Road Infrastructure ppps in Germany: Why Did the f-modell Fail

5.The A model and the V model

As was briefly discussed above, the A model PPP variant was developed to extend existing, but congested, road infrastructures. Originally designed a pilot scheme for a total of four projects, it was replaced from 2009 by a critically redesigned and renamed variant – the V model – which continues to be considered as an alternative to conventional road infrastructure extension projects. Based on a standard BOT model, the A model was officially announced in 2005 and first implemented in 2007. It boasts the following key characteristics:

• 
  Initially, the Federal government assesses whether an extension project may be fit for A Model implementation.
  
• 
  Next, a private concessionaire – either an individual company of a consortium - is selected through a process of collective bidding.
  
• 
  The concession period is set at thirty years.
  
• 
  The concessionaire does not levy road users a toll of its own.
  
• 
  Instead, a percentage of the toll which has been levied on trucks using the specific section of the motorway is passed on to the concessionaire, while passenger vehicles do not contribute to the project costs at all (as they continue to travel toll-free, only being charged a gasoline tax and vehicle tax which are considered general tax revenues and are not earmarked). So, just as is the case with respect to the F model, the concessionaire must bear the traffic volume risk in full (although under the F model, both trucks and passenger vehicles are subject to a user charge!).
  
• 
  In some cases, the Federal government provided seed funding based on competitive tenders.
  

The A model was implemented in four pilot projects:

• 
  Autobahn A8 section Augsburg-Munich (52 kilometers of operation and maintenance, 37 kilometers of extension).
  
• 
  Autobahn A 4 section from the State border of Hessen/Thuringia-Gotha (44 kilometers of operation and maintenance, 25 kilometers of extension).
  
• 
  Autobahn A 5 section Malsch-Offenburg (60 kilometers of operation and maintenance, 72 kilometers of extension).
  
• 
  Autobahn A 1 section Bremen-Hamburg (60 kilometers of operation and maintenance, 41 kilometers of extension).
  

At least one of the four pilot project has turned out to be a substantial commercial failure: The A1 extension with A1 mobil as the concessionaire which represents John Laing Infrastructure Ltd, a British infrastructure and PPP operator, and Johann Bunte Bauunternehmung, a local construction firm. This section represents one of Germany’s most important routes for hinterland traffic to and from Bremen and Hamburg, the locations of the country’s largest seaports. While no details have become publicly available, it is known from leaks to the media that A1 mobil has been actively seeking a renegotiation of its contract with the Federal ministry of Transport since 2009, i.e. only one year after the section was opened for traffic. It is also known that, in 2012, the consortium sought a moratorium with its main creditors as well as a settlement with the Federal government (which never materialized), claiming an additional €124 million in compensation due to unexpectedly low traffic volumes in the wake of the global financial and economic crisis after 2008 (for details see https://www.heise.de/ tp/features/A1-Mobil-PPP-birgt-auch-fuer-Unternehmen-Risiken-3822436.html; https://www. lunapark21.net/oepp-pleite-mit-ansage-der-exemplarische-fall-a1-mobil/). It is noteworthy in this context that the total funding costs slightly exceeded the construction costs of €515 million and that the consortiums projected a pre-tax rate of return slightly higher than thirty per cent (which was accepted by the Federal government in the concession agreement).

A second batch of eight additional A model PPP was initiated in 2011, all of which were based on the revised V model. It largely avoids one of the main risks to PPP concessionaires. More precisely, the crucial difference between the original A model and the V model is the elimination of any direct traffic volume risk for the concessionaire. Instead, it will obtain a so-called availability remuneration which is calculated to reflect the quality and accessibility of the specific Autobahn stretch for users. A bonus-malus components was added to reward or punish any deviation from the pre-agreed availability indicators. However, a small indirect traffic volume risk remains if and when maintenance costs increase due to an unexpected traffic volume increase which is not fully covered by the pre-negotiated availability remuneration.

