Reviving the silk road: evidence from sino-russian trade

Размещено на http://www.allbest.ru/

Reviving the silk road: evidence from sino-russian trade

Introduction

chinese import silk road

The Silk Road Economic Belt and the 21st-century Maritime Silk Road also known as the Belt and Road, (B&R) initiative was unveiled by the Chinese authorities in 2013 and one of its objectives is to improve trade and commercial links between China and Eurasian countries. Since the late 19th century, Russia has invested substantial amount or resources to develop the infrastructure linking its European part to the Far East. Although this infrastructure has been built to increase domestic communication and commerce, it is currently used to serve international trade as well. This paper uses transaction-level data on Chinese imports into Russia to understand to what extent the existing infrastructure is used to ship goods from China to the European part of Russia and identify problems that hamper existing trade flows. The results of the analysis reveal valuable insights that may be taken into account in the process of construction and implementation of new trade routes from China to the West.

Currently, there are three land routes that link China to the European part of Russia. The first one passes through the territory of Kazakhstan, the second one through Zabaikalsk and the third one through Vladivostok, a city located on the south-eastern edge of Russia which can be accessed by both land and sea. The paper primarily focuses on the Chinese imports destined to Moscow because in the presence of new land routes, the distance from China to most Eastern European countries - which are the key part of the B&R initiative - by land, sea and air will be close to the ones to Moscow. The analysis of the data reveals that the volumes of trade shipped by land transportation measured by value are small. This observation is due to the fact that over 70 percent of total shipments are made by air cargo. However, land shipments are quite large in volumes in comparison with sea shipments to the Western parts of Russia.

It is commonly believed that land transportation occupies the middle ground between air and sea transportations in terms of both speed and cost. Thus, it is natural to expect the most expensive goods to be shipped by air, the intermediate ones by land and the cheap ones by sea. In fact, the analysis of this paper confirms that the average value of goods shipped by air is substantially higher than in the other two categories but the differences between goods shipped by land and sea are small. The travel time from Vladivostok to Moscow can take as little as two weeks if trains do not spend much time in multiple stations that they pass through, which is significantly faster than the approximate 38 days that takes to ship a container by sea to the western ports of Russia. However, in practice, cargo may spend significant time in the intermediate train stations and the overall duration may increase up to a month. Security concerns may also discourage firms to use railways as a mode of transportation because long waiting times at the stations increase the probability of theft. These issue should not be neglected in the process of exploitation of the new connection links from China to Europe because most countries located on the planned route have significant institutional problems and perform worse than Russia in international rankings. The mere existence of infrastructure may not be sufficient to substantially increase the volume of land shipments and meet one of the objectives of the B&R initiative. There is already some evidence that recently built railroads in developing countries operate at a fraction of their capacity. See, ”Loans, trains and automobiles,” The Economist, March 22nd 2018. This raises the question of the commercial viability of the project and the ability of developing countries to repay the debts (Hurley, Morris, and Portelance, 2018).

Another interesting pattern documented is that the distance to Moscow and a common border with Russia are an important determinant of provincial exports. In this context it is necessary to emphasize that we are not looking at the Chinese exports to Russia in general but to Moscow specifically which is quite far even from Chinese regions that are located close to the Russian border. This indicates that the construction of new land routes will create new trade opportunities for eastern provinces, such as Xinjiang, and contribute to the development of the region.

The remainder of the paper is organized as follows. Section 2 conducts the literature review. Section 3 describes the data and geographic. Section 4 describes policy implications. Section 5 presents the empirical specification. Section 6 provides results. Section 7 concludes and discuss.

