Paper No. 01-0238

Railroads in Greece: History, Characteristics and Forecasts

John A. Paravantis, Ph.D. Department of Industrial Management, University of Piraeus

80 Karaoli & Dimitriou, 18534 Piraeus, Greece

(jparav@unipi.gr)

Panos D. Prevedouros, Ph.D.1 Department of Civil Engineering, University of Hawaii at Manoa

2540 Dole St., 383, Honolulu, HI 96822, USA

(pdp@hawaii.edu)

Submitted: July 22, 2000

Revised: November 10, 2000 Final Revisions: January 22, 2001

Forthcoming in TRANSPORTATION RESEARCH RECORD 2002 2,000 8 tables/figures/exhibits 5,414 text 7,414 total (max = 7,500)

1 Corresponding author.

2

Abstract Greece was relatively tardy in implementing a national rail network. The Greek Railways Organization, OSE, was formed in 1970 uniting a number of regional railroads, some dating back to 1884. The current rail network in Greece has a length of approximately 2,500 km with standard and metric gauge tracks. Disadvantages of the network include a high percentage of single line track, lack of homogeneity of track gauge, inefficient alignment, absence of intermodal connections and many at-grade intersections. Advantages include double tracks in a significant part of the Athens–Thessalonica axis that serves more than half of Greece’s population, improved operational controls and popular Intercity trains offering speedy service. Factors that affect railroad passenger and freight volumes are competition from airlines and motor carriers, gas price, inflation and GNP. These were used to estimate time-series models. Forecasts indicate increasing passenger and decreasing freight demand. OSE could provide a valuable and viable passenger service if cost-effective expenditure for improving it is possible. The future of freight operations is not hopeful. Additional emphasis may be placed on lines with historical and cultural significance. Further study is needed to assess the sustainability of forecasts vis-à-vis actual market shares and costs.

3

1. Introduction and Historical Summary

This paper investigates information on the past and present of Greek railroads, as currently

represented by the Greek Railway Organization (OSE), beginning with the birth of Greek

urban and regional railroads in the 19th Century. Models were developed to produce

forecasts that attempt to shed light into the future of intercity railroad transportation in

Greece. Market share and cost issues were not addressed in this research. This work may

be of particular interest to decision makers and rail administrators since large amounts of

European funds are currently invested in upgrading major rail lines such as Athens-

Thessaloniki and Athens-Patras.

The paper is structured as follows. A historical review of railroads in Greece is

presented in this section followed by a description of the railroad network in Greece in

section 2. Section 3 presents basic comparisons with railroad networks in Europe. Positive

and negative characteristics of Greek railroads are described in section 4. Section 5 presents

general and operational characteristics as well as economic variables that affect railroad

transportation. Models and forecasts are presented in section 6 followed by a discussion and

the conclusions.

Although major railroad development began in the early 1800s, Greece was tardy in

implementing a railroad network. Greece was reestablished in 1828 following the War of

Independence against the Ottoman Empire which commenced in 1821. This war left the

land barren due to neglect and destruction (1). The sacrifices made in the war of

independence and numerous external loans that Greece had to contract to meet its

obligations delayed the development of railroads compared to other countries in Europe.

The mountainous terrain and sparse population discouraged private enterprises from

developing private railroad networks.

A number of proposals by Feralde (in 1835), Ragavis (in 1843), British interests (in

1859), Baron Louis De Normand (in 1868), Vitalis (in 1869) and Count Moutressi (in 1879)

met with no success. During 1881-1882 railroads were intensely discussed by the Greek

Parliament as the opening of the Suez Canal in 1869 drew world attention on engineering

feats, and the annexation of Thessaly in 1880 brought Greece closer to its Balkan neighbors

and the European mainland. Prime Minister Trikoupis’ address to the Parliament on May 20,

1882 is regarded as the birth date of Greek national railroads (2). Additional lines were

4

added until about 1910 when most of the network was completed. Several of the lines that

form the modern-day network are discussed in the next section.

The Athens – Piraeus transit line has been in operation since 1869 and is the first rail

line to operate in Greece. It has a total length of 25.8 km and a standard track gauge of

1.435 m. Steam trains serviced the line until 1904 when its electrification began.

Electrification was completed in 1954. This metropolitan rail line was the only urban railroad

in operation in Greece until January 2000 when the Athens Metro was put in operation. This

paper focuses on intercity passenger and freight railroads; as such, no further attention is

given to metropolitan rail.

Several old rail lines add aesthetic and cultural value and are connected to national

history (2). Examples of cultural significance include:

• The 22-km narrow-track (0.75 m) route of Odontotos railroad is an Abt-type railroad that uses a rack to move up steep cliffs. It originates at a junction with the Athens – Patras line and terminates at Kalavryta where Greece declared its independence from the Ottoman Empire.

