Friday, January 22, 2016
30 years ago, forecasting port capacity was a fairly straightforward exercise, and a popular one at that, among young students: population, GDP and trade data, together with a simple regression model, would easily do the trick with some excellent results. It should be remembered that, in those days, demand for port services was more or less captive and a cargo destined for, say, Italy would in all likelihood land at an Italian port, as close as possible to the final consignee.
Two things changed this picture since: containerization and the ‘through transport’ concept on the one hand, and the development of extensive land transport infrastructure on the other. Europe, dilapidated by a ruinous war, was being rebuilt, and what better way of doing this than developing your roads and railroads before anything else. If one were to take a cursory look at Europe’s map today, what one would see wouldn’t differ much from a nice dish of Italian spaghetti. Roads and railroads, together with European economic integration, free trade, and the abolition of national borders, opened up the market for port services. Before we knew it, any Asian cargo could in principle reach any European destination, passing through any gateway port around Europe. Port demand was no longer captive and ports started to compete for survival.
To do so, ports had to modernize and develop new capacity; often much more than what was needed. In the 1990s, I remember well, the northern range European ports (from Le Havre to Hamburg) had a collective excess capacity of about 40%. Usually, port capacity is developed in big chunks, each time far ahead of existing demand. A combination of factors can explain this, ranging from the availability of easy and cheap public finance, to managerial egos and port managers’ heartburning desire to leave their footprint in history... One more reason, of course, is economies of scale in port construction: apparently, it’s much cheaper to build two kilometers of quay rather than one.
Excess port capacity, over and above national demand requirements, is created for three more reasons. First is the economies of scale mentioned above, as well as the infamous economic law of Jean-Baptiste Say according to which supply creates its own demand. In other words, once the port is there, the customer is bound to arrive. Unfortunately, this is not always so. The second reason is the footloose nature of the container and its carrier, which may switch ports at the whim of a moment, whenever capacity is scarce and, as a result, the ship may have to wait. Finally, excess capacity is developed in order to capture transshipment traffic; i.e. somebody else’s cargo! As a matter of fact, most of the competition taking place among ports today is for this type of cargo; to capture it, ports often underprice concession fees agreed with terminal operators, and terminal operators, in their turn, underprice terminal handling charges (THC) they charge to carriers (often through hidden or opaque discounts). [by underpricing I refer to port dues and concession fees below the opportunity cost of port land].
Moreover, especially in ports where the management is responsible for its bottom line, more often than not the management tends to cross-subsidize footloose transshipment traffic with captive domestic traffic. In other words, domestic (national) cargoes are penalized through relatively higher prices, compared to transshipment cargo, in the management’s anxiety to capture more of the latter. In a number of cases, some of which already heard in front of a judge (cf. Sarlis vs. MSC at the port of Piraeus) , the preferential treatment of transshipment cargo goes beyond price and it involves other terms of service, such as berth allocations and time-windows. For a number of reasons this is not right and I have always advised affected local shippers to present their case at the national competition authorities or, in case of complacency, directly to the Competition Directorate General of the European Commission.
And here is why this type of cross-subsidization, or price discrimination, between domestic and transshipment cargo is not right:
If sufficient domestic demand for port services does not exist, developing port capacity for transshipment purposes is risky business just because of the footloose nature of the container. Before you know it, you could find yourself with a ghost port in your hands, as it happened recently, for instance, at the Italian port of Taranto, when Evergreen decided to move to Piraeus, in spite of a 60 year-long concession at the Italian port. Years back, I remember quite vividly the crisis that developed at the great port of Singapore, when Maersk decided to move just around the corner, to the Malaysian port of Tanjung Pelepas. The graph above shows that, in the northern range of European ports, transshipment represents, wisely, no more that 40% of their throughput, vis à vis transshipment ports like Malta, Singapore, Damietta or Kingston.