Literature review

A considerable amount of literature is available on the topic of transportation infrastructure in international trade. A number of studies have found that firms engaged in international trade can choose among a large set of modes of transportation to reduce their costs and increase the speed of delivery (Harrigan, 2010; Djankov, Freund and Pahm, 2010; Hummels and Schaur, 2010; Hummels and Schaur, 2013). Harrigan (2010) shows the correlation between the distance to the destination point and the unit values of the products delivered. The author introduces a model of comparative advantage to predict that as country is farther as air shipments of goods are lighter. Hummels and Schaur (2010) provide evidence on uncertainty revealed by past price volatility and show that goods with high price volatility are more likely to be shipped by air. Djankov, Freund and Pahm (2010) estimate the gravity equation for affection of time delays on trade. The authors find that trade is reduced by more than 1% by each additional day of delays of the shipping. Furthermore, Hummels and Schaur (2013) report that the price elasticity of demand influence on firms' choice between air and ocean cargo. Air cargo is faster but maritime shipment is cheaper. The authors conclude that time delays in transportation could increase the value of the good from 0.6 to 2.1 percent. It can therefore be assumed that there is a link between sharp declines in the price of air shipping and rapid growth in the trade. Long time to delivery and expensive transportation can be associated with poor land transport infrastructure and institutional problems. Detmer, Freytag and Draper (2014) find that African trade integration is based upon a comparatively higher share of air cargo relevant products than Southern Africa's trade with industrialized and emerging economies. The reducing of transportation costs could be provide by more liberal market for air cargo services. It will allow Africa to integrate even further. In addition there is one more interesting study in this literature. Bonfatti and Poelhekke (2017) show that containerization could prevent stealing of the fright in Africa. The authors study that transportation infrastructure has been shaped by the location of mines and this has strong effect on the determination of trading patterns in non-mining goods as well. Container shipments provide great protection to the goods in road and maritime transportation, because it's a physical barrier against bad weather, theft, harm and impacts during handling. Furthermore, the adoption of container has significantly contributed to the increasing volumes of international trade (Bernhofen, El-Sahli, and Kneller, 2016). The authors document that one of the major benefits of containerization is that it reduced the pilferage, damage and theft and the insurance costs decreased on 17% per ton.

The role of rail cargo in trade is huge (Gasparik, Mesko, Zitricky, 2010). Rail cargo movement in developing countries is largely by road though several countries have well-developed railway networks (Kahn et al., 2007). With economic reforms and industrialization in the developing world, accompanied by the globalization, freight traffic is increasing with a growing demand for increased speeds and reliability. Moreover, Kahn et al. (2007) confirms that the share of the railways in the total freight traffic has been declining, while the share of roads and air transport has been growing rapidly. The share of freight transport through coastal shipping and inland waterways has not shown an appreciable increase.

In trade infrastructure gravity equation plays a crucial role that relates bilateral trade flows to GDP, distance between counties, common border and other factors that influence on barriers to trade. According to Anderson and Wincoop (2003), national borders in terms of US and Canada transactions reduce volume of trade between industrialized countries by moderate amounts from 20% to 50%. The paper is also related to Blonigen and Wilson (2008), who use US import data to estimate the gravity trade model that reveals that improved port efficiency significantly increases trade volumes. According to the authors port efficiency affects the international trade flows and country-level growth.

In recent years, there has been an increasing amount of literature on modes of transportation, including the cost of delays, value and weight of the freight. Authors show the relationship between the development of the transit links and economic growth of the regions. In addition to this the analysis of Chinese international trade, import and export is sufficiently covered in literature (Yu and Wei, 2012; Bilotkach, Gao, Grimme and Maioli, 2017). Yu and Wei (2012) also used Chinese transaction-level trade data and firm-level production data 2000-2006, to investigate various factors affecting trade processing. Bilotkach, Gao, Grimme and Maioli (2017) examine the data on goods imported into China from Europe and state that air cargo is highly correlated with developments in global trade and global GDP. The authors identify air accessibility index as a measure of air cargo market structure and regress import share by air and shipment price on this index, taking into account product, country, year, and firm specific heterogeneity. Bilotkach, Gao, Grimme and Maioli (2017) showed strong relationships between index and both the share of products shipped by air and shipment price on the other. This paper uses Moscow and Chinese provinces' GDP to estimate export to Moscow. Another essential result is that the easier the air cargo accessibility is, the higher share of import by air. It demonstrates that for larger firms the effects of air cargo accessibility are stronger.