• The 28-km rail line from the city of Volos to the village of Milies in Pelion Mountain is one

of the few 0.60-m narrow track rail lines in the world. This historic route is serviced by steam locomotives and restored wagon-cars and provides access to landscapes of significant sightseeing value.

• The winding route from the city of Drama to the Valley of the Nestos River which

possesses significant ecological importance. The floodplains and associated wetland surrounding the Nestor River are protected by the Ramsar Convention and rail is the only mode that affords views inside the valley.

• Other small routes of historical significance or tourist interest such as Ancient Olympia,

the popular summer resorts of Loutraki and Xylokastro and Nafplio (capital of Greece in the 19th Century).

2. Railroad Networks in Greece

The Greek Railway Organization (OSE) was created in 1970 with Act 674/1970, taking the

lead from the former Railways of the Greek State (SEK). The obligations of OSE were set in

Decree 404/1972, and its Articles of Association in Decree 532/1972. OSE united a number

of separate regional railroad networks, as follows:

• Railways of the Greek State (SEK). Its operation began in the 1904 to 1909 period

and encompassed the standard track gauge Piraeus – Athens – Larissa – northern

5

borders line and other minor lines. As Macedonia and Thrace were reunited with

Greece in the Balkan Wars of 1912-13, additional segments were developed and

united with existing rail lines in northern Greece linking Thessalonica to present time

borders with the Former Yugoslav Republic of Macedonia (FYROM) and Turkey.

These older lines in Macedonia and Thrace were constructed during occupation by

the Ottoman Empire and began operations between 1883 and 1892. Rail lines to the

city of Kozani and Sofia (capital of Bulgaria) were constructed in 1954 and 1963,

respectively.

• Railways of Peloponnese (SPAP). This railroad had a total rail length of 726 km with

metric track gauge of 1.00 m. Segments of this railroad commenced operation in

1884 and the entire network was in operation by 1902. Railways of Peloponnese

became the responsibility of the public sector in 1940 and merged with SEK in 1962.

• Railways of Northwestern Greece (SVDE). This railroad had a single line of 61 km

with metric track gauge of 1.00 m. It commenced operation in 1892 and merged

with Railways of Peloponnese in 1953. This line is no longer in operation.

• Thessaly Railways (STh). This railroad had an initial total rail length of 202 km with

metric track gauge 1.00 m. Segments of Thessaly Railways began operations

between 1884 and 1886. The railroad merged with SEK in 1955.

During the 1980s, Acts 1365/1983 and 59/1985 established OSE as a “social”

company with a Board of Directors. Acts 2366/1995 and 2414/1996 partially removed OSE

from political influence and allowed it to form subsidiary companies that manage European

Union funds (e.g., ERGOSE – OSE Constructions – that currently project-manages the

upgrading of railroad infrastructure throughout Greece.) More recently, Presidential Decree

76/1998 stipulated terms and conditions for issuing a license of operation to railroad

companies in preparation for the scheduled, border-free transportation market within the

European Union (3).

Presently the entire railroad network in Greece consists of 2,507 km of track of

standard and metric gauge. The network encompasses three main lines:

6

• The 594-km standard gauge (1.435 m) Piraeus – Athens – Thessalonica – FYROM

axis which constitutes a section of the European High Speed Network. This axis is in

immediate proximity to areas that contain 54% of the Greek population (4).

• The 220-km long Athens – Corinth – Patras axis with 1.00 m metric gauge.

• The 632-km long standard gauge Thessalonica – Alexandroupolis – Turkey axis.

Passenger service is concentrated on the Patras – Athens – Thessalonica route while

freight is concentrated on lines connecting Thessalonica to northern (FYROM, Bulgaria) and

eastern (Turkey) borders as well to Athens via Larissa. It is anticipated that the railroad line

Patras – Athens – Larissa – Thessalonica will become a high-speed fully electrified line by the

year 2004, connecting nearly two thirds of the country’s population.

3. Comparisons with European Railroad Networks

This section presents four different comparisons: Characteristics of railroad networks of the

15 European countries in the European Union (EU), passenger and freight characteristics of

EU railroads, passenger and freight characteristics of ten Eastern European countries, several

of which neighbor Greece, and characteristics of tracks.

Railroad networks of the 15 European countries are compared in Table 1 (5,6). With

the exception of the Netherlands and Finland, Greece trails its European neighbors in terms

of railroad coverage both of people and area. The Netherlands is behind Greece in terms of

coverage of people, but it is the most densely populated country and it is performing far

better in terms of area coverage. Finland trails Greece in area coverage, but it is the least

densely populated country in the European Union and it performs best in terms of coverage

of people. Greece falls behind countries such as Spain, Portugal and Ireland that share

common characteristics with Greece and are traditionally regarded as similar among member

states in the European Union.

Tables 2a and 2b (7) present passenger and freight characteristics of the 15 EU

countries and 10 Eastern European countries. Greece lags behind both groups of countries.

