Business Rationalisation v/s Virtual Monopolies
Chief Architect's ColumnPosted by Amit Bhagwat Wed, August 12, 2009 23:39:34In my last article, I wrote about the many things that a load-balancing approach to engineering capabilities of the country can do to improve its infrastructure, including railways, in years of slowdown. Within a week of its publication, announcements were made by the British government that the East Coast Main Line - one of the two most significant intercity rail arteries in the UK - would become a state-run enterprise.
The East Coast Main Line (ECML) is important in that apart from connecting the two capitals - London and Edinburgh - its main services connect major population centres, including parts of metropolitan regions of West Yorkshire, South Yorkshire, Greater Glasgow, Tyne & Wear, Teesside, Humberside and Aberdeen, to London, and to one another. Except in the case of Glasgow, which has faster and more frequent services on the West Coast Main Line (WCML), ECML is also the principal (usually also meaning the only) service connecting these regions to London. (refer to various maps such as on http://www.nationalrail.co.uk/tocs_maps/maps/ to better visualise the routes)
Historically, the line has been noted for many high-profile services. It has pioneered innovations & holds world-records in steam technology - the likes of the Mallard and the Flying Scotsman. Doncaster in South Yorkshire, a major junction point on the line, had been a key manufacturing and innovation centre for steam technology. The line also serves the National Rail Museum adjacent to York station - again featuring many historical wonders, mainly in steam technology.
This constant reference to the rather unclean and inefficient technology of the past however also highlights how quickly the rest of the world, and even the rest of England, have moved on and, in many cases, overtaken ECML. In fact, in the last decade or so that I have had the dubious fortune of being tied to ECML services, the line has distinguished itself by slower modernisation of its tracks and rolling stock, consequent limitations to speed and punctuality, and a generally backward attitude of its operator, which happened to be the last rail operator in the UK to ban smoking on its services and has been some distance behind the other operators, in particular the WCML operator, in introducing and rationalising its facilities - ergonomic seats, quiet coach, reliable electric power for small electronics, reliable air-conditioning, even simple features to protect from intense sunlight. Not surprisingly, its last franchisee failed to get its franchise renewed. The new and relatively recent franchisee (since 2007) has not been successful either and is handing the line back to state administration.
On the day that the government announced that it would take control of ECML, the media devoted a fair proportion of its news and features coverage to discuss this event and the experiments with privatisation of key public services that began with Mrs Thatcher’s administration and were allowed to continue by the next two prime ministers, with the current prime minister showing no clear preference. One of those interviewed that day was Ken Livingstone, the last (and first) mayor of Greater London. He had faced a similar situation in context of the London Underground provider Metronet that went into administration putting an extra burden of ~£410M on the taxpayer. Prior to that, in 2002, Britain experienced similar situation with Railtrack - the organisation that was meant to maintain the railway lines.
Mr. Livingstone argued (and wrote the next day): “privatisation is justified on the grounds that the private sector is driven, through the rigour of competition, to be more efficient and more responsive to passengers' needs. This is a fiction in the case of a natural monopoly like a railway. Apart from the brief period of competition among bidders for contracts, there is no day-to-day competition at all – no one is going to build a rival railway line and poach passengers from the private franchisee. They are under no pressure from any competition at all. In such circumstances, it is more rational, and makes more sense in terms of sustaining investment, for rail services to be publicly-owned”
Mr. Livingstone is right to the extent that if a public service is essentially a monopoly at any given point in time and up until a reasonable planning horizon thereafter, then it does not make sense in introducing a system where any money it makes will be siphoned off by financiers rather than being reinvested in the public service; particularly, when it being a necessary public service, it is ultimately the government, using public money, that is responsible to see that the service remains running. So, for example, his statement makes sense when it comes to maintaining an essentially single mesh of a compact railway network, such as in the UK, or a close-knit system such as London Underground. In both cases, Britain briefly and unsuccessfully flirted with privatisation, and both came back under public sector management (though technically, one has later become a company limited by guarantee - a not-for-profit monopoly enterprise, while the other, after a 4-year operation as public-private-partnership (PPP), where the infrastructure is maintained by private companies but it is owned and operated by the local government body Transport for London, has had to bail-out the major private sector partner, thus taking back complete control of 9 out of its 12 lines and putting question mark on viability of the PPP concept) with the taxpayer paying for both fiascos. I wonder however, if it is quite impossible to have fair ongoing competition among train operators, even as competition exists among phone service providers, electricity suppliers, intercity coach operators, etc, who all essentially use the same set of lines, maintained separately, for their services.
