Prensip pou siveye pèfòmans pò: Pwojè ranfòse regleman pò Kap Ayisyen

PRINCIPLES OF MONITORING PORT PERFORMANCE CAP-HAITIEN PORT REGULATORY STRENGTHENING PROJECT February 2018 PRINCIPLES OF MONITORING PORT PERFORMANCE Activity Title: Cap-Haïtien Port Regulatory Strengthening Project Sponsoring USAID Office: Office of Infrastructure Contract Number: AID-521-C-16-00003 Contractor: Nathan Associates Inc. Disclaimer This document is made possible through the support of the American people through the U.S. Agency for International Development (USAID). The contents of this document are the sole responsibility of the author or authors and do not necessarily reflect the views of USAID or the United States government. CONTENTS ACRONYMS 1 INTRODUCTION 2 EVOLUTION OF PORT INDICATORS 5 Growing Interest in Monitoring Port Performance 6 Port Performance Measurement for Benchmarking Against Other Ports 7 Port Authority Administration Indicators 14 Where We Go From Here 17 CONCEPTS AND BEST PRACTICES 19 Assessing Performance 23 Time Accounting System 26 Regulation by Contract 32 Vessel and Gate Operations Reports 34 HAITIAN CONTEXT 38 Current Situation at Cap Haitien Port 38 Examples from Other Haitian Ports 40 Assessment of Proposed Performance monitoring at CHP 43 CONCLUSIONS AND RECOMMENDATIONS 48 Specific Requirements for Monitoring Port Performance at Cap Haitien Port 48 Access to the Major Operations Reports Generated by the TOS 49 Proposed Extended List of Performance Indicators to be Generated by the TOS 49 Defining Adequate Target Values Indicators 52 Calculating Performance Indicators 52 Interim Period 53 Non-Containerized Cargo 53 P R I N C I P L E S O F P O R T P E R F O R M A N C E M O N I T O R I N G | 2 Measuring Port Authority Performance 53 USER MANUAL 55 APN Access to TOS 55 Illustative Examples of Calculating Performance Indicators 55 Fulfilling Regulatory Functions: Monitoring, Enforcement and Taking Corrective Actions 64 ILLUSTRATIONS Figures Figure 1. Port Logistics Chain 3 Figure 2. Deming's (Plan, Do, Check, Act) Wheel of Continuous Quality Management 6 Figure 3. Comparative Regional Berth Productivity Levels by Vessel Size 10 Figure 4. Basic Marine Terminal Stock and Flow 20 Figure 5. Time Accounting System for Port Operations 28 Figure 6. Ship (Vessel) Operation Report 35 Figure 7. Gate Operation Report 36 Figure 8. Ship (Vessel) Operation Report 37 Figure 9. Example of CPS' Crane Productivity 41 Figure 10. Vessel Operation Report 58 Figure 11. Crane Operation Report 59 Figure 12. Gate Operation Report 61 Figure 13. Yard Operation Report 63 Tables Table 1. DEA Scores of 20 Italian Ports 11 Table 2. Regulatory Indicators for Operational Efficiency and Level fo Service 12 Table 3. UNCTAD Port Authority Scorecard Indicators 16 Table 4. European Sea Ports Organization Port Indicators 17 Table 5. Example of Contractual Performance Indicators 3 3 Table 6. Proposed Performance Indicators in Draft PPP Contract 4 4 Table 7. Berth Performance Indicators 5 0 Table 8. Suggested Port Authority Indicators 54 Table 9. Data Items and Performance Indicators 56 ACRONYMS APN Autorité Portuaire Nationale APEC Asia-Pacific Economic Cooperation Forum CHP Cap Haïtien Port CSR Corporate Social Responsibility CY Container Yard DEA Data Envelopment Analysis FA Factor Analysis FTE Full Time Equivalent (for employee status) HA Hectares HHI Herfindahl-Hirschman Index JOC Journal of Commerce LOA Length overall (of a ship) LOS Level of Service MHC Mobile Harbor Crane Mty Empty (containers) PAP Port a Prince PCA Principal Components Analysis PI Performance Indicators PPP Public-Private Partnership PPRISM Port Performance Indicators: Selection and Measurement USAID U.S. Agency for International Development S&F Stock & flow TEU Twenty foot equivalent units TOS Terminal Operation System UNCTAD United Nations Conference on Trade and Development INTRODUCTION The Autorité Portuaire Nationale of Haiti (APN) in the near future for the first time will engage a private sector operator for a container terminal. The terminal will be built in the Port of Cap Haitien (CHP), a relatively small port that also handles bulk and breakbulk cargoes. As the terminal operator will hold a monopoly position for the foreseeable future and hence will not be subjected to competitive influences, the government of Haiti will impose minimum performance standards within the concession contract to ensure a reasonable level of service and productivity in the terminal. While APN will impose these standards, APN is also interested in monitoring other dimensions of operations not under the terminal operator’s control or not necessarily addressed within the contract’s performance provisions. Such monitoring can be used for benchmarking against prior years’ performance or against rival or best practice operators in other ports, or for planning purposes to prepare for capacity expansion. As many relevant performance indicators, or the data for their calculation, are generated through an operator’s terminal operating system (TOS), it is expected that there will be a provision in the concession contract that requires the operator to provide the needed data to APN. And APN may have audit authority to validate the data the terminal operator provides. Indicators are generally derived from the various activities that occur between the port’s entrance buoy and the port (or terminal) gate, constituting the port logistics chain. Figure 1 is a graphic representation of the port logistics chain. It identifies a range of basic services and who is charged for these services – generally, the shipper (the importer and exporter) and carrier (or ship’s agent). The graphic shows the sequence of 11 general activities in the port container terminal and the descriptions of each of these activities. Terminal operators usually employ TOSs to ensure equipment readiness to carry a container and real time tracking of each container’s position in the terminal and stage of customs processing. In this way, containers and equipment are optimally staged, reducing idle time of both equipment and containers. Data