To highlight the low usage of rail in Greece, the outcome of 0.2 billion pass-km per person

for Greece could mean that 6 out of 100 Greeks travel once a year by rail over a distance of

300 km. (This is one of an infinite number of possible descriptions.) In comparison, 10 out

of 100 EU residents travel once a year by rail over a distance of 300 km, and 18 out of 100

Eastern Europeans travel once a year by rail over a distance of 300 km.

7

Railroad track gauges in Greece and similar to Greece (5) are shown below:

Gauge Greece Spain Portugal 1.000 897 km (36.4%) – 307 km (10%) 1.435 1,565 km (63.6%) 471 km (3.6%) – 1.668 – 12,560 km (96.4%) 2,765 km (90%)

It is evident that 1.668 m gauge railroad lines are prevalent in the Iberian Peninsula

which poses a formidable problem to Spain and Portugal when it comes to merging their

lines with European networks. This currently necessitates the transfer of people and goods

at the borders. The low standard track gauge of Spain corresponds to high-speed railroad

lines. In comparison, Greece possesses both 1.435 m and a high percentage of 1.000 m

railroad networks.

4. Positive and Negative Aspects of Greek Railroads

Major disadvantages of Greece’s railroad infrastructure can be summarized as follows (8,9).

• High percentage of single line track. • Lack of homogeneity in track gauge: the difference in track gauge between the

Peloponesse and Athens – Thessalonica lines severely hinders north-south connections, necessitating multiple transfers and reducing the attractiveness to potential customers. The track gauge difference is an important factor for the poor development of railroads in Greece.

• Unsatisfactory alignment: small horizontal curvatures with radii in the order of 300 m and

steep vertical slopes (often up to 2.5%) prevent the development of high speeds, raise fuel consumption and increase travel times.

• Antiquated tracks that hinder the movement of heavy and high-speed locomotives and

rolling stock. • Insufficient rail connections, yards, ramps and freight-handling infrastructure at the

major mainland Greece ports of Elefsina, Kavala, Piraeus, Patras and Thessalonica hinders or precludes marine-rail intermodalism.

• Lack of a modern and spacious commercial station and marshalling yard in the greater

Athens area. The under construction port of Ikonio is designed to address this and the previous shortcoming.

• Existence of 2,587 at-grade intersections with the highway network, often inside cities,

render high-speed movement of trains impossible and is the cause of serious accidents. OSE trains are forced to cross the distance of Piraeus to Athens with a mean speed between 20 and 30 km/h due to numerous at-grade crossings with local streets. These at-grade crossings essentially bisect Athens and result in long delays to both highway and railroad traffic.

8

Advantages of Greece’s railroad infrastructure can be summarized as follows (9).

• Existence and operation of double tracks in a significant part of the main axis Athens – Thessalonica.

• Improved signalization and automation and funds committed for further expansion and updating.

• Popular Intercity trains offer better service to the public (Table 3 provides a comparison). • High frequency of departures.

5. General and Operational Characteristics

Railroads traditionally transport passengers and freight such as coal, grain, oil, chemical

products, pulp, paper and forest products, vehicles, machinery and spare parts. Passenger-

km and freight ton-km show a 50% market share for Greek railroads in both passenger and

freight volumes around 1950, declining to about 11% for passengers and 18% for freight in

1970 and to about 4 to 4.5% for both passengers and freight in 1996 (4). Draft recent data

paint a bleaker pictures with passenger share below 3% and freight share below 1%. The

loss of market share in the last 40 years for Greek railroads is dramatic, especially in the case

of freight where OSE lost almost 90% of ton-km. In contrast, road transport gained nearly

total dominance in passenger transport and more than doubled its share in freight transport,

capturing in both passenger and freight markets more than 95% of total land transportation.

The loss of market share is not exclusive to Greece. Similar highway and railroad

passenger and freight figures prevail in the 15-country European Union as shown below (6).

Notably, in the 1970 to 1997 period, ton-km of freight increased by 107%, but railroad ton-

km decreased by 16%.

PASSENGER (% share) FREIGHT (% share) 1970 1980 1990 1997 1970 1980 1990 1997

Highways 73.6 76.0 78.7 78.5 30.8 33.1 40.6 43.4 Railroads 10.1 8.2 6.5 5.8 21.2 15.2 11.2 8.6

Market research carried out jointly by Olympic Airways and OSE confirmed that

passengers are most annoyed by delays, lack of cleanliness, too many stops, slow speed and

rude behavior of OSE employees (in rough order of preference (10)). More recent market

research carried out between November 1996 and January 1997 showed that freight

customers chose OSE mostly because of competitive pricing and safety (11). On the other

hand, they avoided Greek railroads primarily because of long departure and en-route delays,

9

bureaucracy and rude employee behavior (11). It is widely acknowledged that OSE lags in

terms of service quality compared to its competition.