The possibility of shipping goods from China to Europe gained an increasing attention in various circles. The main driver behind this is the China's B&R initiative but also because there was a commercial interest in using the Eurasian Landbridge rail services. The question has been widely discussed in political circles and in the media. Pomfret (2018) argues that rail Landbridge could reduce trade costs. And as social benefit rail transportation is more environmentally friendly than airfreight. There is also an increasing interest from academics, for example, Li, Bolton, and Westphal (2018). Authors focus on existing railway lines connecting Europe and China and demonstrate that the intercontinental railways have a positive effect on China's exports and imports of particular goods. In case of B&R initiative Maritime Silk Road connect Beijing with economic hubs around the world through the South China Sea, the South Pacific Ocean and the wider Indian Ocean area. But one interesting study describes another maritime route from China to Europe through the Northern Sea. Bekkers, Francois and Rojas-Romagosa (2016) describe commercial viability of the Northern Sea Route and compare with Southern Sea Route. Authors estimate the economic impact of Arctic route and find that it shifts average shipping distances between East Asia and Europe and transportation days by around one third. These reductions lead not only into transportation costs and fuel savings but also to significant transportation time savings that may effectively force trade between East Asia and Europe to change. In the case of rail shipments such approach may be misleading because, as the current study shows, in addition to physical distance, there are significant institutional issues that affect firms' decisions to choose the mode of transportation. In this sense, China's exports to Russia provide a unique setting to understand the scale of these problems.

The originality of this paper is studying the efficiency of transportation modes of different patterns of goods from different provinces of China within the framework of the B&R project.

1. Categories of goods

This paper uses transaction-level customs data from Russia. The dataset contains detailed information that is important for the objectives of the paper. These include origin and destination countries, provinces, firms, goods by HS10, modes of transportation, value and weight. The key points of interest are what kinds of products and how many of them are exported from China. The proportion of each goods' category by value and weight was calculated and then the dominant mode of transportation was identified.

As shown in Table 1, which provides descriptive information on the share of goods exported from China, the Machinery and Electrical equipment yield the largest volume of exported goods, accounting approximately 50 percent of China's total exported goods. Besides this industry three other industries - Textiles, Footwear, Headwear, Raw Hides, Skins, Leather, Furs - account for approximately 20 percent of China's total exported goods.

Another question is how all the goods reach the Russian ports on the borders. The modes of transportation that are of interest in this paper are train, truck, waterway and airfreight. According to the fifth column of the Table 2, it can be seen that Machinery & Electrical equipment are transported by air cargo. It means that predominantly expensive goods are transported by air. Rail transportation is used for stone and glass that is intuitive. Furthermore, the exporters equally choose truck and train in most of the cases. Vegetable Products and Animal Products shipped to Moscow by air cargo because lots of products are perishable goods (Hummles and Schaur, 2013). Otherwise Textiles and Mineral Products are transported by trucks because these goods are lightweight and do not require careful transportation.

Table 1. Ranking of goods by HS 2-Digit

Notes: The first column represents the HS 2-digits code. The second column shows the name of category according to international classification. The third column is the share by value of goods by HS 2-digits over China's total export to Moscow. The fourth column demonstrates the share by weight of goods by HS 2-digits over China's total export to Moscow. The fifth column shows the dominant mode of transportation for each category. After that the research examines the origin province of the goods in China and modes of transportation from. Figure 1 represents the dominant mode of transportation from each province. Coastal provinces such as Shanghai, Hong Kong and Zhejiang choose maritime transportation. Only from one of the nearest provinces Xinjiang goods are exported by land. However the export from Xinjiang is quite low. Beijing is located 670 km from the border of Mongolia and 170 km from the coast and it can be concluded that goods can be transported by sea or by land. However, Figure 1 shows that goods are mostly transported by air from the Beijing. Overall the biggest value of the goods are transported from China to Moscow by air.

Figure 1. Chinese provinces by dominant mode of transportations by value

2. Geographic Setting

Sino-Russian trade presents a unique opportunity to study the transportation mode choice because both of the countries occupy large territories and there are multiple options that can be used to ship goods. The origin location of delivery is China. The destination of goods is Moscow, as it is located far enough from the border with China (the distance between Chinese provinces and Moscow is from 4000 to 9000 km), which allows to use different modes of transportation, as well as directly close to Europe, which allows to reproduce trade routes between China and Europe. In particular, this dataset provides information on the border crossing points. Customs on the Russian border are specified according to the customs code list. According to the geographical location, the existing 224 customs on the Russian border were combined into enlarged geographical entities. Figure 2 describes each of the options that are used to ship goods to Moscow.