In comparison to the economic impacts of domestic traffic, transshipment creates relatively less (local) value added. To draw a parallel, a port handling domestic cargo resembles a city-center hotel, where the visitor will most likely spend money on a number of activities (museums, restaurants, etc.), thus creating considerable value added for the city, vis à vis a highway motel (transshipment), where the traveler would stop there just for a sleep. I often hear my friends in Antwerp telling me that the port of Antwerp, as a result of its multipurpose/labor-intensive character, creates four times more value added than Rotterdam, the latter port being a highly automated/labor-saving one. Thus, I am told, the Belgian taxpayer is much happier to pay taxes for port development than his Dutch counterpart, who has often questioned the social utility of developing more port capacity at Rotterdam. Exceptions to the above do of course exist, and Rotterdam is a good one: the port’s value added is not created simply by the port itself, but by its port cluster, encompassing 50% of Europe’s Asian and North American European Distribution Centers (EDC); a city 50% of whose inhabitants are holders of a foreign passport, just because of the port. But not all ports can realistically aspire to such an enviable situation, developed not today but over a period of 70 years of hard work.
Finally, publicly funded ports are developed exactly in order to have these domestic impacts on business and employment, rather than to steal (transshipment) traffic from their neighbor, particularly in economically interdependent geographical regions such as the European Union. The problem is aggravated when some countries spend public money on port development, while others finance ports privately (UK) and both compete in the same market.
Development of container terminal capacity, including its transshipment potential, will continue unabated; this is normal and, in the long-run, port capacity follows international trade growth. But with one caveat: this infrastructure should be priced (through the appropriate concession fees) in such a way so that investment costs are eventually recovered, irrespective of whether the proceeds from the concession remain with the port, or are returned to its financiers, the latter including also the government. In this way, it does not really matter who finances the investment, i.e. the public or the private sector, as long as the private investor principle applies; i.e. the terms of the financing are not very different from similar private arrangements and, as said, this means that pricing should aim at cost recovery.
Competition on infrastructure is indeed wasteful and governments, like recently that of Italy, have often argued in favor of centralized port infrastructure planning: something we used to do half a century ago, through the various National Port Councils [I remember an advocate of this policy, Francesco Mariani, the President of the Port of Bari, telling me recently that if a global carrier would like to come to Italy, he should only talk to the Minister and it should be he to tell him at which port he should call !]. In today’s Europe, however, something like this wouldn’t only be wrong but, euphemistically, it would be unthinkable: To my mind, the best planners of all are the (well-regulated) markets themselves, together with transparency in the financial flows between port and government.
The role of the public sector in financing container terminals should therefore be limited, and where it exists, or is necessary, it should take place on more or less commercial terms. In this way, limited would also be the risk assumed by the public sector. Competition by neighboring ports, excess capacity and similar concerns should lie only on the shoulders of the concessioners (terminal operators) who should themselves assume such market risks. It doesn’t in this way matter if excess capacity is created. To put it bluntly, if things go well, we will all be raising a glass; if not, well, bad luck. But the national taxpayer should not be bearing the costs (and risk) of private investments, benefiting private users.
The above is easier said than done. If the terms imposed on a private terminal operator, carrier or other, are too onerous, in all likelihood he would be knocking on your competitor’s door. And this is where public policy intervention is necessary. This can take only one form: an understanding that, no matter how terminal investments are financed, container terminals are private goods and their costs should be recovered through user charges; otherwise underpricing is not much different than dumping, sanctioned in many other sectors, including shipbuilding.
It seems the European Commission, in its forthcoming State Aid Guidelines, may be thinking differently, under pressure from powerful EU member states and ports requiring considerable dredging (e.g. river ports). We will soon know the outcome. HH
Tuesday, January 12, 2016
In my post of January 5, 2016, I was writing that big ships impose substantial demands on port capacity, without however paying commensurately for this demand. For instance, I was writing, where before we could accommodate 3 Panamax vessels in a one kilometer berth, today we can host there only two mega vessels of the latest generation (about 400 meters long). Berth utilization obviously goes down. And all this would be fine, as long as carriers were bringing more traffic to the port with their larger vessels. But this doesn’t happen either. As the graph shows, call size is only moderately correlated with vessel size (the red regression line is quite flat). In such a situation, i.e. as carriers demand more berth space without bringing, proportionately, more traffic to the port, it would make a lot of sense to think of a different, revenue-neutral, terminal charging system, based on ship-dimensions rather than $/box.