Exhibit 1 presents the trends in passenger and freight loads as a straight measure

and in terms of passenger-km and ton-km. The decline in passenger volume between 1977

and 1983 coincides with the recovery of democracy in Greece after a 7-year government by

a military junta. Austere economic policies were enacted by the democratic government of

Karamanlis. Those led to economic depression resulting in a reduction in both passenger

travel and freight (imports and exports). The consequences to Greek rail were an acute lack

of modernization, lack of rolling stock, lack of proper maintenance of rolling stock (e.g.,

typically 60-70% of rolling stock was unavailable) and poor performance (e.g., average

speed of 60-65 km/h). These conditions created a particularly fertile ground for motor

carriers.

Overall, tons are relatively high and ton-km relatively low. The latter is largely due to

the short distance between the port of Thessalonica and the northern borders with FYROM

(and the rest of Yugoslavia) and Bulgaria. The decrease in both freight tons and freight ton-

km of 1982 was due to:

• the discontinuation of an important grain transport contract that was competitively awarded to motor carriers,

• a significant decrease in lignite (a type of coal) transport for LARCO, a public sector company that in 1982 ceased operation for 8 months,

• decline in tobacco, fruit and grocery exports, • significant decline in agricultural areas available for sugar beet cultivation, and • decline in construction.

A sharp decline in freight tons occurred after 1993. This is attributed to a Europe-

wide economic recession that affected Greece severely. Other reasons include a decline in

the transportation of sugar and sugar beets (produced by a state company), chemical

products, crude oil, fertilizers and military materials. It is interesting to note that the 1993

decline was compounded by OSE’s inability to transport coal for the BIOLIGNIT Company. In

1995, BIOLIGNIT terminated its contract with OSE and turned to private motor carriers. In

the mid-1990s, imports and exports also suffered greatly due to regional wars in Yugoslavia;

this freight was diverted to trucks which transported it to and from Greece via ferry

connections with Italy.

10

Income from operations appears to grow considerably in the 1990s particularly for

passenger operations. However, when adjusted for inflation, income from passenger

operations is rather stable from the mid 1980s to the late 1990s, whereas income from

freight has experienced a continuous decline since 1975.

The inability of state-funded OSE to determine prices as freely as a private

transportation company would is shown vividly by the following data (12):

Year Passenger Fare Index Freight Tariff Index Inflation Index 1994 100 100 100 1998 103.3 101.5 130.3

Exhibit 2 presents a series of factors which draw a partial picture of the economy and

transportation in Greece. These factors influence railroad trends directly or indirectly. Most

of these factors were used in developing the forecasting models described in section 6.

The per-capita gross national product (GNP) of Greece has been increasing, though

at a decreasing rate, as the country transitioned from a developing to a developed economy.

With the monetary induction of Greece into the European Currency Unit (ECU based on the

euro) on June 19, 2000, Greece surpassed the lower boundary of a fully developed economy.

Its population has been increasing steadily at a modest rate, higher than the average rate for

Europe but lower than the average rate for the U.S.

Tourist arrivals also have increased annually, but at a decreasing rate as crowding

and unavailability of accommodations in peak periods lead to near-saturation, although

cruises and off-season tourism are developing markets. The small reductions in tourist

arrivals in the 1990s are attributable to wars in Yugoslavia (Croatia, Bosnia and Kosovo

conflicts with Serbia). These conflicts eliminated a direct highway connection between

Greece and the huge tourist market of northern Europe and created oversaturation on the

ferry links between Greece and Italy (i.e., capacity-constrained arrivals.)

Traditionally, Greece has been seen as a sea-fearing and agricultural nation.

However, the contribution of agriculture to GNP has stabilized to slightly below 10% in the

1990s. Thirty years ago, it was three times higher. This condition further reduces demand

for rail freight services. In addition, diversification of agriculture into specialized crops (e.g.,

kiwi, strawberries and other perishable, delicate and high-value agricultural products)

11

necessitated the use of motor carriers for speedy transport to export or domestic distribution

centers.

Inflation rates higher than 10% occurred in all but six of the 28 years examined in

this analysis (1970-1998). Inflation was necessary for Greece to reach monetary parity with

its western European counterparts in the Union. Inflation control in the late 1990s was a

precondition for admission to the European Union’s single currency. As a result, the

exchange rate of the drachma vis-à-vis hard currency has suffered between 1980 and mid -

2000. Greece was admitted to the euro-based economy with an equivalency of about 340

drachma to 1 euro, or 360 drachma to U.S. $1. A poor exchange rate affects equipment-

intensive operations such as a railroad that is dependent on foreign technology as it results

in increasingly expensive acquisition of rolling stock, equipment and parts.