The most recent development was the official announcement of “next generation road infrastructure PPP” in late April 2015 which represents an enhancement of the V model variant. A total of ten projects which in total cover 600 kilometers of Autobahnen, two Federal interstate highways and a crossing of the Elbe river (in lieu of a F model solution!), are under consideration for this most recent road infra­structure PPP model. Investment costs are currently being projected at around seven billion Euros. If implemented as designed, private investors will play a substantially larger role in project funding and project management. While this may be economically beneficial in terms of cost savings with respect to project management, comparatively higher funding costs of the private sector will likely remain a potential downside (for a full map of all road infrastructure PPP of the F model, the A model, the V model and the “next generation” variant see https://www.bmvi.de/Shared Docs/DE/Anlage/VerkehrUndMobilitaet/Strasse/oepp-karte.pdf?blob=publicationFile).

6. Critical Assessment and outlook

The potential advantages of the F model over the tra­di­tional tax-based public provision of road infrastructures are obvious. Not only is it the first step of the transition to­wards a user-pays system which should allow for more efficient infrastructure al­lo­ca­ti­on de­ci­si­ons. More­over, it has the potential to partly alleviate – though not nearly completely solve – the prob­lem of chronic underinvestment in much-need­ed road infrastructures in Ger­many due to the bad shape of public finances in general (in­clud­ing the extraordinary fiscal burdens caused by re­uni­fi­ca­tion) and the lack of a legal re­quirement to fully re­in­vest the revenues gen­era­ted by road-transport related taxes in­to the maintenance and ex­pansion of the road net­work in particular. Finally, more wi­de­spread use of the F model might, at least on pa­per due to the strong incentives for pri­vate investors to recoup the costs of their in­vest­ment as quickly as possible, also speed up the completion of eligible road infra­struc­ture projects.

Unfortunately, the practical lessons learned from the two operational F model PPP and the fact that most other proposed projects simply failed to attract any private sec­tor interest at all – especially after the first to show-case projects quickly turned out to be ‘white elephants’ rather than attractive investment opportunities –, gives rise to a pes­si­mi­stic overall assessment of their merits, for the following reasons:

• 
  To begin with, it remains doubtful that any meaningful cost savings compared to the traditional approach could be realised. On the hand, due to their lower ra­tings, the private investors had to raise the necessary capital at higher interest rates than the government could have. On the other hand, contrary to what was ex­pected beforehand, project implementation from the design stage until the of­fi­cial opening took nearly as long as for comparable strictly public infrastructure pro­jects.
  
• 
  More importantly, in order to increase the at­trac­ti­ve­ness of F model PPP for the pri­vate sector, projects risks must be more evenly and fair­ly allocated be­tween the public and the private parties. Due to the sunk cost charac­te­ristics of road in­frastructure investments and the extraordinary long time span over which the – often substantial – investment costs must be recouped, the commercial via­bi­lity requires, with a vengeance, the absence of incalculable political risks. In other words, pri­vate investors should be guaranteed some protection against (or com­­pen­sation for) future political decisions which would have a massive ne­ga­tive im­pact on their investment, e.g. by significantly impacting on traffic volumes. The tax-funded provision of competing toll-free in­­fra­structures after the com­ple­ti­on of the toll project – which actually happened in Lübeck and which effectively killed a third F model PPP, the Strela­sund­que­rung, at a very early stage¹ – is a case in point.
  
• 
  All parties involved massively overestimated future demand levels, resulting in actual traffic volumes falling way short of forecast ones – often by a factor of 100 per cent. In­credibly enough, the private investors failed to take into account some long-stand­ing and well-documented demographic and socio-economic trends which would later turn out to have a strong negative impact on their in­vest­ment. For example, the city of Rostock – like most other East German cities – lost 20 per cent of its in­ha­bi­tants during the first decade after reuni­fication due to two-digit un­employment rates, and a further reduction by ano­ther 15 per cent by 2020 is li­kely (Klingholz/Kröhnert/Olst, 2004). More­over, the number of users of the ferry link which was to be re­placed by the tun­nel, had shrunk from a peak of 3,500 vehicles per day to around 1,000 to 1,300 due to the expansion of the toll-free road network (Brantsch 2004/05, 18).
  
