New university
Economics & Law 2015. № 8-9 (54-55)
ISSN 2221-7347
СТАТИСТИКА
UDC 311.21: 338.47
V. V. Nosov* A.P. Tsypin **
RESEARCH INDICATORS OF RAILWAY TRANSPORT ACTIVITY IN RUSSIA
By using statistical and econometric methods the authors analyzed historical series of indicators ofpassenger and cargo rail turnover in the period from 1956 to 2014 and presented a forecast for the period up to 2016. The results of Chow test on the structural stability of historical series of indicators of passenger and cargo rail turnover has shown the absence of a general trend in historical series of these indices. In such cases, to describe the trends it is not acceptable to apply analytic alignment and it is better to use self-correcting recurrent models that take into account the value of previous levels, and provide an opportunity to obtain reasonably accurate projections of future levels. Forecasting of considered indicators, based on a two-parameter exponential smoothing model, showed a further growth in cargo turnover and a reduction in passenger turnover.
Keywords: railway transport, passenger turnover, cargo turnover, historical series, adaptive prediction method.
Significant territorial space of Russia determined the primary role of the transport system for the country's economic development. In the overall context of transport system development it should be also pointed to the special position of rail transport, so the bulk of goods in the country (more than $ 2 trillion ton-kilometers per year) was transported by rail. Also, the importance of this kind of transportation can be emphasized, pointing to the second place in the carriage of passengers, which is about 140 billion passenger-km. a year and it is second only to air transport.
The analysis of historical series of indicators of passenger and cargo rail turnover plays an important role in assessment of the status and prospects for development of rail transport. At the same time, historical series refer to a set of indicators arranged in chronological sequence spanning long periods of time, in ever-changing political, social and economic conditions in the country.
© Носов В.В., Цыпин А.П., 2015.
DOI: 10.15350/2221-7347.2015.8-9
Носов Владимир Владимирович - доктор экономических наук, профессор кафедры бухгалтерского учета и статистики, Российский государственный социальный университет, Россия.
Цыпин Александр Павлович - кандидат экономических наук, доцент кафедры статистики и эконометрики, Оренбургский государственный университет, Россия.
33
Новый университет
Экономика и право 2015. № 8-9 (54-55)
ISSN 2221-7347
From the presented definition it can be seen that not only time sequence of values of the studied index can be called as historical series, but only on base of this time sequence it will be possible to establish the laws of historical change formations.
The study of the dynamics of rail transport performance on the basis of historical series can be represented in the form of four successive stages:
1. To review the list of official statistical books, directories and yearbooks containing data on rail transport performance.
2. To complete historical series of indicators characterizing rail transport operation.
3. To study the dynamics and selection of similar periods in development of indicators that characterize the work of railway transport in the event of structural instability on the basis of Chow test.
4. To determine the econometric model to assess rail transport performance in the case of structural instability of historical series.
5. To forecast indicators characterizing rail transport operation on the basis of dynamic models.
In the first phase of the study, an important issue in drafting historical series is availability of
statistical data sources, which can be divided into two groups: direct and indirect (see Table 1). A very important fact is that information must be based on the official statistical reporting [1].
Table 1
Sources of information that characterize the work of Railway Transport in Russia
Sources The Soviet period (the Soviet Union) (19171991.) The modern period (RF) (1992-2014)
Direct Statistical books directly characterize the type of activity (industry) - Transport and Communications in the USSR (1933, 1934, 1935, 1936, 1972, 1990, 1991). - Transport and communications (1957, 1967). - Transport and communications in the RSFSR (1990, 1991). - Transport and Communications in the Russian Federation (1992, 1993). - Transport and Communications in Russia (1995, 1996, 1999, 2012, 2014). - Transport in Russia (2002, 2003, 2005, 2007, 2009). - Key indicators of transport in Russia (2004, 2006, 2008, 2010).
Indirect Statistical books of nationwide scale - The economy of the USSR, see "Transport and communications" (1958-1990). - The national economy of RSFSR, see "Transport and communications" (19581990). - Russia in Numbers, section "Transport" (1991) . - Statistical Yearbook, section "Transport" (1992) . - Regions of Russia. Social and economic indicators, section "Transport" (1992).
Information, contained in the above sources for different periods of time, allowed us to move to the next stage - to compare the levels of the studied phenomenon and to transit to historical series.