Greece’s per capita automobile ownership has greatly increased (e.g., 0.03, 0.09,

0.17 and 0.24 autos per capita in 1970, 1980, 1990 and 1997, respectively) although it lags

behind that of other developed countries. For instance, in the late 1990s, the same rate was

near 0.60 in the U.S., 0.50 in Canada, 0.40 in the Netherlands and the U.K., and 0.30 in

Japan. The trend in truck registration is identical to that for passenger cars (one can create

it by simply dividing the passenger cars trend in Exhibit 2 by 2.5.) The inflation-adjusted

price of gasoline per liter has been decreasing from the highs in the 1970s and has been

rather stable throughout the 1990s making cars an increasingly formidable adversary to the

railroad for intercity travel.

6. Models and Forecasts

This section presents an attempt to connect several performance measures of Greek

railroads with external, explanatory variables such as GNP, gas price, vehicle registrations,

domestic airways passenger loads, inflation, exchange rate and so forth. In addition, basic,

short-term forecasts were estimated for the purpose of revealing a likely trend in the

aforementioned railroad performance measures.

Data were gathered for the period 1970-1998; we elected to start our investigation in

1970 because OSE was formed at that year. These data were collected from the Statistical

Bureau of Greece and cross-validated with OSE Annual Reports since 1970. This is the first

time such a thorough compilation of time-series railway data is presented. Most of the

variables considered are depicted in Exhibits 1 and 2. Both ordinary least squares regression

12

(OLS) and auto-correlated regression estimated with maximum likelihood (procedure AREG in

SPSS) were estimated. Nearly 180 models were estimated. The ten best specifications are

presented in Table 3.

The table shows the four dependent variables in the left-most column; they are

railroad passengers, passenger-km, tons, and ton-km. The independent variables column

includes the regression constant (intercept), the explanatory variables and, in the case of

autocorrelated regression, the first order autocorrelation factor AR(1). R2 values are

available for the OLS regression only. A more relevant measure of model fit is the standard

error of estimate (SEE) for the dependent variable; the smaller, the better. It can be seen

that AREG models are superior to OLS models in terms of prediction error. The next two

columns present the estimated coefficients and their associated statistical significance. Most

of the estimates are significant at the 99% level; a few are significant at lower levels. The

models are identified by numbers 1 through 10, and the best for each dependent variable

are plotted in Exhibit 3. The models offer a reasonably tight fit in most instances. The

“Railroad Passengers” graph in Exhibit 3 shows two models, an OLS (model 2) and an

autoregressive one (model 3).

Before looking into the forecasts depicted in Exhibit 3, the independent variables and

their contribution to each model are discussed.

• Rail Passengers (RPASS) is affected negatively by inflation-adjusted gas prices (variable GAS88.) Cheaper gas makes travel by auto more affordable which, in turn, reduces the rail passenger count. Similarly, air travel is a major competitor to rail thus a negative coefficient was estimated for the domestic passenger count for Olympic Airways (variable OLYMPIC). Inflation rate (variable INFLATION) has an effect similar to that of gas price, but the effect of inflation is stronger (as it should be) because it not only affects gasoline, but also other motoring costs such as maintenance, insurance, tolls, etc. The best model is a first order autoregressive one which maintains inflation rate in the specification.

• Rail Passenger-km (RPASSKM) is affected negatively by gas prices and positively by

the GNP trend (variable GNP88$.) The latter is intuitive. A higher GNP (in constant 1988 U.S. dollar terms, per capita) typically corresponds to higher travel activity, some of which is by rail. The best model is a first order autoregressive one which maintains GNP in the specification.

• Rail Freight in metric tons (RTONS) is affected negatively by truck registrations

(variable TRUCKS) and positively by inflation. Both of these outcomes are intuitive. Competition from an increasing pool of trucks reduces the freight carried by rail. Inflation affects rail freight positively because it takes a toll on motor carrier costs and pricing. In contrast, OSE increased freight prices by 1.5% while inflation

13

increased by 30% (see earlier data). The best model is a first order autoregressive one which maintains inflation rate and truck registrations in the specification.

• Rail Freight in metric ton-km (RTONKM) is affected negatively by truck registrations

and positively by gas price. Both of these outcomes are intuitive for reasons which are similar to those discussed above. The best model is a first order autoregressive one which maintains gasoline price and truck registrations in the specification.

As expected, most models suffered from auto-correlation since they were estimated

with time-series data. In addition, many independent variables were highly correlated with

each other. As a result, all but one independent variable was included in the final model in

some cases. For example, high correlation exists between constant GNP and constant

gasoline prices. Exchange rate and passenger cars are also positively correlated since both

of them grew rapidly in the time period examined and they also correlate with the passenger

loads carried by Olympic Airways domestically, for the same reason. Tourist arrivals, GNP

growth rate and exchange rate proved to be weak variables and do not appear in any of the

models. Truck registrations appear in freight models. Passenger car registrations had a

strong negative association with railroad passenger loads but were excluded from the final

specifications due to collinearity with inflation.