Figure 2. Shipping options in Russia

Moscow. The fastest option available to Chinese exporters is to ship goods by air directly to the Moscow's airports.

Saint Petersburg. There are several sea ports in Saint Petersburg. This option is similar to the main maritime trade route that takes place between China and Europe. To reach Saint Petersburg ships need to travel some additional distance compared with the main Northern European ports but this distance is very small when one looks at it in the context of the overall distance from China to Northern Europe. From Saint Petersburg goods need to be shipped to Moscow by land transportation. This land distance is not very large and it is comparable to those from Hamburg to Munich or Vienna.

European ports. Substantial amount of goods shipped to Russia by sea are unloaded at the European ports and then delivered by land. The customs located on the border of Russia with Estonia, Latvia and Belorussia were united into one common point. Lithuania plays an important role in this process because of its location. It can be reached faster by the ships traveling from the West and it is relatively closely located to Moscow. Some companies may prefer to use lorries to directly deliver their goods to Moscow from European ports rather than wait for other vessels to deliver their goods to Saint Petersburg after which they still need to use land transportation to reach Moscow. However, this phenomenon has to do not only with geography but also with port efficiency and limited capacity in Russian ports. An important factor that needs to be taken into account is that Russia is an oil exporting company and container ships need to compete with oil tankers for space.

Black Sea. Goods can be delivered to Moscow via ports located on the Black Sea. According to the data it can be observed that shipments arriving from Ukraine through customs in Bryansk however these shipments are relatively small compared with those arriving directly at Russian ports. For these reason Ukraine is not a separate option but a part of the shipments to the Black Sea. The main Russian port on the Black Sea is Novorossiysk.

Kazakhstan. Xinjiang province located in the North-West of China shares a long border with Kazakhstan. Kazakhstan has a common border with China and Russia and is an important intermediary in trade between the two countries. Favorable geographical position allows to trade on land. For example, Xinjiang province is located in the North-West of China and shares a long border with Kazakhstan. From this province goods can be delivered to Russia through Kazakhstan both by train and by trucks. In Russia territory in the immediate vicinity of Kazakhstan is the customs in Omsk. It should be noticed that this province also shares a short border with Russia but there are no roads in that area because the terrain is very mountainous and both sides of the border are not populated.

Zabaykalsk. Further to the East goods can be shipped to Russia through Zabaykalsk. The trans Mongolian Railway on the one side is connected to the the Chinese city of Jining and on the other side to the Trans-Siberian Railway which makes the delivery of goods to Moscow possible.

Vladivostok. Vladivostok is a coastal city located on the South-Eastern edge of Russia. It can be relatively easily accessed by land from the industrial north-eastern industrial cities of China such as Harbin. Furthermore, goods can be shipped to Vladivostok by sea from the coastal regions of China from where they can be shipped by Trans-Siberian Railway to Moscow. However, goods from China can be delivered to the land in Vladivostok, for example, from the province of Heilongjiang.

Table 2 provides descriptive information on the share of goods imported through different border segments and their characteristics. As can be seen, the largest share of Chinese exports, measured by value, arrive in Moscow by air. For Europe it is only about 28 percent of import measured by value (Hummles and Schaur, 2013). The following largest categories are maritime shipments to Saint Petersburg and to European ports from where they are shipped by land to Moscow. The Black Sea shipments play relatively smaller role by accounting for only one percent in total value. Shipments through Kazakhstan which is a land route are even smaller. The next land route located further to the East, through Zabaykalsk, plays a more important role by accounting for 2.5 percent of total shipment. The volume of shipments passing through Vladivostok is slightly higher.

Table 2. Exports by Routes

Notes: The first column reports the share (in percentages) of each entry point as a share of total shipments measured by value. The second column reports the share (in percentages) of each entry point as a share of total shipments measured by weight. The third column reports the average value (in Russian rubles) divided by weight (in kilograms). The forth column reports the main mode of transportation.