[I have chosen above the word ‘dimensions’ rather than ‘length’, for it is really bay size the critical factor for both crane- and berth productivity. Obviously, it takes twice as much time to unload a container stowed below-deck in row 23, than one loaded on-deck in row 1; and this without considering, of course, the substantial costs of investing in quay cranes with a 70-meter outreach].
Could differential- or variable pricing work, as a new system of Terminal Handling Charges (THC) aiming to offer carriers incentives to improve stowage planning? If the answer is ‘yes’, this could offer carriers faster turnaround times (and time is money), and ports better berth utilization; a truly win-win situation. The issue of stowage planning, these days, is not a theoretical but a real and often troublesome one: Intra-alliance cooperation notwithstanding, in order to fill a larger ship today, a carrier calls at more loading ports than would be warranted by the economics of a hub-and-spoke system, picking up containers even at the last minute before departure. Stowage planning suffers as a result. As one terminal operator commented, “we can do 150 moves per hour for a major independent carrier but only 100 for an alliance ship”. Slot swaps among alliance members do not make things easier either, with shippers booking containers with one company, only to find out that they landed at the other end on the ships of another carrier.
Variable terminal pricing, if it will ever apply in practice, should not become a form of yield management, so successfully, albeit also so frustratingly, applied in aviation (how many times haven’t you tried to book a flight, only to see the price going up while you were trying to do so), but also by the ocean carriers themselves through their market segmentation strategies: pricing strategies directly targeting the individual shipper and his willingness to pay for various logistical add-ons (a more technical post on this will follow soon). As already said, variable pricing should be a ‘revenue-neutral’ charging system, aimed at promoting higher ship and terminal efficiency. After all, yield management could never succeed if applied by an individual port, for, solely by itself, a port has very little pricing leverage vis à vis the footloose container and its carrier. Moreover, any attempt to extract more revenue through yield management would be music to the ears of that port’s competitors, who in all likelihood will leave their own prices unchanged in order to capture the carriers' discontent (competing ports are facing a kinked demand for their services).
The above bring me to another issue, far more important, that of variable pricing, or price discrimination, between local and transshipment cargo: i.e. a cross-subsidization practice that definitely ought to be avoided, if not to be forbidden by public policy intervention. (to be continued…). HH
Friday, January 8, 2016
Due to the regularity, frequency and predictability of port business that containerization brought about in less than 50 years, casual labor has been regularized and port work has become a respectable and well-paid profession. Many people might still remember the old days when, each morning, the dockers would show up at the gate where a tallyman would thumb in those needed for the day, usually with criteria such as age and physical strength (if not with some other criteria of a more ‘pecuniary’ nature…). The rest would go back to the tavern, or to whatever else it was they were doing, waiting for the next day and hoping for the best. During those days, shipping was a very risky and unpredictable business. Once a ship set sail, God only knew if, when and where she would end up to. Often, in banking, risky loans are still called ‘sea loans’.
The new container-handling technology, on the other hand, substituted capital for labor and it was thus strongly resisted, if not sabotaged, by trade unions. Incidentally, the word sabotage derives from the wooden shoes (sabot, or the famous Dutch klomp) workers wore on the sweat-floor, which they were throwing onto the new machines that were stealing their jobs, in the early years of the industrial revolution.
But as port labor was being made redundant, port productivity was increasing by leaps and bounds. The graph I attach shows increases in labor productivity in the 30 years since the introduction of containerization in the 1960s. Labor was cut in half, while tons handled went up 6 times, from the time when a strong young man would swing a sack of flour over his left shoulder and walk up the gangway, to the day when a quay crane driver would do 12 times that much (and of course much more today), with a joystick in hand, sitting somewhere comfortably, even remotely from home through an internet connection. And it was this increase in labor productivity that allowed the fortunate dockers who stayed with the job to enjoy the high salaries -the exceptionally high salaries, some would argue- they do today (if one drives around a modern port today he will see more Mercedes and BMW parked there than in any other part of the city center).