In addition to a visual depiction of model fits, Exhibit 3 presents short-term forecasts

through 2005. These forecasts were prepared by making a number of assumptions based on

reports in The Economist (Jan. 30, 1999, p. 106, May 27, 2000, p. 108) and the OECD

(www.gsu.edu/~wwwrer/Annotated/Global_Markets/greece1.htm). The following growth

rates were used:

• GNP: 3% per annum.

• Inflation: 5% for 1998, 4% for 1999 and 3% for 2000 to 2005.

• Gasoline price: increases are equal to rate of inflation.

• Domestic airline passenger growth: 5% for all years; same as for the 1991 to 1998

period.

• Truck registrations growth: 3.5% for all years; same as for the 1991 to 1998 period.

The forecasts indicate increases in passenger measures and decreases in freight

measures. These outcomes also are intuitive. OSE can provide a valuable and viable

passenger service which should remain competitive given that steps are being taken to

improve service. Examples of such steps include the speedy service on dual rail trucks along

main corridors, the 26 new diesel locomotives which were delivered in 1997, and electric

14

traction which was introduced in 1997 at the Thessalonica – FYROM border line. Roadway

and airport congestion on one hand, and rail efficiency (which enables lower fares) on the

other hand are elements that support demand for passenger service.

The future of freight operations is not hopeful. Strong decreasing trends are

observed which are consistent with observations such as:

• Greece has never relied on heavy industry. As its economy enters a developed state, it is further diversifying into services and high technology industries (computers, telecommunications, etc.) which require little heavy transport capability.

• Agriculture, as shown in Exhibit 2, continues to decrease in significance.

• New pipelines through former Soviet nations carry oil and natural gas directly to

Balkan nations. A rail link to the Mediterranean Sea is becoming less important.

• Military needs are decreasing as local wars and diplomatic efforts appear to have sorted-out several long-standing disputes caused by end of WWII arrangements and by Cold War era policies.

A strong caution must be raised here: This paper did not examine OSE’s fiscal

documents and trends and it did not investigate actual market shares and average cost per

trip. It is possible that OSE’s value of service is both miniscule in terms of share and it is

negated by continuous operating deficits and extraordinary costs for capital improvements.

7. Discussion and Conclusions

Greece is characterized by several peculiarities as a member of the European Union such as

extreme southeastern location in European Union thus low transit volumes, lack of borders

with other European Union members, relatively undeveloped transportation infrastructure,

and relatively low value of commercial exports to other European Union members (5). In an

effort to bridge the gap between Greece and the EU, the Maastricht Treaty of 1991 and

directions adopted in Edinburgh in 1992, encourage and support Greece to invest in new

transportation infrastructure that will facilitate a seamless transportation in the entire

European Union. High-speed, fully electrified rail service is a priority for Greece, especially

along the Patras – Corinth – Athens – Larissa – Thessalonica – northern borders route. The

modernization program of OSE is supported and financed by the Community Supporting

Frame II and the European Cohesion Fund.

Railroad passengers and passenger-km reached a low in the early 1980s and have

been increasing since then except for marked decreases in 1994 and 1998. Our short-term

15

forecasts paint a hopeful picture. It appears that strong attributes of OSE passenger services

are low fares and the provision of fast service to compete with air transport, particularly on

the Athens — Larissa — Thessalonica axis that covers a market of 6 million people.

Freight tons and ton-km exhibit a steady decrease in the last 30 years. Although the

overall picture is consistent with similar trends in most developed economies, low service

quality and tardiness in modernization contributed to Greece’s lagging position among EU

members. Our models show that competition from motor carriers is a strong negative factor

and low rail prices are a positive factor. However, Greece’s location and economy (lack of

heavy industry, low exports, decreasing needs for agricultural product transport) and its

mountainous geography do not bode well for railroad freight transportation.

The overall findings suggest that OSE should concentrate passenger operations.

Additional emphasis may be placed on historical and cultural railroads which, with proper

restoration, marketing and accessibility may provide an supplementary source of income. A

detailed study that includes analysis of market shares (to assess OSE’s contribution to

transport demand) and costs (to assess the true cost of providing existing and improved

service) is needed. This future study should determine whether OSE’s role in Greece’s

transportation system is a productive and cost-effective one.

16

References 1 Matzaridis, P. Concise History of Hellenic Railways [in Greek], 2nd edition, Hellenic

Railways Organization, January 1996. 2 Society of Friends of the Railway. Greek Railways: Their Route from 1869 to Today [in

Greek], Miletos Publications, 1998. 3 Profyllides, B. A Model for Estimating Usage Fees and Prerequisites for Access to

Railway Infrastructure [in Greek]. Research Program Final Report (Phase B), University of Macedonia and Greek Railway Organization, Thessalonica, September 1998.