are collected through the TOS that can be used by the port authority or regulatory body for monitoring terminal operator performance and adherence to tariff and performance standards in concession agreements. Normally, operational performance indicators are tied to the activities that occur within the port logistics chain. P R I N C I P L E S O F P O R T P E R F O R M A N C E M O N I T O R I N G | 3 Figure 1. The Port Logistics Chain 1 While port authorities have an interest in understanding the performance of their terminals, it is in their interest to ensure they perform their port administration responsibilities effectively. Effective resource allocation is the rallying cry for modernized port authorities and stakeholders 1 Based on Kent, Paul E. and Alan Fox, “Is Puerto Limon a Real Lemon? The Impact of Port Inefficiency on a National Economy”, in The International Handbook of Maritime Economics , edited by Kevin Cullinane, 2011. 1) A vessel pilot boards the vessel and guides the vessel’s captain through the port’s entrance channel. If no berth is available, the vessel is assigned to an anchorage area. Tugs then help maneuver the vessel to the berth. Line handlers then tie two lines to the berth. At this point, the ship has incurred charges for navigation or port dues (to cover the cost of dredging the channel and providing lights and buoys for navigation safety), pilotage, tug assist, and line handling. 2) Once the vessel is secured to the berth, the port authority or terminal operator may apply a berthage charge, essentially a “parking” fee usually calculated on the basis of time and vessel size (length). The berthage charge stops when the last line is untied from the ship as it leaves the berth. The charge is applied to the vessel. In some cases, line handling may be incorporated into the berthage charge. 3) Inspection authorities (e.g. defense security, drug enforcement, immigration) may board the ship. Usually cargo is not loaded or unloaded until the authorities have completed their inspections. 4) Longshoremen load or unload cargo using a crane. Containers are “lashed” (secured to the crane) and moved from the ship to the “apron,” the area at the berth set aside for loading and unloading. If a terminal operator’s crane is used, then a charge for crane use is applied to the ship. 5) The port authority may impose a “wharfage” charge, which is intended to cover the use of the apron and other areas of the terminal where the container is moved to or from storage. Wharfage is charged to the shipper (i.e., importer or exporter). 6) The container seal is inspected and a fee may apply, in which it is charged to the shipper. 7) The container is moved to or from an assigned slot in the “yard”. Container storage operations occur in the yard. This area as a whole constitutes “dispatch,” where the container is moved to or from an assigned slot (a space in the yard). Dispatch fees are charged to the shipper. Some ports, particularly smaller ones, may not have a true storage area, but instead have a small area that serves as a buffer between the berth operation and the area behind the berth. In such cases, trucks quickly evacuate containers from the buffer storage area or the container may be loaded onto a truck directly from the vessel for immediate evacuation to an off-dock storage area, sometimes referred to as a satellite storage area. Haiti’s Cap Haitien and the Aleman container terminal in Puerto Limón, Costa Rica have very limited or no buffer storage capacity and hence require prompt evacuation of containers. 8) The container is stored until it is inspected and claimed by the importer. In some ports, off-dock storage areas exist but their relevance decreases as the port increases capacity. 9) The container is inspected by customs and moved onto a truck chassis. 10) Some ports may employ container scanners and a charge for scanning usually applies to the shipper. 11) Gate processing includes weighing the container and reviewing paperwork. Charges for weighing the container apply to the shipper. P R I N C I P L E S O F P O R T P E R F O R M A N C E M O N I T O R I N G | 4 Authorities are keenly aware of the consequences on them of administrative and cost burdens reflected in port authority charges and regulations; stakeholders also demand that resources are properly allocated to ensure effective stewardship of public access infrastructure. Additionally, some port authorities, even if reflecting the landlord model, also provide services to port users, the performance of which the port authority should monitor. In Cap Haitien, for example, both pilotage and tug assist are provided by the port authority. Vessel waiting time for either of these services can have a significant cost consequence to the vessel operator using the port. For container terminal operations, performance indicators are generally universal. This means that the indicators that are germane to the CHP terminal will also be relevant to terminals that APN may concession in future years. So while the focus of this report is on Cap Haitien, the operational indicators we recommend here can be readily applied to other terminals; in some cases, the indicators may also be relevant to non-containerized operations. Chapter 2 of this report presents a general description of performance indicators and some of the methods used to calculate them. There is keen interest, particularly in the post-port reform era, in an ability to compare competitiveness of one port relative to others as well as to gauge the effectiveness of port administration. But as we explain, the popularity of some methods used, especially in the academic arenas, do not provide much guidance to port authorities and terminal operators as to the areas that they need to improve. We assert that the best method used is one that generates indicators that operators, port authorities, and regulators understand. This understanding cannot come from “scores” that are not tied to carrier or terminal activity. Chapter 3 presents a foundation from which port indicators are generated. We briefly describe how