Considering rail transport performance over a long period of time, there is a problem of comparing the studied parameters. The main reasons, not comparable, that the authors had to face in the study are:
- territorial recognition inconsistency as a result of the Soviet Union collapse;
- the lack of information for a number of years [2].
The first problem was solved by the use of information for the period of 1956-1991 years of the Russian Federation, which is approximately comparable to the territory of modern Russia.
The second problem was solved by accounting missing levels based on the book "The national economy of the USSR" by adjusting of rail transport performance across the country on the average share attributable to the RSFSR.
Working with official materials and comparison of levels allowed eventually creating historical series of indicators characterizing the rail transport activity and, above all, such as passenger and cargo turnover (Figure 1; 2).
34
New university
Economics & Law 2015. № 8-9 (54-55)
ISSN 2221-7347
'vDOOOtN^t'vDOOOtN^t'vDOOOtN^t'vDOOOtN^t'vDOOOtN^t'vDOOOtN^t
1Л1ЛЩЩ'О'Л'ОГ'^М^Г'МХХХМ0\ФФа0\ООООО-н^н
0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\G\0\0\0\0\0\CDCDCDCDCDCDCDCD
Fig. 1. Dynamics of cargo rail turnover in Russia
year
According to these data in Figure 1 and 2, it can be concluded that there is a lack of the longterm trend throughout the period under review. It is evident that transition from a command control system to a market economy led to irreversible processes, which influenced negatively the considered indicators.
year
Fig. 2. Dynamics of passenger rail turnover in Russia
So until 1991, there was a stable growth in indicators’ performance, the annual increase in cargo turnover amounted to - 55.18 billion ton-km and passenger turnover - 4.93 billion pass.-km. You can see that the point of change for the trend is the value of cargo turnover in 1992, while for passenger turnover - in 1994, after which the prevailing trend over the years is replaced by a fall, which lasted until the flesh of the 2000s, when cargo turnover "reached" the level 50s, and in passenger turnover of 60-ies of the last century. The difference in points to reverse the trend have studied indicators due to the fact that as a result of the collapse of the Soviet Union to Russia from the former Soviet republics has been actively move to Russian-speaking population.
The consequences of default in 1998 in the Russian economy, as well as the arrival of new management in governing the country has resulted in a radical change in the course of market reforms
35
Новый университет
Экономика и право 2015. № 8-9 (54-55)
ISSN 2221-7347
on a more or less moderate that a positive impact on rail transport, and especially on the cargo turnover.
To confirm assumptions the authors put forward the null hypothesis H0 on structural stability of trend indicators characterizing railway transport operation, using statistical Chow test for this purpose [3] (Table 2).
Table 2
Results of Chow test for structural stability of historical series of passenger and cargo turnover
Historical series Points to reverse the trend F-Fisher statistics
Actual Tabular
Cargo turnover 1992, 1999 140,8 2,28
Passenger turnover 1994, 2000 268,8
According to values of F-Fisher statistics, it was concluded that > ^abular, i.e. the
hypothesis of structural stability under consideration of historical series was rejected, and it confirmed authors’ hypothesis regarding the structural instability of indicators’ dynamics and the need for homogeneous segments of development for building a high-quality model.
Chow test showed no general trend in historical series of considered indicators. Thus, it is impossible to describe the tendency using analytical alignment. For this purpose, it is better to use self-correcting of recurrent models that, characterizing time-varying dynamic features of the dynamics, take into account the value of previous levels, and provide an opportunity to obtain reasonably accurate projections of future levels.
Taking into account the presence of local trends in passenger and cargo turnover, you must use Holt two-parameter model [4], which is generally expressed by the following mathematical formulas:
'yt+T =at+ A
<at=alyt+(\- ax ){at_x + bt_x) (1)
bt = a2(at -at-x) + (1 -a2)bt-i
where yM - a forecast made for a T step forward; a - coefficient of the level number; b - a coefficient of proportionality;
ax, a2, - smoothing constants whose value ranges from 0 to 1.
The main task to be solved before the start of modeling - is a search for optimal values of the smoothing constant. For this purpose, based on exhaustive search of all possible options, using the procedure "Solver" in the RFP «Statistica», the authors chose two models with the following parameters:
- a model for historical series of passenger turnover, with parameters a = 0,9 and
a = 0,2.