4 Profyllides, B. A Technical and Economical Feasibility Investigation of Transportation of Wagons with Reclining Body in the Greek Railway Network [in Greek], Research Program (Phase A), University of Macedonia and Greek Railway Organization, Thessalonica, August 1997.

5 Vasilakos, E. “Transeuropean Transportation Networks of Interest to the Community

and the Role of Greece [in Greek]”, Technical Chamber of Greece, Weekly Bulletin, issue 1772, 1993.

6 European Commission. EU Transport in Figures, Transport Directorate-General in cooperation with Eurostat, July 1999.

7 EU Transport in Figures, Transport Directorate-General in co-operation with Eurostat, European Commission, July 1999.

8 Greek Ministry of Transportation. Roadway, Railway and Air Transportation – Present

Situation and Predictions for the Future, Athens, 1977.

9 Greek Ministry of the Environment. Greece as an Ecological and Cultural Resource: Facts, Actions and Programs for the Protection of the Environment [State of the Environment Report, in Greek], Athens, July 1995.

10 Profyllides, B. Transportation Economics [in Greek], Yahoudi – Yapouli Publications, Thessalonica, 1993.

11 Profyllides, B. Feasibility Study for Establishment by OSE of Transit Company [in Greek], Research Program Final Report (Phase B), University of Macedonia and Greek Railway Organization, Thessalonica, May 1997.

12 Greek Railways Organization. 1998 Annual Report of the Chairman of the Board of

Directors [in Greek], Athens, 1999.

17

Table 1: European Union Railroad Networks

Country

Population (millions)

Area

(1,000 km2)

Density (persons per km2)

Network Length (km)

Network km per 100 000 inhabitants

Network km per

1,000 km2 Austria 8.1 91.9 96 5,643 70 61

Belgium 10.2 30.5 333 3,410 33 112 Denmark 5.3 43.1 122 2,232 42 52

Finland 5.1 338.1 15 5,867 115 17 France 58.6 544 108 31,727 54 58

Germany 82 357 230 38,450 47 108 Greece 10.5 132 79 2,503 24 19 Ireland 3.7 68.9 53 1,945 53 28

Italy 57.5 301.3 191 16,041 28 53 Luxembourg 0.4 2.6 160 274 69 105 Netherlands 15.6 41.5 376 2,805 18 68

Portugal 9.9 91.9 108 2,856 29 31 Spain 39.3 506 78 12,303 31 24

Sweden 8.8 450 20 11,156 127 25 UK 59 244.1 242 16,847 29 69

Europe 373.9 3,236.3 116 154,059 41 48

18

Table 2a: Characteristics of EU Railroads

Table 2b: Characteristics of Eastern European Railroads

COUNTRY1997 RAIL

PASSENGERS(billion pass-km)

1997 RAIL FREIGHT

(billion ton-km)

PASS-KMPER PERSON

PER YEAR

BILLION PASS-KM

PER PERSON

BILLION TON-KM

PER PERSONBelgium 7.0 7.5 688 0.69 0.74Denmark 5.0 1.6 948 0.94 0.30Germany 64.0 72.7 780 0.78 0.89Greece 1.9 0.3 180 0.18 0.03Spain 17.3 11.5 441 0.44 0.29France 61.8 53.9 1055 1.05 0.92Ireland 1.4 0.5 375 0.38 0.14Italy 52.1 23.0 906 0.91 0.40Luxembourg 0.3 0.6 711 0.75 1.50Netherlands 14.4 3.4 925 0.92 0.22Austria 8.3 14.2 1032 1.02 1.75Portugal 4.6 2.2 460 0.46 0.22Finland 3.4 9.9 659 0.67 1.94Sweden 6.3 19.1 711 0.72 2.17UK 34.4 16.9 583 0.58 0.29EU15 average 18.8 15.8 696.9 0.75 0.63

COUNTRY1997 RAIL

PASSENGERS(billion pass-km)

1997 RAIL FREIGHT

(billion ton-km)

PASS-KMPER PERSON

PER YEAR

BILLION PASS-KM

PER PERSON

BILLION TON-KM

PER PERSONBulgaria 4.7 7.4 704 0.56 0.88Czech Republic 7.0 20.7 756 0.69 2.03Slovak Republic 3.1 12.4 575 0.57 2.30Estonia 0.2 4.8 179 0.13 3.20Hungary 6.8 8.1 645 0.67 0.80Lithuania 0.7 8.6 206 0.19 2.32Latvia 1.1 14.0 464 0.44 5.60Poland 20.6 67.7 516 0.53 1.75Romania 13.4 22.0 699 0.59 0.97Slovenia 0.6 2.6 314 0.30 1.30CEC average 7.0 16.7 523.2 0.55 1.60

19

Table 3: Total Travel Times of Alternative Land Transportation Modes

Origin Destination Intercity Bus Automobile Athens

(dist=508 km) Thessalonica (gateway to north)

5 h 50 min 7 h 20 min (80 km/h)

6 h 15 min (95 km/h)

Athens (dist=213 km)

Patras (main port to Italy)

3 h 30 min 3 h 35 min (80 km/h)

2 h 50 min (95 km/h)

Thessalonica (dist=341 km)

Alexandroupolis (near Turkey)

5 h 30 min 6 h 05 min (80 km/h)

4 h 55 min (85 km/h)

20

Exhibit 1. Major productivity statistics of Greek passenger and freight railroads.

