O Ports, Stevedoring Services of America (SSA) and Dubai Ports World Company. Two operators, even from different countries, should have certain similarity in production efficiency if they belong to the same port group. Ideally, an empirical model on port production efficiency should account for these features. Because of data limitation, in this paper we treat the terminal operators as independent decision makers. This is already an extension for existing literatures, which treat ports as decision makers.
The empirical model in this paper is developed under the stochastic frontier framework first expounded by Meeusen and van den Broeck (1977) and Aigner et al. (1977). We use with Cobb-Douglas production function. The input, output and individual characteristics are showed.
Compared with traditional stochastic frontier model, we randomize the intercept of the production frontier to account for individual heterogeneity. This is called the true random effects model in Greene (2005). When the operators are different in their adopted technologies, and such difference is not well controlled, the estimated inefficiency absorbs both the heterogeneity and inefficiency and is thus biased. Unlike the traditional stochastic frontier model using convenient one-parameter distribution form, such as half-normal and exponential distribution, to model random inefficiency, we use a more flexible log-normal distribution with two parameters. More importantly, this specification enables us to interact some observable managerial inputs with the inefficiency to seek for policy implications. Modeling time varying inefficiency is also possible based on this model specification.
The two pictures plot the estimated distribution of efficiency level in different years from the two models. Model 1, in which the individual heterogeneity in production frontiers is controlled only by the observables, clearly overestimates the inefficiency. This suggests that the estimated inefficiency by the model which does not allow individual heterogeneity across operators, in fact, absorbs the individual heterogeneity in production frontier, and thus the distribution of individual level efficiency shifts to the left. Increasing the variance of priors on has certain but insubstantial effects on the estimated inefficiency.
Markov chain Monte Carlo, MCMC
Markov chain Monte Carlo (MCMC) methods (which include random walk Monte Carlo methods), are a class of algorithms for sampling from probability distributions based on constructing a Markov chain that has the desired distribution as its equilibrium distribution. The state of the chain after a large number of steps is then used as a sample from the desired distribution. The quality of the sample improves as a function of the number of steps.
Usually it is not hard to construct a Markov Chain with the desired properties. The more difficult problem is to determine how many steps are needed to converge to the stationary distribution within an acceptable error. A good chain will have rapid mixing—the stationary distribution is reached quickly starting from an arbitrary position—described further under Markov chain mixing time.
Typical use of MCMC sampling can only approximate the target distribution, as there is always some residual effect of the starting position. More sophisticated MCMC-based algorithms such as coupling from the past can produce exact samples, at the cost of additional computation and an unbounded (though finite in expectation) running time.
The most common application of these algorithms is numerically calculating multi-dimensional integrals. In these methods, an ensemble of "walkers" moves around randomly. At each point where the walker steps, the integrand value at that point is counted towards the integral. The walker then may make a number of tentative steps around the area, looking for a place with reasonably high contribution to the integral to move into next. Random walk methods are a kind of random simulation or Monte Carlo method. However, whereas the random samples of the integrand used in a conventional Monte Carlo integration are statistically independent, those used in MCMC are correlated. A Markov chain is constructed in such a way as to have the integrand as its equilibrium distribution. Surprisingly, this is often easy to do.
Multi-dimensional integrals often arise in Bayesian statistics, computational physics and computational biology, so Markov chain Monte Carlo methods are widely used in those fields.
|
|
Port Operator Efficiency Index |
|
General efficiency about port industry
As far as we know, there only a ranking list of TEU now. This provides no enough information for port operation and authority. How efficiency is when the TEU of a port is ranked higher? What attributes the best efficient port has? How difference of performance between the individual port and the group port, such as HPH, DP World, PSA, P&O Ports and SSA? Privatization can lead to higher efficiency? How GDP, trade volume, legal system of a country, ownership, corporate governance, tariff barrier, and deregulation of a port affect the efficiency of port?
Our research and industry practice will answer the questions. Firstly, we will measure and compare the efficiency performance of seaports. Secondly, we examine the relationships between various performance measures and seaport characteristics in order to understand the observed differences in seaport performance.
In all efficiency studies, efficiency index is measured by comparing observed and optimum costs, production, revenue, or whatever the organization is assumed to pursue, subject to the constraints on quantities and prices. The optimal quantity is termed frontier and the efficiency is then the distance between the observed quantity and the frontier.