a stock and flow model, derived from operations research concepts, illustrates the marine terminal’s main components and the flows between these components, representing both a set of productivity and utilization indicators relative to components and flows. We then describe how performance is assessed and how indicators are presented. Using a port time accounting system borrowed from the authors’ contribution to the World Bank’s Port Reform Toolkit , we describe the primary indicators derived from the port time accounting system relative to the terminal’s berth, yard, and gate areas, with particular attention to the indicators relevant to the performance standards in concession contracts. Chapter 4 begins with a review of Haiti’s port operational environments. This includes the facilities in CHP, Port au Prince (PAP), Port Lafito, and Terminal Varruex and the operations reporting systems employed. We then review the indicators that are likely to be included in the CHP concession contract and the ability of these indicators to address the regulatory, administrative, and planning roles of the port authority. Chapter 5 concludes the report with a set of recommendations for monitoring CHP performance and ensuring access to needed data. We then recommend a set of indicators related to berth, yard, and gate performance and the basis for their calculations. We also provide recommendations for indicators for APN to use given its landlord administration role. Chapter 6 provides instructions for data collection and indicator calculation related to the port performance indicator tool that accompanies this report. P R I N C I P L E S O F P O R T P E R F O R M A N C E M O N I T O R I N G | 5 EVOLUTION OF PORT INDICATORS There is an old management adage that declares “ You cannot manage what you do not measure”. This suggests neither the port authority nor the terminal operator can expect to perform at a certain level unless it knows how it is doing and how its decisions may affect performance. From a landlord port authority perspective, this means how effective resources are being utilized to facilitate safe, secure, and efficient environments for port users while maintaining a competitive port system. From the terminal operator’s point of view, the chief concern is to provide efficient vessel and terminal handling services to customers at the lowest possible cost. The need to “measure” for the sake of good management was popularized by the Deming Plan- Do-Check-Act wheel of continuous improvement in an organization. 2 Deming, drawing on the 1920s work of statistician Walter Shewart, advocated for a continuous quality improvement process consisting of a sequence of four repetitive steps to improve organizational performance (Figure 1); as improvements are made, new standards for performance are set. This gave rise to Deming’s total quality management (TQM) concept in the 1980s, engendering keen interest among management executives to ensure their organizations perform well and compete better. This would eventually find its way to the public sector as well. Reinventing government, and the management concepts espoused by Deming and Osborne and Gaebler 3 , among others, would eventually find their way into the port privatization waves of the 1980s and 1990s. 2 Deming’s sequence of steps were defined as follows: Plan -- plan ahead for change and analyze and predict the results; Do -- execute the plan, taking small steps in controlled circumstances; Check: study the results; and Act: take action to standardize or improve the process. See Deming, W. Edwards, Out of the Crisis, Cambridge, Mass. : MIT Press, 2002. 3 See Osborne, David and Ted Gaebler, Reinventing Government – How the Entrepreneurial Spirit is Transforming the Public Sector , Addison-Wesley Publ. Co., 1992. Osborne and Gaebler believed at the time that the United States was undergoing an “American Perestroika”. Like the Soviet version, they believed governments wanted to introduce market forces into monopolistic government enterprises in an effort to reduce spending. The bureaucratic rigidities observed by management gurus Thomas Peters, W. Edwards Deming, and Peter Drucker in private companies were similar to the bureaucratic maladies of government agencies, while both private and public organization structures were bound by strict work rules and centralized command that tied to the Industrial Age. "Hierarchical, centralized bureaucracies designed in the 1930s or 1940s simply do not function well in the rapidly changing, information-rich, P R I N C I P L E S O F P O R T P E R F O R M A N C E M O N I T O R I N G | 6 GROWING INTEREST IN MONITORING PORT PERFORMANCE As public enterprises, port authorities would appear on the radar screen of government reform efforts. These began in the United Kingdom in the 1980s and would spread to Asia and Latin America later in the 1980s and in the following decade. Port reformers justified privatization in part by the promise of improvements in port services as responsibility for providing them was transferred to the private sector. The performance lexicon of the private sector would find their way to government as reformers desired to show lower costs and improved services in the terminals that were now under private sector control. Additionally, though the early public- private partnership transaction experience showed reformers were generally able to avoid transfer of service responsibility to monopoly terminal operators, the scale of investment needed relative to cargo volumes was large; hence, most privatization transactions, while avoiding monopolies, still resulted in oligopolistic environments; the risk in anticompetitive behavior would emerge as a regulatory concern given the very limited number of operators that would vie for the market. So concerns about the competitive playing field while ensuring competitiveness would engender an interest in controlling tariffs and imposing minimum knowledge-intensive society and economy of the 1990s," they wrote, thereby calling for a transformation consisting of flattening hierarchies, decentralizing