- a model for historical series of cargo turnover, with parameters a = 0,9 and a = 0,9.
Having these values of smoothing constants in each model, the sum of squared deviations of mean squares will be the smallest of all possible calculated options.
The forecast of cargo and passenger turnover of railway transport is shown in Figure 3.
36
New university
Economics & Law 2015. № 8-9 (54-55)
ISSN 2221-7347
in in « ю ю « ю t^t^t^t^t^ooooooooooo'.o'.a'.a'.a'.ooooo-^-^-^-^ GsGsGsGsGsGsGsGsGsGsGsGsGsGsGsGsGsGsGsGsGsGsOOOOOOOOO
300.0
250.0
200.0
150.0
100.0
50,0
•D
—■— Cargo turnover, bln. t-km .... Smoothed levels of cargo turnover
—•—Passenger turnover, bln. pass.-km - - - - Smoothed levels of passenger turnover
Fig. 3. Forecast of cargo and passenger turnover of rail transport on the basis of Holt two-parameter model
The forecasts in Figure 3 show that under the influence of prevailing social and economic conditions the dynamics of rail transport in the coming period (2015-2016) will have different directions - cargo turnover in the forecast period is increased, and passenger turnover, by contrast, is reduced.
Turning to obtained forecasts, it is necessary to specify that the maximum value of cargo turnover equal to 2.606 trillion ton-km (1988), in current economic conditions will be achieved only by 2019 (assuming current continue trends). In turn, passenger turnover will continue to decline, and by 2015-2016 will reach the post-war years.
The main objectives to increase cargo turnover:
The main reasons for the decrease in passenger turnover are:
- a low level of population mobility in which the Russian Federation is far behind the developed countries, where rail transport meets only about 14% of total mobility;
- a decrease in the standard of living: the growth rate of spending on rail transport over the past 10 years consistently lagged behind the growth of revenues, population expenditures on goods and services.
- a reduce of the demand for railway transport services and increased amount of household expenditures on use of personal vehicles. Over the past 10 years, the amount of private cars increased by more than 10 million. Passenger car ownership now exceeds the passenger public transport of all kinds;
- limited investment sources;
- preservation of cross-subsidization;
- the lack of long-term decisions on tariff policies and public support;
- partial financing of social obligations by means of budget areas, reduction under the influence of routes and offers of transport solutions in the market;
37
Новый университет
Экономика и право 2015. № 8-9 (54-55)
ISSN 2221-7347
- infrastructure limitations which do not increase the volume of traffic and do not allow using the full speed capabilities of existing rolling stock.
Nevertheless, the forecasts of the Government of the Russian Federation are related to population transport mobility in wake of rising disposable income and increase human transport system to the level of developed countries. In value terms the market of public passenger transport could grow 3.5 - 4 times.
The most significant market growth in absolute terms will be linked to development of economies of the largest metropolitan area of the country, which will be concentrated in the social and economic development. It is estimated that by 2030 the proportion of agglomeration economies, Russia's GDP may reach 57-60%, there will live 35 to 37% of the population.
However, this requires development and supply of new products and services at competitive prices, ensuring the competitiveness of rail transport with major competitors - aviation and road transport.
The main market risks to increase cargo turnover are:
- unfavorable macroeconomic situation, economic stagnation in the coming years;
- the fall in commodity prices;
- advanced development of road and pipeline infrastructure;
- aggressive policy of global and local competitors.
Using these results, there was an attempt to estimate the loss of profit as a result of political and economic changes in the Russian economy. To do this, the authors used "stable" indicators of development segments and on the basis of these parameters the authors assessed linear trends [Studenmund, 2011, 326] and forecasted further development until 2014.
For cargo turnover was obtained following model:
Л956-1991 = 833,63 + 55,18t; R2 = 0,943 (2)
(0 (16,92) (23,76)
According to values of ^-Fisher statistics (Fctual = 564,58 > Ftabular(0,05;1;34) = 4,13). It
indicates the significance of the equation. In parentheses there are the calculated values of the t-test to test the hypothesis on the significance of the equation coefficients. In equation (2), all the parameters are significant because tactual > ttabular(0134) = 1,690 exceeds table value by 10%.