Higher taxes were consequently paid to the government, whose role is to redistribute income and wealth to those who, as a result of technology, have lost their job. There is little doubt, in a civilized society, that those who lose their job must be respectably compensated. And there are laws and strict restructuring procedures for this purpose, which didn’t evolve overnight. In the Netherlands, labor restructuring procedures are amongst the most advanced (and generous) in the world. Sometimes, the restructuring process has been painful; elsewhere, like for instance in Malaysia, it was plain sailing, with more than generous severance packages and golden handshakes (for interested readers, in the mid 90s, and after having looked at more than 100 ports around the world, I wrote 3 reports for the International Labor Organization (ILO) which culminated in my guidelines for successful port privatization -available on academia.edu). Having said this, though, if in today’s world one believes that jobs can be guaranteed at the cost of progress and competitiveness, he is under a big fallacy. Employment today is as much a right as it is an obligation. By this I mean that the worker can no longer sit back and enjoy life as before, but often he has to take life and his future in his own hands, improve, better himself, invest in new competences and skills, or look around for new opportunities even outside his sector; and he should do this sooner rather than later.
The port workers in Rotterdam claim that the port is creating too much new capacity and this will cost jobs. This is not true. Excess capacity does not cost jobs and in Rotterdam, before Maasvlakte II was developed -a project that had been discussed for more than 20 years- the port had lost, in 10 years, 10% of market share to Antwerp and Hamburg, just because it didn’t have enough capacity. The real problem is that the new capacity is highly automated and labor-saving. But this is not bad, and is a fact of life. Since the days of the industrial revolution, capital has substituted for labor and it was only through innovation, implementation of advanced technologies and the consequent increases in labor productivity that we have the advanced economies, societies and welfare systems we are enjoying today (so far…).
The Netherlands is well known to all for its famous polder model: A model where the social partners - employees, employers and Administration- sit around a table and eventually come up with the white smoke of a reasonable win-win solution. Industrial action is not common in this country, to say the least. I am quite confident this will again be the case, and the unfortunate incident of today’s strike, causing millions of euros in damages to the country’s economy, will soon be a regrettable history, allowing this great port, the port of Rotterdam, to continue unhindered being the indisputable leader of the European port industry. HH
Tuesday, January 5, 2016
If you ask an ocean carrier how big a port should be, he would immediately tell you “as big as possible”. Apparently, he wouldn’t want to wait even for a minute if he could help it, and we have well established that, at a port capacity utilization of around 75%, congestion starts to set in. If instead you ask the same question to a port manager, particularly one responsible for the returns on his money, the answer would be “as small as possible”. Obviously, he would love having ships queuing up outside his port, like the good old times, if he could help it. As usual, both need to put some water in their wine: the ship cannot wait, nor however can the port continue spending taxpayer money so that the ship does not do so. This is particularly so in increasingly commercial activities, such as those of transshipment container terminals, whose impacts are not localized in the economies which have borne the brunt of the investment, but are rather dissipated throughout the supply chain, from the country of origin to the country of final destination of the transported goods.
Big ships impose substantial demands on port capacity: A 1 kilometer berth could host 3 panamax ships simultaneously before, but it can accommodate only 2 of the larger containerships of today. As call size is only moderately correlated with vessel size, i.e. as carriers demand more berth space without bringing, proportionately, more traffic to the port, it would make a lot of sense to think of a different, revenue-neutral, terminal charging system based on ship-dimensions rather than $/box. Moreover, OECD and MEL have calculated that the average container, arriving on a larger ship, takes more time to handle and store. In other words, port time per TEU is an increasing function of ship size. The “supply chain” diseconomies of scale of larger ships, including their impact on shippers’ inventory costs, start from the port (graph).
The situation is not helped much by the increasing competition between neighboring ports aimed at stealing transshipment traffic from each other. This is particularly worrying with publicly funded port investments. Clearly there is a need for public policy intervention, at least in Europe which needs to harmonize public spending among its member states, as well as ensure free and fair competition among its ports. I hope to see this taking effect in the forthcoming EU state aid guidelines which should include, finally, indications on what is public port investment and what is investment benefiting identifiable users who should in all honesty bear the cost of its development and maintenance (user pays). HH