decision-making, pursuing productivity-enhancing technologies, and stressing quality and customer satisfaction. Source: Vietze, Johannes , Wikimedia Commons, June 17, 2013, available at: https://commons.wikimedia.org/wiki/File:PDCA_Process.png . Figure 2. Deming’s (Plan, Do, Check, Act) Wheel of Continuous Quality Management P R I N C I P L E S O F P O R T P E R F O R M A N C E M O N I T O R I N G | 7 operational standards, usually in the form of tariff and service level provisions incorporated in concession bid terms and concession contracts or imposed by regulators. Regulatory bodies such as, for example, Peru’s OSITRAN (Supervisory Organization for Public Transport Infrastructure -- Organismo Supervisor de la Inversión en Infraestructura de Transporte Público ), which has the authority to set tariffs and impose service performance standards on terminal operators, would be organized to monitor performance and set tariffs, while many countries (including Peru) impose maximum tariff limits and minimal operational performance standards on concessionaires as part of their contracts. While aimed at ensuring minimal standards are met, port authorities may be assigned performance monitoring responsibility to ensure port users receive services that meet acceptable standards in countries where there are no port sector regulatory agencies. Additionally, performance monitoring allows port authorities to diagnose operational problems, define the risk of congestion, and to formulate strategies for mitigating them. For example, a container terminal operator’s berth occupancy approaching 75% may signal the need to expand the port’s physical capacity assuming the berth operation is efficient; the port authority can also collect and assess certain indicators to determine if the berth operation is not efficient. Finally, port authorities receiving complaints about service quality can further investigate if the complaint has merit given the monitored operational performance level. Port authorities are also interested in determining the impact of policy changes and strategies on the competitiveness of their port system. When governments and port authorities formulate policies and strategies, they need to know whether they had their intended impact or unintended (deleterious) consequences. Performance indicators can be devised to reflect the impact of policy and other government interventions. For example, a change in customs procedures that allows expedient evacuation of an import container to an off-dock terminal can alleviate yard congestion, which if not alleviated hinders berth productivity, which would be reflected in a berth utilization indicator. In a congested environment, a change in government imposed policy for setting container yard storage fees or establishing free storage time can also encourage shippers to evacuate quickly their containers after the free storage period; the shippers’ reaction to the policy changes may be reflected in the average container dwell time indicator. PORT PERFORMANCE MEASUREMENT FOR BENCHMARKING AGAINST OTHER PORTS Another dimension of port performance measurement is in relation to benchmarking against other rival or global standard ports. Port authorities are challenged to get data relevant to operational performance in other port jurisdictions as many data are held confidential by terminal operators. In some cases, port regulators will report certain indicators; the noted OSITRAN regulatory agency is particularly transparent about indicators it collects in relation to performance standards imposed on the two terminal operators in the Port of Callao. Recognizing the difficult data challenge, many researchers and scholars have sought to devise other methods for developing scores that serve as proxies for overall port performance. One popular method is data envelopment analysis (DEA), a technique that generates scores and P R I N C I P L E S O F P O R T P E R F O R M A N C E M O N I T O R I N G | 8 rankings (see Text Box 1). 4 Roll and Hayuth 5 were the first to apply DEA to the port sector. They designed an environment of 20 hypothetical ports using capital and labour as inputs (resources needed for producing outputs) and cargo throughput, service levels, number of ship calls, and customer satisfaction as outputs. They concluded that ports that can accommodate large vessels are more efficient than those that cannot. This conclusion is expected given more cranes can be deployed for larger vessels and the length of lateral crane movements is shorter, hence reducing the average crane cycle time per crane. In evaluating the performance of Brazilian, Argentinean, and Uruguayan ports, Rios and Macada 6 define number of cranes, number of berths, size of terminal area, number of employees, and number of equipment as inputs for measuring performance as represented by outputs, in this case, the number of TEUs handled and the average number of containers moved per ship-hour. They do not consider factors that may distinguish the performance of one port from another, such as domestic versus transhipment container handling, ship size, and the crane technology used, where all these factors can affect productivity. For example, transhipment handling normally has shorter distances between the berth and the container’s first point of rest or between the position of the container in the yard to the berth, as transhipment storage tends to be closer to the berth area than domestic container storage. Additionally, the larger the ship, the more cranes that can be deployed to the vessel. A higher number of gantry cranes reduces the length of lateral crane movements needed to serve the vessels, hence reducing the average crane cycle time per container. Crane technology itself will also have an effect on crane productivity. We can expect a gantry crane to have a productivity level of 25-30 moves per crane-hour, with some terminal operators reporting 40 or moves per crane hour. This compares favorably with mobile cranes and ship’s cranes, which have productivity levels of about 18 and 4-6 moves per crane