Let’s consider this as a business problem we are analysing on behalf of the British government and see if a business functional model emerges. Let’s ask ourselves:
1. What do the passengers need, and are there any obvious solutions?
2. If a genuine competition is likely to provide better value, then what is the best long-term approach to effect this?
Passenger Needs and Some Obvious SolutionsIn simple terms, passengers need quick and comfortable travel at fair cost. So, it is these three factors - speed, comfort and cost - among which trade-offs happen.
Through the last several decades, British railways have neither been particularly fast nor particularly punctual, compared to the European best. While passengers appreciate speed, its lack has not been their biggest turn off, provided that the time stated could be reliably adhered to. Many rail operators, including the East Coast operators, have maintained wide time buffers in their timetables, where the average speed is far lower than the speeds the rail system can safely maintain, simply to avoid being penalised by the government for running late, sometimes for reasons beyond control of the operator. Arguably, a publicly run railway doesn’t need to do this, as there is no one to penalise; and privatisation can sometimes simply lead to sandbagging, rather than genuine efficiency.
While such inflated time tables and trains that still manage to run late, often for bizarre reasons like a driver not turning up, do not make the passengers ecstatic, they bear with them so long as safety is not compromised and so long as they are comfortable while travelling. Unfortunately, they are not always comfortable. Despite poor quality of many services, trains remain overcrowded and rolling stock is often not maintained well enough or not upgraded uniformly in terms of ergonomics, climate control, power supply for low-power electronics, etc. Many of these improvements are possible immediately and at a moderate cost, as is possible the manufacture and leasing of more coaches, particularly while engineering firms are crying out for work. Thanks to the many more stations that once existed on the lines, there is sufficient sectoring and signalling potential that, coupled with already available coach door safety features, will support deploying longer trains immediately, providing a comfortable seat for everyone, while platforms can be lengthened and end-of-station points relocated to improve their operational efficiency. Similarly, it is possible to completely relocate some of the small intermediate stations that mainly serve as access points for cities not on the route.
As for time, passengers need the overall time to be predictable, and reduced where practicable. So, for equivalent comfort, a passenger may prefer travelling by a slower train starting sooner, or, if travelling light, change trains, to reach the destination sooner. Sometimes, a passenger may plan a journey in a number of stages and may want short breaks at intermediate stops, perhaps just to stretch, before reaching the end destination. It would be helpful if these choices could be made without ridiculous discrepancies on the other factors - cost and comforts.
As for cost, a passenger would prefer to get the benefit of competitive costing without being irrationally penalised on the other two factors e.g. paying the same price as others but having to wait for too long, for or on a train, or having to travel on a train that is far less comfortable.
Now, lets explore if we can address Mr Livingstone’s concerns about lack of real competition in what he believes is a natural monopoly. To begin with, let’s analyse how we got where we are today.
An Un-architected Business EvolutionBritish railways have gone through four distinct phases in their evolution:
1. Initially railways were built through private enterprise, in what I described in the last article as irrational Wild West behaviour. Individual entrepreneurs and their financiers went their own way, jealously guarding their railway developments, often using different gauge, developing their own infrastructure – lines, stations and rolling stock, often competing for travel between the same two endpoints, albeit between two different sets of stations and following different railway lines. Where this competition was intense, the railway was anything but a monopoly. Often, such stations were constructed within a few minutes walk from one another. For example, The Great Northern Railway served route between London Kings Cross and Leeds from the 1860s along the route that is today served by ECML. Literally next doors to London Kings Cross was (and still is) London St. Pancras Station (so much so that the London Underground station is actually called Kings Cross St. Pancras). London St. Pancras was the London terminus of Midland Railway. Midland railway started running trains into Leeds Wellington station, next door to the other Leeds station, a decade or so before competitor service began from Kings Cross. St. Pancras remains the London terminus of the Midland Main Line (MML) which is able to, and under the franchise terms infrequently continues to, offer service to Leeds and serves as the main London Eurostar station. While, at the Leeds end, stations underwent many stages of unification and rationalisation, sometimes through company deals and takeovers, to lead to one station, at the London end three naturally viable options remain (Kings Cross, St. Pancras, Euston), all within walking distance of one another and all capable of serving Leeds from London, largely on different lines, and without too awkward a detour.