For passenger turnover was obtained following model:
Л956-1993 = 104,39 + 4,74t; R2 = 0,963 (3)
(0 (30,48) (30,95)
According to values of ^-Fisher statistics (Factual = 957,2 > Ftabular(0,05;1;36) = 4,11). It
indicates the significance of the equation. In parentheses there are the calculated values of the t-test to test the hypothesis on the significance of the equation coefficients. In equation (3), all the parameters are significant because tactual > ttabular(0136) = 1,688 exceeds table value by 10%.
The difference between the forecast and actual values (deviation) represent the desired value gap (loss of profits) (Figure 4).
38
New university
Economics & Law 2015. № 8-9 (54-55)
ISSN 2221-7347
2500
300
2000
J 1500
•s 1000
500
250
200
E
150
ь
100
50
(N(N(N(N(N(N(N(N(N(N(N(N
I Deviation of actual levels of cargo turnover from possible (while maintaining the trend of 1956 - 1991)
- Deviation of actual levels of passenger turnover from possible (while maintaining the trend of 1956 - 1993)
Fig. 4. The value of the backlog of passenger and cargo turnover on the trends established during the Soviet period of economy development
The dynamics that are represented in figure shows significant loss of profits. It is also worth remembering that behind the "dry" statistical calculations there is the real economy.
0
0
Conclusion
Summarizing the research of rail transport activity on the basis of historical series it is possible to make the following conclusions:
- statistical data accumulated in the field of transport allows us to get full (detailed) historical series of basic indicators over an extended period of time (over 50 years);
- dynamics of considered indicators are not stable over time and to describe and build forecasts it is better to use suited adaptive algorithms, in this case, the model reflects the complex qualitative trend of historical series;
- the analysis of historical series of indicators characterizing the operation of railway transportation has shown that in the considered period of time, there are three relatively stable development areas, and the most long-term stable is the first stage where it is possible to notice the rise of the studied parameters;
- forecasting of considered indicators, based on a two-parameter exponential smoothing model, showed a further growth in cargo turnover and a reduction in passenger turnover.
References
[1] Tsypin A.P. Kachestvo oficial'nyh statisticheskih materialov [The quality of the official statistical data]. // Intellekt. Innovacii. Investicii [Intelligence. Innovation. Investment]. - 2013. - №1. P. 88-93.
[2] Nosov, V.V., Kotar, O.K., Kosheleva, M.M., Alajkina, L.N. Novikova, N.A. Assessing effectiveness of insurance premium subsidy in agricultural insurance // Ecology, Environment and Conservation. 2014. - Vol. 20. - № 4. - P. 475-481.
[3] Chow, G.C. Test of equality between sets of coefficients in two linear regressions // Econometrica. -1960. - Vol. 28. - № 3. - P. 591-605.
[4] Brown, G.R. Smoothing, Forecasting and Prediction of Discrete Time Series. - N.Y.: Dover Phoenix Editions, 2004. - 468 p.
39
Новый университет
Экономика и право 2015. № 8-9 (54-55)
ISSN 2221-7347
УДК 311.21: 338.47
В.В. Носов, А.П. Цыпин
ИНДИКАТОРЫ ИССЛЕДОВАНИЯ ЖЕЛЕЗНОДОРОЖНОГО ТРАНСПОРТА В РОССИИ
С помощью статистических и эконометрических методов проанализированы исторические временные ряды показателей пассажирооборота и грузооборота железнодорожного транспорта за период с 1956 по 2014 годы и представлен прогноз на период до 2016 года. Результаты построения теста Чоу на структурную стабильность рядов показателей пассажирооборота и грузооборота железнодорожного транспорта показал отсутствие общей тенденции в исторических временных рядах данных показателей. В подобных случаях для описания тенденции не может применяться аналитическое выравнивание и следует использовать самокорректирующиеся рекуррентные модели, которые учитывают значения предыдущих уровней, и дают возможность получить достаточно точные прогнозы будущих уровней. Построение прогнозов рассмотренных показателей на основе двухпараметрической модели экспоненциального сглаживания показали дальнейший рост грузооборота и снижение пассажирооборота железнодорожного транспорта.
Ключевые слова: железнодорожный транспорт, пассажирооборот, грузооборот, исторические временные ряды, адаптивный метод прогнозирования.
40