hour, respectively. In short, berth productivity, which is the summation of crane productivity per vessel hour, will increase as vessel size increases (Figure 2). Hence, any study of port performance should not ignore differing factors that influence indicators of one port over another, particularly crane technology and vessel size. 4 The technique was developed by Charnes et al, who believed that the efficiency of a number of “producers” can be evaluated to derive comparative efficiency scores. The usual statistical approach to measuring efficiency is to evaluate producers relative to the average producer. But the DEA approach evaluates producers relative to the best producer. Thus, if Producer I generates A(I) units of production with B(I) inputs, then other producers should also be able to do the same if they work efficiently. If Producer II generates A(II) units of production with B(II) inputs, then other producers should be capable of producing at the same efficiency. Producers I, II, and others are then combined to form a composite producer with composite inputs and composite outputs. See Charnes, A., W. W. Cooper, and E. Rhodes, “Measuring the efficiency of decision making units”, European Journal of Operational Research , 1978, Volume 2, pp. 429-444. 5 Roll, Y. and Y. Hayuth, “Port performance comparison applying data envelopment analysis (DEA)”, Maritime Policy and Management , Volume 20, Number 2, 1993, pp. 153-161. 6 Rios, L. R., and A. C. Gastaud Macada, “Analyzing the relative efficiency of container terminals of Mercosur using DEA”, Maritime Economics & Logistics , Volume 8, Number 4, 2006, pp. 331-346, also available at: https://www.researchgate.net/publication/5223629_Analysing_the_Relative_Efficiency_of_Container_Terminals_of_M ercosur_using_DEA . P R I N C I P L E S O F P O R T P E R F O R M A N C E M O N I T O R I N G | 9 We can demonstrate the use of DEA through the following example. Let’s assume operators A, B, and C each have 100 container ships of the same size calling their terminals. Operator A is good at loading/discharge productivity (moves/ship-hour), Operator C has the lowest berth waiting time, and B falls somewhere in between for berth waiting time and loading/discharge productivity. • Operator A: 100 ships, 40 moves/ship-hour, 20 hours berth waiting time • Operator B: 100 ships, 20 moves/ship-hour, 5 hours berth waiting time • Operator C: 100 ships, 10 moves/ship-hour, 0 hours berth waiting time Now, as a DEA analyst, we are like home car builder mechanics who combine parts from different units. Consider Operator A. Clearly no combination of Operators B and C can produce 40 moves per ship/hour with the constraint of 100 ships. Therefore, A is efficient at crane handling and receives an efficiency score of 1.0. On the basis of the above data, we can also see that Operator C is efficient because no combination of Operators A and B can produce 0 berth waiting time in 100 ship calls. Operator C is minimizing berth waiting time more efficiently than any other operator just as Operator A is handling containers more efficiently than the other operators. Operator C focuses on minimizing berth waiting time while Operator A focuses on berth productivity. Operator B would have been more productive if it had reduced its berth waiting time like Operator C and paid more attention to berth productivity like Operator A. Since B was not that productive, it must not be as capable as either A or C and its efficiency score would be below 1.0 to reflect this. The example above can be made more confusing if the DEA model considers unequal values of inputs instead of the constant 100 ship calls—creating a multiple input problem or adding more data points—but the principles hold. Still, the general approach is to consider a range of “production” inputs that produce specific outputs. Text Box 1. Illustrative Application of DEA P R I N C I P L E S O F P O R T P E R F O R M A N C E M O N I T O R I N G | 10 Table 1 provides an example of a set of DEA-generated scores. These are drawn from a DEA application to assess the relative efficiency of 24 Italian ports. 7 The author used liquid and bulk tonnages, number of ships, number of passengers, container moves and TEUs, and sales revenues as input variables for the model and number of personnel, asset value, and operational costs as output variables. Among the author’s conclusions are 1) Italian ports should expand their assets through investment or through mergers and acquisitions, 2) they should focus on the container trades because the container business generates more revenues, and 3) they should constrain the size of their labor force. Though these conclusions are correlated with the scores, they do not provide much practical guidance on the “how” to respond to them, so have no obvious meaning to a terminal operator or a port authority. The scores don’t provide much guidance as to why they are high, or why they are low, or why one port or terminal is ranked higher or lower than other ones. Hence, the score does not provide a direction towards a solution that a terminal operator or port authority needs to tackle a performance constraint. Despite its shortcomings, DEA lends itself to assessing performance on a wide scale and should not be dismissed as a useful technique for comparative port performance. Additionally, data for inputs, regardless of specifics, can be readily available as are data for outputs. Still, the DEA 7 Barros, Carlos Pastana, “A Benchmark Analysis of Italian Seaports Using Data Envelopment Analysis”, Maritime Economics & Logistics , December 2006, Volume 8, Number 4, pp. 347-365. Source: JOC Group Inc., Berth Productivity – The Trends, Outlook and Market Forces Impacting Ship Turnaround Times, July 2014, available at: https://www.joc.com/whitepaper/berth-productivity-trends-outlook-and-market-forces- impacting-ship-turnaround-times-0 . Figure 3. Comparative Regional Berth Productivity Levels by Vessel Size 0 20 40 60 80 100 120 140 10,000+ 7,501-10,000 5,001-7,500 2,501-5,000 2,500 or less Vessel Productivity per Hour (TEUs) Vessel Size Capacity in