• 
  In addition, in all cases the reluctance of potential users to pay for the use of a road was substantially higher than anticipated – though hard to explain ra­tio­nal­ly, given the fact that the additional costs of by-passing the toll road (time, extra fuel etc.) often exceeded the user charge. One possible psychological ex­pla­na­ti­on might be that German motorists have grown accustomed to use the road net­work ‘for free’ – unaware that it has been financed by their taxes, in­cluding road transport-re­la­ted specific taxes such as the car tax and the gas­oli­ne tax. Alter­na­tive­ly, there seems to exist a widespread feeling that due to already the high tax bur­den, any additional charge for using the road network is simply a rip-off with­out any benefits for the user and should therefore be avoided at any cost. It is open to debate whether the inclusion of passenger cars into the Au­to­bahn toll system some time in the future, if it were combined with a com­pen­sa­to­­ry re­duc­tion of the car and/or gasoline tax, might improve the acceptance of road users charges which would, in turn, brighten up the prospects of future F model PPP as well.
  
• 
  It is noteworthy in this context that the aforementioned acceptance problem is severely com­poun­ded by the specific legal rules which re­gu­late the calculation of the toll (and the overall toll structure) the concessionaire is allowed to charge un­der the F model (Kirch­ner, 2007, 3f.). As men­tio­ned above, the formal procedure and legal re­quire­ments differ strikingly from the principles of price regulation in one crucial as­pect: the toll is legally con­si­de­red to be a user fee in the meaning of German ad­ministrative law (Ge­bührenrecht). This severely restricts the pricing strategies the concessionaire is legally allowed to pursue. To be more precise, the toll must include the actual cost of the service offered. Due to the high ratio of fixed to total costs which must be reflected by the calculation, short-run marginal cost pricing – e.g. in the guise of a toll-free passage during the late night period, when demand is ex­tre­me­ly low – would run afoul of the law. By the same token, the concessionaire is not allowed to charge a below-cost in­­tro­duc­tory toll (i.e. to practice penetration pricing) after the opening of the fa­cility in order to attract a large number of users and to familiarize them with the toll system. To conclude, the current legal requirements effectively prevent any mean­ing­ful price dis­cri­mi­na­tion among dif­fe­rent groups of users to exploit their varying will­ing to pay or dif­ferential pricing over the life-cycle of the investment with low tolls initially, followed by higher tolls during later pe­riods.
  
• 
  Finally, there is a potentially harmful conflict of interest between the Federal go­vern­ment on the one hand and the affected Länder government (and/or local go­vern­ment) on the other which may ef­fec­ti­ve­ly thwart a F model solution – which it did at least in the case of the failed Ste­la­sund­que­rung. While under the traditional sy­stem the Federal government provides all the funding with most of the eco­no­mic benefits ac­cruing to locals, under the F model it is the local po­li­ti­cians who have to ‘sell’ the switch-over from a (publicly perceived) free-of-charge road system to a toll sy­stem which primarily affects their local electorate. For obvious reasons, local political and media support for a toll-based solution has so far been lacklustre at best in the case of most proposed F model PPP (and, for that matter, A model PPP as well) (Gawel, 2005, 181).
  

As for the A model, the likely bankruptcy of A1 mobil, the concessionaire of one of the four realized projects, on the one hand casts massive doubt on the ability of the Federal Ministry of Transport in selecting bidders which are capable of profitably operating such project for the full concession period. On the other hand, the rejection of a settlement with the operator was a powerful signal by the government that it will not accept hold-ups from the concessionaires. If a quick replacement could be ensured in case of bankruptcy, this approach would give potential investors stronger incentives to reveal their actual performance in the bidding process. However, the transition from the original A model to the V model has, again, changed the incentive structure in favour of potential contractors by eliminating traffic volume risk for the most part. While this new legal environment makes it easier for companies to calculate their bids more realistically, it remains doubtful that the higher transaction and funding costs involved in PPP implementation as opposed to conventional solutions will render the PPP alternative an economically more beneficial approach. Finally, the “next generation” variant appears to offer a mixed bag of economic effects which, overall may benefit the concessionaires more than the government and the public. The first obvious advantage for the concessionaire would be the elimination of traffic volume risk (to avoid hold-ups and bankruptcies during the thirty year concession period). Second, the proposed stronger role of the concessionaire in securing finding is unlikely to yield in any cost savings to the government and the public due to the typically lower ratings of private as opposed to public debtors in the case of Germany. Third, and last, only the proposed transfer of project planning and management to the concessionaire appear to offer reasonable scope for cost reductions.