The second column provides information on the shares of goods shipped through various border segments measured by weight. It is not surprising that the share of air shipments decreases substantially. It can be noticed, that among other shipping routes, the share of goods passing through Vladivostok is higher but the difference is not as noticeable as in the case of air cargo to Moscow. For other shipping options there are some differences compared with values but overall the picture does not change much. The third column, provides information on the ratio of the average value of a shipment (measured in Russian rubles) divided by weight (kilograms). This measure for Moscow is substantially higher than for other options, indicating that producers of expensive goods face higher waiting costs. The ratio does not differ much for other modes of transportation. This is an interesting result because in usual circumstances land routes provide more rapid transportation and may be used by the producers of medium level of value/weight ratio goods for whom air cargo is too expensive and maritime option too slow. However, in the case of Sino-Russian trade such expectations are not confirmed.

The fourth column provides information on the main mode of transportation. For all border segments the main mode accounts for at least 94 percent of shipments. For Moscow and western ports the transportation mode is quite intuitive and does not deserve much discussion. The situation is more interesting for eastern entry points. In the case of Kazakhstan the main mode of transportation is truck accounting for 94 percent. The second most popular option is train. It should be mentioned that there exist infrastructure that allows to ship goods by train but as the data shows this option is not used intensively. Over the last few years Kazakhstan has invested heavily in its railroad infrastructure and developed a Khorgos Gateway terminal to increase its role as a hub connecting China to the West. Khorgos Gateway is primary dry port for handling trans-Eurasian trains. See, “Khorgos: The New Silk Road's Central Station Comes To Life,” Forbes, February 20th 2017. However, even though there is growth at Khorgos Gateway, the dry port is still smaller capacity than Dostyk port. Most cargo volume that passes from China to Kazakhstan goes through the 50% Russian-owned Dostyk port. In the case of Zabaykalsk the situation is different, over 99 percent of shipments are made by train. Finally in the case of Vladivostok 94 percent of shipments arrive by sea, 4 percent by train and the remaining by truck. Although the predominant mode of transportation to Vladivostok is waterway it should be considered as a mixing between waterway and train because after unloading, these goods travel to Moscow by train and the land distance is significantly longer compared with the sea distance from the Chinese ports to Vladivostok.

Policy implications

The relatively low share of goods exported by rail may seem surprising and requires some discussion. Cargo trains travel from Vladivostok to Moscow in 13-16 days and it takes another 4 days to ship from China to Vladivostok This information is available on the site of transport companies, for example Zheldor Alyans: Cargos Vladivostok - Moscow.. The travel time by sea from the East cost of China to the western ports of Russia takes about 38 days. Cargo trains are faster approximately on 18 days. Of course, if vessels travel directly to Saint Petersburg without any stops it will take less time, however a single manufacturing firm rarely can rent the entire ship. In practice, most firms send one or several containers by a vessel which makes multiple stops on its way. In fact, the container may be delivered to the final destination by a different vessel. Several online sources were used to get these durations. One source is the website of the world's largest shipping company Maersk. Also were used information from VesselFinder and MarineTraffic and track specific vessel traveling from Chinese ports to Russian ones. Based on the numbers presented above, it is clear that the delivery of goods is much faster by train compared with sea which means that there are some factors that deter firms to use train deliveries.

Theft on roads and railroads is a big concern in Russia. In 2009 the Russian Government created a special state enterprise designated to prevent theft from railways. According to the official data that RZD published about thefts from trucks: in the first half of 2016 25 cases of robbery were recorded, 1371 theft and 1095 cases of fraud. In the first 8 months of 2017 more than 300 cases of theft of rail cargo were recoded. See, “Bezopasnost perevozok - prioritetnaya zadacha vedomstvennoy okhrany,” RZD-partner, September 25th 2017. According to Kritskaya (2007) Критская Ю. (2007). 67,8% crimes were committed on railroads, most of which were committed during parking at large freight stations. The most popular goods to theft are Metals (16,7%); Foodstaff (15,1%); Clothes and Footwear (15,0%).