TEUs Europe, Mid East, Africa Asia/Pacific Americas P R I N C I P L E S O F P O R T P E R F O R M A N C E M O N I T O R I N G | 11 scoring does not report on factors of most concern to the industry: efficiency and productivity related to time. More recent research provides some guidance on the indicators that can be generated for benchmarking and concession contract regulation purposes; Kent and Ashar, drawing on the port time accounting system they developed for the World Bank’s Port Reform Toolkit 8 , suggest 8 World Bank, Port Reform Toolkit , “Module 6, Port Regulation: Overseeing the Economic Public Interest in Ports, World Bank”, Second Edition, 2007. Source: Barros, Carlos Pastana, “A Benchmark Analysis of Italian Seaports Using Data Envelopment Analysis”, Maritime Economics & Logistics , December 2006, Volume 8, Number 4, pp. 347-365. Table 1. DEA Scores of 24 Italian Ports P R I N C I P L E S O F P O R T P E R F O R M A N C E M O N I T O R I N G | 12 two sets of indicators for purposes of performance standards in concession contracts in a case study of Colombia’s port system (Table 2) 9 . The port accounting system, described later, accounts for the elapsed times associated with vessel and cargo movements from a port’s entrance buoy through the berth area. Kent and Ashar devised indicators relative to the two categories of operational efficiency and level of service, the first composed of ship, crane, and berth throughput productivity, and the second ship delay, truck delay, and truck turn time. Their reasoning is that these indicators are most indicative of the risk of terminal congestion. They also consider factors that can influence productivity, such as the load and discharge volume per call, crane technology deployed, and cargo type. Importantly, based on interviews with shippers and operators, they provide standards of performance for each of the indicators, as Table 2 shows. Kent and Ashar later extended their research in a test application of their indicators to a selected sample of six Central American commercial gateway ports and three Latin American benchmark ports known to operate at global standards. 10 While they were able to obtain the majority of needed information to determine a limited number of indicators, the process was still somewhat onerous, and hence they conducted additional analysis in an effort to reduce the number of 9 Kent, Paul E. and Asaf Ashar, “Port Performance Indicators for Regulators and Concession Contracts”, paper presented to the International Association of Maritime Economists, Carcavelos, Portugal, July 2010; available at: http://www.asafashar.com/IAME_2010_Article_Performance_Indicators_for_Regulators_Final_Final.pdf . 10 Kent, Paul E., Asaf Ashar, and Gerardo Ayzanoa, “ How Fit Are Central America’s Ports? An Exercise in Measuring Port Performance”, paper presented to the International Association of Maritime Economists, Norfolk, Virginia, July 2014. Table 2. Regulatory Indicators for Operational Efficiency and Level of Service Source: Kent, Paul E. and Asaf Ashar, “Port Performance Indicators for Regulators and Concession Contracts”, paper presented to the International Association of Maritime Economists, Carcavelos, Portugal, July 2010; available at: http://www.asafashar.com/IAME_2010_Article_Performance_Indicators_for_Regulators_Final_Final.pdf . P R I N C I P L E S O F P O R T P E R F O R M A N C E M O N I T O R I N G | 13 indicators collected that would still be reasonably reflective of comparative performance. Using Principal Components Analysis (PCA), which discerns the indicators most representative of port performance, they were able to reduce the number of indicators to two, including ship productivity and crane productivity, greatly easing the process of data collection (Text Box 2). The above discussion focused on the generation of performance indicators that can serve the port authority’s purpose of monitoring terminal operations performance or for gauging the relative Sources: World Bank, Connecting to Compete – Trade Logistics in the Local Economy, The Logistics Performance Index and Its Indicators, Washington, D.C., 2010, available at http://siteresources.worldbank.org/INTTLF/Resources/LPI2010_for_web.pdf ; Asia-Pacific Economic Cooperation (APEC), The Economic Impact of Enhanced Multimodal Connectivity in the APEC Region , APEC Policy Support Unit, Singapore, 2012, available at: http://publications.apec.org/publication- detail.php?pub_id=1028 ); United Nations Conference on Trade and Development (UNCTAD), Review of Maritime Transport, Annual Liner Shipping Connectivity Index (LSCI) Index, Trade Logistics Branch of the Division on Technology and Logistics, Geneva, Switzerland, available at http://unctad.org/en/pages/PublicationWebflyer.aspx?publicationid=380 ). PCA is a commonly used technique in statistics and econometrics. In our port and logistics world. The World Bank’s Logistics Performance Index uses a PCA-weighted average of six raw data series that capture performance along the most important dimensions of logistics activities. Similarly, UNCTAD uses PCA to produce an index based on a set of raw data that covers the main dimensions of international maritime transport. And APEC (the Asia Economic Cooperation forum) used PCA to produce an index of multimodal transport connectivity, this time based on raw performance data for the main modes of transport as well as the logistics sector. PCA takes a number of raw data series and compresses them into one or more summary indicators—“principal components”—using an optimal weighting scheme. Each index is therefore a weighted average of the raw data used to produce it, with weights chosen so that the index accounts for the maximum possible proportion of the variance in the original data series. A PCA application can eliminate some data because it may show a reduced number of data still adequately represents the phenomenon being measured. When the data set is reduced, further analysis can be made simpler and future data set requirements can be reduced without losing important information. There are several