We plot the average efficiency index for port industry from 1997 to 2004. The efficiency index is from 0 to 1. The port will be best in efficiency when the efficiency index is 1. The score is lower when the efficiency is worse. The efficiency index increase from time to time. The increment speed drops down after 2000.
Individual Operator's Efficiency
There is a ranking list based efficiency for all the sampled operators. Just for a sample, we point the location of operator A in the sequence.
|
|
Research output |
|
| ◦ | Featured panel speech, "Outlook of Port Competition in Greater China Region", Session on "Shipping and Ports in Asia", The 4th International Conference of Academic Scholars (ICAS) – Shanghai, China. October 20 – 25, 2005. |
| ◦ | Yan, Jia, Xinyu Sun, and John J. Liu. Technical efficiency of container operators: an Econometric analysis to the world’s major container ports, the Proceedings of the 11th HKSTS Conference, Dec. 2006 |
| ◦ | Host and present at PolyU PPRI Forum 2006 – Logistics Breakout Session, "Regional Ports and Airports Competition Outlook: Performance Benchmarking and Simulation," April 27 – 29, 2006. |
| ◦ | Jia Yan, and John Liu, Plenary session theme presentation, "Global Container Ports Performance Benchmarking," International Association of Maritime Economists (IAME) 2006 – Melbourne, Australia, July 13 - 15, 2006. |
| ◦ | John J. Liu, Trade-based Supply Chain: A Global Trend and The China Factor, IFSPA 2008 |
| ◦ | Oum, T.H., J. Yan, and C. Yu. Ownership Forms Matter for Airport Efficiency: A Stochastic Frontier Investigation of Worldwide Airports. Journal of Urban Economics 64 (2008) 422-435. |
| ◦ | John J. Liu and T.L. Yip. APEC Seminar: Best Practices in Regulation of Ports, Regulation and Port Productivity Overview of Global Port Benchmarking (A Focus on Hong Kong and South China Region), APEC Lima (Peru) 2008 |
| ◦ | Yan, Jia, Xinyu Sun, and John J. Liu. Assessing Container Operator Efficiency with Heterogeneous and Time-Varying Production Frontiers. Transportation Research Part B 43 (2009) 172-185. |
| ◦ | John J. Liu, T.L. Yip and Xinyu Sun, Productivity and Regulation of Global Container Ports: Asia and China Factor, Dokuz Eylul – Hong Kong PolyU Workshop on Maritime Business and Management, 19 – 21 January, 2009, Izmir, Turkey |
| ◦ | T.L. YIP, Xinyu Sun, John J. Liu, Group Competition of Global Ports, IAME 2009 Conference, 26 – 26 June, 2009: Copenhagen, Denmark |
|
|
Project Team |
|
| Project Leader | |
| Prof. John Liu MS, PhD Head, Chair Professor of Maritime Studies, Department of Logistics and Maritime Studies Director, C. Y. Tung International Centre for Maritime Studies The Hong Kong Polytechnic University Consulting, Research and Teaching Interests: Maritime and Supply Chain Logistics, Incentive and Contract, Maritime Rescue Systems, Port Policy and Industrial Organization |
 |
| Members | |
| Dr T. L. Yip BEng, PhD, MBA, CEng, MIMechE, MIEE, MCIM, MHKIM Assistant Professor Department of Logistics and Maritime Studies C. Y. Tung International Centre for Maritime Studies The Hong Kong Polytechnic University Consulting, Research and Teaching Interests: Ports & terminals, Commercial shipping, Transport risk |
 |
| Dr Jia Yan BA, MA, PhD Assistant Professor, School of Economic Sciences Washington State University Consulting, Research and Teaching Interests: Transportation Economics, Applied Microeconomics, Applied Econometrics and Statistics |
 |
| Dr Xinyu Sun BA, MA, PhD Senior Engineer Department of Logistics and Maritime Studies C. Y. Tung International Centre for Maritime Studies The Hong Kong Polytechnic University Consulting, Research and Teaching Interests: Logistics and Supply Chain, Applied Microeconomics, Manufacturing Strategy |
 |
|
|
Contact US |
|
Maybe you are interested in your operator/port's efficiency index. How about your location in the ranking list of efficiency? What's your efficiency
compared with other operator in the same port?
Don't hesitate; contact us to get your own efficiency report.
Please send a request email to
Prof. John Liu
Director of C. Y. Tung International Centre for Maritime Studies
The Hong Kong Polytechnic University
Hung Hom, Kowloon, Hong Kong
Email: lgtjliu@inet.polyu.edu.hk