Using the delivery dates provided above it is possible to conduct an equation of the expected losses from train deliveries.

Theft_pr = Time_Cost * (Maritime - Rail)

Hummels and Schaur (2013) estimate that an additional day in transit is equivalent to an ad valorem tariff of 0.6 to 2.1 percent. In comparison between sea and rail shipments it makes more sense to pick the lower bound of the estimates (0.6 that named Time_Cost) because the most sensitive ones are shipped by air. Further assume that sea shipments take about 18 additional days compared with rail cargo. Calculations of the form implies that there is about 10 percent likelihood that in the case of shipping by rail goods will not arrive in the destination in the desired state.

Empirical specification

In the dataset there is significant variation among Chinese provinces in several dimensions which can be exploited to understand factors determining their trade with Russia and discuss implications for the potential trade in the presence of new routes. To find out how the geographic and economic factors affect the exporting patterns to Moscow and transportation mode choices the provincial gravity equation is used. Gravity equation is estimated with Poisson pseudo-maximum-likelihood (PPML) method. This is the preferable approach because it accounts for zeros and heteroscedasticity in trade data (Silva and Tenreyro, 2006). The authors propose a simple pseudo-maximum-likelihood (PML) estimation technique for gravity equation and use the PML estimator based on the Poisson distribution to quantitatively assess the determinants of bilateral trade flows. In their study the comparison between the PPML estimates and those generated by OLS in the log linear specification, using both the traditional and the fixed-effects gravity equations, showed that the Poisson regression gives less weight to the observations with larger variance, without giving too much weight to observations more prone to contamination by measurement error. The results of this paper also contain the comparison results of OLS estimations with PPML estimations.

Dataset contains information about the address of company which has made the shipment. This information is used to assign each shipment to a province. To implement the gravity equation in the context of this paper, Chinese exports are aggregated at HS 4 level for each province. Overall there are 32 Chinese provinces which make positive shipments (including Hong Kong) to Moscow.

Among geographic factors the distance from the provincial capital to Moscow is considered. The distance from the coast may also affect exporting patterns and transportation mode choice especially in the case of sea shipments. In this paper, a dummy for provinces that share a border with Russia was included. Even if a province shares a border with a Russian region, it is still located far enough from Moscow. For example the nearest Chinese province Xinjiang to Russian border is located 4300 km from Moscow.

Chinese provinces differ in the level of their development and, according to the Linder hypothesis, this factor also affects trade patterns. The main idea behind this hypothesis is that the regions with more similar levels of income trade more intensively with each other. For this reason the absolute difference between the provincial GDP per capita and that in Moscow was included. Another factor to consider are total exports of Chinese provinces. This variable plays an important role because Anderson and van Wincoop (2003) show that in order to obtain unbiased gravity estimates it is essential to include region-specific fixed effects to account for multilateral resistance term. Since each region exports only to one destination, the inclusion of province-level fixed effects is not feasible. The inclusion of total exports can eliminate this problem because it captures all factors that make a specific region more/less export-oriented in general and the focus is on the deviations from this level that are driven by province level factors relevant for the trade with Russia.

The descriptive statistics for variables used in the regressions are presented in Table 3.

Expij =?в0 +в1Exptoti +в2Linderi +в3dMoscowi +в3dCoasti +в4Borderi +лj +еij,

where

· Expij are export to Moscow from province i in good category j;

· Exptoti are total export to the rest of the world;

· Linderi is the absolute difference between GDP per capita in the destination and source;

· dMoscowi is the distance to Moscow;

· dCoasti is the distance to coast;

· Borderi indicator for a common border;

· HS2-level fixed effects (лj).

Table 3. Descriptive statistics

Notes: This table reports the summary statistics of variables that were used in regressions. All variables are in logs. Shipment values are in Russian rubles (logged). Land refers to exports through Kazakhstan and Zabaykalsk. Sea refers to all shipments arriving in Saint Petersburg, European ports and the Black Sea.