2. While Germany, thanks to its central European location and enthusiasm for taking on the rest of Europe all at the same time, had been among the first European nations to organise a comprehensive rail network, the British, when threatened in the first world war, managed to run their railways far better than ever before, under government control, during the war years. That positive experience in the way some 120 railway companies could work together as one network in times of national need (and, apparently not, if left to themselves, their excessive profiteering and bloody-mindedness) led to the Railways Act of 1921. This consolidated the northern routes into two main groupings. The West Coast and Midland trunk routes came under London, Midland and Scottish Railway (LMS), while the East Coast route came under the London and North Eastern Railway (LNER). These were still able to compete, with bigger muscles (and occasionally cooperate, mainly in using common stations, making greater future integration viable), and continued to do so with vigour. So, for example, travelling from London to Leeds using different routes remained viable, while the ticketing remained exclusive to the chosen route.
3. Another world war, temporary nationalisation and associated improved experience later, the case for national ownership became quite strong and the railways were nationalised for good, under the Transport Act of 1947. In some respects, this could have ushered the golden age. There was a single, integrated rail network with cooperating lines, possibility of complete rationalisation of ticketing, and of all revenues going back into public service. Indeed many improvements did happen during this era, though the railways also managed to make colossal losses, mainly due to the inefficient practices, poor accountability and costs of delayed modernisation. A concerted effort to stem the losses was made through a Reshaping plan proposed in the 60s under leadership of Dr. Richard Beeching. Unfortunately, this came out as, and certainly was implemented as, one of those coy administrative ploys that appear to use ‘independent’ investigation to justify premeditated tactics without a thought-through long-term plan (i.e. without anything to do with strategy and architecture). While Dr. Beeching’s logic was quite simple - to get rid of loss-making routes (about quarter of the rail mileage and half the stations) and modernise the rest into a reliable and profitable service, it and its interpretation had several flaws/limitations. To begin with, many of the loss making minor lines could have broken-even with efficient practices, a level of automation, etc and did not need closing. Secondly, these lines fed the trunk lines, even as tributaries feed rivers, and when they were gone, passengers did not necessarily conform to Dr. Beeching’s logic - travel a few tens of miles to the nearest trunk line railhead - in the middle of a crowded urban area - using one’s own car, then park the car there - often at exorbitant parking fee with no one being responsible for its safety, and then pay a hefty price to travel on the modernising truck line. People used their cars all the way instead, the revenues needed for modernisation never quite materialised, many benign proposals of the Beeching report were mothballed and railways continued to struggle, albeit with a smaller volume.
Other factors that worked against modernisation were the outmoded working practices and trade union militancy that made it virtually impossible for a government enterprise to improve. This, more than anything else, made rail privatisation a seemingly attractive option. It was assumed that somehow the same staff would be made to work efficiently when working for private companies and, in any case, any loss of jobs through improved business process or automation would no longer be blamed on the government. Instead, the government would continue to reap tidy franchise fee - because private sector would be far better at management than public sector had ever been.
4. Railways were thus privatised ~1993-97. The privatisation process and governance were often too complex or clumsy and need a lot more than a paragraph to describe. As far as passenger transport and our analysis of virtual monopolies goes, it is sufficient to state that most facilities became monopolies or near-monopolies themselves, however their collective orchestration into passenger transport (the train operating companies, as they are called) was opened to competition for termed-franchise. However, as Mr. Livingstone points out, such once-in-a-while competition is quite limited in its effectiveness and the operator retains the possibility, as opted for by the ECML operator, of defaulting on the franchise payment and pulling out of the contract.