conditions with which the data should conform for a solid PCA analysis (e.g. data should have Gaussian or normal distribution); however, the present data set can still be useful for illustrating a valid approach for reducing data collection efforts and provide the solid basis for ranking terminals relative to operational performance. Text Box 2. The Principal Components Analysis Technique for Reducing Data Requirements P R I N C I P L E S O F P O R T P E R F O R M A N C E M O N I T O R I N G | 14 standing of its terminals among rival and benchmark ports. For comparative performance assessment, research shows that the port authority can focus on two productivity indicators when comparing its performance with other ports, as the PCA analysis demonstrated. However, when benchmarking its own terminals against prior years’ performance, the port authority has the ability to access data for a range of indicators, as long as the data are made available. In the case of private operators, concession and other agreements can compel private operators to provide needed data for indicator calculation; agreements can also subject operators to operational audits to ensure accuracy and reliability of collected data. Port authorities, as administrators, also should monitor their own performance. This suggests an orientation towards resource allocation and management effectiveness. We address the range of indicators that port authorities can use. The ultimate indicators, addressed later in this report, depend on port authority objectives and strategy for performance improvement as well as regulatory requirements that may be imposed by other government agencies. PORT AUTHORITY ADMINISTRATION INDICATORS Businesses tend to measure performance in order to monitor and control their operations, drive improvement, maximize the effectiveness of the improvement effort, achieve alignment with organizational goals and objectives, and to reward and discipline. 11 So they tend to use quantifiable measures indicative of performance over time, which can number in the hundreds. As earlier noted, indicators not only enable organizations to track their performance, but also to compare it against other businesses within its industry. In the context of port reform, port authorities and regulators want to use a set of indicators that allows them to gauge the quality of services transferred to the private sector. But indicators could also be related to any performance standards incorporated into concession contracts, for benchmarking against rival or “best practice” ports for assessing relative competitiveness, or for benchmarking against prior years’ performance to monitor for improvements or declines. While port authorities have an interest in monitoring port operator performance, they also have an interest in following their own performance. A classic source for port authority indicators is the United Nations Conference on Trade and Development (UNCTAD) Port Performance Indicators manual published in 1976 12 , well before the tide of the first privatization wave of the 1980s. As such, port authorities, referred to as operating port authorities, directly managed port operations, with port authority personnel operating equipment and supervising labor gangs. 13 Hence, the UNCTAD manual identifies indicators for port authority 11 Bititci, U. S., Carrie, A., & Turner, T.,“Integrated Performance Measurement Systems: Structure and Dynamics”, Business Performance Measurement: Theory and Practice , A. Neely (Ed.), Cambridge: Cambridge University Press, 2002, pp. 174-197. 12 United Nations Conference on Trade and Development, Port Performance Indicators , May 1976, Geneva. Available at: http://unctad.org/en/PublicationsLibrary/tdbc4d131sup1rev1_en.pdf . 13 Some port authorities (e.g. Karachi Port Trust) in the 1970s also enlisted labor from established labor pools not directly employed by the port authority. Usually, such labor pools included strictly manual labor for lashing containers and carrying sacks and boxes from or to breakbulk vessels. P R I N C I P L E S O F P O R T P E R F O R M A N C E M O N I T O R I N G | 15 administration effectiveness and operations efficiency, focusing on financial performance and resource allocation for the former and operations efficiency for the latter, such as revenue per ton or TEU handled, expenditures per ton or TEU handled, number of employees per unit tons or containers handled, and so on. UNCTAD provided an update on performance indicators in 2016, generally reflecting indicators relevant to the post-port reform era, focusing on four strategic dimensions of port administration, including finance, operations, human resources and market. 14 Interestingly, the market dimension focuses on customer satisfaction, reflecting the renewed interest in keeping customers happy and placating any concerns they may have, though the dimension is also intended to reflect port market share by cargo type, and service effectiveness to customers. 15 UNCTAD proposes developing a scorecard, which is intended to reflect relative standing among several port authorities, the data for which are collected from surveys to port authorities. While comparative data are useful to provide comparisons among a group of port authorities, comparisons can be distorted by the nature of port service offerings. For example, labor costs as a percent of total revenues, cargo dues as a percent of revenues, and so on, can be quite different between ports predominated by container activities and those predominated by bulk or breakbulk cargoes. Even among a population of primarily container ports, indicators can be quite divergent between container ports predominated by transshipment containers and those predominated by domestic or even reefer trades. Still, the indicators can be useful for a port authority to benchmark against prior year results. Table 3 presents UNCTAD’s range of indicators associated with its port (authority) scorecard. UNCTAD also uses indicators from global sources, which can be very useful for monitoring and benchmarking performance of certain activities within the port area but not necessarily under the port authority’s jurisdiction or control, such as customs and liner connectivity offered to port customers. Customs indicators can alert the port authority to constraints influencing operational performance 16 , while liner shipping connectivity can inform the port authority of the extent of options its customers have for importing and exporting their freight. 