Results

The results of the Poisson pseudo-maximum-likelihood (PPML) estimations are presented in panel A of Table 5. Panel B presents the results of OLS estimation for a comparison. The first column displays the results for total imports at HS 4 level, regardless of the mode of transportation. The overall results are in line with basic intuition and literature in general. Total provincial exports are positively associated with exports to Moscow. Deviations from the GDP per capita in the destination negatively affect the volume of exports. The distance to Moscow and the distance to the coast, both have negative effect on exports and a common border with Russia has a positive effect on exports.

The following column presents the results for air cargo shipments to Moscow. The results are very similar to the previous one, the only exception is that the com- mon border dummy is not significant any more. In the third column the dependent variable is the shipments by land which combine shipments through Kazakhstan and Zabaykalsk. In this specification the distance to coast variable becomes statistically insignificant. The size of the estimated coefficient on the distance from Moscow is much larger. These results are intuitive because for land shipments the distance to coast is less relevant and the direct distance to Moscow plays a bigger role.

In the last column the dependent variable is the shipments to Moscow by sea through western ports (St. Petersburg, Europe, Black Sea). One interesting result that emerges is that common border is significant and the estimated coefficient is quite large. Asymmetric entry costs and uncertainty in destination export markets may help to explain this result. Albornoz, Pardo, Corcos, and Ornelas (2012) documented that many firms enter into new markets to test the demand for their goods. In this context, a plausible explanation could be that firms that are located in provinces close to the border face relatively lower entry costs when they enter into regional markets. After the entry they can test the demand for their goods, obtain practical experience how to deal with problems in Russia and establish new business contacts. Using these valuable experience and information they can later enter into more distant and larger markets more easily compared with firms that have no such experience.

Table 5. Determinants of Provincial Exports

Notes: PPML estimations where the dependent variables are the following: (1) total export; (2) shipments to Moscow; (3) shipments through Kazakhstan and Zabaykalsk; (4) shipments arriving at Saint Petersburg, European ports and the Black Sea. Panel A presents regression the results for PPML estimations. Panel B presents the results for OLS regressions. Regressions include HS 2-digit fixed effects. Standard errors are reported in parentheses. * (**) (***) indicates significance at the 10 (5) (1) percent level. The dependent variable does not include shipments to Vladivostok.

Conclusion

The findings of this paper have important implications for the development of the transportation infrastructure and trade policies in China. Results show that there is significant difference in the value of goods shipped by air versus those transported by rail and maritime, which means that there is no competition between air and others. The value characteristics of goods shipped by land and sea do not differ much. The percent of maritime shipments is higher than railway shipments despite the fact that rail is faster. Existing land roads are not used very intensively by Chinese exporters. The reasons could be different: low road infrastructure, institutional problems and etc. Also in the paper the probability of infrastructure problems on the road was calculated. In the case of shipping by rail goods with 10 percent likelihood will not arrive in the destination in the desired state. Institutional and organizational issues may play an important role, in addition to physical infrastructure. The EIU's Risk Briefing service analyzes the cost of bad planning the OBOR initiative project for business profitability in each of the countries. See, “Prospects and challenges on China's `one belt, one road': a risk assessment report,” The Economist, 2015. The authors consider a rating of business risk indicators by scores. The first indicator in this rating is security risk including violent and organized crime. According to the current paper it can be considered that institutional and organizational issues may play an important role, in addition to physical infrastructure. Resources should be invested in the security of the B&R project and, most importantly, in developing institutional infrastructure of land transportation in order to decrease thefts. After the reduction of crime levels on rail and truck shipping, exporters will probably switch their mode from expensive air freight to cheaper land cargo and from slow maritime shipment to fast railway cargo.

The results of gravity equation show that the distance to Moscow, which is far enough even from Chinese regions and a common border with Russia, are important determinant of provincial exports. This demonstrates that the new land road infrastructure will develop new trade opportunities for eastern provinces, such as Xinjiang, and contribute to the development of the region. New land routes will create more trade opportunities for eastern provinces, such as Xinjiang. Xinjiang is the core area of the Silk Road Economic Belt, which shares a border with eight countries including Mongolia, Russia, Kazakhstan, Kyrgyzstan, Tajikistan, Afghanistan, Pakistan and India.