This erratic evolution of British railways gives us some stark characteristics:
1. Although British railways have actually been a tight mesh, usually with many ways of getting from A-to-B, they have been treated, except during war years, as made of separate rail services that have no desire to cooperate, even when under public ownership
2. The ticketing has never been rationalised. So, for example, while the ECML direct train to Leeds is usually the quickest way between London and Leeds, there are times of the day when other trains, such as ECML train to York and then connecting train to Leeds, or WCML train to Manchester and connecting train to Leeds, or MML train to Sheffield, extended to (or changed to) Leeds, can actually get one to Leeds quicker. Similarly, passengers may want to use these alternatives because the Leeds direct train on ECML is using old rolling stock with cramped seats. Unfortunately, such arrangements are not allowed. Even the ECML operator that fails to provide adequate trains to Leeds, and that also operates trains to York that might get one to Leeds more quickly at certain times, has been bloody-minded enough not to allow passengers to travel on the York train, although some those trains might go half empty at those service times and no matter how expensive and flexible ticket might the passenger have purchased
3. Railways, being considered quickest and most comfortable form of city-centre to city-centre transport in most situations, appear to price based on this convenience factor, and not based on the low operational cost they have - a mechanism encouraged by the virtual monopoly franchise system pointed out by Mr. Livingstone, mechanism that would need to be abandoned should genuine competition materialise.
4. While trains look modern and use some automated safety systems, technology has not been used sensibly to create better scheduling and encourage competition.
Of course, technology, used a little sensibly, and driven under the regime of benign “business architecture principles” can make a lot of positive difference and, in fact, break the virtual monopoly that Mr Livingstone, and most passengers, have criticised. We have already appreciated that technology allows use of trains longer than platforms, multiple trains on a platform and tighter scheduling of the trains; and this has indeed been implemented safely on busy lines. What has not happened much, but can happen quite easily with a little principled thinking and simple technology support, is generating genuine everyday competition and breaking the virtual monopoly that the present franchise system gives the train operators on their franchised routes.
Breaking the Virtual MonopolyThe separation of facilities and operation is based on the premise that equal access to facilities (e.g. track use, rolling stock leasing, etc) creates a level playing field where the best managed rail operator will be a favourite to win and retain their share of operations. This means that while some of the facilities may be virtual monopolies, many other facilities and train operations can be open to genuine competition, provided that a thoughtful and benign governance regime is in place that ensures that no services orchestration is allowed to disregard primary objective - of public service. This of course means a principled, well-architected model to begin with, that good governance can then leverage fully upon.
Under such regime, some facilities such as the track network, signalling system, etc, that emerge as natural monopolies, can be maintained as not-for-profit companies, even as the rail route maintainer has become, after the government’s flirtation with a for-profit firm that duly sunk after rewarding its shareholders for some time, when a single rail accident (the Hatfield rail crash), caused by inadequate maintenance capabilities of this profit-making monopoly, made it non-viable.
Some facilities, such as rolling stock leasing, optional station services, value-added train services (e.g. internet on the move, buffet car, etc) and train operations (i.e. using a specific rail line through a specific timetable) can then open up to genuine competition. Track-usage scheduling, including scheduling for maintenance and contingencies, has long been computerised, allowing safe and efficient usage of the rail network. This computerisation also allows logically considering any rail travel, not as a several hundred miles route, but rather as a fairly flexible aggregation of quite small linked sectors, the smallest sector being the segment between two consecutive points on a track (for routing) and consecutive stations (for passengers & goods). This greatly improves possibilities of value through competition.
Let me illustrate some of the possible improvements with the same old example - of travel between London and Leeds.
First, the business principles. Let’s consider just one sensible and rather obvious principle:
Best value and maximum choice for the passenger, without compromising on safety, shall be the guiding principle of British railways.
This principle can logically lead to the following rules.
1. Unless passengers have booked very low priced tickets that require mandatory reservation, for travel on specific seat of specific train, passengers will be allowed to use any viable route from A to B under any one of the following conditions:
a. That gets them to their destination quickest (meaning earliest)
b. That gets them to the destination with fewest changes
c. That they find most comfortable/convenient, without needing to exceed their time or mileage by more than a third, compared to the quickest route.
2. If the passengers booked very low priced tickets that require mandatory reservation, for travel on specific seat of specific train, they may still travel on another service, provided that this other service is not full and they are not taking seat that is reserved or that should be available to flexible ticket holders who have paid higher ticket cost. They may alternatively pay fee for additional reservation on another service, if available, for price not exceeding 50% of their ticket price (modern train management system allows dynamic reservation, even after the train has left its starting point, and thus maximum use of the facility).