17 One advantage to such sources is that they are updated annually or biennially and readily available. 14 United Nations Conference on Trade and Development, Port Management Series , “Volume 4 Port Performance – Linking Performance Indicators to Strategic Objectives”, June 2016, Geneva. Available in English at: http://unctad.org/en/PublicationsLibrary/dtlkdb2016d1_en.pdf ; available in French at: http://unctad.org/fr/PublicationsLibrary/dtlkdb2016d1_fr.pdf . 15 Ibid, p. 12. 16 Good sources indicating customs performance include the World Bank’s Logistics Performance Index where 163 countries are assessed relative to logistics-related criteria, including customs, infrastructure, international shipments, logistics competence, and timeliness. The customs indicator refers to the efficiency of the clearance process. See Arvis, Jean-François, Daniel Saslavsky, Lauri Ojala, Ben Shepherd, Christina Busch, Anasuya Raj, Tapio Naula, Connecting to Compete 2016 Trade Logistics in the Global Economy The Logistics Performance Index and Its Indicators , The World Bank, 2016; available at: https://wb-lpi-media.s3.amazonaws.com/LPI_Report_2016.pdf . 17 The Liner Shipping Connectivity Index shows the extent to which countries are connected to global shipping networks. The United Nations Conference on Trade and Development (UNCTAD) calculates the index based on five maritime components in each country, including the number of ships, their container carrying capacity, maximum vessel size, number of services, and number of companies that deploy container ships in a country's ports. The index P R I N C I P L E S O F P O R T P E R F O R M A N C E M O N I T O R I N G | 16 The European Sea Ports Organization (ESPO) and the European Commission jointly created a port performance indicator reporting system to keep stakeholders informed of the port system’s ability to meet expectations. Over the years, individual countries have adopted different port governance systems, so the indicator project, referred to as PPRISM (Port Performance Indicators: Selection and Measurement), is aimed in part at assessing the effectiveness of the various governance models (see Table 4) while communicating with stakeholders the port system’s economic, social, and environmental impacts. ESPO identifies five categories of indicators, generally intended to illustrate advancements individual port authorities have made to serve customers and support economic growth and European Union policy. For example, the logistics chain and operational category reports on- time performance for both vessels and inland transport modes, vessel turn-around time, and the availability of port community systems to support information sharing among the various port users. The socio- economic category includes number of employees, amount of investment made, and the extent of training provided to port authority employees, all of interest to the port is intended to reflect each country’s position within global container shipping networks. The higher the score, the easier it is to access the global maritime freight transport system. For a description of the Liner Shipping Connectivity Index, see United Nations Conference on Trade and Development, Review of Maritime Transport , UNCTAD, 2016, pp. 41-42, available at: http://unctad.org/en/PublicationsLibrary/rmt2016_en.pdf ; for time series index data, updated annually, see http://unctadstat.unctad.org/wds/TableViewer/tableView.aspx?ReportId=92 . Source: United Nations Conference on Trade and Development, Port Management Series , “Volume 4 Port Performance – Linking Performance Indicators to Strategic Objectives”, June 2016, Geneva, Table 2, p. 22. Table 3. UNCTAD Port Authority Scorecard Indicators Category Indicator EBITDA/revenue (operating margin) Vessel dues/revenue Cargo dues/revenue Rents/revenue Labor cost/revenue Other fees/revenue Tons/employee Revenue/employee EBITDA/employee Labor cost/employee Training costs/wages Average waiting time (hours) Average overall vessel length per vessel (m) Average draft per vessel (m) Average gross tonnage per vessel Average tonnage per arrival Tons per working hour, dry bulk Box per hour, containers Twenty-foot equivalent unit dwell time (days) Tons per hour, liquid bulk Tons per hectare - all Tons per berth meter -- all Quality of port infrastructure Liner shipping connectivity index Burden of customs procedures Female participation rate Gender Equity Finance Human Resources Vessel Operations Cargo Operations World Development Indicator Data P R I N C I P L E S O F P O R T P E R F O R M A N C E M O N I T O R I N G | 17 community. The environmental category provides indicators related to meeting the European Union’s environmental management systems standard, the existence of an environmental monitoring program, and utility resource usage. And, in an effort to monitor market concentration and perhaps market behavior, the market trends and structure category includes market share as well as the Herfindahl-Hirschman Index (HHI) 18 . WHERE WE GO FROM HERE The above discussion illustrates the myriad indicators that can be used to assess port performance. From an operational performance dimension, indicators depend on the intended 18 HHI is a measure of the dominance of firms competing in the same market. HHI is calculated by squaring the market share of each firm competing in the market and then summing the resulting numbers. In the United States, an HHI score of 1,500 and 2,500 indicates the market is moderately concentrated while a score above 2,500 indicates the market to be highly concentrated. Mergers in highly concentrated markets that increase the HHI score by more than 200 points are a