Passengers (000s)

10000

11000

12000

13000

14000

70 74 78 82 86 90 94 98

Passenger-kilometres (million)

1200

1500

1800

2100

70 74 78 82 86 90 94 98

Income from passenger ops. (million drahma)

0

5000

10000

15000

70 74 78 82 86 90 94 98

Income from passenger ops. (const. '88 million)

0

2500

5000

7500

10000

70 74 78 82 86 90 94 98

Metric tons of freight (000s)

1000

2000

3000

4000

70 74 78 82 86 90 94 98

Metric ton-kilometres (million)

0

250

500

750

1000

70 74 78 82 86 90 94 98

Income from freight ops. (million drahma)

0

5000

10000

15000

70 74 78 82 86 90 94 98

Income from freight ops. (const. '88 million)

0

2500

5000

7500

10000

70 74 78 82 86 90 94 98

21

Exhibit 2. Selected vital statistics of Greece.

Per capita GNP in const. '88 U.S. $

4000

5000

6000

7000

8000

9000

10000

70 74 78 82 86 90 94 98

Agriculture's contribution to GNP

0%

5%

10%

15%

20%

25%

30%

35%

70 74 78 82 86 90 94 98

Resident population (million)

8.5

9.0

9.5

10.0

10.5

70 74 78 82 86 90 94 98

Inflation rate (%)

0

5

10

15

20

25

30

70 74 78 82 86 90 94 98

Gasoline price in const. '88 drahma

50

100

150

70 74 78 82 86 90 94 98

Tourist and Visitor arrivals (million)

0

2

4

6

8

10

12

70 74 78 82 86 90 94 98

Passenger cars (million licenses)

0

0.5

1

1.5

2

2.5

3

70 74 78 82 86 90 94 98

Exchange rate: Drahma to U.S. $1

0

50

100

150

200

250

300

70 74 78 82 86 90 94 98

22

DEPENDENT VARIABLE

INDEPENDENT VARIABLE(S) R

2 SEE as % of mean

β 1−α model number

Intercept 15,157.6 0.999GAS88 -20.3 0.994

OLYMPIC -0.0014 0.996

Intercept 13,783.5 0.999INFLATION -76.93 0.996OLYMPIC -0.0008 0.905

Intercept 12,523.2 0.999AR(1) 0.675 0.999

INFLATION -39.94 0.891

Intercept 1,607,254 0.999GNP88$ 60.68 0.973GAS88 -4,107.7 0.992

Intercept 1,131,949 0.995AR(1) 0.752 0.999

GNP88$ 72.98 0.887

Intercept 3,032.2 0.999INFLATION 56.42 0.995

TRUCKS -0.001 0.986

Intercept 3,276.9 0.999AR(1) 0.608 0.998

INFLATION 36.71 0.926TRUCKS -0.001 0.811

Intercept 946,943 0.999TRUCKS -0.544 0.999

Intercept 633,480 0.999TRUCKS -0.447 0.999GAS88 2,773.3 0.993

Intercept 672,674 0.999AR(1) 0.737 0.999

TRUCKS -0.453 0.991GAS88 2,229.8 0.988

RPASS

RPASSKM

RTONS

RTONKM

1

2

3

4

10

0.43

0.45

n.a.

0.39

n.a.

0.37

n.a.

0.71

5

19.2%

15.7%

15.3%

9

6

7

8

Table 3: Estimated Models

13.1%

9.2%

0.8

n.a.

6.6%

6.5%

5.5%

9.5%

7.0%

23

Exhibit 3. Actual and predicted measures.

Railroad Passengers (000s)

10,000

11,000

12,000

13,000

14,000

1970 1975 1980 1985 1990 1995 2000 2005

actual

M2

M3

Railroad Passenger-km (000s)

1,400,000

1,700,000

2,000,000

1970 1975 1980 1985 1990 1995 2000 2005

actual

M5

Railroad Freight (000s metric tons)

1,000

2,000

3,000

4,000

1970 1975 1980 1985 1990 1995 2000 2005

actual

M7

Railroad Freight (000s metric ton-km)

200,000

400,000

600,000

800,000

1,000,000

1970 1975 1980 1985 1990 1995 2000 2005

actual

M10

24

Figure 1: Partial map of Greece showing its entire rail network