Future research of the Sino-Russian trade in the framework of the B&R initiative could include other large destination ports such as Vladivostok, Ekaterinburg or Novosibirsk. Also it is interesting to observe the Chinese imports destined to Russia overall.

References

1. Albornoz, F., H. F. C. Pardo, G. Corcos, and E. Orleans (2012). Sequential exporting. Journal of International Economics, 88, 17 - 31.

2. Anderson, J. E. and E. Van Wincoop (2003). Gravity with Gravitas: A Solution to the Border Puzzle. American Economic Review, 93, 170-192.

3. Bekkers, E., J. F. Francois, and H. Rojas-Romagosa (2017). Melting Ice Caps and the Economic Impact of Opening the Northern Sea Route. The Economic Journal, 128, 1095-1127.

4. Bernhofen, D., Z. El-Sahli, and Kneller R. (2016). Estimating the effects of the container revolution on world trade. Journal of International Economics, 98, 36-50.

5. Bilotkach, V., Gao B., Grimme W. and Maioli S. (2017). Air Cargo Market Structure, Intermodal Competition, and Prices: Evidence from Chinese Imports from Europe. Working paper.

6. Blonigen, B.A. and Wilson W.W. (2008). Port Efficiency and Trade Flows. Review of International Economics. 16, 21-36.

7. Bonfatti, R., Poelhekke, S. (2017). From mine to coast: transport infrastructure and the direction of trade in developing countries. Journal of Development Economics, 127, 91-108.

8. Dettmer, B., Freytag A. and Draper P. (2014). Air Cargo beyond Trade Barriers in Africa. Journal of Economic Integration, 29, 95-138. ?

9. Djankov, S., Freund C. and Pham C. (2010). Trading on Time. The Review of Economics and Statistics, 92, 166-173. ?

10. Gasparik, J., Mesko P., Zitrick V. (2010). Role of rail cargo companies in logistic chains. Railway logistics and rail cargo: proceedings of the 5th SoNorA University Think Tank Conference, 19-25.

11. Harrigan, J. (2010). Airplanes and Comparative Advantage. Journal of International Economics, 82, 181-194.?

12. Hummels, D. L. and G. Schaur (2010). Hedging Price Volatility using Fast Transport. Journal of International Economics 82, 15?25.

13. Hummels, D. L. and G. Schaur (2013). Time as a Trade Barrier. American Economic Review, 103, 2935-59.

14. Hurley, J., S. Morris, and G. Portelance (2018). Examining the Debt Implications of the Belt and Road Initiative from a Policy Perspective. Policy rep., Center for Global Development.

15. Li, Y., K. Bolton, and T. Westphal (2018). The effect of the New Silk Road rail- ways on aggregate trade volumes between China and Europe. Journal of Chinese Economic and Business Studies, Working paper, 1-18.

16. Kahn Ribeiro, S., S. Kobayashi, M. Beuthe, J. Gasca, D. Greene, D. S. Lee, Y. Muromachi, P. J. Newton, S. Plotkin, D. Sperling, R. Wit, P. J. Zhou, 2007: Transport and its infrastructure. Mitigation. B. Metz, O.R. Davidson, P.R. Bosch, R. Dave, L.A. Meyer (Eds.). Climate Change 2007: Contribution of Working Group III to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge: United Kingdom, Cambridge University Press (pp. 323-85).

17. Kritskaya, J.V., (2007). Criminological features theft, the crime is groups in rail transport. Retrieved from http://teoriap-ractica.ru/rus/files/arhiv_zhurnala/2009/1/kritskaya.pdf

18. Pomfret, R. (2018). The Eurasian Landbridge: Linking regional value chains. Retrieved from https://voxeu.org/article/eurasian-landbridge-linking-regional-value-chains

19. Silvia, J. M. C. S. and S. Tenreyro (2006). The Log of Gravity. The Review of Economics and Statistics, 88, 641-658.

20. Yu M. and Wei T., (2012). China's Processing Trade: A firm-level analysis. In McKay H., Song L.(Eds.), Rebalancing and Sustaining Growth in China (pp. 111-148). Australia, Canberra: ANU Press.

Размещено на Allbest.ru