Given that in early days of British railways, route operators were actually quite competitive with more than one viable route offered between most major cities, and given that most of these trunk routes are still available, or can be easily reopened, leading to multiple travel options between the same, or nearby, starting and end points, there is much that modern ICT can do to maximise route use and value to the passenger. For example, we have discussed four viable routes to Leeds from London:
1. From Kings Cross, via Wakefield (current principal route)
2. From Kings Cross, via York
3. From St. Pancras, via Sheffield
4. From Euston, via Manchester
The logical transport rules described earlier will allow travel on any of these routes. Indeed, the destination is the same station - Leeds. The starting points - Kings Cross and St. Pancras are next door to each other (sharing the same London Underground station), while Euston is about 10 minutes walk away (or one stop on London Underground). The three stations – Kings Cross, St. Pancras and Euston were built by the pioneers quite intentionally next to each other, in order to promote competition. Now, with modern ICT, each of these stations can provide real-time departure boards indicating both:
1. What trains are leaving at what time from each of these stations?
2. What options for travel from London to Leeds are available in the near future that would take the passenger to Leeds at the earliest, or within 50 minutes (~30% of quickest travel time) of the earliest? What starting points and changes will these involve? The departure board can thus be sorted by destinations first, followed by arrival times at the destination, making them easy to read
This, with passengers’ familiarity with this hitherto unsuspected facility within a few days, will ensure that they will use the route that they believe provides the most value. The ticketing revenue can then be shared among the operators, based on the route that a passenger has used. All tickets are machine readable, most stations have ticket barriers that read this information, and similar handheld devices can be provided, without excessive cost, to all train managers (ticket-checker, in old money), thus devising the precise proportion of revenue to be shared among the various operators used by the passenger, without the passenger needing to worry about buying the ‘right’ ticket. Moving forward, smart seats can come into being, that read (and hold) tickets, where a passenger can input the point to which s/he intends to travel on that seat, thus providing further input to make dynamic reservations effective. As the rail network provider is a not-for-profit monopoly, it can charge the train operators based on their mileage and average speed allowed on each segment of the route they use. This calculation can provide the basis for dividing ticketing revenue coming from a passenger, who has used more than one train operator in one journey.
Similarly, London Underground, between Kings Cross St Pancras and Euston (or Euston Sq) can be opened up for quicker travel between these stations, and a small fraction of the ticketing revenue can be shared with London Underground, following the same algorithm.
Going forward, it is possible for train operators to develop new routes - any route being logically composed of a series of station-to-next-station segments. England being quite urban with extended stretches of urban conurbations (such as the coast-to-coast Liverpool-Hull region in the north), it is possible to create many direct train routes (e.g. where trains can cross-over between the present day WCML and ECML between Leeds and Manchester). Similarly, passenger service, and not the franchise monopoly, being the guiding principle, it is possible to make the service more logical, where fewer trains will bypass major cities (so, more ECML trains can be routed through Leeds, and WCML trains through Birmingham and Manchester/Liverpool). As it is these large conurbation dwellers who are the biggest users of long-distance trains, such routing can further increase their choices and thereby overall competition.
In conclusion, thanks to the chequered history of railways and of urban development in Britain, modern ICT can make it quite easy to rationalise British railway system while making it sustainable. Of course, ICT is just the enabler. It can not replace sound business principles and good governance. It can only raise the practicability of viable solutions that, until a few decades ago, lacked technology to deliver safely and reliably.
And this, in turn, gives us a few points about Business Architecture. It is meant to:
1. Be based on sound architecture principles which, in turn, direct business rules. Principles exist, more than anything else, to remind the business of its values - what it stands for.
2. Give the business a clear sense of its role in the society - long term sustainable role, unless a business is mandated to have specific term of existence
3. Begin with core function of the business and understanding of essential interlocutors (usually meaning primary customers), then link to services the business provides, and then on to the structure of the business and its role in facilitating those services - in that order.
4. Consider if the business is engineered - formally or otherwise, extract/devise the processes, and examine their rationale and link to core business function
5. Consider constituents of the business as potential contributors to its future. Understand history of the business, if any, and the present state. Appreciate that present constituents of the business do not necessarily belong just to the ‘problem/opportunity space’. Most constituents are legitimate variables and may appear differently in a rationalised solution space.
- Comments(0)http://caoblog.beamsrc.com/#post3
